**** BEGIN LOGGING AT Mon May 21 02:59:59 2012
May 21 06:12:05 <DocScrutinizer> GeneralAntilles: we need a +b cleaning day eventually
May 21 06:13:45 <DocScrutinizer> almost too much for freenode ;-D
May 21 06:19:48 <DocScrutinizer> oops, IPv6 might be static
May 21 06:19:56 <DocScrutinizer> well, let's see
May 21 06:39:38 <ruskie> http://zareason.com/shop/zatab.html
May 21 06:39:39 <archeyDevil> Estel_: I don't like mounting via usb for two reasons. A. It makes a part of the storage unavalible to the device it's self and requires udev on my arch to work with the storage correctly (Which doesn't always work with phones in my experience)
May 21 06:39:58 <archeyDevil> Sicelo: I'll look into it.. Thanks :)
May 21 06:46:48 <Sicelo> archeyDevil: actually Estel_ meant usb-networking :P
May 21 06:47:10 <archeyDevil> Meh.
May 21 06:47:20 * archeyDevil doesn't like teathering through usb :P
May 21 06:47:30 <archeyDevil> dbus-send --type=method_call --dest=org.freedesktop.Notifications /org/freedesktop/Notifications org.freedesktop.Notifications.SystemNoteInfoprint string:"This is a test" 
May 21 06:47:38 <archeyDevil> Anyone know how to make that notification message stay?
May 21 06:48:32 <DocScrutinizer> use SystemNoteDialog
May 21 06:48:43 <DocScrutinizer> instead of SystemNoteInfoprint
May 21 06:49:30 <DocScrutinizer> $dbussend --type=method_call --dest=org.freedesktop.Notifications \         /org/freedesktop/Notifications \         org.freedesktop.Notifications.SystemNoteDialog \         string:"`echo -en "Shutting down 5V supply for USB! Reason:\n$errortext2"`" uint32:0 string:"WARN";
May 21 06:54:53 <archeyDevil> DocScrutinizer: Thanks.. Just one more question... How to get that as a function (shell)?
May 21 06:55:07 <archeyDevil> so mesg "this is a message"
May 21 06:55:25 <archeyDevil> Maybe a different function name.. mesg is taken :P
May 21 06:55:40 <archeyDevil> Just msg then.
May 21 06:56:52 <Sicelo> you can also check out `phone-control` package that does (almost) the  same with less 'typing' :P
May 21 06:57:46 <archeyDevil> Haha
May 21 06:58:04 <archeyDevil> so ssh method to use other parts of the phone?
May 21 06:58:18 <archeyDevil> What repository?
May 21 06:59:42 * archeyDevil can't find the package :(
May 21 06:59:55 <Sicelo> extras, i suppose
May 21 07:00:26 <archeyDevil> apt-get install ...?
May 21 07:00:35 <archeyDevil> E: Couldn't find package phone-control
May 21 07:00:46 * archeyDevil can't find it in the gui either..
May 21 07:01:04 <archeyDevil> even with the search terms "phone" nor "control"
May 21 07:01:39 * archeyDevil installs qtlockscreen :)
May 21 07:01:58 <Sicelo> http://maemo.org/packages/view/phone-control/
May 21 07:03:34 <archeyDevil> It's in extra-devel.. Going to add it :)
May 21 07:06:48 <archeyDevil> Repo added :)_
May 21 07:08:00 <archeyDevil> Nokia-N900:~# apt-get update; apt-get upgrade; apt-get install phone-control
May 21 07:08:02 <archeyDevil> :)
May 21 07:09:13 <archeyDevil> cutetube updated overnight? o.O
May 21 07:09:20 * archeyDevil hopes it works for downloading now :)
May 21 07:21:38 <archeyDevil> how to use phone-control --ledon Sicelo ?
May 21 07:23:53 <DocScrutinizer> http://wiki.maemo.org/Phone_control
May 21 07:26:54 <archeyDevil> o.O Thanks
May 21 07:28:56 <archeyDevil> Possible to change what colors for what events? And create new events? ( For the led )
May 21 07:29:27 <DocScrutinizer> err, the notification led?
May 21 07:29:34 <archeyDevil> Yea.
May 21 07:31:52 <DocScrutinizer> http://wiki.maemo.org/LED_patterns
May 21 07:32:13 <DocScrutinizer> etc/mce/mce.ini
May 21 07:33:38 <archeyDevil> Hmm.
May 21 07:34:13 <DocScrutinizer> http://wiki.maemo.org/LED_Pattern_Editor
May 21 07:44:28 <archeyDevil> There better alternatives to qtlockscreen?
May 21 07:49:33 <Sicelo> gosh.. archeyDevil - careful with apt-get upgrade, especially with devel enable
May 21 07:49:56 <archeyDevil> Sicelo: XD I gave it lower priority.
May 21 07:50:15 <archeyDevil> http://wiki.maemo.org/Extras-devel -- Mentions it
May 21 07:51:25 <archeyDevil> At the end of the page.
May 21 07:52:00 <Sicelo> i know, lol
May 21 07:52:30 <archeyDevil> :)
May 21 07:52:45 * archeyDevil isn't a complete debian noob :P
May 21 07:53:03 * archeyDevil used to run with Debian for his desktop.... Since I tried Arch; No going back :D
May 21 07:53:05 <DocScrutinizer> maemo != debian
May 21 07:53:23 <archeyDevil> DocScrutinizer: Ture. However, apt-get == debian.pkgmgr
May 21 07:53:56 <DocScrutinizer> but maemo / nokia repo is a bit... unusual
May 21 07:54:06 <archeyDevil> True.
May 21 07:54:13 <archeyDevil> Debian freemantle correct?
May 21 07:54:21 <archeyDevil> Or is that the maemo version?
May 21 07:54:31 * archeyDevil doesn't remember freemantle in debian history.
May 21 07:54:39 <DocScrutinizer> fremantle
May 21 07:54:52 <DocScrutinizer> fremantle is maemo5
May 21 07:54:53 <archeyDevil> s,ee,e, 
May 21 07:54:58 <archeyDevil> Okay :)
May 21 07:55:08 <archeyDevil> Maemo going to get out of beta? ;)
May 21 07:55:15 <DocScrutinizer> eh?
May 21 07:55:34 <archeyDevil> http://maemo.org/ The header :P
May 21 07:55:36 <DocScrutinizer> GeneralAntilles kicked somebody for that question like 6h ago
May 21 07:55:38 <archeyDevil> beta
May 21 07:56:21 * archeyDevil wonders why it says beta yet almost no development since 2011 (afaik)
May 21 07:56:24 <DocScrutinizer> maemo.org seems got a bit messed up lately
May 21 07:56:53 <archeyDevil> Okay..
May 21 07:56:55 <DocScrutinizer> anyway that beta never was about maemo, it always been about the website
May 21 07:57:02 <archeyDevil> Ohhh.
May 21 07:57:05 <archeyDevil> Fair enough.
May 21 07:57:54 <DocScrutinizer> and I honestly wonder who messed up maemo.org html again
May 21 07:58:06 * archeyDevil wonders if anyone has openssh on their maemo device and don't have passwords disabled... :P (Very unsecure assuming data plan w\ enough signal)
May 21 07:58:33 <DocScrutinizer> there's no way to use ssh without password
May 21 07:58:45 <archeyDevil> Yes there is..
May 21 07:58:56 <archeyDevil> ssh-keygen; ssh-copy-id etc
May 21 07:59:23 * archeyDevil blocked password login for ssh. Keys required.
May 21 07:59:27 <archeyDevil> key *
May 21 07:59:42 <DocScrutinizer> oops, sorry I misread
May 21 07:59:50 * archeyDevil has keys for all his computers ( and now phone )
May 21 07:59:52 <DocScrutinizer> paswords DISabled
May 21 08:00:10 <eddyb> this is weird, ~/MyDocs is read-only, how did that happen?
May 21 08:00:30 * archeyDevil is root so doesn't notice ;)
May 21 08:00:50 <archeyDevil> drwxrwxrwx   14 user     root        65536 May 21 17:26 MyDocs
May 21 08:01:02 <archeyDevil> eddyb: just chown 700 to the folder.
May 21 08:01:06 <psycho_oreos> eddyb, unclean mount?
May 21 08:01:11 <eddyb> no idea
May 21 08:01:12 * archeyDevil knows that is 777 but fixing that up now.
May 21 08:01:13 <psycho_oreos> that's a bad idea
May 21 08:01:16 <eddyb> why is it vfat?
May 21 08:01:20 <archeyDevil> chmod 700 *
May 21 08:01:36 <psycho_oreos> for users who have windows PC or otherwise non-*nix based PC
May 21 08:01:41 <eddyb> but...
May 21 08:01:52 <eddyb> how do I copy 13GB of files onto it, then?
May 21 08:02:17 * archeyDevil used sshfs
May 21 08:02:47 <archeyDevil> apt-get install openssh; sshfs from desktop. Mounted phone via ssh.
May 21 08:02:52 <psycho_oreos> eddyb, remount it as rw, make sure no other programs are using it.. though I'd still check dmesg
May 21 08:03:09 <psycho_oreos> sshfs won't work even if the fs is mounted as read-only.. not even root can write to it
May 21 08:03:37 <eddyb> I managed to delete some files
May 21 08:03:47 <eddyb> but not copy others
May 21 08:04:24 <eddyb> now, the app (evopedia) can only load files from one of the vfat partitions
May 21 08:04:30 <eddyb> and my files are 13GB
May 21 08:04:39 <eddyb> that won't fit in a vfat partition
May 21 08:04:41 <eddyb> :(
May 21 08:04:45 <archeyDevil> psycho_oreos: True... 
May 21 08:04:54 <archeyDevil> eddyb: df -h
May 21 08:04:58 <eddyb> hmm?
May 21 08:05:20 <archeyDevil> Ehh. dw
May 21 08:05:28 * archeyDevil crys at df w\ busybox
May 21 08:05:56 <archeyDevil> no -T :'(
May 21 08:06:03 <psycho_oreos> ~messybox
May 21 08:06:04 <infobot> messy... err busybox is meant for lean scripting. Regarding all the missing options and immanent limitations (see su) it's not really the interactive shell of choice. A lot of people hate busybox because a lot of system integrators don't understand the difference between busybox and a decent user interactive shell plus unix utils
May 21 08:06:17 * psycho_oreos is certain someone else would agree in here ;)
May 21 08:06:36 * archeyDevil knows it's type not giving read/write mode for the mount
May 21 08:06:36 <eddyb> how the heck has a FAT partition 25GB?
May 21 08:06:58 <archeyDevil> eddyb: Nokia n900?
May 21 08:07:01 <psycho_oreos> I've installed bash along with many other proper utils, though from time to time I do use busybox-power (not busybox standard package) and regular gnutools
May 21 08:07:33 <archeyDevil> eddyb: Nokia N900, comes with that partition like that so it can be unmounted for windows/non-nix systems to mount it via usb/mass-storage.
May 21 08:07:57 <eddyb> but FAT32 has a limit of 4GB
May 21 08:08:03 <archeyDevil> E: Couldn't find package zsh
May 21 08:08:05 * archeyDevil cries.
May 21 08:08:22 <psycho_oreos> for single file sizes yes.. but you that doesn't mean you can make it purely ext3 (though with some hacks)
May 21 08:08:22 <eddyb> user "user" is not allowed to run umount
May 21 08:08:23 <archeyDevil> eddyb: No it doesn't.
May 21 08:08:36 <psycho_oreos> eddyb, sudo or do it as root
May 21 08:08:46 <eddyb> archeyDevil: hmm, I thought it did
May 21 08:08:54 <psycho_oreos> archeyDevil, I think ruskie has zsh via his repo (though I don't think it was stable)
May 21 08:08:54 * archeyDevil wonders which he should have installed? rootsh or gainroot?
May 21 08:08:56 <eddyb> psycho_oreos: that's what sudo says
May 21 08:09:13 <psycho_oreos> eddyb, try getting a root console instead
May 21 08:09:15 <archeyDevil> psycho_oreos: Interesting.. However, not stable.. I'm not using it :)
May 21 08:09:17 <ruskie> yeah I have zsh
May 21 08:09:22 <ruskie> what do you mean stable?
May 21 08:09:25 <archeyDevil> pftt.. What am I talking about.. Gimme zsh :)
May 21 08:09:34 <eddyb> psycho_oreos: how?
May 21 08:10:02 <archeyDevil> psycho_oreos: install rootsh or gainroot (I just asked which is prefured; don't know which, I'm using rootsh here)
May 21 08:10:08 <psycho_oreos> ruskie, *shrugs* I recall you had it under `broken' part of your repository but *shrugs* its been a long while since I last checked
May 21 08:10:12 <psycho_oreos> eddyb, sudo -i
May 21 08:10:27 <archeyDevil> ruskie: Whats your repo info?
May 21 08:10:32 <ruskie> psycho_oreos, the broken stuff is something else ;)
May 21 08:10:37 * archeyDevil is new to maemo btw. ( Not Linux )
May 21 08:10:54 <ruskie> archeyDevil, http://wiki.maemo.org/User:Ruskie
May 21 08:10:57 <ruskie> should all be there
May 21 08:11:05 * archeyDevil is used to Arch Linux ( And other Linux systems such as Debian, Slack, etc. Oh and BSD's too )
May 21 08:11:08 <archeyDevil> :)
May 21 08:11:11 <archeyDevil> ruskie: thanks
May 21 08:11:36 <psycho_oreos> archeyDevil, I'm personally neither :) I just use R&D mode and access gainroot from there if needed be but these days my sudo automatically takes care of that
May 21 08:11:37 <eddyb> user "user" is apparently not allowed to run anything
May 21 08:11:41 <psycho_oreos> ruskie, ahh I see
May 21 08:12:09 <archeyDevil> psycho_oreos: Fair enough.
May 21 08:12:10 <psycho_oreos> eddyb, its a logical security feature.. obviously synonymous with every other regular linux distro out there
May 21 08:12:19 <eddyb> ...
May 21 08:12:30 <archeyDevil> eddyb: app manager, install rootsh or gainroot
May 21 08:12:32 <eddyb> it's the message sudo gives me
May 21 08:12:34 <eddyb> okay
May 21 08:12:48 <archeyDevil> zsh - just don't set it as your default shell
May 21 08:13:00 <archeyDevil> ruskie: What about default for root and have another user toor? :)
May 21 08:13:01 <eddyb> time to wait another half an hour for the app manager to load]
May 21 08:13:31 <archeyDevil> eddyb: XD it's slow.. Once you have installed it just use apt-get to install from the command line.. FAR FASTER :D
May 21 08:14:13 <eddyb> is the password for root the same as for user?
May 21 08:14:16 <ruskie> archeyDevil, I have some stuff in my .ashrc that auto executes zsh on ssh and I have a custom shortcut for the terminal
May 21 08:14:18 <psycho_oreos> eddyb, what about sudo gainroot? I assume you don't have openssh-daemon, rootsh or gainroot installed
May 21 08:14:18 <eddyb> because it seemed to have worked
May 21 08:14:28 <eddyb> I have only openssh
May 21 08:14:30 <archeyDevil> ruskie: Nice :)
May 21 08:14:50 <psycho_oreos> eddyb, if it has worked you should get a root prompt
May 21 08:15:04 <eddyb> hmm, the user password doesn't work for root in sftp
May 21 08:15:26 <eddyb> but it "works" for sudo, except sudo thinks I'm too stupid to touch anything as root
May 21 08:15:35 <archeyDevil> ruskie: haha, My remove list :D
May 21 08:15:36 <psycho_oreos> eddyb, normally iinm it should be the password you set when you installed openssh-client-server
May 21 08:16:06 <eddyb> well, I set a password
May 21 08:16:08 <archeyDevil> ruskie: You have a screencast of your phone (from boot onwards)?
May 21 08:16:09 <psycho_oreos> well I'd change password for root and you'll have ssh/sftp access via root
May 21 08:16:10 <eddyb> and it works for user
May 21 08:16:14 <ruskie> archeyDevil, nope
May 21 08:16:16 <eddyb> and root in sudo
May 21 08:16:19 <archeyDevil> ruskie: D:
May 21 08:16:19 <eddyb> but not root in ssh
May 21 08:16:50 <psycho_oreos> yeah you'll need to set passwd for root.. by default the passwd for root is `rootme'
May 21 08:17:20 <archeyDevil> ruskie: Your maemo device.. Nokia n900?
May 21 08:17:37 <eddyb> rootsh is gainroot, btw
May 21 08:17:46 <ruskie> archeyDevil, yup
May 21 08:17:50 <archeyDevil> ruskie: :)
May 21 08:17:57 * archeyDevil got his Nokia n900 yesterday :D
May 21 08:18:04 <archeyDevil> eBay AU$120
May 21 08:18:07 <psycho_oreos> *shrugs* like I said, I only use R&D mode and gainroot works well
May 21 08:18:49 <psycho_oreos> 2 years back and you would have looked somewhere almost three times that price
May 21 08:19:08 <eddyb> umount/mount as root (with gainroot) worked :D
May 21 08:19:10 <eddyb> thanks guys :D
May 21 08:19:32 <archeyDevil> psycho_oreos: R&D?
May 21 08:19:46 <eddyb> what, 8h :|?
May 21 08:19:49 <psycho_oreos> eddyb, I'd still check dmesg as to why it was mounted ro, I'm certain only if the FS is uncleanly mounted for instance would make it remount as ro
May 21 08:20:27 <psycho_oreos> archeyDevil, research and development.. a mode that can only be enabled via flasher tool (and linux only). Though its not recommended for average/daily use (though personally I really don't care :))
May 21 08:20:39 <psycho_oreos> eddyb, 8h?
May 21 08:20:58 <archeyDevil> psycho_oreos: Haha. So development maemo version?
May 21 08:21:17 <eddyb> psycho_oreos: to copy those 13GB
May 21 08:21:20 <archeyDevil> ruskie: What does your bootmenu change for you?
May 21 08:21:36 <ruskie> archeyDevil, just gives me a shell and ssh though it's untested really
May 21 08:21:38 <eddyb> psycho_oreos: apparently a FAT entry beyond EOF
May 21 08:21:43 <ruskie> I think I tried it once or twice
May 21 08:21:53 <archeyDevil> ruskie: Haha. 
May 21 08:21:57 <archeyDevil> So no X?
May 21 08:22:00 <ruskie> nope
May 21 08:22:05 <ruskie> it's meant to recover the system
May 21 08:22:06 <psycho_oreos> archeyDevil, not so much development but it just adds some fancy verbose messages at boot plus maybe slightly more verbose stuff in dmesg and other things.. again not really recommended for average joe as it may also drain battery
May 21 08:22:21 <ruskie> it also requires fbcon
May 21 08:22:25 <psycho_oreos> eddyb, yeah I'd run fsck imo
May 21 08:23:07 <archeyDevil> ruskie: Yea, fbcon? Isn't that just for framebuffers on the console? ( i.e console background? )
May 21 08:23:33 <psycho_oreos> archeyDevil, R&D mode is also used when one cannot boot their device regularly as it can be used to disable/bypass certain hardware features, such as hw watchdog
May 21 08:24:00 <archeyDevil> psycho_oreos: How to flash the whole phone?
May 21 08:24:07 <psycho_oreos> ~flashing
May 21 08:24:07 <infobot> somebody said maemo-flashing was http://wiki.maemo.org/Updating_the_tablet_firmware
May 21 08:24:12 <psycho_oreos> ^
May 21 08:24:38 <archeyDevil> psycho_oreos: I mean. On the keyboard while booting.. Flash from image on the device as normal phones?
May 21 08:24:54 <archeyDevil> Not giving it custom firmware
May 21 08:25:16 <psycho_oreos> archeyDevil, flashing on the keyboard? you mean like LED flashes on the keyboard?
May 21 08:25:23 <archeyDevil> psycho_oreos: No..
May 21 08:25:41 * archeyDevil means, during boot. wipe the phone so it's the default image ( as from manufacture )
May 21 08:26:19 * archeyDevil just wiped his Nokia n95 by mistake D:
May 21 08:26:22 <archeyDevil> Lol
May 21 08:26:33 <psycho_oreos> well that's done through flashing as pointed above... you can't flash the device once it enters into operational state.. there are certain conditions which has to be met for one to enter flash mode at boot
May 21 08:27:14 <archeyDevil> psycho_oreos: I know.. I was asking during the boot. ( as you press the on button ) What buttons to also hold to flash/reset/wipe the phone?
May 21 08:27:41 <psycho_oreos> n95 is a different case, the last time I flashed it was using nokia suite and that was to update.. there's a software called phoenix flasher which apparently is used by nokia technicians.. and there's firmwares for n95 floating around.. I think cpkb.org is one site where you maybe able to obtain info
May 21 08:27:53 * archeyDevil isn't going to do it now. ( Got all my music; just would be great if I do brick it :P )
May 21 08:28:09 <archeyDevil> psycho_oreos: n95 is too easy to reset/flash/wipe.
May 21 08:28:15 <archeyDevil> n900 on the other hand.. How?
May 21 08:28:20 <psycho_oreos> archeyDevil, for N900 its the letter U, but you hold it down as soon as you see USB icon
May 21 08:28:25 <psycho_oreos> again its noted on the wiki
May 21 08:28:59 * archeyDevil can't find it on the wiki....
May 21 08:29:12 <archeyDevil> N900: While holding the u key, connect the N900 to your computer via USB. Wait until you see a dim screen on your N900 with a Nokia logo in the middle of your screen with an USB icon in the top right corner. Note: You may not need to hold the U key - it seems to automatically flash if the flasher is ready at boot time.
May 21 08:29:17 <archeyDevil> :D
May 21 08:29:20 <archeyDevil> Found it :P
May 21 08:29:27 <archeyDevil> / u 
May 21 08:29:27 <psycho_oreos> I was about to paste that exact same line :þ
May 21 08:29:28 <archeyDevil> :)
May 21 08:29:37 <archeyDevil> psycho_oreos: haha
May 21 08:30:04 <psycho_oreos> archeyDevil, I'd also take note of the paragraph directly below the one you just pasted
May 21 08:30:19 <archeyDevil> psycho_oreos: already read.
May 21 08:30:37 <archeyDevil> Not really much as it's just obvious.
May 21 08:30:42 <archeyDevil> imho
May 21 08:31:17 <psycho_oreos> true but not all people realises that and complain (and you get all these tmo threads)
May 21 08:31:46 <archeyDevil> http://wiki.maemo.org/User:Ruskie - My remove list -- Anything in that list that I'd have to consider removing or not? (Looks legit to me)
May 21 08:32:17 * archeyDevil will leave adobe flash for the pure fact that cutetube doesn't want to work..
May 21 08:32:59 * archeyDevil might try out another alt
May 21 08:33:23 <ruskie> archeyDevil, I use youtube-dl
May 21 08:33:40 <archeyDevil> modest -- Thats the email client right?
May 21 08:33:44 * archeyDevil keeps that :P
May 21 08:34:07 <archeyDevil> Maybe remove the other language packs..
May 21 08:34:27 <archeyDevil> nokiamessaging -- Whats this?
May 21 08:37:12 <archeyDevil> Nokia-N900:~# apt-get install ytube qmltube mytube
May 21 08:37:17 * archeyDevil tests out a few :)
May 21 08:37:25 <archeyDevil> After this operation, 42.1MB of additional disk space will be used.
May 21 08:37:27 <archeyDevil> XD
May 21 08:37:45 <archeyDevil> thats more than my average archlinux update.
May 21 08:38:02 * archeyDevil stops talking alone.
May 21 08:41:49 <archeyDevil> ruskie: ezitext-english-us You removed this.. There other english?
May 21 08:44:19 <ruskie> archeyDevil, I don't use ezitext at all
May 21 08:44:59 <archeyDevil> Okay
May 21 08:45:04 <Sicelo> archeyDevil: flashing using hardware button is not actually flashing.. it's just a deep reset .. most usually seen on symbian devices
May 21 08:52:59 <archeyDevil> What is ezitext?
May 21 08:53:17 <archeyDevil> ruskie: 
May 21 08:54:58 * archeyDevil hates it when ssh stops replying for his phone -.-
May 21 08:59:35 * archeyDevil thanks xargs :D
May 21 09:02:42 <archeyDevil> stdin http://sprunge.us/eCHM?text
May 21 09:02:46 <archeyDevil> wth?
May 21 09:02:55 <archeyDevil> It won't let me uninstall them! D:
May 21 09:04:57 <archeyDevil> :) --yes
May 21 09:12:25 * archeyDevil just removed one very nice feature from maemo :P Just got one desktop/workspace :P
May 21 09:14:58 <archeyDevil> apt-get install aircrack-ng
May 21 09:15:01 <archeyDevil> Ahh.. Whoops
May 21 09:15:22 <psycho_oreos> needs more than just that
May 21 09:16:08 <archeyDevil> psycho_oreos: What? aircrack?
May 21 09:16:50 <psycho_oreos> archeyDevil, yes aircrack-ng needs more than just installing that package alone in order for one to gain benefits
May 21 09:17:22 <archeyDevil> True; What to install though? ( Used to backtrack just having it )
May 21 09:17:49 <psycho_oreos> kernel power + lxp `bleeding edge' drivers
May 21 09:18:02 <archeyDevil> Eh.
May 21 09:18:19 <archeyDevil> psycho_oreos: Serious, needs custom kernel?
May 21 09:18:25 <psycho_oreos> that's just the standard stuff, there's an optional stuff you can install which will allow you to switch between wireless drivers on the fly
May 21 09:18:26 <archeyDevil> And more drivers?
May 21 09:18:40 <psycho_oreos> archeyDevil, for injection support yes..
May 21 09:18:41 <archeyDevil> o.O
May 21 09:18:59 * archeyDevil doesn't really need to inject. But yes, I guess thats a great feature.
May 21 09:19:16 * archeyDevil might get a ubertooth so might be monitoring bluetooth too :)
May 21 09:19:23 <psycho_oreos> aireplay-ng requires injection and aireplay-ng is part of aircrack-ng suite
May 21 09:20:02 <psycho_oreos> there's kismet and bluetooth plugin I think in the repos
May 21 09:20:03 <archeyDevil> psycho_oreos: What do I run to get these extra drivers ( and otf driver switcher? )
May 21 09:20:52 <psycho_oreos> archeyDevil, there's a thread (well a fair few to be frank) that discusses where to get lxp's drivers. From memory I think its http://david.gnedt.eu/
May 21 09:21:30 <psycho_oreos> otf driver switcher was by another bloke, I can't remember the name of the package but I think it was written by ossipena (tmo handle)
May 21 09:21:42 <archeyDevil> http://david.gnedt.eu/blog/2010/12/28/bleeding-edge-wl1251-driver-for-maemo-fremantle/
May 21 09:21:45 <archeyDevil> :)
May 21 09:22:05 <psycho_oreos> you'll still need kp anyway
May 21 09:25:04 <psycho_oreos> and kp adds things like USB hostmode which is directly related to your desires in wanting to attach ubertooth to yer N900
May 21 09:26:08 * archeyDevil might continue this tomorrow. I have to go now.. Cya
May 21 09:26:53 <psycho_oreos> cya
May 21 10:13:58 <narcos> Hey all. I've installed openvpn, and am trying to start it via the command line with "/etc/init.d/openvpn start"
May 21 10:14:17 <narcos> The message I get is "Starting virtual private network daemon: Autostart disabled."
May 21 10:14:37 <narcos> And no VPN seems to come up, and the above messages seems like it should only be a warning
May 21 10:15:15 <MohammadAG> Well considering the script is used for autostart...
May 21 10:15:22 <MohammadAG> pastebin it
May 21 10:16:17 <cityLights> narcos: you should define a vpn acording to the debian guide lines
May 21 10:16:33 <cityLights> where did you put the vpn config files?
May 21 10:17:33 <narcos> MohammadAG: Here's the autostart script, which is just the default one which ships with the installation
May 21 10:18:09 <narcos> cityLights: I can start the openvpn fine manually, "openvpn --config /etc/openvpn.conf"
May 21 10:18:34 <cityLights> right, I remember I had to run it manually
May 21 10:18:44 <cityLights> and fixed the script... let me see
May 21 10:18:54 <narcos> ahhhh
May 21 10:19:11 <narcos> cityLights: That'd be awesome :)
May 21 10:28:22 <cityLights> narcos: I need 1hr to get the script
May 21 10:29:45 <narcos> cityLights: Phew, big config file ;-)
May 21 10:29:59 <narcos> cityLights: But no worries, I'll be around
May 21 10:30:14 <narcos> cityLights: I appreciate your help
May 21 10:37:40 <cityLights> pls check the talk.maemo.org for openvpn, I think I may have posted it
May 21 10:44:06 <narcos> cityLights: Checking...
May 21 10:49:36 <narcos> cityLights: Not finding it...
May 21 10:57:06 <cityLights> ok, keep tuned
May 21 10:58:25 <cityLights> btw, you may use a widget to run the vpn config you want
May 21 10:58:33 <cityLights> there is a openvpn widget
May 21 11:38:17 <narcos> cityLights: Naww I need to do stuff from CLI
May 21 11:43:45 <MohammadAG> andre__: #maemo-bugs's dead right?
May 21 11:44:19 <andre__> MohammadAG, yes
May 21 12:27:11 <cityLights> narcos: I think I just ran it manually from the cli
May 21 12:27:42 <cityLights> is this also your version:
May 21 12:27:44 <cityLights> 85f23f875a45f38c6fcd22d236d67647  /etc/init.d/openvpn
May 21 12:28:01 <cityLights> pls run: md5sum  /etc/init.d/openvpn  and post the line
May 21 12:38:37 <teotwaki> heh
May 21 12:38:40 <teotwaki> Chrome beats IE.
May 21 12:38:45 <teotwaki> http://gs.statcounter.com/#browser-ww-weekly-201121-201221
May 21 12:40:21 <jacekowski> well, google did some pretty heavy marketing
May 21 12:41:46 <chem|st> teotwaki: most downloads from IE - mozilla firefox; most initial searches with bling - mozilla firefox
May 21 12:41:57 <chem|st> ehrm bing
May 21 12:42:17 <jacekowski> firefox is shit
May 21 12:42:36 <jacekowski> i'm not sure if you noticed that firefox is losing users as well
May 21 12:44:09 <chem|st> jacekowski: firefox is eating RAM
May 21 12:44:20 <chem|st> and getting slow...
May 21 12:46:53 <narcos> cityLights: 27b6481561705b9e744da6a6d4b6957b  /etc/init.d/openvpn
May 21 12:46:59 <narcos> Not the same it seems
May 21 12:47:18 <narcos> cityLights: So yeah, I think I'll just do a "openvpn --config /etc/openvpn.conf"
May 21 12:47:30 <cityLights> I think its what I did
May 21 12:47:47 <cityLights> let me post my init
May 21 12:48:36 <narcos> k cool :)
May 21 12:50:34 <narcos> Also, what's the best way to install dhcpd3-server on the maemo? I don't like udhcpd
May 21 12:50:41 <narcos> (N900)
May 21 12:55:56 <cityLights> narcos: http://dpaste.com/750758/
May 21 12:58:38 <jacekowski> chem|st: that's nothing new
May 21 12:58:45 <jacekowski> chem|st: it's been doing that for years
May 21 12:59:32 <Sicelo> narcos: server? iirc dnsmasq is on by default
May 21 13:01:58 <narcos> cityLights: Thanks!
May 21 13:02:41 <narcos> Sicelo: I want to hand out dhcp leases on my hotspot. I'm using udhcpd at the moment
May 21 13:03:03 <Sicelo> N900 will hand the IP's out?
May 21 13:04:05 <narcos> yeah
May 21 13:05:24 <narcos> cityLights: 
May 21 13:05:25 <narcos> Nokia-N900:/etc/openvpn# /etc/init.d/openvpn start
May 21 13:05:26 <narcos> Starting virtual private network daemon: Autostart disabled.
May 21 13:05:36 <narcos> With your config
May 21 13:05:37 <narcos> Hmm
May 21 13:06:20 <Sicelo> dnsmasq can handle that .. examples are here ..http://wiki.maemo.org/Maemo_PirateBox, http://wiki.maemo.org/N900_USB_router
May 21 13:09:48 <narcos> Sicelo: ok thanks
May 21 13:32:25 <narcos> So, for example, I did want to install dhcpd3-server instead of dnsmasq (for existing configs etc) - could I? 
May 21 13:32:32 <narcos> I can't seem to find a .deb
May 21 13:32:59 <Sicelo> definitely not in any repos -- you can roll out your own :P
May 21 13:33:29 <narcos> Sicelo: Not sure what you mean by rolling out my own :)
May 21 13:34:24 <cityLights> he means you can get the source to a development env then make build it
May 21 13:38:58 <narcos> ah
May 21 13:39:00 <narcos> gotcha
May 21 14:20:40 <juozapas> hi i'm looking for n900 charger circuit. I guess it's similar to iphone charger, it uses not only vcc and gnd but data lines too. anyway, maybe someone has any experience with that ?
May 21 14:21:04 <joga> juozapas afaik, you just need to short the data pins
May 21 14:21:49 <joga> I just recently built a ghetto solar charger for usb stuff and just soldered the two middle pins together, it worked at least
May 21 14:22:40 <joga> but to do a fast charging circuit, I don't know about that
May 21 14:22:52 <juozapas> i found that many iphone users put resistors between data lines and vcc gnd
May 21 14:23:33 <Sicelo> afaik it's joga's design which is correct for N900
May 21 14:27:34 <merlin1991> juozapas: iphone is fucked up
May 21 14:27:40 <merlin1991> n900 follows the usb specs
May 21 14:28:02 <merlin1991> thus short that datalines to make the phone recognize your circuit as a "dumb" charger
May 21 14:28:26 <jacekowski> well, as much as i hate to say that, apple way of detecting charger is better
May 21 14:28:45 <juozapas> :/
May 21 14:28:59 <merlin1991> jacekowski: what do they do actually?
May 21 14:29:14 <jacekowski> 2.5V on data lines
May 21 14:29:30 <jacekowski> through iirc 10k resistors from VCC and GND
May 21 14:29:41 <jacekowski> rather than dead short
May 21 14:30:43 <jacekowski> http://www.ladyada.net/images/mintyboost/usb4res.png
May 21 14:30:56 <jacekowski> 2.06V and 2.76V
May 21 14:31:52 <juozapas> fuck i have to make new pcb now..
May 21 14:32:20 <jacekowski> that way you can always add easy signaling later on
May 21 14:32:24 <jacekowski> to signal higher powered charger
May 21 14:42:49 <chem|st> DocScrutinizer51: did you have an eye on tizen? intel and samsung are promising to bring netbooks (in Q3) and smartphones (in Q4)
May 21 14:43:36 <juozapas> is it save to charge other devices (like ebook readers, ipod, camera, gps) with data lines shorted ?
May 21 14:45:27 <flux> tizen was that web2.0 mobile platform?
May 21 14:52:12 <DocScrutinizer51> yay
May 21 14:52:29 <DocScrutinizer51> chem|st: nope
May 21 14:52:32 <GeneralAntilles> ^ we probably shouldn't ban him for more than 24 hours.
May 21 14:52:50 <chem|st> juozapas: it is the USB standard to do so
May 21 14:53:00 <DocScrutinizer51> whom? pupnik?
May 21 14:53:02 <GeneralAntilles> Yeah
May 21 14:53:09 <DocScrutinizer51> he asked for permanent ban
May 21 14:53:10 <chem|st> o.O
May 21 14:53:13 <GeneralAntilles> Did he?
May 21 14:53:26 <DocScrutinizer51> yep actrually
May 21 14:53:36 <DocScrutinizer51> gently and honest
May 21 14:53:45 <GeneralAntilles> Ah well, hopefully he's over it.
May 21 14:54:00 <MohammadAG> haven't seen him for a while
May 21 14:54:03 <DocScrutinizer51> s/gently/polite/
May 21 14:54:04 <infobot> DocScrutinizer51 meant: polite and honest
May 21 14:54:09 <chem|st> juozapas: the n900 is the only device I know not charging, but chargers should have datalines crossed
May 21 14:55:11 <chem|st> flux: LiMo+MeeGo+Intel+Samsung=Tizen
May 21 14:55:38 <GeneralAntilles> OK, keeping the bot bans
May 21 14:56:53 <juozapas> chem|st: what do u mean 'not charging'? if i understand corect it should charge when data lines are crossed
May 21 14:58:17 * DocScrutinizer51 feels reminded at 'orbit cleaning day' as introduced by 'Max Headroom'
May 21 15:01:00 <Estel_> DocScrutinizer,  whos is pupnik? what he did to earn ban? ;)
May 21 15:01:26 <merlin1991> Estel_: be happy not to know :D
May 21 15:01:58 <Woody14619> USB2 standard is 200mA at plugin, which is enough to charge most devices when idle (including N900).  After negotiation (or proper detection, like shorted data pins) amperage draw can go up to 500mA.
May 21 15:02:28 <Estel_> Woody14619,  I always heard of 100 mA @ plugin
May 21 15:02:34 <Estel_> and 500 after neg.
May 21 15:03:05 <Estel_> and after shorted data pins, it can go up to 1.8A
May 21 15:03:06 <Estel_> as per usb dedicated charging pin spec
May 21 15:03:29 <GeneralAntilles> Estel_, he's a Maemo legend.
May 21 15:03:31 <Estel_> negotiation - 500mA, crossed data - 1.8A (up to)
May 21 15:03:37 <GeneralAntilles> Estel_, one of the original emulator guys.
May 21 15:04:07 <Estel_> thats why I asked, I like legends. Less interested in him (or totally not interested), interested in story behind it
May 21 15:04:42 <Woody14619> Estel_, my bad, you are correct, spec is min of 100mA.  Most PCs offer 200mA though.
May 21 15:05:54 <Woody14619> Estel_: Though USB2 max spec is 900mA.  USB3 I believe goes up to 1.8A.
May 21 15:05:55 * Estel_ nods
May 21 15:06:18 <Estel_> arghhh, damn fm radio program, it keeps stucking on heavy swap usage
May 21 15:06:55 <Estel_> I would be very grateful for proper fm radio implementation (utilising allowed passthrough to jack/speakera without working on cpu power)
May 21 15:07:04 <Estel_> No idea why it wasn't ever implemented :/
May 21 15:07:19 <MohammadAG> max is 900?
May 21 15:07:25 <Woody14619> Mind you, shorted data pins means no negotiations, and the 900mA is max negotiable power by USB2 spec.  So...  you could run 5A over it if you wanted in your custom app, but be prepared for it to not work with most chargers. ;)
May 21 15:07:44 <MohammadAG> Nokia's charger outputs 1200mA
May 21 15:07:55 <Estel_> we have one FM radio program with fancy UI, yet bugged as hell
May 21 15:08:10 <MohammadAG> which is?
May 21 15:08:25 <jacekowski> Woody14619: well, usb specs makes a provision of "negotiation" of current by limiting voltage
May 21 15:08:27 <Estel_> Woody14619,  I'm absolutely sure 1.8 A is max as per USB *dedicated charging* port specification
May 21 15:08:47 <Estel_> btw, all dumb chargers with usb port are dedicated charging ports
May 21 15:09:21 <Estel_> ...as usb specified dedicated charging port can't be used for normal usb connection = got data pins shorted solid = it's dumb charger in reality
May 21 15:09:54 <Estel_> ...coming back to radio thread, we got another radio ap in qt, also working on cpu, but only mono due to bugs in qt...
May 21 15:10:13 <Estel_> MohammadAG,  "FM Radio" is bugged as hell
May 21 15:10:23 <Estel_> this one with interface looking like old radio
May 21 15:10:28 <Woody14619> Estel_, while that may be the concesus amoung manufactures, it's not part of the USB2 spec.  (But then neither is shorting the data pins, other then saying that applications must be able to handle that.)
May 21 15:10:30 <chem|st> juozapas: dumb chargers do not negotiate so they have to have crossed datalines for n900s to charge at all
May 21 15:10:35 <Estel_> with kinetic scrolling of frequency
May 21 15:11:19 <Estel_> Woody14619,  I remember reading about "dedicated" (aka dumb) charging port in usb specification
May 21 15:11:23 <Estel_> no idea where, honestly
May 21 15:11:33 <jacekowski> nope
May 21 15:11:45 <jacekowski> there is separate spec for usb charging
May 21 15:11:58 <Woody14619> May have been an addendum at some point, but it's not part of the actual USB spec.
May 21 15:12:10 <Estel_> ... why the hell no one uses capability of our radio module (analog passthrough without involving cpu yet controlable from userland = from program)
May 21 15:12:14 <Estel_> jacekowski,  thanks
May 21 15:12:25 * Estel_ agree
May 21 15:12:47 <jacekowski> look people
May 21 15:12:55 <jacekowski> spec is public
May 21 15:12:55 <Estel_> btw under my desktop I got 20A usb charging port :P
May 21 15:12:56 <jacekowski> just read it
May 21 15:13:01 <jacekowski> http://www.usb.org/developers/docs/usb_20_101111.zip
May 21 15:13:07 <jacekowski> http://www.usb.org/developers/docs/usb_30_spec_022412.zip
May 21 15:13:16 <jacekowski> all docs are there
May 21 15:13:18 <Woody14619> Is there a coper path from the radio output to the jack or speakers?  
May 21 15:13:26 <jacekowski> Woody14619: no
May 21 15:13:30 <Estel_> Woody14619,  sec, I'll find refference link
May 21 15:13:46 <Estel_> unless jacekowski remeber it exactly
May 21 15:14:03 <DocScrutinizer51> usb charging supplement
May 21 15:14:04 <jacekowski> it's in that zip
May 21 15:14:12 <Estel_> jacekowski,  but I'm not talking bullshit about analog passthrough, yep?
May 21 15:14:13 <Woody14619> jacekowski, Thanks. Yes, I've got a copy of that, but have read mainly the parts I've care about for projects.  It's a rather dry/boring read for the most part.
May 21 15:14:32 <jacekowski> Estel_: bit pointless
May 21 15:14:36 <Estel_> why?
May 21 15:14:37 <jacekowski> Estel_: it doesn't use a lot of cpu power
May 21 15:14:49 <jacekowski> as long as you don't involve PA
May 21 15:15:05 <Estel_> but still = shorten battery life using fm radio. and, what's more important to me...
May 21 15:15:14 <Estel_> this thing stutters/freeze on swap usage
May 21 15:15:23 <Estel_> analog passthrough should free us from that
May 21 15:15:44 <jacekowski> well, that's because implementation is shit
May 21 15:15:49 <Estel_> goddamn useless android phones are known for playing radio stutterless even if os hangs up almost completely
May 21 15:16:03 <jacekowski> yes, because implementation isn't so fucked up
May 21 15:16:11 <jacekowski> n900 uses piece of shit called PA
May 21 15:16:24 <Estel_> but they're using passthrough. If i start fm radio on N900 and android, N900 is half second delayed due to processing on cpu
May 21 15:16:27 <jacekowski> with plain alsa it could have been done in 3 lines of code that could have been easily mlocked
May 21 15:16:36 <Estel_> I see.
May 21 15:16:39 <jacekowski> Estel_: again, that's because shit called PA
May 21 15:16:48 <Estel_> ok, but analog passthrough take us out of PA too, yep?
May 21 15:16:58 <Estel_> I see and understand.
May 21 15:16:59 <jacekowski> yes, but that's not required
May 21 15:17:12 <jacekowski> you still want notifications and audio from other apps
May 21 15:17:22 <jacekowski> so you can't just change over
May 21 15:17:26 <Estel_> ok, but implementing passthrough is much more complicated that implementing pa-free on-cpu processing?
May 21 15:17:52 <jacekowski> yes
May 21 15:18:03 <Estel_> I see.
May 21 15:18:27 <Estel_> so what You would propose to *properly* implement fm radio application, witch least hassle required?
May 21 15:18:38 <jacekowski> on n900?
May 21 15:18:43 <jacekowski> you can't
May 21 15:18:45 <Estel_> yea
May 21 15:18:49 <Estel_> hm
May 21 15:19:02 <jacekowski> but in simple word
May 21 15:19:04 <jacekowski> words
May 21 15:19:05 <Woody14619> I propose you go to a dollar store and get a cheep fm receiver... much easer. :)
May 21 15:19:07 <jacekowski> get rid of PA
May 21 15:19:12 <Estel_> passthrough can't coexist with notifications from programs (latter via pulseaudio)
May 21 15:19:24 <jacekowski> change it to alsa builtin DMIX
May 21 15:19:26 <Estel_> Woody14619,  not the point
May 21 15:19:33 <Estel_> I also record fm radio on N900
May 21 15:19:57 <jacekowski> fix all apps that relied on PA directly
May 21 15:20:15 <Estel_> passthrough can't coexist with notifications from programs (latter via pulseaudio) <- it was meant to be question, ? got eaten
May 21 15:20:53 <Woody14619> I generally find most of the carp on FM isn't worth listening to, yet alone recording. :P  It's 70% commercials and 25% top 40.  But to each their own. :)
May 21 15:21:15 <Estel_> Woody14619,  in Poland we got polish radio program 3 with is all different ;)
May 21 15:21:39 <Estel_> for both many polish people and overbroad artist, it's best radio station in the world ;)
May 21 15:21:39 <jacekowski> Estel_: not really
May 21 15:21:47 <jacekowski> what????????
May 21 15:21:52 <jacekowski> stop spreading BS like that
May 21 15:21:57 <Estel_> ?
May 21 15:22:13 <Woody14619> I'm sure it is, if you speak Polish. ;)
May 21 15:22:16 <Estel_> erm, gustas et not discutante I would say
May 21 15:22:20 <jacekowski> 3 is pretty bad
May 21 15:22:28 <Estel_> well, see above
May 21 15:23:27 <Estel_> (as for overbroad artists I mean Tori Amos, Peter Gabriel, old good leo cohen, Archive, zeppelins, old guys from pink floyd etc etc)
May 21 15:23:45 <Woody14619> I mainly use the interent radio for content.  And on some occasions shortwave.  (Though that's saddly going away, what with BBC reducing US coverage and what not).
May 21 15:23:48 <Estel_> all were here in 3's recording studio, and were amazed
May 21 15:24:10 <Estel_> I like when radio speaks to me, instead of shouting playlist on me
May 21 15:24:25 <jacekowski> i hate when they talk on radio
May 21 15:24:27 <Estel_> 3 is quite eclectic
May 21 15:24:38 <Estel_> well, if I need background music, I got my own.
May 21 15:24:52 <Estel_> from radio I expect something more
May 21 15:25:26 <Estel_> personalities, intelligent conversation (also about music), new musical horizons...
May 21 15:25:56 <Woody14619> Well, here in the US, you have three choices:  FM=Top 40 and commercials, almost exclusivly.  AM=Right-wing talk (and maybe PBS/NPR), or pay for XM where you get less commercials but otherwise most of the same content as FM.
May 21 15:26:13 <Estel_> it's quite funny, cause, for example, Piotr Kaczkowski is known only for 3's fans in Poland, yet, one of most respected music journalists abroad, at least in so called "ambitious" music world
May 21 15:26:31 <Estel_> Woody14619,  I know and feel Your pain
May 21 15:27:06 <Estel_> 3 is all about music and personalities. We've got many *real* music/music journalist personalities here
May 21 15:27:24 <Estel_> both from "old good times" and new ones learning stuff from masters
May 21 15:27:37 <Estel_> it's quite a Community around 3
May 21 15:27:52 <Estel_> so, when jacekowski said "i don't like when they speak in radio"
May 21 15:28:08 <Estel_> it reminds me of iphone fans saying "what, opening terminal to do anything?!"
May 21 15:28:19 <Estel_> ;)
May 21 15:28:26 <jacekowski> i want them to play music
May 21 15:28:33 <Estel_> gustas et... etc ;)
May 21 15:28:46 <jacekowski> i only listen to radio in a car and i can't really focus on what they are talking about
May 21 15:28:47 <Estel_> Iphone fans want them to "just work" ;)
May 21 15:29:12 <jacekowski> because i'm focusing on a moron that is behind/in front of me
May 21 15:29:13 <Estel_> I understand. Well, I listen to radio almost all awake time, while doing other things too
May 21 15:29:37 <Estel_> yea, another thing about most polish car drivers - everyone see morons all around on street :P
May 21 15:29:46 <Estel_> while I agree there are many...
May 21 15:30:20 <Estel_> I'm curious why there is so much aggression, hidden or not, in roads. Never seen that in any other country I've been in, even in eghypt and cairo
May 21 15:30:42 <teotwaki> cairo is a country now?
May 21 15:30:50 <Estel_> very funny
May 21 15:31:10 <teotwaki> anyway, road-rage is everywhere. It's not local.
May 21 15:31:12 <Estel_> ...but, considering that it's unofficially 27 000 000 people...
May 21 15:31:22 <Estel_> and poland is 38 000 000
May 21 15:31:24 <Estel_> why not.
May 21 15:31:35 <Estel_> teotwaki,  I know
May 21 15:31:37 <teotwaki> so? London is 15M+, Belgium is 11M+.
May 21 15:31:38 <jacekowski> teotwaki: you've never been to poland then
May 21 15:31:49 <Estel_> ^agree
May 21 15:31:50 <teotwaki> jacekowski: actually, I've driven through Poland.
May 21 15:32:06 <Estel_> I see jacekowski know what I mean ;)
May 21 15:32:20 <jacekowski> things like people ignoring red light, overtaking when it's not safe to do so and things like that are pretty common
May 21 15:32:26 <jacekowski> as in, you see it few times a day
May 21 15:32:47 <jacekowski> i'm doing something around 40000 miles/year
May 21 15:32:55 <teotwaki> and? Grab a taxi in Italy and ask him to get you quickly from one train station to the other -- exactly the same behaviour.
May 21 15:33:06 <jacekowski> teotwaki: nothing like that
May 21 15:33:13 <jacekowski> first thing 
May 21 15:33:17 <Estel_> btw, funny fact - in eghypt, where no one cares about lights etc but everyone know it, You got much more small, non-fatal collisions
May 21 15:33:21 <jacekowski> in poland people don't try to avoid accidents
May 21 15:33:22 <Estel_> although much less fatal ones.
May 21 15:33:47 <Estel_> In countries when You actually expect other to obey the law, much less small, unharmfull colisions, yet more fatal ones
May 21 15:33:47 <teotwaki> Try India.
May 21 15:34:00 <jacekowski> when you see somebody pulls out in front of you
May 21 15:34:07 <Estel_> I suspect it's because on eghypt etc everyone *know* to 100% not trust anyone around
May 21 15:34:13 <teotwaki> Egypt*
May 21 15:34:21 <jacekowski> in normal country it you slow down first trying to avoid accident
May 21 15:34:37 <jacekowski> in poland, first - horn, second - accelerate
May 21 15:34:47 <Estel_> jacekowski,  but teotwaki got a point here, simillar behavior is in italy
May 21 15:34:55 <jacekowski> because they had right of way
May 21 15:35:11 <teotwaki> India is probably the worst place on Earth I've had to drive.
May 21 15:35:13 <jacekowski> Estel_: in italy taxi driver will try to avoid accidents
May 21 15:35:14 <Estel_> ...but polish car drivers tend to think they're "better" than those barbarians in italy or egypt
May 21 15:35:40 <jacekowski> because that means time without a car and no work
May 21 15:35:41 <Estel_> jacekowski,  You're exxagerating it a little. idiots happen, but most people will try to avoid accident :P
May 21 15:35:48 <jacekowski> Estel_: not really
May 21 15:35:56 <jacekowski> Estel_: everywhere else, it's brakes first 
May 21 15:35:57 <Estel_> teotwaki,  agree that india is even worse
May 21 15:36:01 <jacekowski> Estel_: in poland it's horn first
May 21 15:36:10 <Estel_> but, we're - suppposedly - in europe...
May 21 15:36:26 <Estel_> hehe, horn first is really strange thing
May 21 15:36:30 <Estel_> although I agre
May 21 15:36:34 <jacekowski> and people doing 60km/h or more through towns
May 21 15:36:43 <Estel_> if only 60;)
May 21 15:37:00 <Estel_> 80-90 is absolutely common
May 21 15:37:12 <Estel_> 110 is seen 2-3 times a day (some idiot)
May 21 15:37:30 <Estel_> at least here in Poznan
May 21 15:37:45 <Estel_> in warsav it's even worse, sometimes
May 21 15:37:48 <jacekowski> which gives you total stopping distance of 40m+
May 21 15:37:51 <jacekowski> at 60km/h
May 21 15:38:51 <joga> in india I guess the speeds are relatively low
May 21 15:39:01 <teotwaki> joga: beg your pardon?
May 21 15:39:28 <joga> teotwaki I haven't driven there, just judging by the videos I've seen and my friends who have
May 21 15:39:53 <teotwaki> joga: people will happily do 80-100 (kph) on any road, whether the lights on their vehicle are working or not.
May 21 15:39:59 <teotwaki> (at night)
May 21 15:40:27 <joga> but in a crowded, messed up intersection in a city it seems unlikely you can even drive straight for more than a couple of meters
May 21 15:41:17 <Estel_> hehe, fun thing in india is that people crossing road don't care about anything driving there
May 21 15:41:24 <teotwaki> Right outside Pondi, near Auroville, there's a famous crossing -- small dirt road that goes to Avy (actually, might be paved by now) that crosses with a "main" road to Pondi. there's roughly 1 deadly accident a day.
May 21 15:41:32 <Estel_> and people driving dont't care about someone crossing road
May 21 15:41:35 <jacekowski> joga: yeah, but everyone is trying to avoid bumping into other cars
May 21 15:41:42 <teotwaki> http://www.youtube.com/watch?v=FxZxjgKcsPE
May 21 15:41:47 <Estel_> it's a interesting think than no road holocaust happens
May 21 15:42:03 <teotwaki> (I got a +5 Interesting on Slashdot for posting a link to that video...)
May 21 15:42:44 <joga> when I used to live in soviet russia behind our house there was some road where every week at least there was a pile-up with deaths
May 21 15:42:52 <Estel_> btw I remember when guys from polish radio 3 ( ;) ) have gone in motorcycle trip around world
May 21 15:43:09 <joga> nowadays I hear the police there fine people even for not using the signals
May 21 15:43:11 <Estel_> ...and in india...
May 21 15:43:47 <Estel_> 1 guy without legs on platform with wheels tried as hard ase he could (he was quite fast, suprisingly) to get under motorcycle
May 21 15:43:56 <Estel_> as he hope to get money after
May 21 15:43:59 <joga> jacekowski yeah, which - I guess - would keep the speeds there low enough that not *all* of them die :)
May 21 15:44:09 <Estel_> they need to kick him during actual driving
May 21 15:44:19 <Estel_> to not allow him get under wheel
May 21 15:44:26 <joga> how does he stop?
May 21 15:44:54 <joga> oh, nevermind
May 21 15:45:02 <teotwaki> joga: if you think you get fined in India for not using your indicators, think about this: helmets are "recommended", not obligatory.
May 21 15:45:03 <joga> read that the legless guy was attempting to drive the cycle
May 21 15:45:16 <Estel_> no idea, there were 2 of them on 2 motorcycles and one of them needed to kick him while avoiding
May 21 15:45:20 <joga> teotwaki sure
May 21 15:45:22 <Estel_> ahaha
May 21 15:50:51 <MrPingu> merlin1991: Which package do I need for gst-launch-0.10?
May 21 15:51:01 <merlin1991> lemme check
May 21 15:53:42 <merlin1991> MrPingu:  I think its gstreamer-tools
May 21 15:54:32 <MrPingu> Atleast, it downloads something...
May 21 15:56:10 <Estel_> merlin1991,  what is /debian/compat file in source package?
May 21 15:56:23 <Estel_> most packages have "5" there, although I've seen "7" too
May 21 15:56:40 <merlin1991> it's the minimum version of debhelper the package script demands
May 21 15:56:52 <merlin1991> autobuilder runs 5 by default so 5 is perfect
May 21 15:57:39 <Estel_> I see. thanks
May 21 15:58:50 <MrPingu> Yes, it is :) Thanks, but ximagesrc is no element?
May 21 16:11:31 <MrPingu> Seems I dont have libistximagesrc.so, from what I found out in google
May 21 16:11:51 <MrPingu> Maybe I am missing a plugin package?
May 21 16:12:16 <MrPingu> Got to go now, if you know don't hesitate to write it here I will read back the log :)
May 21 16:13:34 <Estel_> Funny, I'm on ad-hoc wifi, listening to fm radio via speakers and browsing tmo via chromim - I'm charging via dedicated charger and barely 150 mA is put into battery, rest consumed by working things
May 21 16:14:03 <Estel_> (battery less than half charged, so it's not rate limiting)
May 21 16:14:07 <freemangordon> ad-hoc wifi?
May 21 16:14:11 <Estel_> yea
May 21 16:14:17 <Estel_> it consumes most power
May 21 16:14:23 <freemangordon> there is where the current is flowing
May 21 16:14:30 <Estel_> exactly
May 21 16:15:12 <Estel_> If I would enable torch, make screen to go max brightness + enable gps and something else, I would be consuming more than charging circuit cn pull
May 21 16:15:18 <Estel_> can*
May 21 16:16:30 <Estel_> I'm on specific situation now - my AP is working in client mode, so I must transfer net via ethernet cable through whole room and corridor, then, my notebook streams wifi to N900 via ad-hoc
May 21 16:17:19 <Estel_> if I would have 2 ap's, one would be wifi client, and another one would be connected via cable and spread net under new name
May 21 16:17:50 <Estel_> (I'm doing this because signal is enough only in one place oh a house, and using pt with boosted transmit power)
May 21 16:18:12 <Estel_> haha, sounds like perfect scenario for stealing wifi bandwidth from someone, lol
May 21 16:18:32 <Estel_> not that I'm doing so
May 21 16:23:55 <ioan> Is moving from a n900 to a n9 a good idea or should I keep the n900?
May 21 16:25:58 <_berto_> ioan: it all depends on what you're looking for
May 21 16:26:18 <_berto_> ioan: the n9 is a smartphone, the n900 is a pocket linux-based computer
May 21 16:26:21 <_berto_> that's how I see it
May 21 16:26:43 <Gh0sty> not giving any money to mokiasoft
May 21 16:26:56 <Corsac> and the n9 has less apps than n900, too
May 21 16:27:04 <Corsac> (in case that matters)
May 21 16:27:06 <Gh0sty> so i keep my n900
May 21 16:27:40 <Gh0sty> lucky i have it with me right now
May 21 16:28:08 <Corsac> well, the n9 is really nice, I dont regret it
May 21 16:28:21 <Gh0sty> in hospital for most of the day ... with nothing to do n900 with internet ...
May 21 16:33:11 <Estel_> chem|st,  could You kill this thread? -> http://talk.maemo.org/showthread.php?p=1210290
May 21 16:37:18 <Sicelo> 17:19 < Estel_> I also record fm radio on N900 <<< that's why I keep my Nokia N-Gage :P
May 21 16:37:26 <Sicelo> ;)
May 21 16:38:43 <Sicelo> btw, Estel_, you tried C Fm radio? seems less memory hungry
May 21 16:43:47 <chem|st> Estel_: done
May 21 16:44:04 <chem|st> Estel_: there is a report button, do you know about it?
May 21 16:55:34 <DocScrutinizer> (If I would enable torch, make screen to go max ...) and do 3G massive data
May 21 16:55:43 <DocScrutinizer> and vibra
May 21 16:55:49 <DocScrutinizer> and video decoding
May 21 16:56:30 <DocScrutinizer> I can't suggest vboost as that's mutually exclusive with charging ;-P
May 21 17:00:42 <Sicelo> :P
May 21 17:02:56 <freemangordon> Estel_, the guy having problems recording in HD on TMO said he has downgraded to CSSU-stable
May 21 17:03:23 <freemangordon> and camera-ui is only in -testing
May 21 17:04:49 <MrPingu> Can you pull more 1200 mAh?
May 21 17:05:10 <MrPingu> I mean my charger says output 1200 mAh
May 21 17:11:48 <MrPingu> merlin1991: Which other gst-packages do you have installed?
May 21 17:13:50 <freemangordon> DocScrutinizer, may I ask you (politely) next time when you question my expertise and experience on driving SW projects to let me at least be a part of the conversation. It would be even better if you have some information on what I do for life, what is my posstion in the company I work for, etc...
May 21 17:14:34 <DocScrutinizer> eh?
May 21 17:14:56 <DocScrutinizer> wazzup?
May 21 17:16:44 <DocScrutinizer> bitter about my notion that's missing some indication of experience in risk management in your acting?
May 21 17:17:51 <freemangordon> bitter? no, just think it is fair if I am included in conversation too
May 21 17:18:14 <DocScrutinizer> sure thing
May 21 17:19:23 <DocScrutinizer> sorry if you feel I acted incorrectly
May 21 17:19:58 <freemangordon> NP, lets move on. DocScrutinizer, according to those logs you somehow were able to read the logs re thumb2 on #beagle
May 21 17:21:09 <DocScrutinizer> only what you pasted
May 21 17:21:19 <freemangordon> that is all :)
May 21 17:22:00 <DocScrutinizer> I supported you in claiming a proper discussion with those who think they know was the right approach
May 21 17:22:34 <DocScrutinizer> the problem is it's a niche case for them
May 21 17:22:56 <DocScrutinizer> so probably way too much effort for zero ROI from their perspective
May 21 17:23:11 <freemangordon> and it the end it turned out(at least that is my understanding) that they don't care/never tried hard to achieve it
May 21 17:23:16 <freemangordon> yep
May 21 17:23:54 <DocScrutinizer> the real pity is all that knowledge wasted
May 21 17:24:27 <freemangordon> BTW Estel_'s words for FUD spreading are not so distant from the reality AIUI
May 21 17:25:06 <DocScrutinizer> which is a sad case that at least allows us to learn from it: never neglect documenting what you do
May 21 17:26:09 <freemangordon> As I can bet both n900s I have, that thumb errata workaround was never enabled on n900 since a month ago. 
May 21 17:26:25 <freemangordon> yeah, sure
May 21 17:29:08 <freemangordon> now, AIUI you(and maybe Jaffa) insist on having a proper testbed. I will think on that, but having in mind the errata description, I don't think it is possible without speciol tools which most probably only TI has.
May 21 17:29:56 <DocScrutinizer> I don't think it's particularly hard to plumb together some assembler that reliably provokes the error
May 21 17:29:57 <freemangordon> BTW, did you read errata description?
May 21 17:30:11 <freemangordon> DocScrutinizer, I think it is impossible
May 21 17:31:00 <DocScrutinizer> why?`
May 21 17:31:02 <freemangordon> if we go into the details, I don't think I am that good in ARM assembler to be able to control BTB/BTAC contents on every cycle
May 21 17:31:18 <DocScrutinizer> aah
May 21 17:31:29 <freemangordon> read the errata description again
May 21 17:31:45 <freemangordon> what conditions are needed, so it to be triggered
May 21 17:31:52 <freemangordon> want a link?
May 21 17:31:56 <DocScrutinizer> sure
May 21 17:32:25 <freemangordon> http://cateee.net/lkddb/web-lkddb/ARM_ERRATA_430973.html
May 21 17:32:55 <DocScrutinizer> if ARM (and THUMB) opcode description doesn't allow to precisely smith a piece of code that invokes the error, then I'd say ARM is crap
May 21 17:33:08 <DocScrutinizer> (the company)
May 21 17:33:37 <freemangordon> DocScrutinizer, there is a pile of debug registers you have access to once attached to jtag ;)
May 21 17:33:57 <DocScrutinizer> I know
May 21 17:34:43 <freemangordon> and it is not the code itself, it is the branch prdictor which does not realize that there are new instructions on the same virtual address
May 21 17:34:46 <DocScrutinizer> and just 6h ago I talked with a colleague about how to exploit the HTB in ARM for a FOSS project
May 21 17:34:55 <freemangordon> HTB?
May 21 17:35:11 <DocScrutinizer> Hardware Treace Block or sth
May 21 17:35:22 <freemangordon> aah, ok
May 21 17:36:07 <DocScrutinizer> a nice IP that recors the last nn k of instructions and register data
May 21 17:36:33 <DocScrutinizer> it can get read out not only via JTAG
May 21 17:36:56 <freemangordon> I realy can't think of anything besides turning u-boot upside-down, hoping I am smart enough to make the correct conclusions based on short description of the errata
May 21 17:37:16 <DocScrutinizer> we're developing a AppProc based debugging tool at my company
May 21 17:37:41 <freemangordon> DocScrutinizer, not entirely true, those can be rad, but only in secure priviledged mode if one of the core pins has the appropriate level on boot
May 21 17:37:48 <freemangordon> *read
May 21 17:37:57 <freemangordon> CP15DISABLE or something
May 21 17:38:24 <DocScrutinizer> quite possible that this HS shit interferes
May 21 17:38:35 <freemangordon> that is for sure
May 21 17:39:06 <DocScrutinizer> even possible the OMAP3430 has no ETB (not HTB, E like Embedded)
May 21 17:41:33 <freemangordon> so, there are 2 options left AIUI. You may accept my logic why errata workarounds work. After all we have quantum theory based almost on pure (i) logical thinking :). BTW I am not sure i've ever tried to.
May 21 17:42:23 <freemangordon> s/i've/you've/
May 21 17:42:23 <infobot> freemangordon meant: so, there are 2 options left AIUI. You may accept my logic why errata workarounds work. After all we have quantum theory based almost on pure (i) logical thinking :). BTW I am not sure you've ever tried to.
May 21 17:44:05 <freemangordon> the other is to wait long enough thumb-testing to have enough users, and if there is no single sigill report for a long enough period of time (2-3 months) to assume that stable. After all we are talking about handheld device here, not a banking core server
May 21 17:45:19 <DocScrutinizer> to me the explanation is pretty clear
May 21 17:45:33 <freemangordon> yes, it is
May 21 17:45:52 <freemangordon> and the workaround is, too
May 21 17:46:48 <DocScrutinizer> you simply run a interworking branch, then replace it by something that's *not* switching from ARM to THUMB or vice versa, and run it again -> BOOM
May 21 17:47:23 <freemangordon> do you think it is so simple?
May 21 17:47:35 <DocScrutinizer> according to that description it is
May 21 17:47:42 <freemangordon> in a multi-process environment?
May 21 17:48:01 <DocScrutinizer> nah
May 21 17:48:17 <DocScrutinizer> you for sure block all interrupts before you run that testbed
May 21 17:50:19 <DocScrutinizer> a testbed needs no X sever to output the results relatime
May 21 17:50:25 <freemangordon> well, I am not so good in omap kernel hacking (so far).  But will think If I could do something in user space. 
May 21 17:50:25 <DocScrutinizer> realtime*
May 21 17:50:43 <freemangordon> yeah, maybe u-boot
May 21 17:51:18 <freemangordon> but again, if we change the conditions, then the testbed is not correct
May 21 17:51:52 <freemangordon> if I don't do context switchws, then the workaround is useless
May 21 17:52:09 <freemangordon> if I do, then I cannot predict when I will hit it
May 21 17:52:11 <freemangordon> see
May 21 17:53:47 <freemangordon> i meant "when i will hit the errara"
May 21 17:59:00 <DocScrutinizer> nope, that's not how a testbed is meant to work and to prove things
May 21 17:59:25 <freemangordon> what we are trying to prove?
May 21 17:59:42 <freemangordon> errata existance or workaround effectiveness?
May 21 17:59:54 <DocScrutinizer> the fact that opcode sequence A, B ,
May 21 17:59:57 <DocScrutinizer> C d
May 21 18:00:11 <DocScrutinizer> breaks on C without your fix
May 21 18:00:16 <freemangordon> as workaround is meant to fix things in multi-process environment
May 21 18:00:18 <DocScrutinizer> and runs with your fix
May 21 18:01:22 <freemangordon> DocScrutinizer, ok, I will try to do some user-mode stuff which provoces it
May 21 18:01:28 <freemangordon> *provokes
May 21 18:01:32 <DocScrutinizer> good
May 21 18:01:34 <DocScrutinizer> :-D
May 21 18:02:03 <freemangordon> but can't promise anything, so you think on your side what will be the next best option :P
May 21 18:03:21 <freemangordon> as it is more than obvious that the only(accesible) expertise left on that matter is here 
May 21 18:05:38 <DocScrutinizer> alas this seems right
May 21 18:06:18 <DocScrutinizer> freemangordon: we seen modest segfaulting like mad when compiled for THUMB, no?
May 21 18:06:31 <freemangordon> yes
May 21 18:06:44 <freemangordon> every minute or so
May 21 18:06:51 <DocScrutinizer> maybe it's possible to reduce modest code to the bare minimum needed for triggering that degfault
May 21 18:07:18 <DocScrutinizer> maybe even running it under gdb and watch it segfault
May 21 18:07:39 <freemangordon> DocScrutinizer, it is not the modest itself, every application behaves like that. ANd it is not only htumb-compiled, arm binaries SIGILL too
May 21 18:08:12 <freemangordon> DocScrutinizer, it is not SEGFAULT, write it down somewhere for further reference :P
May 21 18:08:21 <freemangordon> it is SIGILL
May 21 18:08:48 <freemangordon> I've seen that a lots of times for the last months
May 21 18:08:49 <DocScrutinizer> I'm just saying somewhere inside modest there hast to be that core A,B,C.D opcode sequence that proves your patch effective
May 21 18:09:16 <DocScrutinizer> sure sure sigill
May 21 18:09:59 <freemangordon> CPU tries to execute thumb instruction (i.e. PC%4 !=0) in ARM mode and sigills
May 21 18:10:46 <freemangordon> but it is not a specific instruction sequence that bugs the core, it is some random memory mammping
May 21 18:10:50 <freemangordon> *mapping
May 21 18:11:08 <DocScrutinizer> err, then modest wouldn't run even with your patch?
May 21 18:12:48 <freemangordon> it will, as branch predictor "stale data" is flushed after the context is switched to "modest" and before the control is returned to user space. So on the next branch CPU will read the code and will behave as it should, instead of relying on what BP thinks
May 21 18:13:40 <freemangordon> BTW patch is from TI, not mine
May 21 18:14:04 <freemangordon> MY part is actually enabling it by calling the secure monitor
May 21 18:15:12 <DocScrutinizer> even the better - then we only need to show your part actually enables the patch
May 21 18:15:32 <freemangordon> oooh, that is easy :D
May 21 18:15:36 <DocScrutinizer> as I'm absolutely sure TI tested the patch for effectiveness
May 21 18:15:49 <freemangordon> DocScrutinizer, why you don't read what I wrote
May 21 18:16:33 <DocScrutinizer> err, I do
May 21 18:16:47 <Estel_> DocScrutinizer,  in this case You don't :P
May 21 18:17:06 <freemangordon> DocScrutinizer, what I did is to run u-boot and to get the contents of ACR. We need IBE bit set. It is not
May 21 18:17:15 <Estel_> freemangordon said ages ago that negators are forcing himto prove that TI err5a does what it does
May 21 18:17:39 <Estel_> he proved long time ago that his method to enable Ti's workaround is working
May 21 18:18:16 <Estel_> and all concerns from You ( DocScrutinizer ) , dm8tbr, Jaffa etc were based on doubts that TI errata does what it promises
May 21 18:18:31 <Estel_> i.e. testbed would only be meant to prove the latter
May 21 18:18:42 <Estel_> (checking if TI does what promises with errata)
May 21 18:18:49 <DocScrutinizer> an erratum never promises anything, it's simply an exiasting annoyance
May 21 18:18:54 <freemangordon> DocScrutinizer, patch code is here http://kerneltrap.org/mailarchive/git-commits-head/2009/5/3/5622724 , follow me and I will explain why what we have in KP does not work and how I make it work
May 21 18:18:58 <Estel_> exactly
May 21 18:19:24 <Estel_> but but insisting on testbed, You wanted mechanism to prove that errata isn't fcked up itself
May 21 18:19:24 <freemangordon> the patch consist of two parts, first is setting IBE bit in ACR
May 21 18:19:33 <Estel_> nevermind
May 21 18:19:50 <freemangordon> +	mrc	p15, 0, r10, c1, c0, 1		@ read aux control register
May 21 18:19:50 <freemangordon> +	orr	r10, r10, #(1 << 6)		@ set IBE to 1
May 21 18:19:50 <freemangordon> +	mcr	p15, 0, r10, c1, c0, 1		@ write aux control register
May 21 18:19:53 <freemangordon> sorry
May 21 18:19:54 <Estel_> freemangordon is saying more interesting stuff than me, so bear with him ;)
May 21 18:20:23 <freemangordon> DocScrutinizer, the part of the patch which sets IDE bit in ACR does not work on n900
May 21 18:20:40 <DocScrutinizer> :nod:
May 21 18:20:51 <freemangordon> as ACR is not accessible in unsecure priviledged mod (i.e. kernel)
May 21 18:20:57 <DocScrutinizer> :nod:
May 21 18:21:01 <freemangordon> same for u-boot
May 21 18:21:20 <Estel_> ...in meantime, I got question about program (well, just set of interconnected scripts) I'm preparing for repositories, so just tell me when You two ends and I'll ask, as I'm ashamed to break such conversation) ;)
May 21 18:21:55 <freemangordon> A month or so ago I found an SMC call (that is a call that transfers the control to ROM code) which is used for setting ACR contents
May 21 18:22:07 <freemangordon> And it is desctibed in TRM by the way
May 21 18:22:12 <freemangordon> *described
May 21 18:22:43 <DocScrutinizer> fine, so I stae I think your patch is fine, for the scope of this erratum
May 21 18:22:47 <DocScrutinizer> state*
May 21 18:23:18 <freemangordon> so I patched u-boot https://gitorious.org/~freemangordon/u-boot-shr/freemangordons-u-boot/commit/409fec178945067233629a471653361b927c05f0
May 21 18:23:46 <Estel_> when I told You same thing You negated it :( just joking, I was using term secure mode, which probably wasn't so clear to You ( DocScrutinizer )
May 21 18:23:48 <DocScrutinizer> now we only got this rumour there are two errata and allegedly mutually exclusive. We'd need to find who invented that and if he was right or wrong
May 21 18:24:08 <freemangordon> aah, that rumor, yeah
May 21 18:24:19 <Estel_> afaik 2nd errata is in bin space
May 21 18:24:21 <freemangordon> give me a second and I will explain (with the links)
May 21 18:24:27 <freemangordon> Estel_, no
May 21 18:24:29 <Estel_> and 3th errata is irrelevant if 1st errata is in place
May 21 18:24:42 <Estel_> Ok, swicthing to listening mode :P
May 21 18:24:43 <freemangordon> just give me some time
May 21 18:24:49 <DocScrutinizer> Estel_: a patch isn't an erratum
May 21 18:25:00 <DocScrutinizer> erratum=bug, patch=fix
May 21 18:25:02 <Estel_> DocScrutinizer,  true, true
May 21 18:25:27 <Estel_> in meantime, when freemangordon searchs, I'll ask my small question ok?
May 21 18:25:33 <Estel_> first one is rather simple
May 21 18:25:57 <freemangordon> DocScrutinizer, the errata which we are told is mutually exclusive (and its number is stated in logs on #beagle dm8tbr is talking about) is 687067
May 21 18:26:07 <Estel_> I would like to "unleash" (in sh script) notification that stays until that line of code is done
May 21 18:26:20 <Estel_> (showing processing circle)
May 21 18:26:42 <freemangordon> DocScrutinizer: that is what I found about it, not Cortex-a8, but the same OMAP3 core http://cache.freescale.com/files/32bit/doc/errata/IMX50CE.pdf
May 21 18:27:01 <freemangordon> search the pdf for 687067
May 21 18:28:38 <freemangordon> and read the errata description. My understanding is that someone has misread that as setting the IBE bit would lead to errata triggering, and because part of the previous workaround is setting the IBE bit, thus mutual exclusivity
May 21 18:28:42 <Estel_> (i.e. re-enabling swap, and popping-up notification about it via maemo notifications, that stays until lines of code responsible for reswapping aren't finished)
May 21 18:29:55 <freemangordon> DocScrutinizer, which is nonsense, as to really trigger errata 687067, one have to issue "invalidate Branch predictor by MVA"
May 21 18:33:33 <freemangordon> along with having IBE bit enabled. But workaround for the previous errata does not issue "invalidate Branch predictor by MVA", but "Invalidate entire branch predictor array."
May 21 18:34:26 <javispedro> untrue
May 21 18:34:40 <freemangordon> so errata 687067 is never triggered. 
May 21 18:34:53 <javispedro> it is the same thing
May 21 18:35:25 <javispedro> behind the scenes
May 21 18:35:27 <freemangordon> javispedro, those are 2 diffrent operations, check on ARM side, coprocessor c7, cache operations
May 21 18:35:35 <DocScrutinizer> (My understanding is that someone...) yes
May 21 18:35:50 <javispedro> I know, they have a one bit difference iirc
May 21 18:36:03 <freemangordon> and it is clearly stated that errata is triggered by "invalidate by MVA"
May 21 18:37:11 <freemangordon> javispedro: "3. An invalidate Branch predictor by MVA operation is executed"
May 21 18:38:16 <javispedro> this is not an errata
May 21 18:38:19 <javispedro> it is actually intentional
May 21 18:38:41 <DocScrutinizer> anyway 687047 explains that a proper initialized to 0  "L1 System Array Debug Register 0" will "fix" the erratum, just as does IBE=0
May 21 18:38:56 <freemangordon> DocScrutinizer, we don't have acces to that
May 21 18:39:01 <freemangordon> *access
May 21 18:39:57 <freemangordon> javispedro, if you have some additional documentation, please share it. What is not errata? And why is that intentional, what intention?
May 21 18:40:21 * mva >_> freemangordon
May 21 18:40:53 <javispedro> "debug"
May 21 18:41:01 <javispedro> the entire acr is for debug stuff
May 21 18:41:04 <javispedro> *acr register
May 21 18:41:33 <freemangordon> and how is that related?
May 21 18:41:40 <javispedro> ibe
May 21 18:42:03 <javispedro> you are not supposed to use it :) 
May 21 18:42:10 <freemangordon> aaah, ok
May 21 18:42:11 <DocScrutinizer> ???
May 21 18:42:14 <freemangordon> :)
May 21 18:42:23 <javispedro> in any case
May 21 18:42:26 <freemangordon> javispedro, where did you get that from?
May 21 18:42:47 <javispedro> there is a true official way to disable the branch predictor iirc
May 21 18:42:54 <DocScrutinizer> yes
May 21 18:42:54 <freemangordon> and btw we are talking about upstream patch, not some invention of mine
May 21 18:43:40 <freemangordon> javispedro, disable the whole BP? why should anyone want to do that?
May 21 18:44:04 <freemangordon> and how is that related to workaround working/not?
May 21 18:44:14 <Estel_> andp again, please provide sources of Your informations, javispedro
May 21 18:44:38 <DocScrutinizer> if any possible
May 21 18:44:40 <Estel_> s/andp/and
May 21 18:45:05 * Estel_ is wondering why s/ / never works for him
May 21 18:45:16 <freemangordon> you are missing the last /
May 21 18:45:22 <DocScrutinizer> because it's s/ / /
May 21 18:45:23 <Estel_> thanks
May 21 18:46:18 <Estel_> guys, again, could please someone tell me how to invoke notification, that is persisting until operation in that line of code (which it's attached to) ends?
May 21 18:46:48 <Estel_> for example, sh script that re-enables swap - I would like maemo notification (yellow bar) to pop-up and stay until it proceeds
May 21 18:46:53 <javispedro> freemangordon: was thinking to check if the crashes stop
May 21 18:47:06 <Estel_> saying "Re-enabling swap..."
May 21 18:47:21 <DocScrutinizer> Estel_: no known method to close a notification programmatically
May 21 18:47:33 <Estel_> damn, Saturn is using it on Cleven
May 21 18:48:02 <Estel_> during searching for surrounding networks
May 21 18:48:13 <Estel_> it says "scanning for networks"
May 21 18:48:41 <Estel_> and anomated circle is inside this yellow notification bar - this whole thing dissapear after search is complete
May 21 18:48:44 <freemangordon> javispedro: I (and not only me) am able to boot and use ubuntu 12.04 with those fixes. Without them it crashes(SIGILL) every 15-20 seconds
May 21 18:49:01 <Estel_> if it take 5 seconds, it persist for 5 seconds. If it take 30, it stays 30 sec
May 21 18:49:09 <Estel_> will need to ask saturn I suppose
May 21 18:49:59 <DocScrutinizer> 687067 is a rather unlikely_to_happen erratum
May 21 18:50:11 <Estel_> 2nd question, much *more*important for actual operation of my "program", so please, bear with me
May 21 18:50:26 <Estel_> I want to create a script, meant to be used only once...
May 21 18:50:29 <DocScrutinizer> nevertheless we should fix it before moving to a system that *might* trigger it
May 21 18:51:01 <Estel_> that sources config filled by user, and patches /event.d/rcS-late accordingly
May 21 18:51:15 <Estel_> replacing given line with other content
May 21 18:51:38 <freemangordon> javispedro, have in mind you will need latest u-boot and KP49 or custom build KP50 with workaround for 430973 enabled, as it is disabled in KP50
May 21 18:51:38 <Estel_> could anyone tell me, how such script should look like?
May 21 18:51:48 <freemangordon> DocScrutinizer, again, it is never triggered
May 21 18:51:59 <Estel_> of course, initially I'm sourcing config file with values, then using those values to prepare proper lines
May 21 18:52:19 <DocScrutinizer> please discuss this with javispedro, it seems he thinks different
May 21 18:52:33 <DocScrutinizer> as I'm away for dinner now - sorry
May 21 18:52:35 <javispedro> I... don't, actually
May 21 18:52:46 <Estel_> then, I want script (run as root) to find certain line in rcS-late, and replace it with line defined in my script
May 21 18:52:51 <javispedro> if it's working, what was the problem then?
May 21 18:53:05 <freemangordon> DocScrutinizer is not convinced :D
May 21 18:53:29 <javispedro> you will still see random failures in jitted code
May 21 18:53:35 <Estel_> javispedro,  the thing is that many - otherwise respectable - people think that errata fix is not applicable to N900
May 21 18:53:40 <Estel_> no matter what
May 21 18:53:44 <javispedro> well
May 21 18:53:45 <DocScrutinizer> I am the problem, I want to know it really got *fixed* and not simply moved elsewhere, like an alignment error in data section
May 21 18:53:56 <Estel_> without any reasoning, beside some FUD ;)
May 21 18:53:58 <freemangordon> btw I have to go, otherwise I risk my GF leaving :D
May 21 18:54:02 <javispedro> the errata imposes a small cycle penalty on context switch
May 21 18:54:03 <Estel_> DocScrutinizer,  isn't the problem
May 21 18:54:19 <javispedro> s/the errata/the fix
May 21 18:54:32 <javispedro> well, _comparatively_ small
May 21 18:54:39 <freemangordon> javispedro, yes, but the point is that we have 20-40 % reduction in codesize
May 21 18:54:43 <Estel_> dm8tbr,  and some random clueless arm employee was more of a problem
May 21 18:55:08 <freemangordon> and having in mind RAM size on n900...
May 21 18:55:27 <freemangordon> we could get rid of most of the swapping
May 21 18:55:29 <Estel_> ...it would make me happy penguing
May 21 18:55:41 <javispedro> that is the part I don't believe
May 21 18:55:55 <freemangordon> which part?
May 21 18:56:00 <dm8tbr> Estel_: stop being a jerk. I tried to help and the conversation between gordon and mru didn't go beyond chitchat.
May 21 18:56:00 * DocScrutinizer waves
May 21 18:56:13 <Estel_> I don't think 20-40% would get us out of swapping, but I belive It would reduce it
May 21 18:56:20 <Estel_> s/It/it/
May 21 18:56:20 <infobot> Estel_ meant: I don't think 20-40% would get us out of swapping, but I belive it would reduce it
May 21 18:56:23 <javispedro> cya DocScrutinizer 
May 21 18:56:25 <freemangordon> but of course
May 21 18:56:43 <freemangordon> there is no way we could run only in RAM, but still
May 21 18:56:56 <Estel_> dm8tbr,  not necessary, when I discutted with You last time, You were making stupid *facedesk*s and other emoticons
May 21 18:57:04 <Estel_> without any counter-reasoning
May 21 18:57:06 <freemangordon> 20-30 (or even more) MB of free RAM will make a difference
May 21 18:57:39 <freemangordon> BTW I got to go too, happy arguing :P
May 21 18:57:59 <Estel_> and I was using same argumentd as freemangordon today, yet using less technical wording, as I'm much less experienced. Still, merit was understandable for anyone with willing other thing making laughs
May 21 18:58:19 <Estel_> freemangordon, see ya
May 21 18:58:22 <dm8tbr> Estel_: I do not think I need to discuss this with you. If anyone else here wants to, that's fine, but you have been hostile from the get got, so please just go put me on ignore and stop being a jerk. KTXBAI
May 21 18:58:49 <Estel_> I see no hostility on discussing, I'm just stating facts. anyway, it's no problem for me
May 21 18:59:04 <Estel_> and I don't need excuses, if this was meant to be it ;)
May 21 18:59:14 <dm8tbr> FOAD *plonk*
May 21 18:59:14 <Estel_> lets just move forward
May 21 18:59:52 <Estel_> BTW, last time You've used this face on a desk when I was talking about possible source of FUD around errata, so...
May 21 18:59:55 <Sicelo> ~foad
May 21 18:59:55 <infobot> foad is probably \"fuck off and die\". Considered by many to be impolite.
May 21 19:00:06 <Estel_> take Your time, it's EOT for me
May 21 19:00:08 <Estel_> haha
May 21 19:00:20 <Estel_> Sicelo, n1
May 21 19:00:59 <Sicelo> n1? i just didn't know that foad thingie
May 21 19:00:59 <Estel_> I think he meant face on a desk, though
May 21 19:01:00 <Estel_> still, it's his face, so, let him enjoying it.
May 21 19:01:05 <Sicelo> oh
May 21 19:01:08 <Sicelo> :D
May 21 19:01:16 <Estel_> but it mixed nicely with discussion
May 21 19:01:29 <Estel_> merlin1991,  ping
May 21 19:04:14 <Sicelo> just one question from me regarding this Thumb thing: say we all eventually embrace it, etc and we start using thumb-compiled stuff ... what about the stuff which can't be thus recompiled, eg bme?
May 21 19:05:06 * Sicelo doesn't even really know what this Thumb is (besides being some property of the CPU)
May 21 19:06:06 <Estel_> theoretically, there is a speed disadvantage when running non-thumb compiled code AND errata fix applied, but, it's hardly noticeable in real life
May 21 19:06:23 <Estel_> and speed gain due to less ram used by core components is much bigger advantage
May 21 19:06:40 <Estel_> thumb2 is just a flag during compiling, from developer side
May 21 19:07:05 <freemangordon> Sicelo, it will continue to work as before "thumb era"
May 21 19:07:17 <freemangordon> there is no need to re-compile everything
May 21 19:07:35 <Estel_> freemangordon,  You've lost girlfriend? haven't beed suspecting You'll return so quickly ;)
May 21 19:08:01 <freemangordon> Estel_, luckily she got a phonecall, still talking :D
May 21 19:08:08 <Estel_> :D
May 21 19:08:25 <Estel_> wouldn't You mind to help such a noob as me with small thing?
May 21 19:08:29 <freemangordon> so actually I didn't return, nut still not left
May 21 19:08:36 <Estel_> question, actually
May 21 19:09:11 <Sicelo> i see. 
May 21 19:09:24 <freemangordon> Sicelo, thumb 2 instructions are 16bit (2 bytes) while ARM are 32 bits(4 bytes)
May 21 19:09:28 <Estel_> I'm finishing small package that I wqant to upload into repos. Last thing that is keeping me from it is that I would like to spare editing of rcS-late to users by hand
May 21 19:09:40 <freemangordon> sed?
May 21 19:09:51 <Estel_> I want to create a single-run script, that will read config file...
May 21 19:10:12 <Estel_> and find line in rcS-late + replace it with thing I prepare
May 21 19:10:20 <Estel_> (using variables from config)
May 21 19:10:25 <Estel_> I know how to source etc 
May 21 19:10:32 <Estel_> just how to properly sed
May 21 19:10:46 <Estel_> rcS-late (or any other existing script)
May 21 19:10:53 <Estel_> I'm big noob here
May 21 19:11:23 <Estel_> so example of hyphotetical line of code would help much
May 21 19:12:01 <Estel_> lets say I want to find a "foo" line in rcS-late and replace it with "bar is great" 
May 21 19:12:02 <freemangordon> Estel_, I am not the best one to ask re shell scripts, but AFAIK you should use sed to do a inline replacement, along with leaving your "signiture" somewhere in the script to avoid double replacements. Sorry, I think it is better to ask DocScrutinizer or merlin1991
May 21 19:12:24 <Estel_> ok, thanks a lot for suggestion anyway
May 21 19:12:30 <Estel_> I'll try messing with seed or ask them
May 21 19:12:52 <Estel_> tried that, but DocScrutinizer was too much on thumb thing and merlin1991 is absent, will try to catch them
May 21 19:12:58 <freemangordon> *sed
May 21 19:13:07 <freemangordon> not seed :P
May 21 19:13:07 * Estel_ is going to read about sed
May 21 19:13:12 <Estel_> :P
May 21 19:13:25 <freemangordon> you'd better not, justr ask google
May 21 19:13:29 <freemangordon> *just
May 21 19:13:35 <Sicelo> sed -e -i s/<original>/<new>/ <filename>
May 21 19:13:44 <freemangordon> :nod:
May 21 19:13:45 <Sicelo> i think it's like that
May 21 19:13:52 <freemangordon> why -e ?
May 21 19:14:19 <freemangordon> abyway, bb
May 21 19:14:24 <freemangordon> *anyway
May 21 19:14:32 <Estel_> oh god, sed is the thing I'm using to correct lines here? I said I'm noob
May 21 19:14:38 <Estel_> see ya, thanks!
May 21 19:14:39 <Sicelo> yeah, lol.. remove the 'e'
May 21 19:16:36 <Sicelo> anyway, reading is best .. ;)
May 21 19:16:57 <Raimu> :P
May 21 19:16:58 <Woody14619> Estel_, technicaly, sed is based on ed, a line editor the came before vi.  (s = streem)
May 21 19:17:07 <Raimu> ...whoops, wrong window. Sorry.
May 21 19:17:47 <Estel_> Woody14619,  nice to know
May 21 19:18:11 <Estel_> hm, I've searched about it, but can't find what freemangordon meant by "kind of signature" to avoid double-replacing
May 21 19:18:23 <Woody14619> Also note, that most ed commands still work in vi when in ed mode (using :)  So :s/orig/new/ works in vi as well.  (one per line, append a g to the end to have it happen multiple times per line)
May 21 19:18:23 <DocScrutinizer> yeah. ed anybody?
May 21 19:18:26 <DocScrutinizer> :-P
May 21 19:18:56 <DocScrutinizer> Woody14619: probably s in sed wasn't for stream but for silly
May 21 19:18:59 <DocScrutinizer> ;-)
May 21 19:19:11 <Sicelo> lulz
May 21 19:19:13 <DocScrutinizer> ed is really a monster
May 21 19:19:37 <javispedro> Estel_: you are being way overoptimistic
May 21 19:19:39 <Woody14619> lol :)  Yes, ed sucks, but was a lot nicer than puch cards, which was the other option at the time.
May 21 19:19:40 <DocScrutinizer> edlin and whatnot else
May 21 19:19:49 <javispedro> let's assume for a moment that thumb actually produces a 40% reduction in code size, which I don't believe
May 21 19:20:01 <Woody14619> .oO(and vi doesn't work well on tele-type devices, which lack screens...)
May 21 19:20:10 <DocScrutinizer> indeed
May 21 19:20:13 <javispedro> use procps to measure the text resident size of all processes in the system
May 21 19:20:29 <javispedro> iirc, on a fresh booted n900 it was around 50MiB
May 21 19:20:40 <javispedro> note that this is overly pessimistic as this countes shared pages twice
May 21 19:20:47 <javispedro> s/twice/more than once
May 21 19:21:02 <Woody14619> edlin was the crappy MS nock-off of ed.  It would be like compairing the "NOKlA" brand "n900" against the one from NOKIA.... ;)
May 21 19:21:16 <javispedro> you will save 25MiB _hard_ maximum 
May 21 19:21:26 <javispedro> the real number will probably be way less than that
May 21 19:21:37 <Estel_> javispedro,  any way I could measure text resident size of processes I've enabled now?
May 21 19:21:57 <Estel_> what's the point of measuring it for fresh system
May 21 19:21:59 <javispedro> use procps, grep man for text resident size
May 21 19:22:09 <Estel_> on fresh system You hardly use swap
May 21 19:22:20 <javispedro> fremantle preloads, you know
May 21 19:22:22 <Estel_> problem is when You use qt things
May 21 19:22:25 <Estel_> yes.
May 21 19:22:47 <DocScrutinizer> and on system with huge uptime, memory gets used by data, not progamtext
May 21 19:22:48 <javispedro> a fresh system here has already a bunch of all the most used apps
May 21 19:22:51 <Estel_> but printfood for not-preloaded things, yet thumb compiled, will drop to, yes?
May 21 19:23:06 <Estel_> s/to/too
May 21 19:23:07 <javispedro> "yes"
May 21 19:23:08 <DocScrutinizer> the size of code stays the same
May 21 19:23:09 <Estel_> s/to/too/
May 21 19:23:10 <infobot> Estel_ meant: s/too/too
May 21 19:23:35 <DocScrutinizer> haha
May 21 19:23:40 <Woody14619> :)
May 21 19:23:41 <Estel_> I mean that, during usage of memory-heavy things, we main gain more
May 21 19:23:43 <Estel_> :)
May 21 19:23:51 <DocScrutinizer> no
May 21 19:24:04 <DocScrutinizer> memory heavy things are heavy on data, not on code
May 21 19:24:13 <Estel_> byw 25 MB would be more than 10% ;)
May 21 19:24:21 <Estel_> true
May 21 19:24:32 <DocScrutinizer> btw code never gets swapped out
May 21 19:24:33 <Estel_> (in fact it depends, but generally true)
May 21 19:24:56 <javispedro> DocScrutinizer was missing a "to swap" there.
May 21 19:24:56 <Estel_> so it's even more important to have it with reduced size ;)
May 21 19:25:01 <DocScrutinizer> useless to swap out code, as it's already on 'disk'
May 21 19:25:05 <javispedro> aka code never gets swapped out to swap :)
May 21 19:25:34 <Estel_> all after all we get more ram available for data that don't need to be swapped out
May 21 19:25:39 * DocScrutinizer waves again
May 21 19:26:00 <Estel_> considering ramzswap was a way to gain a little ram at cost of CPU
May 21 19:26:01 <Estel_> cycles
May 21 19:26:02 <javispedro> Estel_: as said, this is assuming 40% code reduction (which I don't believe), and 0% loss in performance (which I don't believe either)
May 21 19:26:13 <Estel_> and now we can gain a little ram without any cost (almost)
May 21 19:26:21 <Estel_> I think it's worrth effort
May 21 19:26:23 <javispedro> well
May 21 19:26:27 <javispedro> it's an effort.
May 21 19:26:39 <Estel_> everything is :)
May 21 19:27:04 <javispedro> for the interesting programs, it will for sure not be as simple as toggling -mthumb
May 21 19:27:04 <Estel_> cdreating HD video recording for N900 was an huge effort too, and I enjoy it very much
May 21 19:27:14 <javispedro> think inline asm, etc.
May 21 19:27:21 <Estel_> hm.
May 21 19:27:40 <Estel_> last two lines should be forwarded to freemangordon, I've no idea, frankly
May 21 19:27:46 <javispedro> also, I don't remember if interworking is enabled =)
May 21 19:27:49 <Estel_> would be interested in his comment about it
May 21 19:28:24 <Estel_> btw I mean that even "wasting" cpu cycles for compression using ramzswap is a win, most of the times, on system with low ram
May 21 19:28:30 <Woody14619> javispedro, I believe the 40%... The stats on that are pretty firm.  But it's not a huge ammount.  Still, if it's available, even getting back 15-20Mb would make a difference for some people, usage dependant.
May 21 19:28:37 <Estel_> so such "almost" free ram gain is a real deal
May 21 19:29:15 <javispedro> Woody14619: 40% would mean that ~40% of the code doesn't need the extra upper arm registers
May 21 19:29:46 <javispedro> still possible and quite certain for some programs, but I have my doubts if that will be the case for "everything"
May 21 19:30:02 <Estel_> only one way to check that
May 21 19:30:05 <Estel_> for sure
May 21 19:30:12 <Estel_> actually trying :)
May 21 19:30:52 <javispedro> feel free
May 21 19:31:18 <Estel_> cssu devel sounds great place for that
May 21 19:31:32 <Estel_> s/sounds/seems like/
May 21 19:31:32 <infobot> Estel_ meant: cssu devel seems like great place for that
May 21 19:33:58 <Woody14619> Which is believable, and even testable.  I'm betting it compiles a little differntly, but most would compile fine.  It is an effort, yes.  But if there's a working patch for the issue, it may be worth it for some.
May 21 19:34:21 <Estel_> as for sed
May 21 19:34:38 <Estel_> ok, but how will it allow me to replace multi-line things in config?
May 21 19:34:48 <Estel_> s/config/file/
May 21 19:34:49 <infobot> Estel_ meant: ok, but how will it allow me to replace multi-line things in file?
May 21 19:35:16 <Estel_> lets say I want my script to replace 2-liner in rcS-late with 7-liner
May 21 19:35:30 <Estel_> Sicelo, DocScrutinizer, Woody14619?
May 21 19:36:24 <Sicelo> that's for the greats :P
May 21 19:36:43 <Sicelo> i think you will use a bunch of \n though 
May 21 19:37:56 <Estel_> probably right
May 21 19:41:21 <Woody14619> yup yup.  Or, if you're feeling really self-damaging, look at awk. :)
May 21 19:41:30 <Woody14619> work becons
May 21 19:45:07 <Estel_> hm, smth bad
May 21 19:45:43 <Estel_> sed s/original\n3liner\ncontent/new\n2linder/
May 21 19:45:47 <Estel_> doesnt work
May 21 19:46:02 <Estel_> it should replace 3 lines
May 21 19:46:12 <Estel_> "original 3liner content"
May 21 19:46:33 <Estel_> with 2 lines "new 2linder" (typo, should be liner, but whatever)
May 21 19:46:45 <Estel_> any ideas?
May 21 19:47:37 <Sicelo> ~jrtools
May 21 19:47:38 <infobot> methinks jrtools is http://wiki.maemo.org/User:Joerg_rw/tools
May 21 19:47:58 <NIN101> I am no sed pro according  to my experience multilne replacements with sed are PITA.
May 21 19:48:00 <Estel_> it says "unmatched "/"
May 21 19:48:16 <Estel_> agree. what's better alternative?
May 21 19:48:50 <Estel_> (unmatched "/" probably mean it wasn't able to find matching original content due to \n's inside)
May 21 19:49:49 <Sicelo> you used the "?
May 21 19:51:22 <Estel_> ?
May 21 19:51:31 <Estel_> could You elaborate more?
May 21 19:52:55 <Sicelo> look..i'm worse than even NIN101 on this :P
May 21 19:53:51 <Sicelo> i was just asking if you remembered to use the quote marks in appropriate places .. sed -i "s/.../.../"
May 21 19:54:38 <Estel_> of course forget about it, thanks a lot
May 21 19:54:42 <Estel_> will try now
May 21 19:55:00 <Estel_> btw request here
May 21 19:55:11 <Estel_> does anyone have unmodified rcS-late?
May 21 19:55:25 <Estel_> could You please paste it to zerobin or any pastebin?
May 21 19:58:20 <Sicelo> lol..me doesn't even know where that stays :P
May 21 20:00:57 <Estel_> now sed is working for me and I can replace one linders properly
May 21 20:01:11 <Estel_> yet, no way of using \n allows me to replace multi-liners
May 21 20:01:17 <Estel_> i've tried:
May 21 20:01:24 <Estel_> line1\nline2
May 21 20:01:32 <Estel_> line1 \nline2
May 21 20:01:40 <Estel_> line1\n line2
May 21 20:01:48 <Estel_> line1 \n line2
May 21 20:01:52 <Estel_> ...al to no avail
May 21 20:02:00 <Estel_> s/al/all/
May 21 20:02:01 <infobot> Estel_ meant: ...all to no avail
May 21 20:02:11 <Estel_> any ideas, anyone? Woody14619?
May 21 20:02:13 <Sicelo> where's this rsS-late
May 21 20:02:26 <Estel_> /etc/event.d/rcS-late
May 21 20:02:33 * Sicelo is lazy to do a find
May 21 20:02:34 <Estel_> critical file during booting
May 21 20:02:50 <Estel_> file isn't problem here I'm testing it with test file
May 21 20:03:13 <Estel_> problem is with replacing multi liners with multiliners
May 21 20:03:16 <Sicelo> oh, i thought you wanted it pasted? :\
May 21 20:03:22 <Estel_> ah sorry
May 21 20:03:24 <Estel_> of course
May 21 20:03:31 <Estel_> forget, but still need it
May 21 20:03:33 <Estel_> :P
May 21 20:04:16 <Estel_> I could invoke sed multiple times, but it still doesn't allow me to change number of lines to greater
May 21 20:05:29 <Estel_> (I could replace 7 lines with less, for example, 5, by replacing unnecesary lines with spaces...)
May 21 20:05:40 <Sicelo> http://paste.debian.net/170489/
May 21 20:05:41 <Estel_> (but can't change 5 lines to 7)
May 21 20:05:44 <Estel_> thanks a lot
May 21 20:06:01 <Estel_> so now only sed problem remains
May 21 20:06:56 <NIN101> Estel_ 
May 21 20:07:03 <NIN101> system-services
May 21 20:07:11 <NIN101> this package contains /etc/event.d/rcS-late
May 21 20:07:46 <Estel_> thanks.
May 21 20:08:14 <Estel_> anyway, problem with correct way for multi-lines replacing with arbitrary lines number still persist
May 21 20:08:36 <Estel_> I want to replace 1 line with multiliner :/
May 21 20:09:54 <Estel_> DocScrutinizer,  ping
May 21 20:10:00 <Sicelo> http://www.refining-linux.org/archives/27/20-Multi-line-sed-search-and-replace/
May 21 20:10:07 <Estel_> oh
May 21 20:10:17 <Estel_> thanks!
May 21 20:10:23 * Estel_ is going to read
May 21 20:12:08 <DocScrutinizer> Estel_: use diff / patch
May 21 20:12:34 * DocScrutinizer away again, cursing his own life
May 21 20:12:49 <Estel_> thanks...
May 21 20:13:57 * Sicelo gets ready for bed
May 21 20:14:34 <Sicelo> this channel rocks.. 98% of my linux knowledge comes from here :P
May 21 20:24:33 <GNUtoo-desktop> DocScrutinizer, hi I didn't understand the errata either, does it fix the issues like when you switch from arm to thumb and it illegal instructions because of that?
May 21 20:26:30 <DocScrutinizer51> it seems the branch prediction queue isn't getting properly cleaned in some cases when using a opcode (bx?) that is switching instruction set
May 21 20:27:06 <GNUtoo-desktop> ok I understand that(branch prediction)
May 21 20:27:19 <GNUtoo-desktop> it's fixed in 3.4?
May 21 20:27:46 <GNUtoo-desktop> because we have firefox that illegal instructions on SHR
May 21 20:28:02 <GNUtoo-desktop> and it's related to arm->thumb interworking switch
May 21 20:28:03 <DocScrutinizer51> it's fixed in more recent chips
May 21 20:28:10 <GNUtoo-desktop> ok
May 21 20:28:29 <GNUtoo-desktop> is it fixed in the gta04 chip?
May 21 20:28:42 <DocScrutinizer51> not on n900 omap soc though
May 21 20:28:57 <DocScrutinizer51> dunno, probably
May 21 20:29:03 <GNUtoo-desktop> ok
May 21 20:29:05 <GNUtoo-desktop> thanks
May 21 20:29:21 <DocScrutinizer51> check revision
May 21 20:29:37 <DocScrutinizer51> ask nikolaus
May 21 20:33:42 <DocScrutinizer51> r0p1 ?
May 21 20:33:54 <DocScrutinizer51> til r2pq ?
May 21 20:34:11 <DocScrutinizer51> it's in one of the links freemangordon posted
May 21 20:34:55 <Estel_> ping Hurrian
May 21 20:35:23 <DocScrutinizer51> Estel_: there are lesser-effort approaches to safe 15, 20 Mb of memory
May 21 20:35:34 <Estel_> like?
May 21 20:38:03 <Estel_> DocScrutinizer, ^^^
May 21 20:39:12 <DocScrutinizer51> forex look into all the preloaded stuff. Look at that 25MB file containing ALL translations
May 21 20:39:57 <Estel_> it's nice, we can mix it up with thumb2 and get 30-40 Mb of memory :)
May 21 20:40:16 <Estel_> but it's data, not code, yep?
May 21 20:40:18 <DocScrutinizer51> htop is your ftiend, and your common sense your shrpest weapon
May 21 20:40:36 <Estel_> unused translation must be amongst few things that get swapped out
May 21 20:40:54 <Estel_> common sense tells me that it's not mutualy exclusive with thumb2
May 21 20:40:59 <Estel_> btw, question
May 21 20:41:00 <DocScrutinizer51> WUT? you won't gain 20
May 21 20:41:10 <DocScrutinizer51> ooh, all together
May 21 20:41:31 <Estel_> swapon -a in rcS-late, it enables swap
May 21 20:41:39 <Estel_> how does it know where it is?
May 21 20:41:46 <Estel_> searches automatialy emmc?
May 21 20:41:52 <Estel_> and enable first found swap?
May 21 20:42:17 <DocScrutinizer51> look at autogenerated fstab
May 21 20:42:29 <Estel_> i.e. checks all mmcblk0p* and enable first found swap, or have it hardcoded?
May 21 20:42:31 <Estel_> I see
May 21 20:43:03 <Estel_> emr, so my guess that if user have swap on /dev/mmcblk0p758, swapon -a will also enable it
May 21 20:43:12 <Estel_> ?
May 21 20:43:42 <Estel_> no requiment for it to be on /dev/mmcblk0p3?
May 21 20:44:37 <jacekowski> well kinda
May 21 20:44:42 <jacekowski> it would have to be in fstab
May 21 20:44:50 <jacekowski> and fstab is regenerated quite often
May 21 20:54:09 <Estel_> jacekowski,  for my use case it's invoked from rcS-late, so it's during boot
May 21 20:54:27 <Estel_> no chances for missing fstab generation ;)
May 21 20:57:56 <Estel_> DocScrutinizer,  dumbie question again... while working with sed, how to distinguish "/" being part of line to be parsed, from "/" being part of sed command?
May 21 20:58:33 <Estel_> i.e. sed -i "s/home/user/newhome/newuser/"
May 21 20:58:37 <DocScrutinizer51> use arbitrary other delimiyer
May 21 20:58:46 <Estel_> use what?!
May 21 20:58:56 <DocScrutinizer51> like @
May 21 20:59:16 <Estel_> erm, any examples? I fail to get it
May 21 20:59:19 <DocScrutinizer51> s@foo/bar@fsck@
May 21 20:59:25 <Estel_> oh
May 21 20:59:43 <Estel_> and sed is bright enough to know what is delimiyer and what not
May 21 20:59:51 <Estel_> well, You've made it so simple :)
May 21 20:59:56 <DocScrutinizer51> yep
May 21 20:59:58 <Estel_> thanks a lot...
May 21 21:00:29 <DocScrutinizer51> yw
May 21 21:01:21 <merlin1991> just spent 3 hours swapping a laptop display
May 21 21:01:25 <merlin1991> fun times :D
May 21 21:02:00 <DocScrutinizer51> just spent 3 ours cursing and swearing
May 21 21:02:10 <DocScrutinizer51> as I lost my EC card
May 21 21:02:23 <DocScrutinizer51> real fun times
May 21 21:02:39 <merlin1991> well I cursed a lot too
May 21 21:02:49 <merlin1991> first time the display cbale wasnt in properly 
May 21 21:03:00 <merlin1991> then I didn't put in the keyboard properly
May 21 21:03:03 <merlin1991> :D
May 21 21:03:45 <DocScrutinizer51> yesterday I started to rip N9
May 21 21:04:07 <DocScrutinizer51> after watching some semi official youtube
May 21 21:05:25 <DocScrutinizer51> that suggests to 'do not re-use! discard!' half the phone
May 21 21:05:35 <DocScrutinizer51> I stopped that
May 21 21:05:57 <DocScrutinizer51> lol
May 21 21:06:27 <Estel_> DocScrutinizer,  those "do not use" iss bullshit
May 21 21:06:40 <Estel_> they say it about everything, from camera module to screws
May 21 21:06:47 <DocScrutinizer51> aemi
May 21 21:06:54 <DocScrutinizer51> semi
May 21 21:06:54 <Estel_> same apply for "official" instructions for N900
May 21 21:07:13 <Estel_> just use common sense as sharpest tool ;)
May 21 21:07:41 <DocScrutinizer51> yep
May 21 21:08:12 <DocScrutinizer51> instructions are for shops to ensure customers get 'like new' quality
May 21 21:08:14 <MrPingu_> Indeed, they say not to use these screws again
May 21 21:09:13 <MrPingu_> Btw, evening sirs
May 21 21:09:30 <Estel_> hi MrPingu
May 21 21:09:40 <Estel_> hey, experts, little question here again :P
May 21 21:09:48 <Estel_> take look at this please:
May 21 21:09:49 <Estel_> http://sebsauvage.net/paste/?15f3715baa9167de#wEO/AFX/OSRy+Pxue+sXooVRrWqxbpFdpTr+qRYPBTg=
May 21 21:10:03 <Estel_> I would like to scavenge this code, although, change it in a way
May 21 21:10:20 <Estel_> that I first source config file, where main swap and backup swap is defined
May 21 21:10:44 <Estel_> and then, instead of searching microsd for swap type partition, like in original code
May 21 21:11:14 <Estel_> I would like it to check if main swap (as defined in config) is available, and in other case, enable backup swap
May 21 21:11:39 <Estel_> (reasons for main swap being unavailable main be lack of microSD card, or corrupted, or whatever)
May 21 21:11:51 <Estel_> any idea how to achieve that?
May 21 21:12:39 <Estel_> by experts I mean DocScrutinizer, merlin1991 or anyone interested and able to help ;)
May 21 21:13:28 <DocScrutinizer51> n900 here
May 21 21:13:49 <DocScrutinizer51> no way I'll check that now
May 21 21:15:39 <DocScrutinizer51> I'd guess blkinfo or what it's called is your friend
May 21 21:16:21 <merlin1991> but invoke it with the flash cache command or you'll have real fun
May 21 21:16:44 <DocScrutinizer51> blkid
May 21 21:17:05 <DocScrutinizer51> merlin1991: ??
May 21 21:17:12 <DocScrutinizer51> sznc?
May 21 21:17:14 <DocScrutinizer51> sync?
May 21 21:17:36 <merlin1991> blikid -g
May 21 21:17:41 <merlin1991> err blkid -g
May 21 21:18:11 <merlin1991> otherwise he can have swap partitions that are long gone
May 21 21:18:22 <DocScrutinizer51> yup
May 21 21:18:48 <DocScrutinizer51> :-)
May 21 21:19:07 <MrPingu_> merlin1991 : What other gst packages do you have? From what I found of googling I miss libgstximagesrc.so so I might miss a whole package? :P
May 21 21:19:45 <merlin1991> MrPingu_: I only ever used tha todo screenshots
May 21 21:21:05 <merlin1991> but there's the package load-applet-xiamgewsrc
May 21 21:21:10 <merlin1991> might contain what you want
May 21 21:23:02 <MrPingu_> I installed gst-tools, earlier today thank you for that but all pipes are giving me the following error: No element "ximagesrc". Using the commands from the link you gave
May 21 21:23:12 <MrPingu_> Will look at it ;)
May 21 21:30:09 <Estel_> DocScrutinizer,  merlin1991, thanks.
May 21 21:30:17 <Estel_> blkid -g gives me nothing, though
May 21 21:30:26 <Estel_> only blkid alone gives me results
May 21 21:30:46 <merlin1991> you need to run blkid -g && blkid
May 21 21:31:02 <merlin1991> blkid -g only does the garbage collection, but does not display anyhting
May 21 21:31:04 <merlin1991> lulzworthy
May 21 21:32:15 <Estel_> I see, sorry
May 21 21:32:23 <Estel_> I think I'm a little tired :P
May 21 21:32:44 <Estel_> preparing first 100% own package in life for last few hours :P
May 21 21:39:21 <Estel_> merlin1991,  this line in original script:
May 21 21:39:24 <Estel_>       swapmicro=$(echo "$drivelist" -n | grep Id=82 | awk '/mmcblk1/ {print $1}')
May 21 21:39:33 <Estel_> says, in human language
May 21 21:40:22 <Estel_> "grep some value from drivelist output, and put it as value for swapmicro. Otherwise, leave swapmicro empty
May 21 21:40:47 <Estel_> later we have:
May 21 21:40:48 <Estel_>       if [ "$swapmicro" ]; then
May 21 21:41:07 <Estel_> which is either true if swapmicro got assigned proper value
May 21 21:41:11 <Estel_> or false, if not
May 21 21:41:27 <merlin1991> so where's your problem? :D
May 21 21:41:54 <Estel_> I would like to rephrase first mentioned line to do blkid -g && blkid and to grep location of swap
May 21 21:42:02 <Estel_> I would like to rephrase first mentioned line to do blkid -g && blkid and to grep location of swap
May 21 21:42:10 <Estel_> (definedd earlier)
May 21 21:42:17 <Estel_> or to leave it empty
May 21 21:42:25 <Estel_> which would make 2nd line false
May 21 21:42:40 <Estel_> for the life of me I can't get it working properly
May 21 21:42:53 <Estel_> any ideas how that line should like, i.e. what instead of:
May 21 21:43:05 <Estel_>       swapmicro=$(echo "$drivelist" -n | grep Id=82 | awk '/mmcblk1/ {print $1}')
May 21 21:43:19 <Estel_> having example line, would make it easier for me to learn...
May 21 21:44:28 <Estel_> forgive that I'm asking probably easy things
May 21 21:45:09 <Estel_> lets say I've, from sourced config file, a declaration: swap0=/dev/mmcblk1p2
May 21 21:45:11 <merlin1991> swapmicro=$(blkid -g && blkid  |  grep mmcblk1.*swap | awk 'print substr($1, 1, length($1) -1)')
May 21 21:45:28 <Estel_> oh god
May 21 21:45:40 <Estel_> so I make it:
May 21 21:45:54 <merlin1991> that should find the first swap on an sd card
May 21 21:46:18 <Estel_> yea, but I must set it to explicitly use swap provided via config
May 21 21:46:30 <MrPingu_> What are you cooking, Estel_ ?
May 21 21:46:40 <Estel_> i.e. I got declaration swap0=/dev/mmcblk1p2
May 21 21:46:53 <merlin1991> that's even easier
May 21 21:47:01 <Estel_> MrPingu,  solution for fool-proof anti-swap fragmentation
May 21 21:47:55 <Estel_> almost finished, merlin1991 is *very* kind to help me with hardest part, a script that will adjust people's rcS-late to enable proper swap on boot, or go failsafe to backup swap
May 21 21:48:23 <MrPingu_> sounds for me, I was/am too lazy to search/make swap refresh script :P 
May 21 21:48:27 <Estel_> swapmicro=$(blkid -g && blkid  |  grep $swap0.*swap | awk 'print substr($1, 1, length($1) -1)')
May 21 21:48:28 <Estel_> ???
May 21 21:48:41 <merlin1991> that sounds about right
May 21 21:48:43 <MrPingu_> Stil using the event.d method :P
May 21 21:49:22 <Estel_> merlin1991,  as I don't understand this line fully: doesn't it still search for first swap on microsd unnecessary?
May 21 21:49:34 <merlin1991> only that it doesn't work xD
May 21 21:49:39 <Estel_> :P
May 21 21:49:40 * merlin1991 investigates
May 21 21:50:08 <merlin1991> ah yeah blkid -g must move one line up
May 21 21:50:17 <Estel_> remember that I've put $swap0 there, which is earlier declared as, for example, /dev/mmcblk1p2
May 21 21:52:25 * merlin1991 plays a lil with awk
May 21 21:56:18 <merlin1991> swapmicro=$(blkid  |  grep $swap0.*swap | awk '{print substr($1, 1, length($1) -1)}')
May 21 21:56:42 <merlin1991> hm still not valid
May 21 21:56:42 <merlin1991> ffs
May 21 21:56:54 <Estel_> huge thank You! so, I need to put blkid -g a line above?
May 21 21:56:58 <Estel_> hm
May 21 21:57:06 <merlin1991> ah wait it works
May 21 21:57:10 <merlin1991> I'm only the master of pastefile
May 21 21:57:15 <merlin1991> and yeah blkid -g a line above
May 21 21:57:19 <Estel_> ;)
May 21 21:57:40 <merlin1991> s/pastefile/pastefail/
May 21 21:57:50 <Hurrian> Estel_?
May 21 21:58:46 <Estel_> Hurrian,  this example of rcsFlate editing to contain failsafe swap enabling is Your work, or from something else?
May 21 21:58:54 <Estel_> I would like to give proper credits
May 21 21:59:08 <Estel_> merlin1991,  so, now, if I put on next line:
May 21 21:59:10 <Estel_> if [ "$swapmicro" ]; then
May 21 21:59:12 <Hurrian> it's from the swapmicro script, lemme google the author (another guy on tmo)
May 21 21:59:34 <Estel_> it will be "dev/mmcblk1p2" as $swapmicro
May 21 21:59:43 <Estel_> if it exist
May 21 21:59:49 <Estel_> or false?
May 21 22:00:00 <merlin1991> /dev/mmcblk1p2
May 21 22:00:11 <Estel_> yes, sorry forget initial /
May 21 22:00:12 <merlin1991> the awk line is to strip the : from the output
May 21 22:00:21 <Estel_> ah, I see
May 21 22:00:22 <merlin1991> (err pipe not line)
May 21 22:00:50 <MrPingu> Yay, I fixed the gst-tools ;)
May 21 22:01:02 <Estel_> later I have "else", that enables backup swap
May 21 22:01:41 <merlin1991> btw I'd even use swapmicro=$(blkid  |  grep $swap0.*TYPE=\"swap\" | awk '{print sub
May 21 22:01:41 <merlin1991> str($1, 1, length($1) -1)}')
May 21 22:01:48 <Estel_> huge thanks merlin1991, really
May 21 22:02:03 <Estel_> hm?
May 21 22:02:03 <Hurrian> Estel_ : http://talk.maemo.org/showpost.php?p=846082&postcount=84
May 21 22:02:04 <merlin1991> otherwise you could end up trying to swap on a partitions that has the genious label of swap but not the type :D
May 21 22:02:11 <MrPingu> What I did was enabling SDK and TOOLS repo pull plugins-good-extra 
May 21 22:02:21 <Estel_> :D
May 21 22:02:29 <MrPingu> Thanks for the  help merlin1991 ;)
May 21 22:02:42 <merlin1991> err, yes :D
May 21 22:02:57 <merlin1991> fsckd putty
May 21 22:03:12 <merlin1991> was supposed to be swapmicro=$(blkid  |  grep $swap0.*TYPE=\"swap\" | awk '{print substr($1, 1, length($1) -1)}')
May 21 22:04:39 <Estel_> I'm using it with pleasure :)
May 21 22:04:55 <Estel_> soon I should have it finished, will try to upload into -devel...
May 21 22:05:08 <merlin1991> also I wouldn't do a simple if [ $swapmicro ] but if [ -n $swapmicro ]
May 21 22:05:27 <Estel_> right, ok.
May 21 22:05:28 <merlin1991> it's safer to check against non zero length than implicit "false"
May 21 22:05:48 * Estel_ nods
May 21 22:06:37 <MrPingu> I won't mind testing, as soon as you packaged and uploaded it, Estel_ 
May 21 22:06:57 <merlin1991> MrPingu: so when can we see the result of your work? :D
May 21 22:07:42 <Estel_> Hurrian,  thanks
May 21 22:07:48 <Estel_> MrPingu,  also thanks ;)
May 21 22:08:00 <DocScrutinizer> Estel_: you're tired? *I* got BSE - o my brain transmuted to extremely concentrated vacuum
May 21 22:08:13 <Estel_> :P
May 21 22:08:16 <Hurrian> Estel_, are you pushing it as a replacement rcS-late?
May 21 22:08:52 <Estel_> no, just as set of script for swap refreshing purposes, which also take user configurable swap locations and adjust rcs-late accordingly
May 21 22:08:58 <Hurrian> ah
May 21 22:09:00 <Hurrian> alright
May 21 22:09:02 <Estel_> so, it's kinda multitool ;)
May 21 22:09:06 <MrPingu> As soon as I got a proper vid of cleven and faircrack and like ;) 
May 21 22:10:15 <DocScrutinizer> pondering about this shitty vendor machine is built in a way to make fools with bse forget their credit card, since it offers the goods before you remove the card. Pondering still when I was 20m away from machine. 3h later I noticed I forgot my credit card
May 21 22:10:50 <Estel_> in fact I've also developed small notification system that iform user about success, or, in case of failure, report about exact step that failed
May 21 22:10:57 <Estel_> via maemo notification system
May 21 22:10:59 <DocScrutinizer> I expect a ROTFL at very least
May 21 22:11:15 <Estel_> DocScrutinizer,  was about to write lol
May 21 22:11:16 <Estel_> ;)
May 21 22:11:26 <Estel_> but I suspect it's big problem for You
May 21 22:11:44 <merlin1991> well dunno I suppose it's more annoying than a real problem
May 21 22:11:51 <merlin1991> you have to lock the fsckd card and get a new one
May 21 22:11:51 <Estel_> ah, now I know why bank machines want me to take card first
May 21 22:11:55 <merlin1991> and pay a lil for that
May 21 22:12:03 <Estel_> :P
May 21 22:13:03 <DocScrutinizer> it's big problem not for the card but for my infallibly accurate brain rotting to pile of crap with this job
May 21 22:14:44 <DocScrutinizer> last time something like this happened to me was 15 years ago, on Valium which my doc thought was a good prescription against my stomach problems
May 21 22:15:26 <GeneralAntilles> Hooray, brain rot!
May 21 22:15:31 * GeneralAntilles drools.
May 21 22:15:35 <DocScrutinizer> card been locked since 4h now
May 21 22:15:52 <MrPingu> I will post a YT link here, for sure ;) Anyway it can take some weeks as I actually have exams and go on holiday with classmates directly after :P
May 21 22:16:15 <GNUtoo-desktop> ok
May 21 22:16:30 <DocScrutinizer> btw the lock phone call voice machine is such a fucktard as well, I already was about to power up my MTHELs
May 21 22:17:37 <DocScrutinizer> +49911116116
May 21 22:17:50 <DocScrutinizer> FSCKU
May 21 22:18:15 <DocScrutinizer> I buolt better voice mailboxes 20 years ago
May 21 22:18:19 <DocScrutinizer> built*
May 21 22:18:42 <merlin1991> reminds me of the day my server @ 1and1 went down
May 21 22:18:50 <merlin1991> phone system was totally fsckd
May 21 22:18:58 <DocScrutinizer> 1&1 NUUUUUUKE°!
May 21 22:19:00 <merlin1991> I ended up in sales on the technical support hotline
May 21 22:19:00 <GeneralAntilles> DocScrutinizer, tell them to get off your lawn.
May 21 22:19:16 <merlin1991> DocScrutinizer: nuked them quickly after that, now with hetzner :)
May 21 22:19:26 <Estel_> merlin1991,  look at:http://sebsauvage.net/paste/?d5c04fea4770a39c#GwkIL3tpY+hOMrFyP8ouZBGYsUb4srSo2VePMNOsGs4=
May 21 22:19:28 <DocScrutinizer> hetzner better
May 21 22:19:28 <Estel_> please
May 21 22:19:36 <Estel_> it doesnt work
May 21 22:19:42 <Estel_> I'm also pastefail?
May 21 22:20:32 * merlin1991 fires up ssh again
May 21 22:20:40 <DocScrutinizer> Estel_: bracket arguments into '
May 21 22:21:07 <DocScrutinizer> err, not for args with $env
May 21 22:22:19 <DocScrutinizer> yayayay
May 21 22:22:42 <DocScrutinizer> t900:/usr/sbin# blkid  |  grep $swap0.*TYPE=\"swap\" | awk '{print substr($1, 1, length($1) -1)}'
May 21 22:22:44 <DocScrutinizer> /dev/mmcblk0p3
May 21 22:23:32 <DocScrutinizer> NB I haven't defined $swap0 before
May 21 22:25:14 <DocScrutinizer> t900:/usr/sbin# swap0=/dev/mmcblk1p3
May 21 22:25:15 <DocScrutinizer> t900:/usr/sbin# blkid  |  grep $swap0.*TYPE=\"swap\" | awk '{print substr($1, 1, length($1) -1)}'
May 21 22:25:17 <DocScrutinizer> t900:/usr/sbin#
May 21 22:26:00 <MrPingu> reminded me of closing an internet connection of my passed away aunt, you were always put in wait. Then I called the customer support : I want to buy "something" and directly I got a real man at the other end. 
May 21 22:26:19 <DocScrutinizer> Estel_: maybe your "script" (sorry for ") works perfectly, just the system doesn't do what you expect it to do
May 21 22:26:21 <merlin1991> why the fuck does the -n check succed even when the string is a zero string
May 21 22:27:18 <Estel_> oh god, I've stepped on nail left by my son on floor (part of toy with hammer)
May 21 22:27:25 <DocScrutinizer> jr@halley:~> help [
May 21 22:27:26 <Estel_> my roon is like bloodbath now
May 21 22:27:32 <Hurrian> ouch.
May 21 22:27:35 <DocScrutinizer>  This is a synonym for the "test" builtin, bu
May 21 22:27:37 <Estel_> s/roon/room/
May 21 22:27:38 <infobot> Estel_ meant: my room is like bloodbath now
May 21 22:28:06 <DocScrutinizer> PUNX RULE
May 21 22:29:07 <merlin1991> DocScrutinizer: any particular reason why a if [ -n $something ] would be true when $something = "" ?
May 21 22:29:13 <Estel_> DocScrutinizer,  this is not script, it's snippet from main script
May 21 22:29:21 <DocScrutinizer>       -n STRING
May 21 22:29:22 <DocScrutinizer>          STRING      True if string is not empty.
May 21 22:29:36 <Estel_> I pasted it just for fix purposes, whole tool is much more complicated
May 21 22:30:06 <merlin1991> DocScrutinizer: by "" I mean string but empty
May 21 22:30:10 <DocScrutinizer> yay, complicificatificated for the winning
May 21 22:30:24 <Estel_> merlin1991,  another problem is that after executing it I haven't good swap prompt that device is invalid
May 21 22:30:28 <DocScrutinizer> merlin1991: I'm just c&p quoting
May 21 22:30:30 <Estel_> but swap usage screen
May 21 22:30:31 <Estel_> ah
May 21 22:30:35 <Estel_> cause it's empty
May 21 22:30:36 <DocScrutinizer> consider non-printable
May 21 22:30:39 <Estel_> so it used empty
May 21 22:31:07 * merlin1991 rewrites awk part
May 21 22:31:09 <Estel_> so there is another problem too - I actually have /dev/mmcblk1p3
May 21 22:31:17 <Estel_> so why it was empty anyway?
May 21 22:31:24 <Estel_> blkid list it too
May 21 22:31:49 <merlin1991> Estel_: for me it works when I have the partition
May 21 22:32:00 <merlin1991> but it also does a false positive when I don'T have it
May 21 22:33:12 <DocScrutinizer> /sbin/swapon $(grep $swap0.*TYPE=\"swap\" | awk '{print substr($1, 1, length($1) -1)}') ) || swapon /dev/mmcblk0p7
May 21 22:33:42 <merlin1991> DocScrutinizer: consider http://sebsauvage.net/paste/?9083b5a04c9dcf29#jrRTQiNpgT/QzFdHeT1+tUreMuCD5nJgQLCEtMemrUk=
May 21 22:33:44 <DocScrutinizer> beat that
May 21 22:34:02 <merlin1991> for some reason I always end up in the "yes" tree
May 21 22:34:14 <merlin1991> arf
May 21 22:34:20 <merlin1991> stupid term
May 21 22:34:49 <merlin1991> http://sebsauvage.net/paste/?35f7fb495a6af862#ENYcrbkOaa9WTWHmS6dubJgokvOOHobmPKb1c8MVTkQ=
May 21 22:35:17 <DocScrutinizer> echo ">>$swapmicro<<" $(echo $swapmicro|od)
May 21 22:36:04 <merlin1991> 0000000 000012 0000001
May 21 22:36:07 <merlin1991> dafuq?
May 21 22:36:25 <DocScrutinizer> toldya ;-P
May 21 22:36:34 * merlin1991 wonders wft that is
May 21 22:36:42 <DocScrutinizer> utf8?
May 21 22:37:02 <DocScrutinizer> don't trust blkid
May 21 22:37:08 <merlin1991> nah it's awk
May 21 22:37:10 <merlin1991> fu awk
May 21 22:37:19 <DocScrutinizer> lol, not awk
May 21 22:37:23 <DocScrutinizer> MESSYBOX!!!
May 21 22:37:34 <merlin1991> hm not awk aswell
May 21 22:37:40 <merlin1991> but messybox somewhere
May 21 22:38:24 <DocScrutinizer> honestly, make better regex!
May 21 22:38:46 <merlin1991> heh why? :D
May 21 22:38:47 <Estel_> erm, I'm here but in listening mode
May 21 22:39:32 <DocScrutinizer> "\(\/dev\/mmcblk[0-9]p[0-9]\)"
May 21 22:39:50 <teotwaki_> mwhahahaha
May 21 22:39:52 <merlin1991> that makes no sense
May 21 22:39:57 <merlin1991> that would match anything again
May 21 22:40:10 <merlin1991> but I want to find the line that has preset /dev node and is swap
May 21 22:40:14 <merlin1991> otherwise discard all
May 21 22:40:15 <teotwaki_> on some inventor programme, some dude came round, showing off a prototype of a gaussian canon
May 21 22:40:33 <teotwaki_> and went "I can make a perpetual movement machine with this, free energy, woo"
May 21 22:40:56 <teotwaki_> everytime anyone asked a question, he went "Look at the shiny ball, it's attracted by magnets, look, it moves by itself"
May 21 22:41:01 <DocScrutinizer> what does (grep $swap0.*TYPE=\"swap\" | awk '{print substr($1, 1, length($1) -1)} ?
May 21 22:41:05 <teotwaki_> and they all went wild.
May 21 22:41:23 <teotwaki_> how fucking retarded do you have to be.
May 21 22:42:17 <DocScrutinizer> awk '{print substr($1, 1, length($1) -1)} gets you any shitee that blkid might spew out
May 21 22:42:25 <teotwaki_> people having a hard time with regexp? Need help?
May 21 22:42:36 <merlin1991> check blkid output against devnode in $swap0 + the type of swap and get rid of the : in the  blkid output since I was lazy to write a better echo $swap0 if found solution :D
May 21 22:42:56 <DocScrutinizer> teotwaki_: ME? nope!
May 21 22:43:07 <teotwaki_> not you, I know :)
May 21 22:44:09 <merlin1991> hm I could use the return code of grep instead of that silly solution
May 21 22:45:08 <teotwaki_> merlin1991: what are you trying to achieve?
May 21 22:45:15 <merlin1991> madness
May 21 22:45:15 <merlin1991> ;)
May 21 22:45:18 <teotwaki_> sparta?
May 21 22:45:19 <merlin1991> nah I'm htere I think
May 21 22:46:07 * DocScrutinizer throws some Haloperidol pills into channel - madness for everybody
May 21 22:46:20 <merlin1991> this is getting silly
May 21 22:46:27 <DocScrutinizer> indeed
May 21 22:46:34 <teotwaki_> merlin1991: just explain your problem.
May 21 22:46:37 <DocScrutinizer> I'm *extremely* silly today
May 21 22:46:38 <merlin1991> http://sebsauvage.net/paste/?a5530899c8625f49#pZeeI7ZulBKoRo8yatvq/S95sthp1ma7GcdmtjHOM3M=
May 21 22:46:48 <merlin1991> and still the bloody utf header in $swapmicro :D
May 21 22:46:58 <DocScrutinizer> LOL
May 21 22:47:11 <teotwaki_> I don't have a device at the ready
May 21 22:47:16 <teotwaki_> so don't know what that does -- walk me through it
May 21 22:47:29 <merlin1991> teotwaki_: basically  Idefine a devnode
May 21 22:47:38 <teotwaki_> aye
May 21 22:47:41 <merlin1991> then use the blkid to get all valid partitions with types and whatnot
May 21 22:47:49 <teotwaki_> ok
May 21 22:47:59 <merlin1991> grep for the devnode I want, under the condiditon that it is swap
May 21 22:48:11 <merlin1991> and if yes it should swap otherwise use the fallback
May 21 22:48:12 <DocScrutinizer> blkid|od -xc
May 21 22:49:15 <merlin1991> DocScrutinizer: echo $whyistihishappening | od
May 21 22:49:22 <merlin1991> -> 0000000  005077
May 21 22:49:23 <merlin1991> 0000002
May 21 22:49:30 <teotwaki_> blkid | grep "$swap0.*TYPE=.swap." && swapmicro=$swap0
May 21 22:49:32 <DocScrutinizer> 0780922907318931709^4275^ß839
May 21 22:50:02 * DocScrutinizer kicks himself
May 21 22:50:20 <teotwaki_> if [ -n "${swapmicro}" ]
May 21 22:50:53 <DocScrutinizer> echo -n "${swapmicro}"|od -xc
May 21 22:51:27 <teotwaki_> or
May 21 22:51:37 <teotwaki_> blkid | grep "$swap0.*TYPE=.swap."
May 21 22:51:43 * DocScrutinizer blames messybox, as it's always OK to do that
May 21 22:52:00 <teotwaki_> if [ $? -eq 0 ] ; then swapmicro="$swap0"
May 21 22:52:05 <teotwaki_> ; fi
May 21 22:52:10 <DocScrutinizer> indeed
May 21 22:52:33 <merlin1991> teotwaki_: thanks for the -n "$something"
May 21 22:52:35 <merlin1991> :)
May 21 22:52:37 <DocScrutinizer> if terse form doesn't work, get verbose!
May 21 22:52:42 <merlin1991> avoids messybox, yay
May 21 22:53:20 * DocScrutinizer hates terse code
May 21 22:54:20 <DocScrutinizer> especially those c wankers who think it's cute to pack 7 'normal' lines of code into one statement
May 21 22:54:33 <merlin1991> Estel_:  http://sebsauvage.net/paste/?c77ddc4d9e0ae234#WdfCcHYUz3Cs9E2npsUTWIhuBLiMrfz1CSpfDmtXOgs=
May 21 22:55:23 <merlin1991> DocScrutinizer: ash evalueates left to right?
May 21 22:55:25 <teotwaki_> or, if you want to get funky: if [ `blkid | grep -c "$swap0.*TYPE=.swap."` -gt 0 ]; then echo $swap0 | od; fi
May 21 22:55:37 <DocScrutinizer> merlin1991: usually r-t-l
May 21 22:55:38 <Estel_> merlin1991,  thanks
May 21 22:55:54 <DocScrutinizer> at least for pipes etc
May 21 22:56:02 <merlin1991> DocScrutinizer: that answer is not verbose enough for me :D
May 21 22:56:25 <DocScrutinizer> a|b|c; c will get invoked first
May 21 22:56:29 <teotwaki_> wut?
May 21 22:56:40 <teotwaki_> how does that work, considering c gets fed the input of b?
May 21 22:56:54 <DocScrutinizer> it's a fifo, dear teotwaki_ 
May 21 22:57:03 <teotwaki_> ?
May 21 22:57:11 <merlin1991> well what I want to ask if the last line in my latest paste only runs the swapon $swap2 if the first bunch failed
May 21 22:57:32 <merlin1991> or actually does other nasty things I didn't even think of :D
May 21 22:57:59 <DocScrutinizer> t900:/usr/sbin# set -vx
May 21 22:58:00 <DocScrutinizer> t900:/usr/sbin# date|cat|od
May 21 22:58:02 <DocScrutinizer> date|cat|od
May 21 22:58:03 <DocScrutinizer> + cat
May 21 22:58:05 <DocScrutinizer> + od
May 21 22:58:06 <DocScrutinizer> + date
May 21 22:58:08 <DocScrutinizer> 0000000  072524 020145 060515 020171 031062 030040 035061 032060
May 21 22:58:31 <DocScrutinizer> so not rtl either
May 21 22:58:39 <DocScrutinizer> honestly nfc
May 21 22:58:54 <DocScrutinizer> mid to outside?
May 21 22:59:52 <merlin1991> Estel_: my last script seems to work 
May 21 23:00:02 * merlin1991 hates messybox scripting
May 21 23:00:19 * DocScrutinizer hates messybox
May 21 23:00:22 <DocScrutinizer> period
May 21 23:00:50 <Estel_> I'm investigating it, don't think I'm such bastard to say thanks and go away after You, DocScrutinizer and teotwaki did such work :D
May 21 23:01:02 <Estel_> In fact, I'm going to bless You ;)
May 21 23:01:18 <Estel_> just testing it now in many situations
May 21 23:02:58 * DocScrutinizer headdesks a bit more - listens to remnants of what been a brain making funny noises
May 21 23:05:36 <Estel_> it works perfectly, thanks a lot :D
May 21 23:05:43 <Estel_> reallyp it saved my day :P
May 21 23:06:03 <Estel_> + I got moral satisfaction ;) that my problem wasn't so trivial to solve
May 21 23:06:08 <Estel_> (joke)
May 21 23:06:48 <Estel_> I'm going to integrate it now with rest... Only few long but working things with sed, and I'll have this bastard ready for -devel
May 21 23:07:24 <DocScrutinizer> honestly, for 'complicated multiline edits' use patch
May 21 23:07:44 <DocScrutinizer> nm, patch: not found
May 21 23:07:55 <teotwaki_> DocScrutinizer: interesting...
May 21 23:07:59 <DocScrutinizer> F U messybox!
May 21 23:08:11 <DocScrutinizer> teotwaki_: ????
May 21 23:08:16 <merlin1991> hehe good old patch missing
May 21 23:08:20 <teotwaki_> DocScrutinizer: cat, od, date
May 21 23:08:28 <DocScrutinizer> interesting to watch me go insane, or what?
May 21 23:08:47 <DocScrutinizer> aah
May 21 23:08:52 <DocScrutinizer> indeed
May 21 23:09:04 <DocScrutinizer> weird stuff
May 21 23:09:07 <teotwaki_> $ echo "foo" | grep foo
May 21 23:09:07 <teotwaki_> + echo foo
May 21 23:09:07 <teotwaki_> + grep --colour=auto foo
May 21 23:09:23 <merlin1991> you probably remember our fun times in #maemo-ssu thining about how to patch conversations css :D
May 21 23:09:40 <DocScrutinizer> maybe I do
May 21 23:10:20 <DocScrutinizer> anyway logicals like &&, || are stictly ltr
May 21 23:10:37 <Estel_> exactly
May 21 23:10:38 <Estel_> so
May 21 23:10:43 <Estel_> as path is missing
May 21 23:10:46 <DocScrutinizer> for pipes though things change sometimes
May 21 23:11:02 <Estel_> can I use sed inside script with argument containing variables?
May 21 23:11:09 <DocScrutinizer> sure
May 21 23:11:11 <Estel_> I mean:
May 21 23:11:19 <DocScrutinizer> it's shell expanding $env
May 21 23:11:24 <DocScrutinizer> not sed
May 21 23:11:51 <Estel_> variables inside sed argument will be properly filled, instead of thing replacing my line with $swap0
May 21 23:11:53 <Estel_> :P
May 21 23:11:55 <Estel_> I see
May 21 23:12:05 <Estel_> a relief, I think
May 21 23:12:09 <DocScrutinizer> shell expands everything except things 'enclosed in ticks'
May 21 23:12:18 <teotwaki_> single quotes
May 21 23:13:28 <DocScrutinizer> Estel_: set -vx tells you the real invocation with parameters
May 21 23:13:43 <DocScrutinizer> [2012-05-22 01:09:07] <teotwaki_> $ echo "foo" | grep foo
May 21 23:13:45 <DocScrutinizer> [2012-05-22 01:09:07] <teotwaki_> + echo foo
May 21 23:13:46 <DocScrutinizer> [2012-05-22 01:09:07] <teotwaki_> + grep --colour=auto foo
May 21 23:13:55 <Estel_> yea, of course
May 21 23:13:58 <DocScrutinizer> obviously grep is an alias here
May 21 23:14:09 <Estel_> sed -i "1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@" /etc/event.d/rcS-late
May 21 23:14:21 <Estel_> do You like this short command? :P
May 21 23:14:28 <DocScrutinizer> no
May 21 23:15:06 <Estel_> errors inside or just gustas et non discutante?
May 21 23:15:23 <DocScrutinizer> too drunk to focus on that mess ;-P
May 21 23:15:30 <Estel_> :PPPPP
May 21 23:15:51 <teotwaki_> anyway, I'm out
May 21 23:15:53 <teotwaki_> 'night people
May 21 23:16:00 <DocScrutinizer> o/
May 21 23:16:05 <Psi> does anyone know what impedance the n900 headphone out is designed to drive?
May 21 23:16:32 <DocScrutinizer> Estel_: it has bugs as well
May 21 23:16:40 <DocScrutinizer> unpaired "
May 21 23:16:56 <DocScrutinizer> "Failed to enable paging partition."
May 21 23:17:03 <DocScrutinizer> or 
May 21 23:17:11 <DocScrutinizer> "1!N; s@      /sbin/swapon -a || echo "
May 21 23:17:32 <DocScrutinizer> or
May 21 23:17:45 <DocScrutinizer> "@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@"
May 21 23:17:54 <Estel_> I see, but wtf it doesngt work
May 21 23:18:01 <Estel_> yea
May 21 23:18:09 <Estel_> how to overcome it?
May 21 23:18:18 <DocScrutinizer> escape the " inside
May 21 23:18:27 <Estel_> by?
May 21 23:18:30 <DocScrutinizer> \
May 21 23:19:23 <Estel_> but but
May 21 23:19:30 <Estel_> inside contain an escaped " already
May 21 23:19:32 <Estel_> I mean
May 21 23:19:48 <Estel_> TYPE=\"swap\"
May 21 23:19:55 <DocScrutinizer> so what?
May 21 23:19:56 <Estel_> it must get inside like that
May 21 23:20:06 <Estel_> so how to distinguish it from my escaped "
May 21 23:20:29 <DocScrutinizer> then use ' instead of " for the sed argument
May 21 23:20:44 <DocScrutinizer> or even better, use $env
May 21 23:21:10 <Estel_> sed -i '1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@' /etc/event.d/rcS-late
May 21 23:21:11 <Estel_> like that
May 21 23:21:17 <Estel_> ah, define argument as env
May 21 23:21:27 <DocScrutinizer> yup
May 21 23:21:33 <Estel_> like forsed=sed -i "1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@" /etc/event.d/rcS-late
May 21 23:21:40 <Estel_> wrong
May 21 23:22:19 <Estel_> forsed=1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@
May 21 23:22:23 <Estel_> correct?
May 21 23:22:36 <Estel_> then sed -i $forsed
May 21 23:22:50 <Estel_> then sed -i $forsed /etc/event.d/rcS-late
May 21 23:22:52 <Estel_> actually
May 21 23:22:55 <DocScrutinizer> forsed='1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@'
May 21 23:23:07 <Estel_> ah, shouldn't drop the '
May 21 23:23:11 <Estel_> I understand
May 21 23:23:13 <DocScrutinizer> or sth like that
May 21 23:23:39 <DocScrutinizer> escaping is rather tricky on shell
May 21 23:23:40 <Estel_> it its $env are the ' at beginning and end needed still, got ya
May 21 23:23:54 <Estel_> s/got ya/got that/
May 21 23:23:54 <infobot> Estel_ meant: it its $env are the ' at beginning and end needed still, got that
May 21 23:24:48 <DocScrutinizer> discard 1!N;
May 21 23:25:00 <DocScrutinizer> it confuses hell outa me
May 21 23:25:43 <cehteh> mhm
May 21 23:25:46 <DocScrutinizer> iscard leading whitespace
May 21 23:26:23 <cehteh> busybox or bash?
May 21 23:26:43 <DocScrutinizer> discard overly comprehensive matchstring, replace large parts by .*
May 21 23:28:21 <DocScrutinizer> s@swapon -a.*@...
May 21 23:28:35 <DocScrutinizer> err
May 21 23:28:54 <Estel_> doesn't work
May 21 23:28:55 <DocScrutinizer> s@/sbin/swapon -a.*@...
May 21 23:29:13 <DocScrutinizer> even
May 21 23:29:19 <DocScrutinizer> s@ */sbin/swapon -a.*@...
May 21 23:29:32 <Estel_> why should I do that
May 21 23:29:52 <Estel_> when I want to swapon only pre-defined backup swap
May 21 23:29:53 <DocScrutinizer> readability
May 21 23:30:35 <DocScrutinizer>  /sbin/swapon -a || echo .Failed to enable paging partition\..
May 21 23:30:49 <DocScrutinizer> dot is a special symbol
May 21 23:31:26 <DocScrutinizer> . matches one arbitrary char
May 21 23:31:45 <DocScrutinizer> X* matches X and XXXXXXXXXXXXXXXXXX
May 21 23:31:56 <DocScrutinizer> .* matches everything
May 21 23:32:06 <DocScrutinizer> \. matches a dot
May 21 23:33:11 <DocScrutinizer> u? matches zero or one 'u'
May 21 23:33:14 <DocScrutinizer> iirc
May 21 23:33:18 <DocScrutinizer> I'm tired
May 21 23:33:20 <DocScrutinizer> n8
May 21 23:35:40 <DocScrutinizer> Estel_: use a textfile and grep to test your regex
May 21 23:36:14 <DocScrutinizer> then copy that regex to first half od sed s§<matchstring>§
May 21 23:36:28 <DocScrutinizer> s/od/of/
May 21 23:36:28 <infobot> DocScrutinizer meant: then copy that regex to first half of sed s§<matchstring>§
May 21 23:57:22 <Estel_> DocScrutinizer,  thanks, although I don' understand 3/4 of what You're telling me
May 21 23:57:35 <DocScrutinizer> ask
May 21 23:57:44 <Estel_> sorry for dissaperance, got notification about new mail from modest and N900 just got into frozen state
May 21 23:57:54 <DocScrutinizer> lol
May 21 23:58:00 <Estel_> due to swap fragmentation, many shit opened, + fm radio
May 21 23:58:09 <Estel_> interesingly enough, I was able to ssh
May 21 23:58:10 <DocScrutinizer> ~hail modest
May 21 23:58:11 * infobot bows down to modest and chants, "I'M NOT WORTHY!!"
May 21 23:58:33 <Estel_> but N900 was only repeatedly turning keyboard lights on and off
May 21 23:58:43 <Estel_> I killed radii, modest etc via ssh
May 21 23:58:52 <Estel_> screen changed accordingly to taskbar, but no change
May 21 23:59:05 <Estel_> enraged, I killed browserd with no effect
May 21 23:59:20 <Estel_> waited a while, and due no change, killed - still via ssh - xorg
May 21 23:59:26 <Estel_> to force proper shutdown :P
May 21 23:59:34 <DocScrutinizer> hihi
May 21 23:59:45 <Estel_> anyway
May 21 23:59:59 <Estel_> erm, about ask, I'll start with semi-simple question:
May 22 00:00:59 <DocScrutinizer> I'll start with semi-asshat answer: man 7 regex :-P
May 22 00:01:24 <Estel_> :P
May 22 00:01:28 <Estel_> will read it
May 22 00:01:30 <Estel_> through
May 22 00:01:38 <Estel_> http://sebsauvage.net/paste/?ead3931098e1bfa9#GZHCNnklBf2N8Fm2v0AeV5+st28zLC6jxoVcSBSA/Bc=
May 22 00:01:44 <Estel_> why this shit doesn't work?
May 22 00:01:55 <Estel_> command itself is ok, I mean sed
May 22 00:02:23 <Estel_> I want sed to insert my (merlin's ;) ) beautiful complicated commands into rcS-late
May 22 00:02:38 <Estel_> orally or anally, no mater. but sed is driving me mad
May 22 00:02:46 <DocScrutinizer> try sed -i "$forsed"
May 22 00:02:53 <Estel_> as You can see by bullshit 'm writing
May 22 00:02:55 <Estel_> dafuq?
May 22 00:03:12 <Estel_> ah
May 22 00:03:15 <Estel_> will try
May 22 00:03:23 <Estel_> but it will make it sit inside "''"
May 22 00:03:33 <Estel_> as forsed got ' at end and beginning
May 22 00:03:47 <DocScrutinizer> no, it doesn't
May 22 00:03:55 <DocScrutinizer> I think
May 22 00:04:59 <DocScrutinizer> jr@halley:~> forsed='1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@'
May 22 00:05:00 <DocScrutinizer> jr@halley:~> echo ">>$forsed<<"
May 22 00:05:02 <DocScrutinizer> >>1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@<<
May 22 00:05:55 <DocScrutinizer> sorry, gotta sleep
May 22 00:06:12 <Estel_> no problem, good night
May 22 00:06:20 <Estel_> now it works but give fckd results
May 22 00:06:40 <Estel_> inserts $swap0 instead of it's content
May 22 00:06:47 <Estel_> + do other weird shit
May 22 00:07:47 <Estel_> instead of replacing line, it makes it:
May 22 00:08:01 <Estel_> blkid | grep $swap0.*TYPE="swap"       /sbin/swapon -a || echo "Failed to enable paging partition."      /sbin/swapon -a || echo "Failed to enable paging partition." /sbin/swapon $swap0 || /sbin/swapon $swap2
May 22 00:08:08 <Estel_> while effect should be:
May 22 00:08:51 <Estel_> blkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon /dev/mmcblk1p2 || /sbin/swapon /dev/mmcblk0p7
May 22 00:09:12 <Estel_> blkid | grep $/dev/mmcblk1p2*TYPE=\"swap\" && /sbin/swapon /dev/mmcblk1p2 || /sbin/swapon /dev/mmcblk0p7
May 22 00:09:14 <Estel_> in fact
May 22 00:09:36 <Estel_> ~kill sed
May 22 00:09:36 * infobot shoots a charged pseudograviton gun at sed
May 22 00:10:07 <Estel_> merlin1991,  You're probably asleep already?
May 22 00:10:10 <DocScrutinizer> http://tldp.org/LDP/abs/abs-guide.pdf
May 22 00:11:11 <DocScrutinizer> Estel_: shell expands, unless enclosed in ' '
May 22 00:11:14 <merlin1991> Estel_: nah still here
May 22 00:11:37 <Estel_> thanks godness. Oh, and DocScrutinizer is lunatic
May 22 00:11:38 <DocScrutinizer> '$foo' -> $foo
May 22 00:11:50 <DocScrutinizer> "$foo" -> bar
May 22 00:12:08 <Estel_> so how to format this fckd sed command
May 22 00:12:22 <Estel_> merlin1991,  it gives me and DocScrutinizer headache
May 22 00:12:36 <DocScrutinizer> foo="this ->\" is an doubletick"
May 22 00:12:40 <Estel_> or just me, who knows
May 22 00:12:49 <merlin1991> dunno I never did sed stuff so far :D
May 22 00:12:54 <Estel_> :P
May 22 00:13:19 <DocScrutinizer> echo $foo -> this " is an doubletick
May 22 00:13:26 <Estel_> so how would You make script that find line in rcS-late and replace it with our beautiful work
May 22 00:13:31 <Estel_> (i.e. script cooked earlier)
May 22 00:14:17 <Estel_> DocScrutinizer,  but inside script there is also escaped part required
May 22 00:14:31 <Estel_> I'm just lost how to make sed interpret it properly
May 22 00:14:48 <DocScrutinizer51> inside escapes escapes neewd to be escaped
May 22 00:14:50 <DocScrutinizer51> lol
May 22 00:14:56 <Estel_> 0_o
May 22 00:14:59 <Estel_> *TYPE=\"swap\"
May 22 00:15:06 <Estel_> this part need to be passed to edited file
May 22 00:15:09 <Estel_> not escaped
May 22 00:15:12 <merlin1991> reminds me of how to regex match a \ in java
May 22 00:15:28 <DocScrutinizer51> echo "\"\\\\""
May 22 00:15:29 <merlin1991> Patter.Compile("\\\\"); <-- awesome stuff right there
May 22 00:16:03 <Estel_> DocScrutinizer,  wtf?
May 22 00:16:09 <Estel_> forsed='1!N; s@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@'
May 22 00:16:18 <Estel_> Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@'
May 22 00:16:25 <Estel_> wtf
May 22 00:16:29 <Estel_> sorry, wrong quote
May 22 00:17:52 <Estel_> http://sebsauvage.net/paste/?c4e3aa202e012ba1#gh42uDO/Tt63XcVfUsHKexKkJvQMvWJ2cVcQITvKHR4=
May 22 00:17:56 <Estel_> whats wrong with this
May 22 00:18:04 <Estel_> i.e. should this shit should look
May 22 00:18:18 * Estel_ is head desking
May 22 00:18:35 <Estel_> I just *must* finish this shit today and upload *madness*
May 22 00:18:42 <Estel_> i mean tonight :P
May 22 00:18:57 <Estel_> then I can ~kill sed
May 22 00:19:06 <DocScrutinizer51> WRONG approach
May 22 00:19:17 <Estel_> probably, but no patch in busybox
May 22 00:19:28 <DocScrutinizer51> if it gets tough today, ship next week
May 22 00:19:41 <Estel_> ah, it's what You mean. Yea, I know
May 22 00:19:57 <Estel_> but seriously, it's damn last thing before packaging :P
May 22 00:20:05 <Estel_> literally last
May 22 00:20:18 <DocScrutinizer51> nope
May 22 00:20:44 <DocScrutinizer51> last thing is let it ripe 3 days, and some 2 friends test it
May 22 00:20:57 <Estel_> need to package it before that
May 22 00:21:02 <Estel_> packaging != repos
May 22 00:21:24 <Estel_> would be silly to let it test 3 days then make bug while packaging
May 22 00:21:41 <DocScrutinizer51> aure
May 22 00:21:49 <Estel_> btw, I'm using main part for a year or so. it's just that fuckin one-time-run script that needs finishing
May 22 00:22:22 <Estel_> it's all because I decided to make really easy to use package, without instruction to manually edit rcS-Late before use
May 22 00:22:52 <Estel_> user just provide locations in config , and run one timer. It handles updating rcS-late
May 22 00:23:16 <Estel_> then, You can eternally use what is main part of this package - well-tested rock solid scripts to refresh swap
May 22 00:23:40 <Estel_> and of course this "small addition" to make rcS-late edit is driving me mad
May 22 00:23:50 <Estel_> but, seriously, do You have idea how to fix that?
May 22 00:23:58 <Estel_> (see last zerobin)
May 22 00:27:54 * Estel_ trying different approaches to fight with sed
May 22 00:32:20 * DocScrutinizer51 jumps out of bed again
May 22 00:32:56 <DocScrutinizer51> Estel_: it's definitely are questionable idea to edit rcS-*
May 22 00:33:23 <Estel_> no problem, this package is also for that
May 22 00:33:33 <Estel_> and doesn't do it automaticaly
May 22 00:33:47 <Estel_> user can do it by hand if willing to or not do at all
May 22 00:33:57 <Estel_> (yet, many things wont work if not done)
May 22 00:34:26 <Estel_> it just contain instruction to edit config file and run adjust-rcS-late.sh script only once
May 22 00:35:08 <Estel_> I'll also add line writing counter from iostat to have info about data written to swap
May 22 00:35:16 <Estel_> anyway, it's onlhy one sane approach to it
May 22 00:35:26 <Estel_> other would be to use /event.d/ script
May 22 00:35:32 <Estel_> but it would slow down boot
May 22 00:35:40 <Estel_> as swap would be enabled by rcS-late
May 22 00:35:52 <Estel_> then disabled and enabled (other) via scfript in event.d
May 22 00:35:59 <Estel_> idiotic + time consuming
May 22 00:36:46 <Estel_> btw I use rcS-late editing approach for ages it doesn't bring any problems
May 22 00:37:00 <Estel_> if done correctly
May 22 00:37:14 <Estel_> and I swant to spare people bootlops due to typos while editing manualy
May 22 00:38:26 <DocScrutinizer51> learn about upstart and sysV init. then better find a way that doesn't involve editing core initscripts
May 22 00:39:30 <DocScrutinizer51> insert a own inotscript that runs before rcS-late and edits fstab accordingly
May 22 00:40:08 <DocScrutinizer51> add a script to undo swap mounting done by rc+-late
May 22 00:40:37 <DocScrutinizer51> editing core stuff is EVIL
May 22 00:53:48 <Estel_> honestly, something that mess with fstab just to cheat rcS-late is more EVIL
May 22 00:54:32 <Estel_> if I have to choose between holy way editing script that got executed once on boot and then is just a file sitting on disk...
May 22 00:54:51 <Estel_> ...or fstab, on which many things could rely...
May 22 00:55:00 <Estel_> I preffer option 1
May 22 00:55:17 <Estel_> anyway, fighting with sed goes better and better, I think I'm close to finding it
May 22 01:00:38 <freemangordon> damn, there was an earthquake here :(
May 22 01:21:19 <Estel_> freemangordon,  not fatal one I hope?
May 22 01:21:39 <Estel_> sed -i \'$forsed\' /home/user/original-rcS-late
May 22 01:21:50 <Estel_> sed: unsupported command '
May 22 01:21:56 <freemangordon> no, nothing serious, just a little frightening
May 22 01:21:56 <Estel_> what the fuckin fuck?
May 22 01:21:59 <Estel_> I see
May 22 01:22:08 <Estel_> here= where?
May 22 01:22:15 <freemangordon> 5.6-6 or something
May 22 01:22:23 <Estel_> not very light either
May 22 01:22:28 <freemangordon> Sofia
May 22 01:23:07 <Estel_> it apparently talks about \', because if i change it to \", it complains about unsupported command "
May 22 01:23:26 <Estel_> hm, should't Sofia be out of earthquakes zone?
May 22 01:24:05 <Estel_> still, I've no idea how the fuck it interprets this ' as command, as without it, it complains about unmatched @ (content of $forsed)
May 22 01:24:11 <Estel_> ~kill sed
May 22 01:24:12 * infobot shoots a hyper-charged anti-photon gun at sed
May 22 01:24:15 <Estel_> ~kill sed
May 22 01:24:21 <Estel_> ~kill sed
May 22 01:24:21 * infobot shoots a hyper-charged anti-quark gun at sed
May 22 01:24:27 <Estel_> ~kill sed
May 22 01:24:31 <Estel_> ~kill sed
May 22 01:24:31 * infobot shoots a ionized proton gun at sed
May 22 01:24:52 <Estel_> why, ah why busybox doesn't contain "patch"
May 22 01:25:05 <Estel_> ~suicude sed
May 22 01:25:15 <Estel_> ~die sed
May 22 01:25:15 * infobot murders sed, takes two shots to the head then crumples to the ground, lifeless...
May 22 01:25:49 <Estel_> ~annihilate sed
May 22 01:25:49 * infobot bakes a cake and gives it to sed
May 22 01:25:53 <Estel_> LOL
May 22 01:26:05 <Estel_> I'm fighting with this shit for few hours already
May 22 01:26:09 <Estel_> 3 or 4
May 22 01:26:19 * Estel_ headdesk
May 22 01:26:35 <Estel_> s/headdesk/headdesks
May 22 01:26:38 <Estel_> s/headdesk/headdesks/
May 22 01:26:39 <infobot> Estel_ meant: s/headdesks/headdesks
May 22 01:26:44 <Estel_> fu
May 22 01:26:44 <beford> hah
May 22 01:27:21 <Estel_> no, seriously, I'm out of ideas
May 22 01:27:29 <Estel_> how to make sed properly interpret:
May 22 01:28:53 <Estel_> sed -i 's@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@' /etc/event.d/rcS-late
May 22 01:29:04 <Estel_> I've properly escaped scapes inside escapes
May 22 01:29:07 <Estel_> which gave me
May 22 01:29:16 <Estel_> with usage of $env
May 22 01:29:28 <Estel_> forsed="s@      /sbin/swapon \-a || echo \"Failed to enable paging partition.\"@blkid \-g\nblkid | grep $swap0.*TYPE=\\\"swap\\\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@"
May 22 01:29:41 <Estel_> sed -i \'$forsed\' /home/user/original-rcS-late
May 22 01:29:51 <Estel_> ...yet this unholy shit still doesn't work
May 22 01:30:03 <Estel_> any idea, anyone? pleeease?
May 22 01:30:12 <Estel_> ~murder sed
May 22 01:30:13 * infobot shoots sed in his sleep
May 22 01:31:08 <Estel_> ~assasinate sed
May 22 01:31:16 <Estel_> ~rape sed
May 22 01:31:16 * infobot takes sed behind the WallMart and makes a few grunts and screams
May 22 01:31:36 <Estel_> ~sodomize sed
May 22 01:31:42 <Estel_> ~seed sed
May 22 01:32:12 <Estel_> no really. I'm out of today, this shit just ain't gonna work for now :(
May 22 01:32:22 <Estel_> s/of/for/
May 22 01:32:23 <infobot> Estel_ meant: no really. I'm out for today, this shit just ain't gonna work for now :(
May 22 01:32:40 <beford> hold on
May 22 01:32:44 <Estel_> ?
May 22 01:33:02 <beford> do you have the original file?
May 22 01:33:11 <beford> i mean, pastebin it so I can see and give it a try
May 22 01:33:26 <Estel_> I'm little angry, as it's last part of work needed to upload bigger package
May 22 01:33:28 <Estel_> ok, but
May 22 01:33:34 <Estel_> do You mean script
May 22 01:33:39 <Estel_> or file to be edited
May 22 01:33:43 <Estel_> by sed
May 22 01:33:47 <beford> the file to be edited, and the line you wnat to change
May 22 01:33:49 <beford> want*
May 22 01:33:57 <Estel_> ok, pastebinning
May 22 01:35:17 <Estel_> file to be edited
May 22 01:35:20 <Estel_> http://sebsauvage.net/paste/?2c697dcc486362c1#G25BFmY3qhTXSyDCr1vbnYnf+3E70eumrD0bDnzKEXo=
May 22 01:37:06 <Estel_> editing script (well, make shit version, as it's part of complex script set, normally)
May 22 01:37:38 <Estel_> http://sebsauvage.net/paste/?24a1e04d862b42df#hRgaptuujC3WAviw/DIjNEuftrPmbATtGyPPD51ZNZ4=
May 22 01:38:03 <Estel_> ..and, finally, line to be editd and how it should look after edit:
May 22 01:38:45 <beford> line 52 
May 22 01:40:07 <Estel_> http://sebsauvage.net/paste/?b1bf09d1dfcf58f9#R233nu+06ujjLg8dRINFkmwFvzL5oVSc2a8kH6PqOKg=
May 22 01:40:32 <Estel_> yea, 52 to be edited
May 22 01:40:38 <Estel_> and replaced by 2 lines
May 22 01:40:57 <Estel_> already managed to get info about working multi-lining in sed etc...
May 22 01:41:20 <beford> isn't it unsafe to edit system scripts though?
May 22 01:41:45 <Estel_> last things is script that will read config file (i replacedd this by declaring values, in makeshift script) and modify rc$-late accordingly
May 22 01:41:52 <Estel_> it's unsafe to make typos during edit
May 22 01:42:16 <Estel_> although, I've almost finished program (set of scripts) mean't to easy refreshing of swap
May 22 01:42:30 <Estel_> + it allows to set arbitrary swaps on boot
May 22 01:42:52 <Estel_> for both tasks editing rcS-late is necessary and wanted
May 22 01:43:14 <Estel_> I want to save editing rcS-lated manually for end users
May 22 01:43:29 <Estel_> and prepare script that, once config if filled by user, will do it for them
May 22 01:43:39 <beford> automagically yea
May 22 01:43:45 <Estel_> would be trivial with patch, but patch isn't part of busybox
May 22 01:44:14 <Estel_> not exactly automagically, user need explicitly invoke adjust-rcS-late.sh script ;)
May 22 01:44:40 <Estel_> program manual and description in package manager will state to edit config and run script once
May 22 01:44:57 <Estel_> problem - can't force sed to work properly
May 22 01:45:07 <Estel_> huge thanks if You manage to find it
May 22 01:45:30 <Estel_> btw,set -xv bit is to show output in console, so You see shat actually is passed to sedd by sh
May 22 01:45:52 <Estel_> many tricky things with " and \ that need to be escaped from shell or sed
May 22 01:46:07 <Estel_> (in order to get into file that's edited)
May 22 01:46:21 <Estel_> thats why my current workshop version looks like that:
May 22 01:46:47 <Estel_> ...what You found in adjust-rcS-late that I've zerobin'ned before
May 22 01:46:53 <Estel_> it *almost* work
May 22 01:47:19 <Estel_> arguments are escaped correctly etc... just freaking sed is shitting about unsupported argument '
May 22 01:47:52 <Estel_> if I replace \'$forsed\' with \"$forsed\"...
May 22 01:48:08 <Estel_> it complains about "unsupported option "
May 22 01:48:14 <Estel_> unsupported "
May 22 01:48:30 <Estel_> so it's clearly related to \'$forsed\'
May 22 02:01:48 <Estel_> beford,  hard times with sed, eh?
May 22 02:03:08 * beford tries hard
May 22 02:05:58 * Estel_ too
May 22 02:10:54 <Estel_> I don't get whysed cares for content between @ @ @
May 22 02:11:13 <Estel_> if it would think it's not proper syntax, it would complain for unmatched @
May 22 02:11:41 <Estel_> yet, it doesn't so he knows it's command to sed
May 22 02:11:49 <Estel_> so why he fuckin cares about inside?
May 22 02:12:01 <Estel_> I must escape -a inside swapon -a
May 22 02:12:12 <Estel_> or sed complains about unknown option a
May 22 02:12:48 <Estel_> but, if this shithead know that it's string for replace - and he know, as he don't complain about unmatched @ - why he keeps bitching about it?
May 22 02:13:38 <Estel_> after 5 hours of fighting with it I'm getting murderous instincts against sed and it's creators
May 22 02:20:06 <cehteh> heh
May 22 02:20:19 <cehteh> sorry i am too tired else i may help :P
May 22 02:21:33 <luke-jr> https://dl.dropbox.com/u/726501/recaptcha.png
May 22 02:23:20 <cehteh> i must be some machine .. i almost never manage to deciper recaptchas
May 22 02:42:09 <Estel_> beford,  how hacking this is going? failure here :(
May 22 02:42:57 <Sc0rpius> what are we hacking?
May 22 02:44:59 <Sc0rpius> did you fix your sed script?
May 22 02:46:02 <Estel_> Sc0rpius,  nope
May 22 02:46:03 <Estel_> sed is
May 22 02:46:10 <Estel_> ~murder sed
May 22 02:46:10 * infobot shoots sed in his sleep
May 22 02:46:19 <Sc0rpius> sed is great
May 22 02:46:23 <Estel_> yea?
May 22 02:46:26 <Sc0rpius> let's see what you're doing wrong
May 22 02:46:29 <Estel_> wanna change Your mind?
May 22 02:46:31 <Estel_> thanks
May 22 02:46:39 <Estel_> I'll send You zerobins
May 22 02:47:16 <Sc0rpius> sed -i 's@      /sbin/swapon -a || echo "Failed to enable paging partition."@blkid -g\nblkid | grep $swap0.*TYPE=\"swap\" && /sbin/swapon $swap0 || /sbin/swapon $swap2@'
May 22 02:47:21 <Sc0rpius> that line doesn't work as you expect, right?
May 22 02:47:23 <Estel_> file to be edited:
May 22 02:47:25 <Estel_> http://sebsauvage.net/paste/?2c697dcc486362c1#G25BFmY3qhTXSyDCr1vbnYnf+3E70eumrD0bDnzKEXo=
May 22 02:47:35 <Estel_> editing script (well, make shit version, as it's part of complex script set, normally):
May 22 02:47:46 <Estel_> http://sebsauvage.net/paste/?24a1e04d862b42df#hRgaptuujC3WAviw/DIjNEuftrPmbATtGyPPD51ZNZ4=
May 22 02:47:55 <Estel_> ..and, finally, line to be editd and how it should look after edit:
May 22 02:48:03 <Estel_> http://sebsauvage.net/paste/?b1bf09d1dfcf58f9#R233nu+06ujjLg8dRINFkmwFvzL5oVSc2a8kH6PqOKg=
May 22 02:48:04 <beford> i thinkg i did it
May 22 02:48:07 <beford> think
May 22 02:48:07 <beford> xD
May 22 02:48:09 <beford> not sure
May 22 02:48:22 <Estel_> If You manage to fix it I'll call You sedking
May 22 02:48:27 <Estel_> beford,  dafuq?
May 22 02:48:32 <Estel_> lemme check
May 22 02:48:37 <Estel_> paste it please :D
May 22 02:48:42 <Sc0rpius> ok let's see
May 22 02:48:52 <beford> http://paste2.org/p/2031236
May 22 02:49:52 <Estel_> beford,  ommiting -i is necessary?
May 22 02:50:32 <Sc0rpius> it should add two lines?
May 22 02:50:34 <beford> oh I ommited it to avoid having to restore the original file
May 22 02:50:40 <beford> it should work anyway
May 22 02:50:40 <Sc0rpius> replace swapon line for two lines?
May 22 02:50:42 * beford checks
May 22 02:51:43 <Sc0rpius> your script works, you just need to scape | and $ lemme show you:
May 22 02:52:07 <Estel_> Sc0rpius,  yea, two lines
May 22 02:52:20 <Estel_> sorry I'm checking... hm?
May 22 02:52:30 <beford> mine works with -s
May 22 02:52:32 <beford> I mean -i
May 22 02:52:33 <Estel_> scape | and $?
May 22 02:54:47 <Sc0rpius> sed "s@/sbin/swapon \-a || echo \"Failed to enable paging partition.\"@blkid \-g\n\tblkid \| grep $swap0.*TYPE=\\\"swap\\\" \&\& /sbin/swapon \$swap0 \|\| /sbin/swapon \$swap2@" file
May 22 02:54:52 <Sc0rpius> that one works for me (file is your file).
May 22 02:55:25 <Estel_> beford,  for some reason don't find necessary line, yet at least, doesn't throw errors
May 22 02:55:35 <Estel_> Sc0rpius,  lemme check...
May 22 02:55:40 <Sc0rpius> check, it'll work
May 22 02:56:00 <Estel_> I belive, I'm checking what was wrong, trying to learn
May 22 02:56:21 <Sc0rpius> $things in the second part of the sed are variables
May 22 02:56:27 <Sc0rpius> you didn't escape those
May 22 02:56:35 <Sc0rpius> I can do things like this:
May 22 02:56:39 <beford> I though he wanted to escape them?
May 22 02:56:43 <beford> thought*
May 22 02:56:45 <Estel_> right, see it
May 22 02:56:55 <Sc0rpius> sed s/I'll print just this:(.*)+/$0/
May 22 02:56:59 <beford> http://sebsauvage.net/paste/?b1bf09d1dfcf58f9#R233nu+06ujjLg8dRINFkmwFvzL5oVSc2a8kH6PqOKg=
**** ENDING LOGGING AT Tue May 22 02:59:58 2012