**** BEGIN LOGGING AT Sun Oct 06 02:59:58 2013
Oct 06 03:00:33 <psycho_oreos> lol, a measly 100b space saved.
Oct 06 03:01:43 <DocScrutinizer05> ii  tracker                                             0.6.95-35maemo1+0m5+0cssu0                 All-in-one indexer, search tool and metadata database.
Oct 06 03:01:46 <DocScrutinizer05> how humble
Oct 06 03:02:19 <DocScrutinizer05> and none of those friggin tracker-* tools seems to support -V
Oct 06 03:02:59 <psycho_oreos> o.O
Oct 06 03:03:25 <psycho_oreos> I thought there might have been a way somewhere with those individual tracker-* binaries.
Oct 06 03:13:30 <DocScrutinizer05> meh, they save a few 100 bytes in binaries of >50k by leaving out --help and --version support
Oct 06 03:13:44 <DocScrutinizer05>   /usr/lib/locale/locale-archive 29763056 bytes
Oct 06 03:14:08 <DocScrutinizer05> my first linux system: <20MB, *incl swap*
Oct 06 03:14:29 <DocScrutinizer05> looks sabe, doesn't it?
Oct 06 03:14:33 <DocScrutinizer05> sane even
Oct 06 03:16:31 <psycho_oreos> Whoa, less than 20MB, that is very small.
Oct 06 03:22:07 <DocScrutinizer05> it wasn't small at that time
Oct 06 03:23:38 <DocScrutinizer05> and it came with complete set of manpages
Oct 06 03:23:50 <psycho_oreos> Yeah, technology of course would change sporadically.
Oct 06 03:23:52 <DocScrutinizer05> for all that could have manpages at that time
Oct 06 03:24:08 <psycho_oreos> Must have also very limited set of userland tools I'd gather.
Oct 06 03:24:41 <DocScrutinizer05> well, it was several hundred cmdline tools anyway
Oct 06 03:24:58 <DocScrutinizer05> X11 wasn't too exciting though
Oct 06 03:26:04 <DocScrutinizer05> and a nightmare to configure, on that iirc cyrrus chipset
Oct 06 03:26:49 <psycho_oreos> Haha I remember having to manually write out X11 config files awhile back. That was fun, along with meddling with (what was then) proprietary kernel video driver.
Oct 06 03:27:03 <DocScrutinizer05> yes
Oct 06 03:27:35 <DocScrutinizer05> at that time you even had to binary-patch the vmlinuz to tell what drive to mount / from
Oct 06 03:27:42 <DocScrutinizer05> or sth like that
Oct 06 03:28:25 <DocScrutinizer05> of course you also could've built your own kernel, which took only one day
Oct 06 03:29:04 <DocScrutinizer05> once you managed to boot to linux at all
Oct 06 03:29:18 <psycho_oreos> I never had to binary-patch. Though running on vanilla kernels was heck a lot of fun. One bad mishap with trying to boot the kernel, you get kpanic. I've seen at numerous times VFS not SYNCED or something like that.
Oct 06 03:31:04 <DocScrutinizer05> there been a cmdline tool to patch the kernel, if only I could remeber how it been called. I think you needed it to make kernel boot from HDD instead of floppy
Oct 06 03:31:17 <psycho_oreos> I also had to deal with rpmdb whenever it decides to chuck a spaz (broken installs for instance, requiring rpmdb to be rebuilt). This was at the time when Debian folks had apt-get for instance whilst the Redhat folks had to make do with apt-rpm or experimental yum (not available when I started rh8 but later became available).
Oct 06 03:32:21 <psycho_oreos> I also recall at the time there was also ways to boot linux from DOS shell as well.
Oct 06 03:32:38 <DocScrutinizer05> yep, loadlin
Oct 06 03:32:48 <DocScrutinizer05> .exe
Oct 06 03:32:54 <psycho_oreos> Hahah yeah.
Oct 06 03:33:00 <DocScrutinizer05> those were the times
Oct 06 03:33:53 <jonwil> Gentoo makes things so easy, even a kernel upgrade is fairly simple these days
Oct 06 03:34:00 <psycho_oreos> Compiling 2.4.x kernel was loads more interesting, there was far more text to show compared to the much more cleaner (and potentially organised) 2.6.x kernel. 2.5.x kernel were for those wanting to be on bleeding edge at the time.
Oct 06 03:34:40 <DocScrutinizer05> meh, 1.x kernels FTW! ;-P
Oct 06 03:35:19 <psycho_oreos> My only gripe with gentoo is that if only they also allow binary packages, I'm not particularly fond of having to compile everything from scratch when I want the machine to be up and running. Though notably I've heard numerous times that gentoo maintainers take more care as opposed to arch maintainers.
Oct 06 03:35:51 <psycho_oreos> Heh 1.x kernels. I've never experienced that. I guess I was still in windows land at the time.
Oct 06 03:36:42 <DocScrutinizer05> must be like 20 years ago
Oct 06 03:37:36 <psycho_oreos> Even if it was 20 years ago, I remember DOS days things that somewhat haunt me was the command prompt back then.
Oct 06 03:37:53 <psycho_oreos> Good old 5 1/4 inch drives iinm.
Oct 06 03:38:22 <DocScrutinizer05> yeah
Oct 06 03:38:37 <DocScrutinizer05> but before that came minix (Tanenbaum)
Oct 06 03:38:55 <psycho_oreos> Loud and clunky {2,3}86 processor.
Oct 06 03:39:12 <DocScrutinizer05> and before or during that time I used another *great* unixoid system: amigaOS
Oct 06 03:39:49 <psycho_oreos> minix was probably what linux was initially based on.
Oct 06 03:40:02 <DocScrutinizer05> not rally afaik
Oct 06 03:40:08 <DocScrutinizer05> maybe inspired by
Oct 06 03:40:15 <DocScrutinizer05> really*
Oct 06 03:40:18 <psycho_oreos> I think amiga is still functioning lol, amiga64 or whatever it is now. They still charge people for the OS.
Oct 06 03:40:47 <psycho_oreos> Ahh I stand corrected. I thought linux was a clone of minix that was in a bid to mimick unix.
Oct 06 03:41:53 <DocScrutinizer05> http://de.wikipedia.org/wiki/Minix_(Betriebssystem)#Streit_um_Linux_und_Minix
Oct 06 03:43:32 <DocScrutinizer05> 1992 Tanenbaum claimed that linux would soon be obsolete and replaced by a better system like GNU Hurd - LOL
Oct 06 03:46:59 <DocScrutinizer05> I tried to port minix to Amiga1000 in 1988, but it wasn't really feasible, thanks to the missing MMU in MotorRoller CPU
Oct 06 03:48:43 <DocScrutinizer05> oh, Linus used Minix as a development environment to write linux
Oct 06 04:00:04 <psycho_oreos> I sort of got carried away in the wikipedia loop (yet again).
Oct 06 04:02:54 <psycho_oreos> Yeah though otherwise, minix is largely different from linux. hurd would probably be similar to minix's ideal goals in being non-monolithic kernel.
Oct 06 08:57:26 <Wizzup> hm... how programmable is the n900 usb mode?
Oct 06 08:59:22 * Wizzup is thinking of implementing some kind of yubikey
Oct 06 09:00:41 <psycho_oreos> I don't know about yubikey (though I have heard of it or at least something like it awhile ago). Though there was ps3 jailbreak program which was also ported to N900 iinm.
Oct 06 09:01:38 <kerio> Wizzup: very, i think
Oct 06 09:01:55 <kerio> just write a usb gadget module
Oct 06 09:01:56 <Wizzup> so, would it be possible to make the n900 act as a HID device?
Oct 06 09:02:13 <Wizzup> (that's a key question)
Oct 06 09:02:18 <Wizzup> kerio: I see
Oct 06 09:07:27 <Wizzup> http://talk.maemo.org/showthread.php?t=62071 this confirms it I think
Oct 06 09:14:39 * Wizzup will have fun with this
Oct 06 09:14:42 <Wizzup> kerio: thanks for the pointer
Oct 06 09:15:39 <kerio> i can give you another one! (char *)0xbabecafe
Oct 06 09:16:07 <Wizzup> I prefer a cafebabe
Oct 06 09:17:54 <Wizzup> so I guess I can make my laptop create a usb gadget as well? (to ease debugging...)
Oct 06 09:23:46 <kerio> mh, i don't know
Oct 06 09:23:53 <kerio> Pali: what's a good way to debug a usb gadget module?
Oct 06 09:50:38 <Pali> if your laptop usb chip support peripheral mode, you can
Oct 06 09:51:04 <Pali> but I did not see any pc motherboard with usb peripheral mode
Oct 06 09:51:18 <Pali> anything is host mode only
Oct 06 09:51:49 <Wizzup> ah, I see
Oct 06 09:51:54 <Wizzup> but n900 does :) 
Oct 06 09:52:07 <Pali> way to debug kernel module? start kernel in qemu (which can connect to gdb server) and debug it in gdb
Oct 06 09:52:28 <Wizzup> hm, no, I don't want to debug a kernel module
Oct 06 09:52:31 <Pali> debugging life kernel - I think only printk
Oct 06 09:52:42 * Wizzup needs to setup a maemo sdk+emulator first anyway
Oct 06 09:52:45 <Pali> and what you want to debug?
Oct 06 09:52:52 <Wizzup> Pali: I want to build the gadget usb module and create usb gadgets
Oct 06 09:53:23 <Wizzup> Pali: Although I guess the usb gadget module already exists... :) 
Oct 06 09:53:29 <Wizzup> but something like gadgetfs
Oct 06 09:53:47 <Pali> I never worked with gadgetfs
Oct 06 09:54:04 <Pali> so it is up to you
Oct 06 09:54:17 <Pali> use maemo sdk or cross compiler
Oct 06 09:54:47 <Pali> unload all gadget drivers (g_*) and load yours
Oct 06 09:54:52 <Pali> and see what happen
Oct 06 09:55:48 <Wizzup> ok
Oct 06 09:55:50 <Wizzup> thank you
Oct 06 09:56:06 <Wizzup> [I will play with this later]
Oct 06 11:37:39 <int_ua> reinob: ping
Oct 06 14:18:33 <janemba> d
Oct 06 18:49:43 <Sicelo> one of the times Maemo drives me crazy is when you connect N900 in usb storage mode, and MyDocs doesn't get exported. you're told the filesystem is busy. howeverer, you're not running anything at the time which is using MyDocs. you check with `lsof` and the command ends peacefully. nothing returned.
Oct 06 18:50:29 <Pali> lsof on maemo not working correctly
Oct 06 18:50:35 <Pali> fuser too
Oct 06 18:50:44 <Pali> thanks messybox!
Oct 06 18:51:13 <Sicelo> most times i reboot the device.. but like right now, i really with to avoid it
Oct 06 18:51:49 <Pali> something is using MyDocs and lsof show you messy output...
Oct 06 18:52:13 <Defiant> uhm did anyone succeeded in using a 3rd party charger for n900?
Oct 06 18:52:20 <Pali> because lsof implementation is messy and is part of messybox
Oct 06 18:53:08 <Sicelo> lsof returns nothing at this moment..
Oct 06 18:53:33 <Defiant> aah I see pin 2 and 3..
Oct 06 18:53:34 <Pali> so it is messy :-)
Oct 06 18:53:40 <Sicelo> Defiant: if you shot its data pins somehow
Oct 06 18:53:44 <Sicelo> *short
Oct 06 18:53:58 <Sicelo> Pali: reboot? 
Oct 06 18:54:15 <Pali> you can try rebooting...
Oct 06 18:54:34 <Pali> or closing all applications
Oct 06 18:54:37 <Sicelo> aw, i was really hoping to avoid that
Oct 06 18:54:45 <Sicelo> all applications are closed :P
Oct 06 18:54:47 <Pali> specialy camera, mediaplayer...
Oct 06 18:55:06 <Pali> or you can look into /proc/*/fd/
Oct 06 18:55:24 <Pali> and try to find process which has opened something in MyDocs
Oct 06 18:55:34 <ecc3g> seems a lot of recent chargers indeed short the data pins together but are lacking in other ways...
Oct 06 18:55:39 <Pali> or check /proc/* pwd
Oct 06 18:55:51 <Defiant> ecc3g: what other ways?
Oct 06 18:58:54 <ecc3g> mostly bad electronics - can't supply the full current
Oct 06 18:59:02 <Defiant> doh
Oct 06 21:09:31 <DocScrutinizer05> that's silly
Oct 06 21:19:40 <DocScrutinizer05> must be really crappy chargers when they can't deliver the 1000mA the N900 needs
Oct 06 21:21:37 <ShadowJK> Those single-transistor smps
Oct 06 21:35:14 <ShadowJK> http://www.youtube.com/watch?v=T88ej64aXUM
Oct 06 22:26:03 <DocScrutinizer05> LOL
Oct 06 22:26:33 <DocScrutinizer05> efficiency 15%
Oct 07 00:23:11 <ecc3g> argh.  should I be mad that some sites don't recognize microb as a phone browser?
Oct 07 00:24:56 <psycho_oreos> Not really, you do realise that N900 was really targeting at a niche market. Those who still hold onto their N900 should probably be a bit more lenient in the light of common handhelds making it's mark on the internet.
Oct 07 00:25:51 <ecc3g> need to get an user agent switcher I guess
Oct 07 00:26:10 <ecc3g> I wonder if I should make it do Android or Symbian, though. heh
Oct 07 00:27:02 <psycho_oreos> Try either and see which one works out better. I'd say android would cover more sites than symbian, then again.
Oct 07 00:27:48 <ecc3g> Well, since microb is xulrunner, and both android default and symbian are webkit...
Oct 07 00:28:27 <psycho_oreos> There are a few other browsers available for N900, though notably old/flaky.
Oct 07 00:28:43 <ecc3g> I guess in that respect, saying my phone is ios wouldn't matter... it's also webkit
Oct 07 00:30:34 <psycho_oreos> There was also macuco (I think that's how it's spelt) browser which virtually mimicks iOS behaviour.
Oct 07 00:31:40 <ecc3g> it's that silly flash... argh.
Oct 07 00:32:06 <ecc3g> though gnash still isn't quite usable, has someone tried porting it to n900?
Oct 07 00:32:42 <Wizzup> haven't tried, but it works okay for some things on my odroid (also arm)
Oct 07 00:32:50 <Wizzup> but I doubt the n900 has enough ram / is fast enough
Oct 07 00:33:35 <ecc3g> gnash itself is slow, but I suppose at this point getting behavior correct is more important than execution speed
Oct 07 00:33:46 <ecc3g> theoretically it shouldn't be any slower than real flash :D
Oct 07 00:34:00 <ecc3g> (which, then again, is slow too...)
**** ENDING LOGGING AT Mon Oct 07 02:59:58 2013