**** BEGIN LOGGING AT Fri Jul 08 02:59:58 2016
Jul 08 03:45:22 <Linkandzelda> is it true that moving swap to sd card speeds up things? is there a way to test if the microsd is actually faster? i can imagine some cases were it would be a bottleneck
Jul 08 06:48:47 <freemangordon> Linkandzelda: yep, it is true. But make sure you have good SD card with fast random IO, like sammy pro cards
Jul 08 06:49:12 <KotCzarny> uh, wrong channel
Jul 08 06:49:24 <freemangordon> hmm?
Jul 08 06:49:40 <KotCzarny> i've answered Linkandzelda on wrong channel
Jul 08 06:49:43 <KotCzarny> just noticed
Jul 08 06:50:00 <freemangordon> ah :)
Jul 08 09:49:00 <DocScrutinizer05> freemangordon: Linkandzelda: http://paste.opensuse.org/43972960
Jul 08 09:50:44 <DocScrutinizer05> swap on uSD is _not_ per se faster than eMMC, it's a 4bit interface vs 8bit on eMMC and clock rate is comparable. What speeds up stuff is when you unebtangle storage access between swap and e.g. any read/write IO on rootfs
Jul 08 09:51:19 <DocScrutinizer05> unentangle*
Jul 08 09:55:42 <DocScrutinizer05> what's also very relevant (and not at all focused in my simple tests above) is write performance which is highly dependent on controller strategy and architecture (how much buffer RAM, page size, handling of truncate for empty pages that don't need erase, **erase performance**, etc pp)
Jul 08 09:56:21 <DocScrutinizer05> a page erase may take as long as 0.1s
Jul 08 09:57:26 <DocScrutinizer05> so it's very desirable behavior when the controller does background page erase to always have a sufficient number of clean pages that can get written to without prior erase
Jul 08 09:58:34 * DocScrutinizer05 idly wonders if maemo mounts the MMCs with "discard" option, if even the kernel knows that option already
Jul 08 09:59:26 <DocScrutinizer05> IroN900:~# mount|grep mmc
Jul 08 09:59:28 <DocScrutinizer05> /dev/mmcblk0p2 on /home type ext3 (noatime,nodiratime,errors=continue,commit=1,data=writeback)
Jul 08 09:59:29 <DocScrutinizer05> /dev/mmcblk0p1 on /home/user/MyDocs type vfat (rw,noatime,nodiratime,noauto,nodev,exec,nosuid,utf8,uid=29999,shortname=mixed,dmask=000,fmask=0000,rodir)
Jul 08 09:59:31 <DocScrutinizer05> /dev/mmcblk1p2 on /media/mmc1p2 type ext3 (rw,noatime,nodiratime,noauto,nodev,exec,nosuid,data=writeback)
Jul 08 09:59:32 <DocScrutinizer05> /dev/mmcblk1p1 on /media/mmc1 type vfat (rw,noatime,nodiratime,noauto,nodev,exec,nosuid,utf8,uid=29999,shortname=mixed,dmask=000,fmask=0000,rodir)
Jul 08 10:01:25 <DocScrutinizer05> FFS! Linux IroN900 2.6.28-omap1  vs  https://wiki.ubuntuusers.de/SSD/TRIM/  >>Online Discard  ext4  Kernel 2.6.33<<
Jul 08 10:01:44 <DocScrutinizer05> freemangordon: ^^^
Jul 08 10:03:09 <DocScrutinizer05> Batched Discard   ext2, ext3,   Kernel 2.6.38
Jul 08 10:04:05 <DocScrutinizer05> now THAT qualifies for a sound reason to finally update kernel in CSSU
Jul 08 10:15:24 <DocScrutinizer05> I mean, flash storage without truncate/discard? a nogo
Jul 08 10:15:40 <DocScrutinizer05> aka TRIM
Jul 08 10:18:26 <DocScrutinizer05> freemangordon: http://www.sammypro.com/ ???
Jul 08 10:19:07 <DocScrutinizer05> I still think genuine SanDisk are prolly the best cards
Jul 08 10:20:09 <DocScrutinizer05> or are you talking about Samsung?
Jul 08 10:20:47 <DocScrutinizer05> http://bgr.com/2016/07/07/fastest-256gb-microsd-card-samsung-ufs/
Jul 08 11:10:52 <Bitnotri> So guys I'm coming back to n900 after a long hiatus. I love writing notes on it and that's mainly what I've been doing with it, making it my ultra-portable journal.
Jul 08 11:11:20 <Bitnotri> However in the past I used to read books, browse web and have a wide array of apps on it which are no longer usable
Jul 08 11:11:36 <Bitnotri> What should I do to make n900 run as smoothly as possible in 2016?
Jul 08 11:12:01 <Bitnotri> What OS is the best to put on n900 right now?
Jul 08 11:15:10 <zGrr> Harmattan+CSSU
Jul 08 11:15:21 <Bitnotri> Thanks
Jul 08 11:15:54 <DocScrutinizer05> harmattan? hardly
Jul 08 11:16:08 <DocScrutinizer05> Bitnotri: 
Jul 08 11:16:10 <DocScrutinizer05> ~cssu
Jul 08 11:16:11 <infobot> i heard cssu is http://wiki.maemo.org/Community_SSU, or (Community Seamless Software Update)
Jul 08 11:16:39 <Bitnotri> What would you say is the best option DocScrutinizer05?
Jul 08 11:16:42 <DocScrutinizer05> Bitnotri: forget harmattan
Jul 08 11:16:51 <DocScrutinizer05> Bitnotri: see above
Jul 08 11:17:59 <DocScrutinizer05> a fresh reflash often also works wonders
Jul 08 11:18:06 <Bitnotri> What about some projects such as DebianN900?
Jul 08 11:18:13 <DocScrutinizer05> what about it?
Jul 08 11:18:30 <Bitnotri> Would it be a better option?
Jul 08 11:18:38 <DocScrutinizer05> depends
Jul 08 11:18:55 <DocScrutinizer05> I don't see any advantage over a easydeb chroot
Jul 08 11:18:58 <Bitnotri> What factors should I consider?
Jul 08 11:20:03 <DocScrutinizer05> there's no other OS release than maemo fremantle that's better adapted to support every aspect of the hw platform
Jul 08 11:20:22 <Bitnotri> I've reflashed my device in the past, tried to install those community ssus. I may have done the job wrong, however the problems I've head previously with the web browser in particular didn't disappear. It may be different now so it may be worth the shot
Jul 08 11:20:58 <Bitnotri> Have you tried harmattan or other distributions? How did you like it?
Jul 08 11:20:58 <DocScrutinizer05> web brosing nowadays is a task that even i5 PC with 8GB RAM can suck
Jul 08 11:21:21 <DocScrutinizer05> that's not the browsers' fault but the culprit are idiot "webdesigners"
Jul 08 11:21:43 <Bitnotri> Yeah, but even stripped of css and js would be good
Jul 08 11:21:45 <DocScrutinizer05> harmattan doesn't work on N900
Jul 08 11:22:02 <Bitnotri> Period?
Jul 08 11:22:09 <DocScrutinizer05> a complete wrong hint
Jul 08 11:22:31 <DocScrutinizer05> maemo harmattan is for Nokia N9, NOT N900
Jul 08 11:22:53 <Bitnotri> what do you think about debian900?
Jul 08 11:23:08 <DocScrutinizer05> nothing, I never seen a reason to try it
Jul 08 11:23:41 <Bitnotri> Ok, thanks. Would it be hard to reverse if I tried it?
Jul 08 11:24:10 <DocScrutinizer05> _if_ I needed to run apps than can't run under plain Fremantle since they need nifty new features only found in newer debian versions, I'd go for a easydeb chroot for that purpose
Jul 08 11:24:33 <DocScrutinizer05> reverse?
Jul 08 11:24:39 <DocScrutinizer05> you mean go back to maemo?
Jul 08 11:24:44 <Bitnotri> Yup
Jul 08 11:24:49 <DocScrutinizer05> baby simple, do
Jul 08 11:24:56 <DocScrutinizer05> ~lazyflashing
Jul 08 11:24:57 <infobot> lazyflashing is probably http://wiki.maemo.org/Updating_the_tablet_firmware#The_Lazy_Approach
Jul 08 11:25:15 <DocScrutinizer05> you also could install dualboot via uBoot
Jul 08 11:25:29 <DocScrutinizer05> to have debian and maemo as boottime alternatives
Jul 08 11:25:45 <Bitnotri> Ok, thanks for all the help, I've got a lot of reading and installing to do :)
Jul 08 11:25:56 <DocScrutinizer05> yw :-)
Jul 08 11:26:54 <DocScrutinizer05> note that with backupmenu you also can create comprehensive backup of your complete maemo system, to roll back to that very state later on
Jul 08 11:27:00 <DocScrutinizer05> ~bm
Jul 08 11:27:00 <infobot> rumour has it, backupmenu is http://talk.maemo.org/showthread.php?t=63975, or one-click install handling dependency issues: http://maemo.cloud-7.de/maemo5/et_al/HAM-catalogs/BM.install
Jul 08 11:27:33 <Bitnotri> Oh, that seems really useful. Thank you ;)
Jul 08 11:27:39 <DocScrutinizer05> yw :-)
Jul 08 11:28:06 <DocScrutinizer05> also not 
Jul 08 11:28:12 <DocScrutinizer05> ~multiboot
Jul 08 11:28:13 <infobot> well, multiboot is http://maemo.org/packages/view/multiboot/, or **DEPRECATED*, see ~maemo-multiboot
Jul 08 11:28:21 <DocScrutinizer05> ~maemo-multiboot
Jul 08 11:28:22 <infobot> it has been said that maemo-multiboot is deprecated, and a horrible hack.  PROBLEMS WITH NITDROID/MULTIBOOT? reflash rootfs&kernel aka COMBINED
Jul 08 11:28:52 <DocScrutinizer05> also note*  but what I wrote was as good ;-)
Jul 08 11:29:14 <DocScrutinizer05> maybe useful, unless you never touched cmdline:
Jul 08 11:29:18 <DocScrutinizer05> ~jrtools
Jul 08 11:29:18 <infobot> it has been said that jrtools is http://wiki.maemo.org/User:Joerg_rw/tools
Jul 08 11:30:02 <Bitnotri> Nah, I'm an active ubuntu user so I can handle cli ;)
Jul 08 11:30:03 <DocScrutinizer05> for CSSU I recommend Testing
Jul 08 11:30:45 <DocScrutinizer05> don't go for cssu-thumb until you learned what exactly it is and which are the implications of using it
Jul 08 11:30:53 <Bitnotri> ok
Jul 08 11:31:25 <DocScrutinizer05> ~uboot
Jul 08 11:31:26 <infobot> N900 uBoot is a siamese twin binary [uBoot+stockMaemoKernel] that resides in kernel NAND partition /dev/mtd3 aka "kernel". You can't uninstall it, rather you'll nuke it when you flash/install another kernel like stock maemo kernel or powerkernel. To start other than stock maemo kernel via uBoot, you have to provide the according kernel image files, or http://talk.maemo.org/showthread.php?t=81613
Jul 08 11:32:46 <DocScrutinizer05> and yes, you might be interested in powerkernel, when you want to go fancy with your N900
Jul 08 11:33:17 <DocScrutinizer05> it comes with support for a lot of goodies like ext4, netfilters, whatnot else
Jul 08 11:33:39 <DocScrutinizer05> ~kp
Jul 08 11:33:39 <infobot> rumour has it, kp is http://talk.maemo.org/showthread.php?t=94287
Jul 08 11:33:54 <Bitnotri> Would powerkernel help with n900 performance?
Jul 08 11:34:00 <DocScrutinizer05> not much
Jul 08 11:34:11 <DocScrutinizer05> cssu will help
Jul 08 11:35:19 <Bitnotri> did you overclock your n900?
Jul 08 11:35:37 <DocScrutinizer05> just 2h ago I noticed N900 kernel do4esn't support TRIM for flash storage, which is a terrible flaw regarding performance, especially for swap
Jul 08 11:35:44 <DocScrutinizer05> nope
Jul 08 11:35:52 <DocScrutinizer05> overclocking is basically useless
Jul 08 11:36:08 <Bitnotri> Yeah, so I thought
Jul 08 11:36:16 <DocScrutinizer05> you can overclock less than factor 2 which is still almost worthless for user experience
Jul 08 11:36:36 <DocScrutinizer05> and the real performance hogs are swap
Jul 08 11:36:57 <DocScrutinizer05> and generally RAM shortage and storage IO bottleneck
Jul 08 11:37:01 <Bitnotri> Can anything be done about it?
Jul 08 11:37:18 <DocScrutinizer05> get a Neo900 ;-) - in 9 months 
Jul 08 11:37:32 <DocScrutinizer05> thumb helps about RAM shortage a bit
Jul 08 11:37:36 <DocScrutinizer05> ~thumb
Jul 08 11:37:36 <infobot> somebody said cssu-thumb was <Doc_Scrutinizer05> [thumb2 microb] indeed seems to render like mad, subjectively, or http://talk.maemo.org/showpost.php?p=1220597, or http://wiki.maemo.org/Community_SSU/Thumb
Jul 08 11:38:10 <DocScrutinizer05> thumb code is significantly smaller than ARM instruction set
Jul 08 11:38:13 <Bitnotri> neo900 seems like won't come out at all, it's been moving so slow for such a long time
Jul 08 11:38:28 <DocScrutinizer05> yes, it's hurting me 
Jul 08 11:38:41 <DocScrutinizer05> we suffer one roadblock after the other
Jul 08 11:39:34 <Bitnotri> Such it is without a major support
Jul 08 11:39:35 <DocScrutinizer05> now Nikolaus declared ultimate bail out on PCB layout, so the complete workflow and toolchain needs a redesign/re-evaluation
Jul 08 11:40:19 <DocScrutinizer05> we're this very moment pondering to migrate fron cadsoft eagle to KiCad
Jul 08 11:41:02 <Bitnotri> So you're actively working on neo900?
Jul 08 11:41:02 <DocScrutinizer05> as if there were no other urgent tasks pending
Jul 08 11:41:06 <DocScrutinizer05> sure
Jul 08 11:42:03 <DocScrutinizer05> http://irclog.whitequark.org/neo900/2016-07-07  and https://neo900.org/git/?p=misc;a=log;pg=1
Jul 08 11:48:07 <DocScrutinizer05> Bitnotri: feel free to /join #neo900
Jul 08 11:49:15 <Bitnotri> Ok, I'm just not sure if I'd be able to contribute anything :P
Jul 08 11:49:21 <DocScrutinizer05> np
Jul 08 11:49:31 <DocScrutinizer05> every lurker is welcome too
Jul 08 11:51:00 <DocScrutinizer05> actually considering that IRC #neo900 is our most up-to-date and realtime communication channel with the highest traffic, I'm happy about each prospect customer having a look in there to notice the project is not entirely dead
Jul 08 11:51:32 <Bitnotri> Ok, I'll definitely check it out then
Jul 08 11:51:44 <Bitnotri> Does any web browser work reasonably well on your n900?
Jul 08 11:51:53 <DocScrutinizer05> we suck terribly on lack of a decent communications officer who could update the webpage with new info, answer threads in forum and email, etc
Jul 08 11:52:08 <DocScrutinizer05> (browser) tbh nope
Jul 08 11:52:40 <DocScrutinizer05> it's sufficent for important selected tasks like online banking, checking my share-service at maemo.cloud-7.de etc
Jul 08 11:53:15 <DocScrutinizer05> but then I'm possibly not the right person to ask since I'm not much html-centric at all
Jul 08 11:53:44 <DocScrutinizer05> I use xterm more often than microB (or any browser), on N900
Jul 08 11:53:58 <Bitnotri> ok
Jul 08 11:54:20 <Sicelo> microb good enough for me. i don't really use any other browser
Jul 08 11:54:37 <DocScrutinizer05> for a quick fixing of the WLAN modem router in my favorite pub, when the owner messed it op one more, microB is just sufficently working
Jul 08 11:55:00 <DocScrutinizer05> Sicelo: same here
Jul 08 11:55:43 <Sicelo> also supports client certificates just fine .. haven't found another one on N900 that has that :)
Jul 08 11:55:56 <DocScrutinizer05> watching YT videos is something I'm actually missing. Evcen while I learned now there's cutetube2 
Jul 08 11:56:31 <Bitnotri> Oh I remember that app, it was really nice
Jul 08 11:58:30 * DocScrutinizer05 - on a completely unrelated sidenote - points at (http://maemo.cloud-7.de/maemo5/session-log_enable-catalogs_README.txt and) http://maemo.cloud-7.de/maemo5/usr/local/sbin/enable-catalogs
Jul 08 11:58:43 <Sicelo> it still works fine Bitnotri (cutetube2)
Jul 08 11:58:43 <DocScrutinizer05> ooh, and ~jrrepos
Jul 08 11:58:51 <DocScrutinizer05> ~jrrepos
Jul 08 11:58:51 <infobot> rumour has it, jrrepos is http://maemo.cloud-7.de/maemo5/et_al/HAM-catalogs/
Jul 08 11:58:54 <Sicelo> then again .. i'm not big on YT either
Jul 08 12:12:04 <Linkandzelda> DocScrutinizer05: thanks for the test. i have a 32gb sandisk pro xchc uhs1 or something. will look into moving swap to it and see what happens. i noticed when streaming internet radio the device slows down a whole lot and i get the feeling its swaping a lot
Jul 08 12:13:25 <Sicelo> what bit rates for your streams? mine doesn't slow down 
Jul 08 12:15:41 <DocScrutinizer05> iotop ftw
Jul 08 12:15:56 <DocScrutinizer05> iotop and htop
Jul 08 12:16:10 <Linkandzelda> Sicelo: 192 kbps
Jul 08 12:16:26 <Linkandzelda> when streaming it shows puleaudio using between 14-20% cpu
Jul 08 12:16:40 <DocScrutinizer05> yeah, pukeaudio sucks
Jul 08 12:18:32 <Linkandzelda> and of course battery drops massive
Jul 08 12:18:47 <DocScrutinizer05> Linkandzelda: note that N900 neither can use UHS nor xdhc
Jul 08 12:19:15 <Linkandzelda> didnt think so, but it would it be able to use the standard functionalities right?
Jul 08 12:19:23 <DocScrutinizer05> but odds are a UHS capable uSD is actually fast enough to run at max speed N900 can do
Jul 08 12:19:29 <DocScrutinizer05> yes
Jul 08 12:19:38 <Linkandzelda> my thoughts too
Jul 08 12:19:52 <Linkandzelda> and is it safe for me to run the same tests you just did?
Jul 08 12:20:00 <DocScrutinizer05> iirc that would be 40MHz clock at 4bit bus width
Jul 08 12:20:14 <DocScrutinizer05> yes
Jul 08 12:20:29 <DocScrutinizer05> those tests are considered safe by me
Jul 08 12:20:33 <DocScrutinizer05> no warranty
Jul 08 12:21:13 <DocScrutinizer05> my uSD is actually a little dusty already
Jul 08 12:21:21 <DocScrutinizer05> might easily be 6 years old
Jul 08 12:21:38 <Linkandzelda> ive got another process using cpu here
Jul 08 12:21:53 <DocScrutinizer05> **kill it**  hehehe
Jul 08 12:22:05 <Linkandzelda> mafw-dbus-wrapper mafw-gst-renderer
Jul 08 12:22:15 <DocScrutinizer05> that's audio playback
Jul 08 12:22:36 <Linkandzelda> using between 7-14% cpu
Jul 08 12:22:58 <Linkandzelda> leaving the average while streaming and doing nothing being 50% usage, thats high i guess
Jul 08 12:23:05 <DocScrutinizer05> http://paste.opensuse.org/35212726
Jul 08 12:23:24 <Linkandzelda> i see
Jul 08 12:24:21 <DocScrutinizer05> http://wstaw.org/m/2016/07/08/plasma-desktophc2277.png
Jul 08 12:24:46 <DocScrutinizer05> that's idle system
Jul 08 12:25:13 <DocScrutinizer05> during audio playback mafw and polypaudio will be CPU hogs
Jul 08 12:25:25 <Linkandzelda> any tweaks for pulse?
Jul 08 12:26:24 <DocScrutinizer05> all tweaks so far fired back afaik
Jul 08 12:26:53 <Linkandzelda> ah, shame
Jul 08 12:26:55 <Linkandzelda> well its ok
Jul 08 12:26:55 <DocScrutinizer05> recompiling that shit without debug output might help a little
Jul 08 12:27:39 <Linkandzelda> can you explain what those dd commands are doing exactly?
Jul 08 12:27:42 <DocScrutinizer05> many peaple tried to fix suspected flaws in PA, it necer panned out to the better iirc
Jul 08 12:28:27 <DocScrutinizer05> dd copies raw chunks of 1Mbyte data from physical device (if=) to nirvana (of=/dev/null)
Jul 08 12:29:09 <DocScrutinizer05> no partitions, no filesystem involved. Using raw IO without buffering (iflag=direct)
Jul 08 12:31:09 <DocScrutinizer05> without iflag=direct you still will see effects from regular filesystem buffering which may hold parts of the data to read in RAM buffers
Jul 08 12:31:40 <DocScrutinizer05> at least all this is how I hipe stuff works ;-)
Jul 08 12:31:44 <DocScrutinizer05> hope*
Jul 08 12:32:14 <Linkandzelda> fair enough, as long as there is no loss then thats all that matters, right hehe
Jul 08 12:32:16 <DocScrutinizer05> you may go for larger amount of test data, with count=500 or even 1000
Jul 08 12:32:40 <DocScrutinizer05> multiply by bLOCKsIZE= which is 1 Mbyte
Jul 08 12:33:08 <DocScrutinizer05> so bs=1M count=1000 will read 1GB data
Jul 08 12:33:22 <Oksanaa> What's up with kernel-power upgrade? /just curious, aptitude does not cite changelog/
Jul 08 12:33:48 <DocScrutinizer05> same for bs=1000M count=1, but that might allocate insane size of program internal buffer in RAM
Jul 08 12:34:40 <DocScrutinizer05> Oksanaa: is there any new update supposed to be around?
Jul 08 12:35:05 <Oksanaa> Well, I just did aptitude update upgrade and got it
Jul 08 12:35:50 <DocScrutinizer05> then maybe that's a good instance why you shouldn't do upgrade ;-)
Jul 08 12:36:17 <fishbulb> hey I've got some version of fmms that doesn't work
Jul 08 12:37:10 <Linkandzelda> DocScrutinizer05: on an off note, when i type "root" and enter pass, it seems to put me in busybox, since dd gave a syntax error and mention busybox. after i typed "bash" then tried dd again its working as expected. ive got shells set in /etc/passwd for both user and root to /bin/bash. any way i can fix that issue?
Jul 08 12:37:12 <Oksanaa> hmm, how so? /cssu-thumb-testing if it helps/ fmms? cool, let's check if I have it installed, did once half-receive mms with it
Jul 08 12:37:32 * DocScrutinizer05 idly curses his current IRC client version that doesn't (yet) support notes attached to users
Jul 08 12:37:52 <fishbulb> I think it came from a dev repo
Jul 08 12:39:11 <Linkandzelda> DocScrutinizer05: or actually not, it seems to put me in bash the first time but still references the dd binary from busybox, or something, im not sure
Jul 08 12:39:30 <DocScrutinizer05> Linkandzelda: tbh I never bothered. I always considered it safer and not sufficently annoying to stray from doing "root; bash" each time I log in
Jul 08 12:39:35 * Oksanaa has fmms installed, yes
Jul 08 12:39:49 <fishbulb> ok but it won't be the same as mine
Jul 08 12:39:55 <DocScrutinizer05> Linkandzelda: `which dd`
Jul 08 12:39:55 <Linkandzelda> i see
Jul 08 12:40:11 <fishbulb> is KotCzarny here?
Jul 08 12:40:18 <fishbulb> he wrote fmms right
Jul 08 12:40:20 <Oksanaa> fishbulb: fmms 1.3.4, according to dpkg
Jul 08 12:40:30 <Linkandzelda> DocScrutinizer05: /bin/dd
Jul 08 12:40:35 <DocScrutinizer05> IroN900:~# which dd
Jul 08 12:40:37 <DocScrutinizer05> /usr/bin/gnu/dd
Jul 08 12:41:09 <DocScrutinizer05> IroN900:~# ls -l /bin/dd
Jul 08 12:41:10 <DocScrutinizer05> lrwxrwxrwx 1 root root 7 2014-12-06 12:39 /bin/dd -> busybox
Jul 08 12:41:18 <Linkandzelda> why lol
Jul 08 12:41:21 <Linkandzelda> http://pastebin.com/K0CPa3cd
Jul 08 12:41:24 <DocScrutinizer05> IroN900:~# echo $PATH
Jul 08 12:41:26 <DocScrutinizer05> /usr/local/sbin:/usr/local/bin:/usr/sbin/gnu:/usr/sbin:/usr/bin/gnu:/usr/bin:/sbin:/bin:/usr/bin/X11
Jul 08 12:41:38 <Oksanaa> KotCzarny was idle for 5h 52min, so should return soon-ish
Jul 08 12:42:02 <Linkandzelda> oh seems im missing stuff in path
Jul 08 12:42:08 <DocScrutinizer05> no, frals wrote fmms
Jul 08 12:42:16 <DocScrutinizer05> that's where the 'f' is from
Jul 08 12:42:30 <fishbulb> history is fascinating
Jul 08 12:42:43 <fishbulb> is frals ever here?
Jul 08 12:42:57 <DocScrutinizer05> o.O
Jul 08 12:43:25 <DocScrutinizer05> pro hint: TAB usually does nick completion
Jul 08 12:43:42 <fishbulb> but here as in present
Jul 08 12:44:12 <DocScrutinizer05> /whois frals
Jul 08 12:44:23 <Oksanaa> frals: was  idle     : 5 days 15 hours 35 mins 58 secs
Jul 08 12:45:02 <fishbulb> yerp.
Jul 08 12:45:42 <Oksanaa> Send email to Nick Leppänen Larsson, aka frals .of the mysterious. frals .dot. se ?
Jul 08 12:46:04 <fishbulb> nah
Jul 08 12:46:13 <DocScrutinizer05> however please don't expect to get a reasonable response to >>hey I've got some version of fmms that doesn't work<< which isn't even any question
Jul 08 12:46:28 <fishbulb> I'm already not going to bother him by email
Jul 08 12:47:12 <DocScrutinizer05> Linkandzelda: about "missing in PATH" see ~jrtools
Jul 08 12:47:20 <DocScrutinizer05> .bashrc in there
Jul 08 12:47:57 <DocScrutinizer05> though you're of course free to pick any other approach to set the PATH of your liking
Jul 08 12:48:22 <Oksanaa> Okay... Maybe, find something specific on how it misworks. Haven't tried to fmms myself, though. Have enough cellular-provider troubles without it
Jul 08 12:48:34 <DocScrutinizer05> actually I probably shouldn't recommend doing that for user:user .bashrc
Jul 08 12:49:13 <DocScrutinizer05> using any other than messybox as default shell for user:user is known to have potential to fubar your system
Jul 08 12:49:36 <DocScrutinizer05> so does changing default PATH for user
Jul 08 12:50:12 <DocScrutinizer05> there are a few messybox-isms in the xinit or whatever, that possibly lead to a bootloop
Jul 08 12:50:39 <DocScrutinizer05> funny enough only for user, not for root
Jul 08 12:50:55 <Linkandzelda> DocScrutinizer05: yea i have that .bashrc, but its weird
Jul 08 12:51:08 <Linkandzelda> when connecting via ssh/loading xterm it doesnt use it
Jul 08 12:51:31 <DocScrutinizer05> http://paste.opensuse.org/75522475
Jul 08 12:51:44 <DocScrutinizer05> ssh uses other login method
Jul 08 12:52:05 <Linkandzelda> can i find where the .bashrc it uses is at?
Jul 08 12:52:14 <Linkandzelda> since i get the same behaviour from xterm
Jul 08 12:53:25 <DocScrutinizer05> it's a nice exercise for an afternoon to read through the manpages for bash, shell, login, ssh(d) etc, to figure which startscripts are used with which shell and type of login
Jul 08 12:53:58 <Linkandzelda> fair enough
Jul 08 12:54:24 <DocScrutinizer05> I was pretty surprised to find messybox source .bashrc
Jul 08 12:54:55 <DocScrutinizer05> which sucks in so many regards
Jul 08 12:55:11 <DocScrutinizer05> well, in short: messybox sucks
Jul 08 12:55:20 <DocScrutinizer05> ~messybox
Jul 08 12:55:21 <infobot> messy... err busybox is meant for lean scripting. Regarding all the missing options and immanent limitations (see su, passwd, nice, ps, diff as used by mc...) 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
Jul 08 12:56:27 <DocScrutinizer05> thus busybox-power is a pretty incorrect approach
Jul 08 12:57:16 <DocScrutinizer05> and I strongly deprecate busybox-power
Jul 08 12:57:20 <Linkandzelda> DocScrutinizer05: found what the "root" script does, sudo gainroot --use-su
Jul 08 12:57:33 <DocScrutinizer05> yes
Jul 08 12:58:00 <DocScrutinizer05> and gainroot is funny too
Jul 08 12:58:01 <Linkandzelda> HISTFILE=/root/.ash_history /bin/sh
Jul 08 12:58:10 <Linkandzelda> its got a typo in /usr/bin/gainroot ?
Jul 08 12:58:24 <DocScrutinizer05> no, why?
Jul 08 12:58:34 <Linkandzelda> it should be bash_history i thought
Jul 08 12:58:45 <DocScrutinizer05> no, please not
Jul 08 12:58:52 <Linkandzelda> i see now it has its own path here in gainroot
Jul 08 12:59:35 <DocScrutinizer05> I'm more than happy with "bash" being the _only_ entry in ~/ash_history
Jul 08 13:00:20 <Oksanaa> Old kernel version: 2.6.28.10-power51, and new one is power53. What the difference?
Jul 08 13:00:29 <DocScrutinizer05> http://paste.opensuse.org/93685597
Jul 08 13:00:39 <DocScrutinizer05> ~kp
Jul 08 13:00:40 <infobot> hmm... kp is http://talk.maemo.org/showthread.php?t=94287
Jul 08 13:01:11 <Linkandzelda> is it dangerious to edit the PATH in /usr/sbin/gainroot ?
Jul 08 13:01:36 <DocScrutinizer05> yes, possibly
Jul 08 13:02:27 <DocScrutinizer05> some fubar packages may depend on gainroot
Jul 08 13:04:02 <Linkandzelda> would just like to add the paths you have from your regular path so it can pick up the non-bb dd
Jul 08 13:06:23 <DocScrutinizer05> try it, but I strongly recommend to do a BM backup prior to that
Jul 08 13:06:32 <Linkandzelda> ok i will do that
Jul 08 13:06:57 <DocScrutinizer05> so you could restore the original gainroot file if shit hits the fan
Jul 08 13:07:27 <DocScrutinizer05> well, for that purpose prolly any old backup will do#
Jul 08 13:07:54 <DocScrutinizer05> as long as you restore only the gainroot file
Jul 08 13:09:17 <DocScrutinizer05> Linkandzelda: you know all gnu cmds * are also available via /usr/bin/g*
Jul 08 13:09:23 <DocScrutinizer05> gdd in this case
Jul 08 13:09:51 <DocScrutinizer05> IroN900:~# which gdd
Jul 08 13:09:53 <DocScrutinizer05> /usr/bin/gdd
Jul 08 13:09:54 <DocScrutinizer05> IroN900:~# ll `which gdd`
Jul 08 13:09:56 <DocScrutinizer05> lrwxrwxrwx 1 root root 6 2014-12-06 12:40 /usr/bin/gdd -> gnu/dd
Jul 08 13:10:16 <Linkandzelda> yes, but i'll forget to use g too often
Jul 08 13:10:24 <DocScrutinizer05> it sucks :-)
Jul 08 13:12:11 <Oksanaa> Ouch... Is it that difficult for maemo repositories to upgrade perl to something higher than 5.8? /just trying to install mosh/
Jul 08 13:14:23 * Oksanaa should someday just go and feed various source packages into auto-builder. Well, they may need optification beforehand...
Jul 08 13:16:07 <Oksanaa> Has no gui > needs minimum of modification. But optification is really necessary
Jul 08 13:17:16 <DocScrutinizer05> optification can get done semi-automatically
Jul 08 13:18:52 <DocScrutinizer05> IroN900:~# maemo-optify-
Jul 08 13:18:53 <DocScrutinizer05> maemo-optify-auto-opt.sh        maemo-optify-firstboot.sh       maemo-optify-lib.sh             maemo-optify-make-mountbind.sh 
Jul 08 13:20:28 <DocScrutinizer05> IroN900:~# less `which maemo-optify-auto-opt.sh `
Jul 08 13:20:30 <DocScrutinizer05> IroN900:~# less /etc/maemo-optify-auto-opt.conf
Jul 08 13:22:19 <Oksanaa> Is it recommended to put "/usr/bin" in here? Why not?
Jul 08 13:23:03 <DocScrutinizer05> because that's exactly the way to nuke your system
Jul 08 13:24:19 <DocScrutinizer05> everything in /usr/bin/ that's not a symlink is probably there because it must not get optified
Jul 08 13:24:36 <DocScrutinizer05> ~optification
Jul 08 13:24:36 <infobot> optification is a inventive duct tape workaround to reclaim space in fs root, done due to the fact the systeminit *and* partitioning is FUBAR,  http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Packaging,_Deploying_and_Distributing/Installing_under_opt_and_MyDocs, or ""OMG - I wish they looked into FHS and moved /usr to eMMC"", http://www.pathname.com/fhs/pub/fhs-2.3.html#PURPOSE2 bullet1,2 and fhs-2.3.html#PURPOSE16 dot3"
Jul 08 13:25:39 <DocScrutinizer05> everything in /usr/bin/ that's not a symlink is probably also something that should have been installed in /bin instead of /usr/bin to start with
Jul 08 13:26:18 <DocScrutinizer05> dunno which fool decided to pit stuff needed during early boot into /usr/*
Jul 08 13:26:23 * Oksanaa grumbles that qtoctave should have no problems being optified
Jul 08 13:26:23 <DocScrutinizer05> put*
Jul 08 13:27:38 <DocScrutinizer05> it's clearly a duty of the distro maintainers to move stuff from /usr/bin to /bin when it is needed during early boot. See FHS
Jul 08 13:28:46 <Oksanaa> Hmm, would be nice. What open-source packages are used during early boot and have things in /usr/bin or /usr/sbin?
Jul 08 13:30:43 <DocScrutinizer05> it's plain ridiculous to first build stuff for --prefix=/usr/bin, then go "OMG!! we need that during early boot, so let's move /usr/bin to rootfs where it's available during early boot. Then we optify everything in /usr/bin that is _not_ needed "
Jul 08 13:31:16 <DocScrutinizer05> ((What open-source packages are used during early boot and have things in /usr/bin or /usr/sbin?)) check for symlinks
Jul 08 13:31:36 <DocScrutinizer05> everything not a symlink is incorrectly put into /usr/
Jul 08 13:33:58 <DocScrutinizer05> find /usr/bin -type f -ls
Jul 08 13:34:33 <DocScrutinizer05> vs   find /usr/bin -type l -ls
Jul 08 13:34:51 <Oksanaa> Let's see : Calendar.launch (closed-source?), mceledpattern (! totally not needed for boot ?)
Jul 08 13:35:41 * DocScrutinizer05 pukes at /usr/bin/setkeycodes -> /bin/busybox
Jul 08 13:36:04 <DocScrutinizer05> though it's prolly even formally ok
Jul 08 13:37:35 <Oksanaa> fmtx_client , and I should move my personal /usr/bin/randomscripts elsewhere (like, into aliases in /etc/profile or something)
Jul 08 13:37:38 <DocScrutinizer05> mceledpattern is only 35k
Jul 08 13:37:54 <DocScrutinizer05> which is a candidate for optification, but not mandatory
Jul 08 13:38:44 <Oksanaa> x11vnc and qtoctave are huge, around 1MB each?
Jul 08 13:39:15 <Oksanaa> Xorg and python2.5, too
Jul 08 13:39:20 <DocScrutinizer05> http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Packaging,_Deploying_and_Distributing/Installing_under_opt_and_MyDocs  >>Developers are encouraged to make good use of them, specially for applications requiring more than 500 KB, including dependencies.<< err *cough*
Jul 08 13:39:44 <DocScrutinizer05> Oksanaa: huh?
Jul 08 13:40:10 <DocScrutinizer05> you won't go happy with optifying python or X11
Jul 08 13:40:38 <Oksanaa> x11vnc is not needed for boot, is it?
Jul 08 13:40:43 <DocScrutinizer05> nope
Jul 08 13:41:03 <DocScrutinizer05> but Xorg is, I guess
Jul 08 13:41:10 <DocScrutinizer05> python 2.5 too
Jul 08 13:41:16 <Oksanaa> Xorg is system-needed, though. Not going to move it around on my device
Jul 08 13:41:51 <DocScrutinizer05> also python *is* optified afaik
Jul 08 13:42:09 <DocScrutinizer05> it comes with its own otification by bindmount
Jul 08 13:42:22 <Oksanaa> -rwxr-xr-x    1 root     root       1028632 May 18  2010 python2.5
Jul 08 13:42:34 <DocScrutinizer05> mount|grep py
Jul 08 13:43:27 <Oksanaa> lib python is optified, bin isn't
Jul 08 13:44:35 <DocScrutinizer05> yep, right:  df -h `which python2.5`
Jul 08 13:45:22 <Oksanaa> # ls -laSr /usr/bin/ | grep -v lrwx < the command I used for looking at not-optified /usr/bin
Jul 08 13:45:53 <DocScrutinizer05> hmmm
Jul 08 13:46:07 <DocScrutinizer05> why not find /usr/bin -type l -ls
Jul 08 13:48:24 <Oksanaa> find doesn't like the options, just prints out Usage
Jul 08 13:48:33 <DocScrutinizer05> err -type f
Jul 08 13:48:44 <DocScrutinizer05> meh, get decent tools!
Jul 08 13:48:49 <DocScrutinizer05> ~messybox
Jul 08 13:48:58 <infobot> messy... err busybox is meant for lean scripting. Regarding all the missing options and immanent limitations (see su, passwd, nice, ps, diff as used by mc...) 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
Jul 08 13:49:40 <DocScrutinizer05> apt-get findutils
Jul 08 13:50:08 <DocScrutinizer05> apt-get install findutils-gnu; even
Jul 08 13:50:44 <DocScrutinizer05> X-P  -->
Jul 08 13:50:46 <DocScrutinizer05> IroN900:~# apt-cache search findutil
Jul 08 13:50:47 <DocScrutinizer05> busybox - Tiny utilities for small and embedded systems
Jul 08 13:50:49 <DocScrutinizer05> findutils-gnu - utilities for finding files--find, xargs
Jul 08 13:51:14 <DocScrutinizer05> "Tiny SILLY FUBAR utilities..."
Jul 08 13:51:34 * Oksanaa has findutils-gnu installed
Jul 08 13:52:00 <DocScrutinizer05> then maybe try gfind /usr/bin -type f -ls
Jul 08 13:52:13 <DocScrutinizer05> or fix your PATH
Jul 08 13:54:08 <Oksanaa> Works, gfind. Output fairly same. So, x11vnc, qtoctave, and qtjs and gpg and some others
Jul 08 13:54:36 * Oksanaa finds the file list to be long
Jul 08 13:56:05 <Oksanaa> Could they be added to this /etc/maemo-optify-auto-opt.conf ? And when would it apply?
Jul 08 13:57:12 * Oksanaa would only try x11vnc and qtoctave, for start; do not recognise others well enough
Jul 08 13:59:02 <Oksanaa> osso_pdfviewer, modest and mediaplayer also catch eye
Jul 08 14:07:46 <DocScrutinizer05> ((Could they be added)) basically yes  ((when would it apply)) during next boot -- however it may render your device locked in a bootloop
Jul 08 14:10:24 * DocScrutinizer05 suggests to profile boot process to find all the binaries needed during early boot (before mounting /opt) and then doing a major sanitation of this whole mess: move needed packages /usr/*7* -> /*/*, move /usr -> eMMC
Jul 08 14:11:22 <DocScrutinizer05> I honestly had hoped for *NOKIA* doing that with pr1.3, but nooooo
Jul 08 14:13:25 * DocScrutinizer05 ponders to write maemo-antioptify* which does the exact opposite of optify: move all packages from /usr/?bin/* to /?bin/* that are _not_ subject to optification
Jul 08 14:14:39 <DocScrutinizer05> create symlinks for all those packages in the mointpoint dir of /usr, for best measure.
Jul 08 14:15:14 <DocScrutinizer05> once /usr gets mounted, those symlinks would get covered
Jul 08 14:16:02 <DocScrutinizer05> we could have similar symlinks in /usr/*/ for any scripts that use absolute paths to commands
Jul 08 14:18:57 <DocScrutinizer05> could all get handled during early boot, immediately before /usr gets mounted to its final mointpoint dir, in that maemo-antioptify-auto-noopt.sh script
Jul 08 14:20:31 <DocScrutinizer05> mount /usr /to /mnt/usr so you can copy the needed binaries from /user/?bin|lib|whatnot to rootfs and create symlinks in /mnt/usr/*/* and /usr/*/* 
Jul 08 14:20:42 <DocScrutinizer05> s/copy/move
Jul 08 14:21:21 <DocScrutinizer05> then remount eMMC usr from /mnt/usr to /usr
Jul 08 14:22:13 <DocScrutinizer05> finally delete antioptify-auto-noopt.sh (selfdestruct ;-D ) since it's never again needed
Jul 08 14:22:49 <DocScrutinizer05> replace maemo-optify-auto-opt.sh by a oneliner:  #!/bin/sh
Jul 08 14:23:03 <DocScrutinizer05> since that is also not needed anymore from now on
Jul 08 14:23:58 <DocScrutinizer05> optification my ass!
Jul 08 14:25:30 <DocScrutinizer05> I had a reason to add ""OMG - I wish they looked into FHS and moved /usr to eMMC"" to the ~optification factoid
Jul 08 14:34:26 <Oksanaa> Troublesome, especially from deb-dpkg-file-tracking point of view. Will these packages be easy to upgrade, after such filesystem movements?
Jul 08 18:02:49 * Sicelo is happy with the nokia battery he got for his N900
Jul 08 18:03:33 <Sicelo> LMD = 1573mAh 
Jul 08 19:05:42 <KotCzarny> sicelo: feelin awesome already?
Jul 08 19:14:45 <Sicelo> i'm happy .. i know i could have bought a better battery via ebay, but that's really out of my reach
Jul 08 22:57:42 <frals> someone pinged? ^^
Jul 08 23:10:27 <DocScrutinizer05> frals: yeah, somebody claimed to have a "not working fmms version". nevermind I'd say
Jul 08 23:11:41 <DocScrutinizer05> I didn't hear any question at all, anybody could answer
Jul 08 23:13:32 <DocScrutinizer05> frals: http://mg.pov.lt/maemo-irclog/%23maemo.2016-07-08.log.html#t2016-07-08T15:36:17
Jul 08 23:15:49 <DocScrutinizer05> http://mg.pov.lt/maemo-irclog/%23maemo.2016-07-08.log.html#t2016-07-08T15:46:12
Jul 09 02:03:47 <Linkandzelda> DocScrutinizer05: was able to fix the issue with /bin/dd vs the gnu path version with ssh login to user and root as well as xterm login. i added "source ~/.bashrc" into .profile for /home/user and /root/ now it executes a source on loading of each shell and the paths are set to use gnu before the others, no more busybox binaries or root; bash;
Jul 09 02:44:27 <Vajb> Sicelo: LMD = 3221mAh ;)
**** ENDING LOGGING AT Sat Jul 09 02:59:58 2016