**** BEGIN LOGGING AT Mon Apr 04 02:59:58 2016
Apr 04 03:11:24 <liuhoujiang> ls
Apr 04 06:12:18 <azkay> Anyone worked with http stuff in qt?
Apr 04 08:31:10 <ceene> i think there's gonna be a lot of n900 for sale at the end of 2016
Apr 04 08:34:31 <bencoh> :(
Apr 04 08:34:53 <ceene> i haven't taken a decision yet, but with coderus out of whatsapp development
Apr 04 08:35:04 <ceene> i don't think i'll be able to continue
Apr 04 08:35:19 <ceene> i think i could implement textsecure on our current version
Apr 04 08:35:24 <ceene> and it may work fine until 2016
Apr 04 08:35:40 <ceene> but after that, i don't even think development will be possible at all, with the deprecation of java based implementations (s40)
Apr 04 08:36:50 <Vajb> :(
Apr 04 08:40:25 <Vajb> is there any other working im client left after that?
Apr 04 08:40:39 <ceene> there's telegram
Apr 04 08:40:41 <ceene> jabber
Apr 04 08:40:44 <ceene> via telepathy
Apr 04 08:40:56 <ceene> there's pidgin, although i haven't found it to be very stable
Apr 04 08:42:14 <Vajb> yes i have pidgin. Tried to get fb chat working via it, but didn't try anything else
Apr 04 09:33:54 <auenf> people still use whatsapp?
Apr 04 09:38:29 <bencoh> some do, yeah
Apr 04 10:21:22 <Maxdamantus> thatwasp
Apr 04 10:21:36 <Maxdamantus> Wait, that's not right.
Apr 04 11:33:07 <Wizzup> Is it possible to force the usb to host mode in the n900 kernel? I think I set the kernel to host mode, but the dts needs a change i think
Apr 04 11:33:18 <Wizzup> probably in usb_otg_hs node, value     mode = <2>;
Apr 04 11:33:26 <Wizzup> but I'm not sure what other mode I should try
Apr 04 12:03:49 <DocScrutinizer05> I massively doubt that dirty hack we invented for hostmode in N900 will *ever* go upstream
Apr 04 12:04:26 <DocScrutinizer05> it also will never be useful for any other device
Apr 04 12:07:33 <jon_y> there is no detection mechanism?
Apr 04 12:08:24 <DocScrutinizer05> not any useful one for the MUSB core
Apr 04 12:08:42 <jon_y> hardware limitation?
Apr 04 12:08:46 <DocScrutinizer05> yes
Apr 04 12:08:53 <jon_y> ok
Apr 04 12:09:26 <jon_y> I wish everything was upstreamed, get the latest linux 4.6-rc2 on the phone :)
Apr 04 12:09:42 <DocScrutinizer05> you can't get the MUSB core to enter hostmode when the signal "ID grounded" doesn't come up the ULPI from PHY
Apr 04 12:10:17 <jon_y> well, at least get nice things like F2FS for flash storage
Apr 04 12:10:31 <DocScrutinizer05> N900 PHY doesn't support the software emulation mode for that signal
Apr 04 12:11:30 <DocScrutinizer05> thus N900 never can have true hostmode, I implemented a emulation of hostmode, sort of
Apr 04 12:11:54 <DocScrutinizer05> with creative abuse of the MUSB core debug modes
Apr 04 12:13:56 <DocScrutinizer05> Wizzup: so your answer is: it's not possible to force the USB to hostmode in N900 (hardware)
Apr 04 12:15:32 <jon_y> forever stuck on 2.6.28
Apr 04 12:16:13 <DocScrutinizer05> well, you can port H-E-N kernel patches to 4.x
Apr 04 12:16:17 <jon_y> I can't remember if n900 even does ext4
Apr 04 12:16:24 <DocScrutinizer05> it does
Apr 04 12:16:31 <jon_y> oh ok
Apr 04 12:16:45 <jon_y> ext4 is pretty all-rounder even on flash
Apr 04 12:17:16 <DocScrutinizer05> (H-E-N) but don't hold your breath for that dirty hack ever going upstream
Apr 04 12:17:17 <jon_y> lemme guess, camera and wifi doesn't work?
Apr 04 12:17:27 <bencoh> wifi does iirc
Apr 04 12:17:33 <bencoh> camera broke at some point
Apr 04 12:17:42 <jon_y> ok, how about telephony?
Apr 04 12:17:51 <bencoh> http://elinux.org/N900
Apr 04 12:18:17 <jon_y> put grsec in it for fun
Apr 04 12:18:30 <DocScrutinizer05> telephony is a particularly sad topic on N900
Apr 04 12:18:50 <DocScrutinizer05> in kernel update context
Apr 04 12:19:39 <jon_y> opengl is alo never going to work on mainline?
Apr 04 12:19:40 <DocScrutinizer05> even when we might get a working cmtspeech device, the complete control of modem is basically out of sight afaik
Apr 04 12:20:48 <jon_y> feature page also says alsa isn't upstreamed yet?
Apr 04 12:20:54 <DocScrutinizer05> though ofono and FSO did implement the needed ISI command stack so it *might* work when somebody ports that stuff
Apr 04 12:21:45 <bencoh> jon_y: http://natisbad.org/N900/n900-custom-kernel.html
Apr 04 12:22:15 <bencoh> DocScrutinizer05: actually, telephony isn't that bad on the kernel side
Apr 04 12:22:34 <jon_y> nice
Apr 04 12:23:06 <jon_y> feature wise, how does power5x compare to grsec version?
Apr 04 12:23:09 <bencoh> DocScrutinizer05: kernel drivers are here, and some people hacked userland a bit to have it work
Apr 04 12:24:15 * jon_y doing kernel work as a job irl
Apr 04 12:25:26 <DocScrutinizer05> if you're looking for a domain to work on: camera and freq-governor
Apr 04 12:25:47 <DocScrutinizer05> clock generation at large
Apr 04 12:26:09 <jon_y> unfortunately, I doubt PCI(e) stuff is applicable for phones
Apr 04 12:26:25 <DocScrutinizer05> camera is tricky since the bus is muxed between two sensors
Apr 04 12:26:59 <bencoh> DocScrutinizer05: you should talk to pali and freemangordon 
Apr 04 12:27:01 <bencoh> woops
Apr 04 12:27:03 <bencoh> jon_y: ^
Apr 04 12:27:08 <jon_y> :)
Apr 04 12:27:30 <jon_y> I know next to nothing for ARM architecture
Apr 04 12:27:36 <jon_y> or OMAP3
Apr 04 12:27:52 <bencoh> if you're used to kernel dev I'm sure this wont be a big deal :)
Apr 04 12:28:02 <jon_y> k
Apr 04 12:28:49 <jon_y> whats the best way to debug code if you don't have a guinea pig n900?
Apr 04 12:30:41 <bencoh> good question, I've never dived into it because I only have one as well
Apr 04 12:31:26 <jon_y> I'm pretty impressed with grsec, using it on my workstation in the office
Apr 04 12:31:53 <jon_y> especially runtime disabling of USB
Apr 04 12:32:01 <jon_y> badusb is bad
Apr 04 12:32:18 <DocScrutinizer05> jon_y: qemu
Apr 04 12:32:36 <DocScrutinizer05> iirc even bootloader works in qemu
Apr 04 12:32:38 <jon_y> last I checked n900 qemu wasn't mainlined
Apr 04 12:32:44 <jon_y> or was it
Apr 04 12:33:07 <DocScrutinizer05> ask freemangordon
Apr 04 12:33:22 <DocScrutinizer05> he and pali do that all the time
Apr 04 12:33:22 <jon_y> freemangordon: is it in mainline qemu?
Apr 04 12:33:53 <jon_y> not sure what feature can be developed for it
Apr 04 12:34:22 <DocScrutinizer05> [2016-03-27 Sun 19:55:48] <freemangordon> with that change there is no more error in qemu, going to test on the device
Apr 04 12:44:17 <zGrr> moin 
Apr 04 12:58:33 <Wizzup> DocScrutinizer05: I don't think your answer is relevant for mainline
Apr 04 12:58:39 <Wizzup> I didn't mention that I was working on mainline
Apr 04 12:58:46 <Wizzup> but I don't think your answer is correct or relevant here
Apr 04 12:59:18 <Wizzup> bbiab.
Apr 04 13:03:36 <DocScrutinizer05> Wizzup: which answer? the answer that N900 *cannot* do hostmode? that's relevant for each N900 no matter which OS
Apr 04 13:04:23 <DocScrutinizer05> H-E-N is a dirty hack and about as close to hostmode as it gets, but it's not a correct complete hostmode
Apr 04 13:04:57 <DocScrutinizer05> I should know, I "invented" it
Apr 04 13:05:17 <jon_y> file a patent :)
Apr 04 13:05:33 <DocScrutinizer05> patent how to shoot your own foot?
Apr 04 13:06:02 <jon_y> sure, it is a patent-able concept
Apr 04 13:06:16 <DocScrutinizer05> patent how to drive in a car where manuf forgot to mount 4th wheel?
Apr 04 13:06:38 <Wizzup> DocScrutinizer05: there is a musb controller
Apr 04 13:06:42 <ceene> seems like a pretty useful thing indeed
Apr 04 13:06:48 <Wizzup> and it seems like you can do hostmode with h-e-n
Apr 04 13:06:48 <DocScrutinizer05> Wizzup: so what?
Apr 04 13:06:54 <DocScrutinizer05> no
Apr 04 13:06:58 <Wizzup> so if you the usb node into host mode in the kernel, and set it, it should work
Apr 04 13:07:01 <Wizzup> Pali also said it worked.
Apr 04 13:07:09 <DocScrutinizer05> you can't
Apr 04 13:07:10 <Wizzup> I just want to reproduce what he did
Apr 04 13:07:29 <DocScrutinizer05> OHMY!! pali did what I shown him
Apr 04 13:07:38 <DocScrutinizer05> it's FAKE!
Apr 04 13:07:50 <DocScrutinizer05> this is no hostmode, this is a dirty hack
Apr 04 13:07:54 <Wizzup> Please stop.
Apr 04 13:08:01 <Wizzup> This is not productive to my original question at all
Apr 04 13:08:06 <DocScrutinizer05> aha
Apr 04 13:08:20 <Wizzup> I wasn't asking about the maemo kernel hack
Apr 04 13:08:32 <DocScrutinizer05> I tld you N900 cannot do hostmode, so that H-E-N hack I invented will never go upstream
Apr 04 13:08:48 <Wizzup> I would agree that it cannot do otg
Apr 04 13:09:11 <DocScrutinizer05> and there's no other way than H-E-N hack to get anything remotely resembling hostmode on N900
Apr 04 13:10:00 <Wizzup> musb has been changed a lot for many socs, and the device tree nodes allow you to specify the mode in which musb code operates
Apr 04 13:10:13 <Wizzup> I would be very surprised if it is not possible, but I'll wait for Pali to pop by
Apr 04 13:10:19 <Wizzup> IIRC he said it worked for him some time ago
Apr 04 13:10:21 <DocScrutinizer05> the MUSB core hardware stae machine doesn't enter hostmode unless signalled a "ID-pin grounded" from PHY. N900 phy can't do that
Apr 04 13:10:30 <Wizzup> You can easily wire that up in the dts I believe.
Apr 04 13:10:36 <Wizzup> I also hacked that together in olimex boards
Apr 04 13:11:03 <DocScrutinizer05> again, Pali just re-implemented H-E-N which been invented and developed by a team lead by me
Apr 04 13:11:05 <Wizzup> either way, I have to go back to some meetings
Apr 04 13:11:16 <DocScrutinizer05> no you can't
Apr 04 13:11:19 <Wizzup> do you realise I am only talking about mainline?
Apr 04 13:11:32 <DocScrutinizer05> do you realize I only talk about N900
Apr 04 13:11:37 * Wizzup sighs
Apr 04 13:11:41 <DocScrutinizer05> HARDWARE
Apr 04 13:11:41 <Wizzup> I'll wait for pali to come by
Apr 04 13:11:45 <Wizzup> Yes, I read that
Apr 04 13:11:56 <DocScrutinizer05> do whatever you want, pali will tell you to ask me
Apr 04 13:12:10 <Wizzup> I don't care if the 'hen hack' can be mainlined or not, it still seems very doable from sw, as the 'hack' proved
Apr 04 13:13:42 <DocScrutinizer05> it IS A HACK that does no real hostmode
Apr 04 13:14:37 <DocScrutinizer05> yes it could work on all mentorgraphics MUSB cores, but it can't do speed negotiation and automatic ENUM
Apr 04 13:14:59 <DocScrutinizer05> that's a hw limitation of MG MUSBcore
Apr 04 13:15:21 <Wizzup> ack
Apr 04 13:15:53 <DocScrutinizer05> and for proper hw design that has ID pin wired (or knows the software emu for "ID grounded" in PHY) it's absolutely useless
Apr 04 13:16:50 <DocScrutinizer05> anyway you won't force MUSB core into hostmode by any software means other than via PHY emulation of "ID grounded"
Apr 04 13:17:29 <DocScrutinizer05> what H-E-N does  is debug mode which is a pathetic fake of a hostmode
Apr 04 13:17:55 <Wizzup> yes, but that debug entry is no longer available
Apr 04 13:18:00 <Wizzup> I think, in mainline kernel
Apr 04 13:18:03 <Wizzup> what is what I wanted to ask Pali
Apr 04 13:18:08 <Wizzup> what -> that
Apr 04 13:20:11 <DocScrutinizer05> it never will go upstream I guess
Apr 04 13:20:50 <DocScrutinizer05> it's way too dirty and hacky and fringe case only applicable to one particular device
Apr 04 13:21:19 <DocScrutinizer05> and it doesn't belong into kernel really, unless you can't do proper hostmode
Apr 04 13:22:14 <DocScrutinizer05> so this is a specific N900 kernel hack
Apr 04 13:22:57 <DocScrutinizer05> iirc *particularly* Pali's implemetation that's way more kernel-centric than mine which does most stuff in userland
Apr 04 13:24:45 <DocScrutinizer05> Pali's hostmode tries to automate what user does in MohammadAG's H-E-N control gui with my kernel hack
Apr 04 13:25:09 <DocScrutinizer05> iirc
Apr 04 13:25:15 <DocScrutinizer05> never looked into it
Apr 04 13:25:29 <DocScrutinizer05> in RL
Apr 04 13:26:05 <DocScrutinizer05> here you are: http://talk.maemo.org/showthread.php?p=685367#post685367
Apr 04 13:47:15 <kerio> yeah but surely we could get the debug entry in mainline
Apr 04 13:47:19 <kerio> without *doing* stuff with it
Apr 04 13:47:31 <kerio> so that a clean-ish implementation of H-E-N would work
Apr 04 13:48:37 <jon_y> /sys/kernel/debug/hen/mode? :)
Apr 04 14:00:44 <DocScrutinizer05> kerio: ack
Apr 04 14:01:42 <DocScrutinizer05> that would basically be the more early implementations of original H-E-N kernel patches
Apr 04 14:03:04 <DocScrutinizer05> though there are still a few hacks that are unrelated, since the H-E-N hostmode uses OTG software which for example explicitly forbids USB hubs in upstream version, without my dirty hacks
Apr 04 14:03:24 <DocScrutinizer05> same for power negotiation during ENUM
Apr 04 14:04:02 <DocScrutinizer05> actually that's completely unsolved/neglected in H-E-N, we simply shut it off
Apr 04 14:05:02 <DocScrutinizer05> OTG specs say that devices negotiate a power requirement of iirc up to 100mA(!)
Apr 04 14:05:21 <DocScrutinizer05> N900 hostmode has that disabled
Apr 04 14:05:49 <DocScrutinizer05> it simply ignores any power requirement announced by attached peripheral
Apr 04 14:06:33 <DocScrutinizer05> ...since you (driver) don't know about external VBUS supplies via Y-cable anyway
Apr 04 14:07:16 <DocScrutinizer05> so that's actually pretty impossible to solve in kernel space, and even a hack massively breaks the specs of USB
Apr 04 14:08:02 <DocScrutinizer05> in H-E-N booston script takes care about that stuff
Apr 04 14:08:25 <DocScrutinizer05> it checks if booster has an errorcondition
Apr 04 14:08:47 <DocScrutinizer05> like overcurrent or batery undervoltage or overheating
Apr 04 14:10:53 <DocScrutinizer05> booster however is in power/regulators domain, not USB domain, so in kernel you get nasty dependencies and possibly even layering conflicts
Apr 04 14:13:40 <kerio> if only nokia didn't drop OTG just to release early ._.
Apr 04 14:13:48 <DocScrutinizer05> indeed
Apr 04 14:13:58 <kerio> also if only they didn't fuck up the musb port solder
Apr 04 14:14:23 <DocScrutinizer05> well, they did a hw choice with PHY1707 which isn't made for hostmode
Apr 04 14:14:54 <DocScrutinizer05> and that was afaik necessary because of a SiErr in TWL4030
Apr 04 14:15:03 <DocScrutinizer05> which fucked up charging
Apr 04 14:16:09 <DocScrutinizer05> so s/TWL4030 PHY/1707 PHY/  --> no sw ID pin handling  --> no proper hostmode
Apr 04 14:17:00 <kerio> :(
Apr 04 14:17:10 <kerio> well they could've found something else
Apr 04 14:17:11 <DocScrutinizer05> prolly you could port forward a lot of needed kernel bits for hiostmode from N810 kernel
Apr 04 16:09:10 <freemangordon> jon_y: no, you need linaro version
Apr 04 16:10:29 <freemangordon> jon_y: it is a bit tricky to install it on newer distros, but I use it here on Ubuntu 14.04 so it is not impossible
Apr 04 16:26:19 <xes> is anyone able to decode PDU data in /var/spool/sms/incoming/ blobs? Do we have any tool for this in maemo's repo?
Apr 04 16:36:15 <DocScrutinizer05> xes: I once wrote such tool, nfc where it went to
Apr 04 16:39:23 <DocScrutinizer05> http://maemo.cloud-7.de/maemo5/usr/local/bin/smscb.py
Apr 04 16:39:32 <DocScrutinizer05> o/ afk
Apr 04 16:43:47 <xes> DocScrutinizer05: thanks
Apr 04 17:02:12 <xes> DocScrutinizer05: ^^ this listens and decodes the cell brodcast arrays of hex. Isn't it?
Apr 04 19:18:02 <Pali> hi! if somebody uses wpa-eap/enterprise on n900 I have something interesting!
Apr 04 19:18:08 <Pali> fixed autoconnect support!
Apr 04 19:18:12 <Pali> it is in cssu-devel: https://talk.maemo.org/showthread.php?p=1502805#post1502805
Apr 04 19:49:39 <DocScrutinizer05> xes: yes
Apr 04 19:49:59 <DocScrutinizer05> PDUs are supposed to be similar though
Apr 04 19:50:26 <DocScrutinizer05> CB is multicast SMS
Apr 04 19:50:53 <DocScrutinizer05> Pali: \o/
Apr 04 19:51:25 <Wizzup> freemangordon, Pali: great to see the audio problem being fixed in some form (saw the patch)
Apr 04 19:53:19 <DocScrutinizer05> xes: CB has a channel number though
Apr 04 19:53:39 <DocScrutinizer05> you want to ignaore that for plain SMS, I guess
Apr 04 19:54:26 <DocScrutinizer05> there's also some message type which I dunno if I took care about that
Apr 04 19:55:28 <DocScrutinizer05> anyway the plain PDU 8/7 encoding is similar
Apr 04 19:55:48 <DocScrutinizer05> rather: identical
Apr 04 20:00:33 <Pali> Wizzup: hm? which patch?
Apr 04 20:00:45 <Wizzup> Pali: "Nokia N900 - audio TPA6130A2 problems"
Apr 04 20:00:49 <Pali> ah, right
Apr 04 20:01:05 <Wizzup> p.s. have you gotten ofono to work? and/or the modem/3g?
Apr 04 20:01:19 <Wizzup> I've been struggling to get data to work on mainline, but that is not using maemo, but gentoo rather
Apr 04 20:01:20 <Pali> ofono should work
Apr 04 20:01:36 <Pali> pavelm already uses voice calls...
Apr 04 20:02:13 <Wizzup> I saw that. But didn't see how he did it.
Apr 04 20:02:18 <Wizzup> I
Apr 04 20:02:24 <Wizzup> I'd like to try it as well.
Apr 04 20:03:13 <Pali> maybe you can contant him and ask
Apr 04 20:03:16 <Wizzup> ACK
Apr 04 20:10:17 <DocScrutinizer05> xes: http://wiki.maemo.org/Phone_control#Send_SMS  http://talk.maemo.org/showpost.php?p=558430&postcount=57
Apr 04 20:12:50 <bencoh> Wizzup: you can find some (scarse) info on his blog and his libcmtspeech repository
Apr 04 20:14:42 <freemangordon> Wizzup: that audio problem exists only when device is rebooted from stock kernel, so it is not really a show-stopper
Apr 04 20:14:56 <Wizzup> yes, I know, just happy to see the progress
Apr 04 20:15:36 <freemangordon> yeah. the next one going to be solved is that device cannot be rebooted/powered down
Apr 04 20:15:48 <bencoh> Wizzup: https://www.gitorious.org/libcmtspeechdata/libcmtspeechdata.git/
Apr 04 20:16:10 <freemangordon> bencoh: hmm, not really :)
Apr 04 20:16:23 <bencoh> freemangordon: not really what?
Apr 04 20:16:37 <Wizzup> freemangordon: That doesn't work? I think it works for me
Apr 04 20:16:46 <freemangordon> bencoh: https://github.com/community-ssu/cmtspeech
Apr 04 20:17:02 <Wizzup> bencoh: ack - thanks
Apr 04 20:17:14 <bencoh> freemangordon: yeah that's "ours", but looks like he hacked it a bit
Apr 04 20:17:56 <freemangordon> could be, but "ours" is as close to what we have in fremantle as possible. and this is supposed to work with bb5 in n900
Apr 04 20:18:15 <Wizzup> bb5?
Apr 04 20:19:20 <freemangordon> Wizzup: the modem in n900
Apr 04 20:19:25 <Wizzup> ack
Apr 04 20:20:55 <Wizzup> I still have toruble probing n900-modem successfully
Apr 04 20:21:06 <Wizzup> but I'll search for some doc first
Apr 04 20:25:45 <bencoh> I still wonder why they didn't choose ofono for maemo
Apr 04 20:25:52 <Wizzup> Maybe it wasn't ready at the time
Apr 04 20:26:16 <bencoh> especially considering http://www.remlab.net/op/maemo2009.pdf
Apr 04 20:27:54 <bencoh> (hmm this thing is actually pretty nice to get the big picture of how telephony works on maemo)
Apr 04 20:33:27 <DocScrutinizer05> freemangordon: >>that audio problem exists only when device is rebooted from stock kernel // cannot be rebooted/powered down<< now that's sounding interesting for me :-D
Apr 04 20:33:46 <DocScrutinizer05> both
Apr 04 20:34:03 <DocScrutinizer05> both sound like sticky states in hw
Apr 04 20:34:28 <freemangordon> DocScrutinizer05: no, both are bugs in mainline kernel :)
Apr 04 20:34:40 <DocScrutinizer05> first maybe modem PCN interface initialization, second about boot
Apr 04 20:34:50 <DocScrutinizer05> PCM
Apr 04 20:35:47 <DocScrutinizer05> well, a audio bug that only appears when you warmboot into mainline kernel from stock kernel is for sure a kernel bug, but a very inbteresting one
Apr 04 20:36:06 <freemangordon> DocScrutinizer05: (second one) http://www.spinics.net/lists/linux-usb/msg138479.html
Apr 04 20:36:10 <DocScrutinizer05> ta
Apr 04 20:36:52 <freemangordon> DocScrutinizer05: (audio bug) https://lkml.org/lkml/2016/3/26/22
Apr 04 20:37:06 <DocScrutinizer05> double-ta
Apr 04 20:37:46 <DocScrutinizer05> (first one) so the device doesn't start up again after shutdown, but it doesn't really finsih shutdown?
Apr 04 20:38:15 <freemangordon> it doesn't finish the shutdown, because there is a kernel oops in the process
Apr 04 20:38:16 <DocScrutinizer05> then it's off my radar of interesting bugs
Apr 04 20:46:07 <jonwil> hi
Apr 04 20:47:48 <freemangordon> jonwil: hi!
Apr 04 20:48:38 <jonwil> Any interesting news in the n[eo]900 community?
Apr 04 20:49:54 <DocScrutinizer05> hi jonwil! long time no see :-)
Apr 04 20:50:28 <DocScrutinizer05> jonwil: maybe interesting: Devuan seems very interested in Neo900
Apr 04 20:51:12 <jonwil> Devuan is?
Apr 04 20:51:14 <DocScrutinizer05> jonwil: plus the "usual" good news, like Pali and freemangordon constantly improving kernel support
Apr 04 20:51:23 <DocScrutinizer05> devuan.org
Apr 04 20:51:27 <jonwil> ok
Apr 04 20:51:29 <jonwil> good to know kernel is getting better
Apr 04 20:51:56 <DocScrutinizer05> /join #devuan
Apr 04 20:52:06 <kerio> jonwil: devuan is debian with sysvinit-core
Apr 04 20:54:54 <DocScrutinizer05> no, devuan is debian without systemd cancer
Apr 04 20:55:14 <DocScrutinizer05> with freedom of choice instead
Apr 04 20:55:40 <freemangordon> Pali: while we are at it, is there anything else to be fixed in mainline (besides that div-by-zero I cannot pull myself together to work on)?
Apr 04 20:56:01 <kerio> Pali: why can't uboot boot a kernel+initrd without the uboot header? :(
Apr 04 20:58:56 <Pali> freemangordon: non working sre's version of bluetooth driver :-)
Apr 04 20:59:00 <Pali> or camera/isp
Apr 04 20:59:11 <Pali> (front camera is already in mainline)
Apr 04 20:59:13 <DocScrutinizer05> IOW devuan is what maemo must get based on, now that debian is systemd-infested
Apr 04 21:00:11 <kerio> why not openbsd
Apr 04 21:00:12 <freemangordon> Pali: hmm, maybe upgrade github to 4.6-rc2 without sre's patches, so if it doesn;t work we will know who to blame
Apr 04 21:00:50 <freemangordon> Pali: as I fixed video-bus-switch driver and both cameras give green only output
Apr 04 21:01:25 <freemangordon> but I am not sure if the problem is not in sre's patches
Apr 04 21:02:14 <Pali> I do not want to remove those patches from tree so we do not lost them
Apr 04 21:02:23 <Pali> but you can easily run git rebase -i
Apr 04 21:02:31 <Pali> and remove them from your local repo
Apr 04 21:02:35 <freemangordon> you won;t lose them they are in the other branches
Apr 04 21:03:00 <freemangordon> Pali: wasn't pavel working on BT driver?
Apr 04 21:03:09 <Pali> not anymore
Apr 04 21:03:18 <freemangordon> oh. any idea why?
Apr 04 21:03:20 <Pali> his driver was removed from mainline
Apr 04 21:03:36 <Pali> I still have copy in -n900 branches
Apr 04 21:03:43 <Pali> that version is working
Apr 04 21:03:56 <Pali> but not acceptable for mainline anymore
Apr 04 21:03:56 <freemangordon> any idea wht needs to be done for upstreaming?
Apr 04 21:04:05 <Pali> using sre version :-)
Apr 04 21:04:13 <freemangordon> which doesn't work?
Apr 04 21:04:16 <freemangordon> :D
Apr 04 21:04:22 <Pali> that uses in kernel omap uart driver
Apr 04 21:04:37 <Pali> basically it is rewritten to use "correctly" bluetooth api
Apr 04 21:04:43 <freemangordon> is it in mainline?
Apr 04 21:04:49 <Pali> sre's version not
Apr 04 21:05:01 <freemangordon> any clue if he is going to send patches?
Apr 04 21:05:02 <Pali> sre has it in his tree
Apr 04 21:05:07 <Pali> not going
Apr 04 21:05:17 <Pali> as he did not touched for more months
Apr 04 21:05:20 <Pali> now hacking n950
Apr 04 21:05:27 <Pali> not n900
Apr 04 21:05:35 <freemangordon> hmm, maybe we should ask him what are his plans
Apr 04 21:06:34 <freemangordon> anyway, /me is going afk, night
Apr 04 21:13:19 <DocScrutinizer05> is it a question of hiring him to continue his N900 work?
Apr 04 21:15:45 <jonwil> Pali: Where did that libicd-network-wlan code/git repo come from?
Apr 04 21:18:16 <Pali> from gitorious or garage
Apr 04 21:23:23 <jonwil> Time to update some of these packages on my N900 :)
Apr 04 21:25:31 <jonwil> btw it seems like my new set of root CAs in maemo-security-certman is working great
Apr 04 21:26:12 <bencoh> do you have an automated build process for that?
Apr 04 21:26:17 <jonwil> Only issue is people who have wrong certificates stored in the microb certificate store and are unable to simply wipe the certificate store like I did (i.e. they have manually added certs)
Apr 04 21:26:47 <DocScrutinizer05> diff?
Apr 04 21:26:49 <jonwil> I have a set of steps that could (in theory) be automated
Apr 04 21:27:21 <jonwil> I pushed the commits to CSSU Git a couple months ago
Apr 04 21:27:23 <ecc3g> clear
Apr 04 21:27:51 * ecc3g finds a /
Apr 04 21:28:09 <DocScrutinizer05> whatever that means
Apr 04 21:29:15 <DocScrutinizer05> jonwil: could you do a (sort|)diff between user cert store and devault stock cert store, to find the certs user added manually and insert them to the new store in post-install?
Apr 04 21:30:01 <jonwil> There is probably a way to remove the bogus certificates somehow but I haven't found out how
Apr 04 21:30:30 <DocScrutinizer05> hey, I'm sure somebody can help with that
Apr 04 21:30:56 <jonwil> The NSS people haven't been able to help really
Apr 04 21:30:59 <DocScrutinizer05> if not (maybe it's broken), go the dif&add path
Apr 04 21:31:31 <jonwil> Basically the problem we have is this:
Apr 04 21:32:10 <jonwil> When you visit site xyz on the stock set of certificates, it gets sent a certain certificate for a particular CA and stores it in the certificate database for some reason
Apr 04 21:32:43 <jonwil> Then when you visit the same site on the new set of certificates, it gets a totally different certificate chain but the bogus stored certificate causes it to fail
Apr 04 21:33:18 <jonwil> So we need a way to loop all the certificates in the certificate store, identify these automatically cached/added certificates and remove them without removing anything manually added
Apr 04 21:33:28 <jonwil> any kind of diffing wont help here
Apr 04 21:33:51 <kerio> jonwil: just delete the whole shebang
Apr 04 21:33:56 <kerio> ez
Apr 04 21:34:14 <jonwil> but then people loose certificates they manually added and need to keep
Apr 04 21:34:22 <jonwil> and may not be able to re-add for various reasons
Apr 04 21:34:45 <kerio> what was the issue again?
Apr 04 21:34:49 <kerio> miscached intermediates?
Apr 04 21:34:54 <jonwil> yes
Apr 04 21:35:00 <kerio> causing which error, exactly?
Apr 04 21:37:08 <jonwil> I cant remember off the top of my head
Apr 04 21:37:21 <jonwil> speaking of which, there is a new set of mozilla root CAs out which I will import into CSSU git now :)
Apr 04 21:38:48 <kerio> can we get some chacha20-poly1305 up in this
Apr 04 21:39:30 <kerio> idk how updateable NSS by itself is
Apr 04 21:39:35 <kerio> probably not much
Apr 04 21:41:39 <jonwil> It should be possible to update it, we have code for all the things that talk to it as far as I know
Apr 04 21:41:45 <jonwil> so we can make changes if we need dto
Apr 04 21:41:47 <jonwil> need to
Apr 04 21:42:01 <DocScrutinizer05> http://dowse.eu/
Apr 04 22:01:33 <jonwil> ok, newest mozilla root CA certs are in cssu git now :)
Apr 04 22:13:40 <Wizzup> :)!
Apr 04 22:27:03 <jon_y> freemangordon: what are the complications with using linaro?
Apr 04 22:27:20 <jon_y> I'm familiar with rebuilding toolchains
Apr 04 22:28:42 <jonwil> damn, I just remembered why I hate compiling OpenSSL. It takes hours to do it :(
Apr 04 22:34:40 <xes> ~~
Apr 04 22:35:12 <xes> hours?
Apr 04 22:38:39 <kerio> 10.77s here
Apr 04 23:45:10 <DocScrutinizer51> I guess joga is building on target
Apr 04 23:45:23 <DocScrutinizer51> jonwil:
Apr 04 23:46:03 <DocScrutinizer51> xchat nick completion sucks
Apr 04 23:53:39 <jonwil> no, I am not building on target, I am building on my scratchbox dev vm
Apr 04 23:53:55 <jonwil> it didn't take as long as I thought it would
Apr 04 23:54:10 <jonwil> the shuny new Core i5 I bought a few months ago may have helped there...
Apr 04 23:57:48 <jonwil> ok, thats got the latest versions of the CSSU git repos I care about onto my phone :)
Apr 05 00:10:48 <DocScrutinizer05> n8
**** ENDING LOGGING AT Tue Apr 05 02:59:58 2016