**** BEGIN LOGGING AT Thu Jan 24 02:59:59 2013 Jan 24 06:30:01 Hehe: http://www.engadget.com/2013/01/24/canonical-ubuntu-smartphone-core-apps/ Jan 24 06:30:34 * RST38h wonders when the "you develop - we say thanks and sell" business model has become so popular Jan 24 06:39:47 suckers Jan 24 06:40:41 became popular when Shuttleworth decided to order his Canonical to build that me-too-windows Jan 24 06:40:57 Hasn't THAT started with Gnome though? Jan 24 06:41:07 as if Canonical had invented linux desktop Jan 24 06:41:40 RST38h: gnome wasn't always awful Jan 24 06:41:41 the me-too-windows, yes. That's been g*UULP Jan 24 06:41:57 khm: I do not mean awful, I mean me-too-windowsness part Jan 24 06:41:59 ingo molnar brought the microsoft cloning mission Jan 24 06:42:24 Doc: Canonical did invent the first practically usable linux desktop Jan 24 06:42:38 no, not ingo Jan 24 06:42:47 ~lart RST38h Jan 24 06:42:48 * infobot says "boot to the head" and knocks RST38h over Jan 24 06:42:52 Doc: Without having to endlessly fiddle with it and/or squint at twm windows Jan 24 06:42:57 miguel de icaza Jan 24 06:43:03 he's the microsoft follower Jan 24 06:43:19 RST38h: also not really true. there were a few before ubuntu. Jan 24 06:43:24 * DocScrutinizer05 used KDE when it was still Jan 24 06:43:26 khm: Not really Jan 24 06:43:27 ubuntu just had the money to market it Jan 24 06:43:44 OK and gnome never been heard of outside those mothern countries Jan 24 06:43:44 * RST38h tried many linux distros. Ubuntu was the firstone that installed and worked. Jan 24 06:43:51 RST38h: yes, really. Jan 24 06:43:53 ok Jan 24 06:44:05 I guess we'll just take your experience as global fact Jan 24 06:44:13 s/mothern/northern/ Jan 24 06:44:14 DocScrutinizer05 meant: OK and gnome never been heard of outside those northern countries Jan 24 06:44:20 khm: My experience is the basis of my personal opinion Jan 24 06:44:35 thanks for letting me know Jan 24 06:44:49 meanwhile, other people had different experiences. Jan 24 06:44:56 khm: You can, of course, argue against it, but it is usually not fruitful Jan 24 06:45:22 khm: Yeah, I am familiar with linuxoids' "it works for me, and your hands are growing out of your ass" clause Jan 24 06:45:24 of course not. I have nothing to gain by showing you that you're wrong. Jan 24 06:45:33 SuSE installed just fine since Rev6 here, maybe since Rev5 Jan 24 06:45:37 with KDE Jan 24 06:45:37 now you're attacking a straw man. Jan 24 06:46:02 Doc: I have been getting official boxed SuSE kits from SuSE itself since version 1 Jan 24 06:46:29 Doc: NONE of them EVER worked on my laptop(s). And I even know why, but it does not make things right. Jan 24 06:46:45 RST38h: there never been a ver1 Jan 24 06:46:50 DocScrutinizer05: RST38h has used every piece of software released by every human being since the Altair 8800 Jan 24 06:46:55 they started with 4.4 iirc Jan 24 06:47:09 Doc: On the desktop, SuSE worked but required about 2-3 times more fiddling than FreeBSD, for Tentacled's sake Jan 24 06:47:21 what he is saying now is the absolute unfiltered truth and anyone who disagrees with him is a lying elitist terrorist. Jan 24 06:47:31 Doc: Well, whatever first boxed edition they started selling. Still have 5-6of those boxessomewhere Jan 24 06:47:34 shhh kids Jan 24 06:47:39 anyway Jan 24 06:47:41 the point is Jan 24 06:47:55 ubuntu wasn't the first, and my favorite part of gnome will always be gconf Jan 24 06:48:06 they loved the windows registry so much they made their own Jan 24 06:48:11 and rubbed some xml on for good measure Jan 24 06:48:28 indeed Jan 24 06:48:38 Doc: [do abstain from starting me on Redhat or some other abomination, as I got to enable Redhat bootup on some upcoming Intel chips] Jan 24 06:49:11 RST38h: please post a complete list of things nobody should mention Jan 24 06:49:36 shhh kids Jan 24 06:50:00 daddy had no breakfast coffee yet Jan 24 06:50:12 don't you have an iv? Jan 24 06:50:54 Moo, MohammadAG Jan 24 06:51:14 MohammadAG: It Lives!!!! Jan 24 06:51:17 :-D Jan 24 06:53:53 philhug: thanks for your mail Jan 24 06:54:59 SpeedEvil: could you please find that link about the voltage/Amperes and waveshape for that electric painkiller again? Jan 24 06:55:05 * DocScrutinizer05 needs it Jan 24 07:45:19 talk.maemo.org. 26619 IN A 188.117.59.209 Jan 24 07:50:30 any one who uses dyndns on the n900? Jan 24 08:00:11 yay Jan 24 08:00:31 interesting idea, never occurred to me Jan 24 08:02:13 as long as you are not behind NAT, it would be nice Jan 24 08:30:04 Doc: So, what is the current deal about maemo repositories? The official one no longer answers. Jan 24 08:30:11 Is there a replacement now? Jan 24 08:30:22 ~mirrors Jan 24 08:30:22 mirror is, like, http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 24 08:30:33 there's a TMO post which explains all the things Jan 24 08:30:46 Is there a PERMANENT repository now? Jan 24 08:30:56 Not a mirror of the old repository, I mean? Jan 24 08:31:08 that I do not know Jan 24 08:31:32 That is why I have been asking DocScrutinizer Jan 24 08:32:04 hmm? Jan 24 08:32:46 ~repo-down Jan 24 08:32:47 methinks repo-down is http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-20.log.html#t2013-01-20T02:08:42 Jan 24 08:34:26 Aha, thanks. This has been 4 days ago though. Any news since then? =) Jan 24 08:35:26 Too many users then...hmm Jan 24 08:37:56 we seen FW again crumpling on the ground this morning, we took down stage.m.o again and gave FW a boot, since then *.m.o excl repo is back to normal Jan 24 08:38:15 plans are to expose repo.m.o directly to the internets now Jan 24 08:38:32 without FW, or with dedicated traffic shaping Jan 24 08:38:49 so it will no longer tear down whole *.m.o Jan 24 08:39:38 we tweaked FW yesterday, but this changed situation only from "instantly down as soon as stage gets enabled" to "down after 90min" Jan 24 08:41:10 first crash after 90min got "fixed" by a reboot, seems Eero kept stage.m.o up and eventually it again crashed reso came to a grinding halt Jan 24 08:41:48 Ok, the summary then: Still messed up beyond any use ? Jan 24 08:41:56 repo yes Jan 24 08:42:10 rest should be fine right now Jan 24 08:42:42 DocScrutinizer05: how many connections do you expect to r.m.o per second? Jan 24 08:42:50 except the other usual problems, with mail delivery, login/account-creation, autobuilder-down, etc Jan 24 08:43:06 warfare: honestly, nfc Jan 24 08:43:13 :] Jan 24 08:43:26 warfare: there are 40k devices out there, allegedly Jan 24 08:43:29 Any way I can help with anything server related? Jan 24 08:44:24 DocScrutinizer05: You have done a perfect job, thx. Looking forward to see how repos will survive wild internet :) Jan 24 08:44:26 40k connections shouldn't kill a firewall, as apt on maemo serializes connections. Jan 24 08:44:35 If Eero was seeking for help, I guess he would come here and ask. So all I can do is occasional forward of terse but brilliant remarks to him Jan 24 08:45:51 [2013-01-24 08:57:24] anyway, I need to get back to my usual duties, we'll see about the best way to bypass that cranky FW as soon as fesible Jan 24 08:46:03 [2013-01-24 08:58:01] at least repository alone should be easy to move to the public side (and rate limit just that one port) Jan 24 08:47:30 warfare: since for now it's not "our" call to manage the whole stuff, I hope for Nemein to get things sorted asap Jan 24 08:52:03 DocScrutinizer05: Ok :) Jan 24 08:59:48 btw 15min ago I fired up a mail about general login/register problems in *.m.o, ML delivery issues, autobuilder/drop Jan 24 09:16:25 please evrybody check http://wiki.maemo.org/Migrating_to_Community-driven_Infrastructure and holler resp /query me, if you know about any maintainer in duty. We urgently need to gather that info and contact maintainers Jan 24 09:16:53 Yes, I know it's a bit strange that not even council has an idea who's maintaining what Jan 24 10:02:12 Hi maemoers. Does anyone have advice on interacting with Bluetooth from an N900 via Python? Jan 24 10:02:48 I can use hcitool, and parse the output, but would be neater to use some APIs Jan 24 10:05:32 hola Jan 24 10:06:58 narcos: thp might know, but he's not on the channel, but you can try to /msg him Jan 24 10:07:10 merlin1991: Cool, thanks Jan 24 10:10:23 narcos: afaik bluemaemo is written in python, you may want to get the sources and check Jan 24 10:10:36 merlin1991, freemangordon: look at irc log about mce Jan 24 10:11:14 http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-23.log.html#t2013-01-23T23:22:27 Jan 24 10:12:38 freemangordon: Checking it out now, thanks Jan 24 10:12:47 Pali: do we really need mce rewritten now? Jan 24 10:13:03 Pali: i.e. are there any bugs to be fixed? Jan 24 10:13:26 I want to remove /proc/component_version from full maemo system Jan 24 10:13:29 to use upstream kernel Jan 24 10:13:46 also mce has hardcoded some /sys paths which was changed in upstream kernel Jan 24 10:13:51 Pali: did you boot maemo with upstream? on the device? Jan 24 10:14:01 not yet Jan 24 10:14:16 well, I think mce is your last problem :) Jan 24 10:14:30 what about GSM and GPS? Jan 24 10:15:37 GPS is not hard problem if we have working phonet0 interface Jan 24 10:15:46 there is gpsd plugin for that Jan 24 10:16:08 also simple console application which print gps packets from phonet0 Jan 24 10:16:58 Pali: I'd rather focus on glibc/kernel in CSSU. Now when wiki is up (?) again. And leave upstream kernel for a better times :) Jan 24 10:17:26 look at gps2.c: https://wiki.maemo.org/N900_GPS_Reverse_Engineering Jan 24 10:17:58 nice :) Jan 24 10:18:16 does ssi/hsi work? Jan 24 10:18:27 On maemo with 2.6.28 kernel it worked Jan 24 10:18:37 I mean with upstream Jan 24 10:18:39 it using AF_PHONET Jan 24 10:18:44 I know Jan 24 10:18:48 not tested Jan 24 10:19:02 there is problem with ssi/hsi on 3.8 Jan 24 10:19:05 because it was not in 3.5 Jan 24 10:19:15 ssi driver oopsing Jan 24 10:19:17 hmm, what problem? Jan 24 10:19:21 oops :) Jan 24 10:19:33 Pali: which patch? Jan 24 10:19:48 when loading omap_ssi.ko Jan 24 10:20:13 that was working with 3.5, no oopses. at least with the patches I used Jan 24 10:20:37 but userspace was giving errors Jan 24 10:20:58 it was unable to open/init/whatever the modem Jan 24 10:21:12 also it oopsing when loading g_nokia driver Jan 24 10:21:17 but usb network working Jan 24 10:21:24 also mass storage Jan 24 10:21:39 is g_nokia upstream? Jan 24 10:22:07 yes Jan 24 10:22:11 but problem is in musb Jan 24 10:22:21 hmm, strange then Jan 24 10:22:24 or somewhere in usb code Jan 24 10:22:37 btw I have patch which adding mass storage mode to g_nokia Jan 24 10:22:42 Pali: did you build it as module? (musb) Jan 24 10:22:48 no Jan 24 10:22:51 ok Jan 24 10:22:54 when I tried that it not worked Jan 24 10:23:06 that is why ask Jan 24 10:23:13 *I ask Jan 24 10:23:32 patch for g_nokia was rejected because only nokia can update g_nokia.ko driver in upstream Jan 24 10:23:44 what? why? Jan 24 10:23:51 read https://lkml.org/lkml/2013/1/19/165 Jan 24 10:27:50 * narcos finds http://hackgnar.com/article/slides-libraries-and-tutorials-from-my-defcon-and/ Jan 24 10:28:51 Pali: :( Jan 24 10:29:08 BTW it seems all of the kernel maintainers own n900 :D Jan 24 10:29:17 which is good Jan 24 10:31:41 Pali: so, what do you try to do? Push 3.8 (or 3.9) in CSSU? elaborate please. Jan 24 10:32:16 no, to cssu we can push only some subset of kernel-power Jan 24 10:32:25 :nod: Jan 24 10:32:28 to not break other parts Jan 24 10:32:36 freemangordon, I will push kernel-power git repo to garage Jan 24 10:32:42 ok Jan 24 10:32:47 last deb packages are on TMO thread Jan 24 10:32:57 autobuilder still not working Jan 24 10:33:08 Pali: BTW any signs when autobuilder will be up again? Jan 24 10:33:32 I did not received any message Jan 24 10:33:50 nice, git push to garage not working... Jan 24 10:33:58 remote: sudo: no tty present and no askpass program specified Jan 24 10:34:06 ! [remote rejected] master -> master (pre-receive hook declined) Jan 24 10:34:12 error: failed to push some refs to 'ssh://drop.maemo.org/git/kernel-power' Jan 24 10:34:24 damn Jan 24 10:34:33 Pali: move to gitorious Jan 24 10:34:52 I will push it to cssu kernel to kernel-power branch Jan 24 10:34:57 ok Jan 24 10:35:00 sounds sane Jan 24 10:38:41 Hmm, can we get Python 2.7 on Maemo? Jan 24 10:41:02 freemangordon, pushed: https://gitorious.org/community-ssu/kernel-cssu/trees/kernel-power Jan 24 11:03:25 freemangordon, now I merged kernel-power to kernel-cssu master Jan 24 11:03:29 look if merge is ok Jan 24 11:03:45 then we can start selecting patches for kernel in cssu Jan 24 11:04:29 Pali: here you are, pal! Jan 24 11:04:45 Pali: many thanks for your excellent answer re GPG key Jan 24 11:04:45 DocScrutinizer05, so? Jan 24 11:05:03 DocScrutinizer05, do we need updated GPG key? Jan 24 11:05:08 tablets-dev is down Jan 24 11:05:09 I'd like to forward or even publish it to the general community Jan 24 11:05:14 catalogue is down too Jan 24 11:05:26 so downloads.maemo is comming too.... Jan 24 11:05:45 that's unfortunate but completely out of our range of responsibility Jan 24 11:06:11 I guess we have backups, not much more we can do for now Jan 24 11:06:19 ok Jan 24 11:06:34 DocScrutinizer05, can nokia change expiration of GPG key? Jan 24 11:06:53 seems the can, and have, on their repo Jan 24 11:06:55 and can nokia release new FW via PC-suite? Jan 24 11:07:05 they didn't answer that Jan 24 11:07:36 DocScrutinizer05, I looked at well know public gpg keyservers and nokia gpg key was not updated Jan 24 11:07:52 rather they asked how to update key on device as well, and if that would "re-establish chain of trust and put repo back to work" Jan 24 11:08:21 DocScrutinizer05, updating public gpg key on devic need user interaction Jan 24 11:08:33 well, we for sure can suggest to them to update the key on gpgkeyserver Jan 24 11:08:38 $ apt-key adv blabla... Jan 24 11:08:46 yes, *I* know that Jan 24 11:09:21 but they should also update DEB packages which has that gpg key included Jan 24 11:09:25 I'm swamped with work so haven't found the time yet to forward your last answer to them Jan 24 11:09:40 because if somebody reinstall that DEB packages from repository, new key can be deleted... Jan 24 11:10:00 all deb packages are in that mail Jan 24 11:10:11 Pali: did you discuss with other guys over at #maemo-ssu about that magic never-expiring key on our devices? Jan 24 11:10:23 no Jan 24 11:10:30 please do Jan 24 11:10:37 but maemo repositry has that key Jan 24 11:11:00 also I think maemo cssu-stable repo has never expiring Jan 24 11:11:09 nokia might use that key to sign the repo, and thus put everything back to operation "magically" - or not Jan 24 11:11:43 DocScrutinizer05, nokia should not change gpg key on their repository Jan 24 11:11:49 it can be bigger problem Jan 24 11:11:58 please discuss it with other guys, I don't want to send Nokia incomplete info Jan 24 11:11:59 Pali: why? Jan 24 11:12:12 can we move that to ssu chan please? Jan 24 11:12:19 freemangordon, because updating HAM config file is hard Jan 24 11:12:39 but in HAM are no expiration dates Jan 24 11:12:47 can we move that to ssu chan please? Jan 24 11:12:47 Pali: we have a non-expired key on the devices, why resigning the repo with it should bring problems? Jan 24 11:12:49 only fingerptints of gpg keys Jan 24 11:12:52 sure Jan 24 11:13:00 ok Jan 24 11:15:38 fun stuff, the cheap chinese battery vi____ recommended states that [something in chinese]: >2400µAh Jan 24 11:16:30 Pali: the key used for maemo extras won't work for ssu/mr0 and ssu/apps Jan 24 11:16:34 different trust level Jan 24 11:16:37 they need to use the maemosw key Jan 24 11:16:49 kerio, move to #maemo-ssu Jan 24 11:19:03 wut wut. Jan 24 11:19:07 my ears are burning Jan 24 11:19:24 vi____: your 1600mAh battery Jan 24 11:19:29 the weird orange one Jan 24 11:19:33 aaah yes. the orange monster. Jan 24 11:19:34 i bought it too, just arrived Jan 24 11:19:49 if it's really 1600mAh, i'll be impressed Jan 24 11:19:51 Ah, the last I looked they seemed to have dissappeared. Jan 24 11:19:56 ^spelling Jan 24 11:22:00 well, unless my bq chip lies...that is what I got. Jan 24 11:23:10 how many times have you calibrated? Jan 24 11:23:15 about 3 Jan 24 11:23:19 fair 'nuff Jan 24 11:23:48 first reading was ~1700 Jan 24 11:23:57 second dropped to about ~1620 Jan 24 11:24:06 third ~1602 Jan 24 11:24:09 or there abouts. Jan 24 11:24:15 hm Jan 24 12:46:59 When reaching calibration threshold, bq assumes there's still 6% remaining. For Japod atleast it's more like 1%, so you get 5% "lie" in there Jan 24 12:49:46 DocScrutinizer05, did you got my email about garage? Jan 24 12:50:13 now I found another problem: I cannot push to git kernel-power garage repository Jan 24 12:50:42 error message is here: http://mg.pov.lt/maemo-irclog/latest.log.html#t2013-01-24T12:33:58 Jan 24 12:52:04 ~seen romaxa Jan 24 12:52:08 romaxa is currently on #maemo (1d 13h 50m 10s), last said: 'freemangordon: no via browser-eal interface'. Jan 24 12:53:02 Pali: [2013-01-24 09:59:48] btw 15min ago I fired up a mail about general login/register problems in *.m.o, ML delivery issues, autobuilder/drop Jan 24 12:54:10 DocScrutinizer05: thanks ! :) Jan 24 12:56:09 DocScrutinizer05, problem is that it is not possible to change project members permission on garage Jan 24 12:56:21 I forwared you mail about it Jan 24 12:57:03 Pali: [2013-01-24 09:59:48] btw 15min ago I fired up a mail about general login/register problems in *.m.o, ML delivery issues, autobuilder/drop Jan 24 12:57:19 guess what it been about, and what I attached Jan 24 12:59:07 * DocScrutinizer05 idly wonders when check comes through and if they will pay me good for that stuff. Turned into a 60h/week job recently Jan 24 13:00:04 meh, actually more like 90 Jan 24 13:39:42 Meanwhile: http://www.engadget.com/2013/01/24/nokia-microsoft-payments/ Jan 24 14:15:17 Pali: time_to_empty_now isn't accessible if bq27k isn't calibrated, is that intended? Jan 24 14:16:00 kerio, yes time_time_empty reported by bq27x should be reported only if chip is calibrated Jan 24 14:16:35 Pali: BTW did you fix the applet bars when BME is used? Jan 24 14:16:51 I did not look at it yet Jan 24 14:16:55 ok Jan 24 14:18:44 Pali: BTW did you fix the version check that has been mistakenly implemented instead of a check on the existance of bme_RX-51? Jan 24 14:28:54 I added the merlin mirror , but when I run apt-get update I get hash sum mismatch error . What am I doing wrong ? Jan 24 14:29:40 pavi: use skeiron mirror instead... Jan 24 14:29:49 thedead1440: ok Jan 24 14:35:22 some repo's are just named as fremantle while some as fremantle-1.3 , why differ? Jan 24 14:36:18 My N900 is GSM dead while I am trying to download the stardict to atleast use it as a dictionary ;) Jan 24 14:36:19 pavi: if you're looking just for extras-devel then this should be the best option: http://extras-devel.merlin1991.at/ Jan 24 14:36:53 merlin1991: ahh the merlin of the mirror , thanks Yeah I was looking for extras-devel Jan 24 14:37:44 repos still behind the busy firewall? Jan 24 14:40:07 pavi: some packages work on all versions of fremantle. some require specific versions of libraries that differed between releases. Jan 24 14:40:34 khm: ok Jan 24 14:40:52 pavi: histerical raisins, pretty much Jan 24 14:44:35 Ho ho ho http://www.phoronix.com/scan.php?page=news_item&px=MTI4MjI Jan 24 14:44:59 And then... http://www.techweekeurope.co.uk/news/alan-cox-leaves-intel-linux-105279 Jan 24 14:46:32 Lenart KILLED ALAN COX! Jan 24 14:47:37 lennart?? Jan 24 14:47:47 poettering?? Jan 24 14:47:53 There is only one Jan 24 14:48:07 I mean, only one whomade it into verbs Jan 24 15:04:26 freemangordon: is locking at 500MHz harmful? Jan 24 15:05:31 garage.maemo.org is down? Jan 24 15:07:00 No route to host Jan 24 15:19:33 also wiki is down Jan 24 15:20:00 DocScrutinizer05 ^^^ Jan 24 15:20:02 kerio; not significantly, but why? Omap3 switches freq so fast, not like Omap2 :) Jan 24 15:21:17 ShadowJK: to use as much energy as possible Jan 24 15:21:31 although i'd also need a busyloop Jan 24 15:22:27 WAAAAAH! >>Couldn't authenticate against garage. (DB problem)<< Jan 24 15:24:00 ShadowJK: what the hell, the power consumption is going *down* :s Jan 24 15:24:29 i'm calibrating the orange battery Jan 24 15:35:04 while true: do A=1 ; done Jan 24 15:35:12 busyloop in busybox Jan 24 15:38:04 while true; do true; done Jan 24 15:38:32 or sleep 10000 Jan 24 15:38:44 while :; do true; done Jan 24 15:39:00 for some reason busybox shell wont take shorter forms than that it seems Jan 24 15:39:19 and now i'm worried about overheating :s Jan 24 15:45:15 ShadowJK: what was your test of a good battery again? Jan 24 15:45:30 because i'm not entirely sure that this battery will last for long Jan 24 15:45:49 Pali: pong Jan 24 15:46:24 freemangordon: Pali: btw we have made irc channel related to embedlite on irc.mozilla.org #embedlite Jan 24 15:47:10 romaxa, last month I sent you email about microb, did you received it? Jan 24 15:48:22 [general notice] garage.maemo.org down, other services (except tmo) operational as long as you don't try to log in Jan 24 15:48:54 DocScrutinizer05: why don't you send it as an actual NOTICE? Jan 24 15:49:04 Pali: just found it, I guess I read it and forgot to answer ;( Jan 24 15:49:12 Can anyone tell me what work is being done to improve microb and who is doing it? Jan 24 15:50:08 kerio: e.g. because of several users have notices on ignore, and those who watch my nick won't get highlighted as well Jan 24 15:50:28 and it probably doen't even show up in chanlog Jan 24 15:51:02 Pali: is that about tablet-browser-ui redistro/ Jan 24 15:51:15 DocScrutinizer05: you can notice the channel Jan 24 15:51:27 romaxa, yes Jan 24 15:51:33 some clients don't like that, though Jan 24 15:51:35 kerio: ORLY? Jan 24 15:51:52 what does redistro mean? Jan 24 15:51:54 actually, only mIRC, that i know of Jan 24 15:53:07 Pali: I think you can distribute new binary, no sources :( Jan 24 15:53:24 meanwhile: http://lwn.net/SubscriberLink/533687/2adb64ca055d936a/ Jan 24 15:53:38 niiiice Jan 24 15:53:40 I thought tablet-browser-ui was closed source and therefore something we couldn't change Jan 24 15:53:40 romaxa there is problem with binaries too, we had cssu meating abou t it Jan 24 15:53:46 and therefore didn't need to redistribute Jan 24 15:53:56 i.e. we change tablet-browser-engine (the open bits Jan 24 15:54:05 Pali: for browser with sources I would suggest to join to embedlite development with QML in order to make everything fine Jan 24 15:54:13 and redistribute that Jan 24 15:54:20 [general notice] garage.maemo.org down, other services (except tmo) operational as long as you don't try to log in. ETA 90min Jan 24 15:54:29 but leave the UI as-is unless someone wants to reverse engineer all that mess Jan 24 15:54:52 jonwil: it's not that easy, with the most recent gecko Jan 24 15:54:58 * DocScrutinizer05 has some RL affairs pending, and waves Jan 24 15:55:24 So is the issue that we need to write a new UI that is capable of talking to the newest gecko? Jan 24 15:55:35 jonwil: more likely, a wrapper Jan 24 15:55:45 but really, tablet-browser-ui needs a rewrite Jan 24 15:57:42 Remember though that anything we do has to remain compatible with the other things that use the browser Jan 24 15:57:47 rtcom-messaging-ui in particular Jan 24 15:58:01 Or are you saying we rewrite that beast too? Jan 24 15:58:22 we'd still need a compatibility wrapper or something similar to make those work as they did Jan 24 15:58:33 keeping two geckos in ram at the same time is not feasible Jan 24 15:59:08 yeah true Jan 24 15:59:09 :nod: Jan 24 15:59:43 but latest coomit for embedlite is something "make gtk build finaly work" Jan 24 16:00:16 I am building that right now and will check, yesterday it was segfaulting Jan 24 16:01:10 I guess it is possible to rewrite microb-eal and use new gecko Jan 24 16:01:47 though according to romaxa it is not that easy :) Jan 24 16:01:48 :-) Jan 24 16:03:02 romaxa: Seems there are some changes in fremantle browser, I built dummy-eal and tried to start browser it dummy as engine. Jan 24 16:03:09 It was not loaded Jan 24 16:03:37 is microb-eal usage somehow hardcoded? Jan 24 16:05:18 ok, so the way to go for the short term (IMO) is to keep the dbus interface exposed by tablet-browser-daemon (aka browserd) exactly the same and replace all the bits underneath it. That should keep tutorial-home-applet, rtcom-messaging-ui and tablet-browser-ui 100% working as-is Jan 24 16:05:42 jonwil: in theory, yes Jan 24 16:06:15 though we need to find how exactly all these parrts interact as the info on garage seems outdated Jan 24 16:06:16 I think from memory the only consumers of Gecko are tutorial-home-applet, rtcom-messaging-ui, tablet-browser-ui and nokiamaps Jan 24 16:06:38 Or I am too stupid to understand it (the info on garage) :D Jan 24 16:07:23 nokiamaps links directly to browser-eal and doesn't go through browserd (for reasons I dont understand) Jan 24 16:08:03 jonwil: romaxa said conversations do the same (iirc) Jan 24 16:08:53 jonwil: is it clear to you what browser-eal and microb-eal do? Jan 24 16:09:07 and what loads what and when? Jan 24 16:09:54 my understanding is that browser (UI) tells browserd (via dbus) to load a specific engine. is that correct? Jan 24 16:10:37 Let me check some things and get on that in a sec Jan 24 16:10:39 freemangordon: I think last moment we stopped keeping it work with other engines... so possibly something got broken Jan 24 16:10:42 kerio; difference between load/idle voltage. Difference in voltage (in Volts) divided by Difference of current (in Amps). < .3 is really good Jan 24 16:11:28 romaxa: I see. But in theory replacing microb-eal would do the job, ain't? Jan 24 16:11:38 freemangordon: in theory yes Jan 24 16:11:43 great Jan 24 16:12:07 [general notice] garage.maemo.org recovered. ETA 90min been incorrect, sorry ;-) Jan 24 16:14:16 romaxa: don;t get me wring, I am all for having working qml embedlite on n900, but qml is overkill on top of already memory hog gecko. the best that could happen is if we can somehow built libxul to support both qml and gtk UI and to 1. replace microb-eal and 2. have qml interface for those who don;t like microb Jan 24 16:14:20 *wrong Jan 24 16:14:41 don't get him wring, too Jan 24 16:14:46 that's also important Jan 24 16:16:21 having old gecko connected to one of the most important part of the system (convesation ui) does not leave many choices Jan 24 16:16:30 AIUI Jan 24 16:16:31 /notice is a really badly named thing. In the RFC, it's described as being same as normal message, except that bots must not respond or take any action to notice, and that bots must use notice and not msg. The rationale being to avoid bots replying to bots and flooding eachother. It's unfortunate that murc started treating it the way it does :) Jan 24 16:18:11 freemangordon: in that case I guess you need to make microb-eal compiling against latest gecko.... I guess that is not extremely hard to do Jan 24 16:18:26 ShadowJK: thanks for clarifying Jan 24 16:18:38 romaxa: with the except it is closed source :D AFAIK Jan 24 16:18:48 freemangordon: microb-eal is open Jan 24 16:18:54 yes it is Jan 24 16:18:56 is it? great Jan 24 16:19:14 aah, why then I am wasting time with bummy-eal? Jan 24 16:19:16 freemangordon: most of the browser stack behind browser-eal is open Jan 24 16:19:30 romaxa: thanks a lot Jan 24 16:20:03 AFAIK everything pulled in by tablet-browser-daemon (including tablet-browser-daemon itself) is open except for plugins Jan 24 16:20:47 freemangordon: even RenderDocument API which microb using for SHMEM rendering still available http://mxr.mozilla.org/mozilla-central/source/dom/interfaces/base/nsIDOMWindowUtils.idl#1076 Jan 24 16:21:07 great, will try Jan 24 16:21:33 romaxa: could embedlite be used? Jan 24 16:21:39 freemangordon: all sources were available in https://vcs.maemo.org/svn/browser/mozilla/trunk/ but not sure what is going on now Jan 24 16:22:21 garage/repos/wiki are down afaik Jan 24 16:22:27