**** BEGIN LOGGING AT Thu Aug 22 02:59:58 2013 Aug 22 07:46:44 DocScrutinizer05: oh i know; that was just an example of my dried up humor. ;-P Aug 22 11:35:23 shouldn't syncevolution be in extras? I can't find it, but i have the repo "http://repository.maemo.org/extras/".. Aug 22 11:37:29 hi all Aug 22 11:37:53 Is there software that can sava battery life? Aug 22 11:37:56 or show what take more power battery? Aug 22 11:53:47 i gather from this page http://maemo.org/packages/view/syncevolution-frontend/ that syncevolution is only in -extras and -devel, so never mind Aug 22 11:54:23 but why do the repos use expired certs? Aug 22 11:58:16 benedikt: which repo? Aug 22 11:58:25 downloads.maemo.nokia.com? Aug 22 11:58:44 because nokia's gpg key expired and nokia does not want to fix it Aug 22 12:09:19 Pali: yep. are there maemo/hildon hosted repos? Aug 22 12:09:55 servers *.nokia.com are nokias's Aug 22 12:10:08 obviously :) Aug 22 12:10:10 servers *.maemo.org are maemo Aug 22 12:10:54 we can fix maemo repos but not nokia's Aug 22 12:12:03 my sources.list has repository.maemo.org for extras and downloads.maemo.nokia.com for the rest. are there *.maemo.org servers for them that i can replace the nokia servers with? Aug 22 12:13:58 nokia.com has OVI and system packages Aug 22 12:14:07 other extras are on maemo Aug 22 12:14:26 are the nokia repos being updated? Aug 22 12:14:31 Pali: hi Aug 22 12:15:25 Pali: lets continue ke-recv here, it is too much of an effor to post about it on TMO :) Aug 22 12:17:48 benedikt: no, they refused Aug 22 12:17:54 freemangordon: hi Aug 22 12:18:24 Pali: are there any plans on mirroring it and eventually moving to Aug 22 12:18:27 *.maemo.org? Aug 22 12:18:47 benedikt: do not know Aug 22 13:05:34 benedikt: how would we do that? All users had to install a package manually to get the new repo locations. that's not viable Aug 22 13:06:03 DocScrutinizer05: who controls the maemo images these days? Aug 22 13:10:14 images? Aug 22 13:10:38 what exactly do you refer to? Aug 22 13:11:14 Probably Flash-ROM-Images Aug 22 13:11:18 DocScrutinizer05: http://tablets-dev.nokia.com/ # these. but it is .nokia.com so i'm guessing nokia Aug 22 13:11:51 hah, that's actually us Aug 22 13:12:09 to control them Aug 22 13:12:35 that way you could at least get new devices to automatically pull from repos you control yourselves.. but thats not a lot.. Aug 22 13:12:44 yes, feasible Aug 22 13:13:07 might even be possible to push a packages to -extras that updates the sources.list.. Aug 22 13:13:40 we pondered that. It seems there's not a singe package in extras that everybody has installed Aug 22 13:13:57 single* Aug 22 13:14:03 well. thats a shame. Aug 22 13:16:16 btw, benedikt: you should NOT use sources.list for anything Aug 22 13:16:44 DocScrutinizer05: i know. ive used debian for more than a decade, it's just simpler to refer to it as sources.list than the full pathname Aug 22 13:17:43 [2013-08-21 23:03:26] anyway if you *really want to mess with HAM and apt catalogs/repos on file level, this is what you're looking for: http://maemo.cloud-7.de/maemo5/etc/apt/sources.list.d/hildon-application-manager.list and http://maemo.cloud-7.de/maemo5/etc/hildon-application-manager/catalogues Aug 22 13:17:44 [2013-08-21 23:04:11] everything else apt reƶated gets overwritten by HAM eventually Aug 22 13:18:03 HAM? Aug 22 13:18:34 seems like a good reason to uninstall the application manager then Aug 22 13:18:47 [2013-08-21 23:05:12] or is ignored by HAM but not by fapman and apt-get, which makes for rather confusing behaviour that nobody really appreciates Aug 22 13:19:09 DocScrutinizer05: nah, sources.list proper gets used by HAM and not modified Aug 22 13:19:24 kerio: cool, noted. Aug 22 13:19:25 HAM's .list file obviously gets overwritten at least on every catalog modification Aug 22 13:19:34 but also not displayed in "catalogs" I guess Aug 22 13:19:35 if not every time HAM starts Aug 22 13:19:40 they're also displayed in catalogs Aug 22 13:20:33 still it's nonsensical to edit it, since otherwise why is mine 0 bytes? Aug 22 13:21:59 it's kinda like editing resolv.conf Aug 22 13:22:56 resolv.conf has 127.0.0.1 and nuttin else, per definition. and sources.list is empty, per definition Aug 22 13:25:56 benedikt: I don't think uninstalling HAM is a great idea Aug 22 13:26:50 it won't instantly bring the system down, but for sure you're begging for trouble when you do Aug 22 13:29:40 benedikt: and since we're already close to that topic: please note that apt-get upgrade and particularly apt-get dist-upgrade and autoremove, and prolly even -f sometimes, are extremely dangerous and deprecated, since they all tend to fubar your maemo system Aug 22 13:30:02 maemo is not a plain debian when it comes to apt, repos, packaging Aug 22 13:30:23 that's why you want to keep HAM Aug 22 13:32:28 for quite a few things Nokia invented very own concepts orthogonal to standard linux resp debian Aug 22 13:34:46 other stuff they implemented their own personal flavour of some functions that later on got also into upstream linux but implemented differently. See sudoers and update-sudoers and include directive Aug 22 13:36:07 duh, that would make for a very nice CSSU project: get a newer sudo that knows about "include " Aug 22 13:36:26 nuke update-sudoers Aug 22 13:38:08 ooh, benedikt: without HAM no proper installing or updating of any CSSU Aug 22 13:40:06 nothing wrong with using apt-get install for getting new packages, but *updates* you should leave to HAM and possibly auto-update Aug 22 13:40:13 CSSU? Aug 22 13:40:17 ~cssu Aug 22 13:40:18 i guess cssu is http://wiki.maemo.org/Community_SSU Aug 22 13:40:38 I strongly suggest *stable Aug 22 13:40:48 didnt know that. might be why i broke maemo last time Aug 22 13:41:44 yes, ignoring those idiosyncrasies of maemo is a too easy way to mess stuff up Aug 22 13:42:38 actually you shouldn't even do updates with HAM as long as extras-devel (or -testing) is enabled Aug 22 13:43:24 use apt? Aug 22 13:43:37 not all packages that would get updated to latest version in extras-devel are safe. every now and then we see a really rogue one in there Aug 22 13:44:12 wait what Aug 22 13:44:28 just befor you said i should only update with HAM, Aug 22 13:44:30 ? Aug 22 13:44:31 jon_y: nah, cherrypick package by package. Then it doesn't matter if you use apt-get install or HAM Aug 22 13:44:41 ok Aug 22 13:45:31 if i have extras-devel and/or -testing, should i update with apt-get or HAM? what if i have either of those? Aug 22 13:45:47 but a apt-get upgrade or "install/update all" in HAM is most likely fatal when extras-devel is enabled, sooner or later Aug 22 13:46:21 having both extras-devel and extras-testing doesn't really make any sense Aug 22 13:46:54 since all packages in testing went through devel Aug 22 13:47:47 a package goes extras-devel -> extras-testing -> extras Aug 22 13:48:07 extras is considered "safe" Aug 22 13:49:13 testing basically only for *real* testers who want to help evaluating the packages and make them go to extras or reject them due to flaws they have Aug 22 13:49:56 devel is bleeding edge that's not tested at all, rogue or dangerous or fubar stuff *will* be there Aug 22 13:50:25 amog pretty stable other packages that simply never got promoted to extras Aug 22 13:50:32 among* Aug 22 13:51:01 right, but how does one update safely? Aug 22 13:51:01 wasnt sure if packages went first to -devel or -testing Aug 22 13:51:49 you simply do update via HAM with extras-testing and exzras-devel repo disbaled Aug 22 13:52:49 for -testing and -devel you do cherrypicking for the one or two ;) packages you got from there, after you checked the changelog and supplementary info about the new version of that package Aug 22 13:53:52 it is cumbersome, but then extras-devel never been meant for general usage Aug 22 13:54:38 recently somebody uploaded some X11 related package to extras-devel that broke the whole system Aug 22 13:55:23 aiui every user would have installed it as soon as you do upgrade on extras-devel via apt or HAM Aug 22 13:56:02 libc6 eventually appeared on extras-devel Aug 22 13:56:24 incompatible broken qt versions appeared on extras-devel Aug 22 13:56:34 heh Aug 22 13:57:13 using extras-devel like a normal repo is a warranty for doom, if not today then maybe next month Aug 22 13:58:29 you enable extras-devel, install (or update) your package(s) you're interested in, and immediately disable the catalog again Aug 22 14:00:21 DocScrutinizer05: yeah, that sounds like the safe way to do it Aug 22 14:00:28 :-) Aug 22 14:05:26 the added benefit: HAM is much less slow as long as extras-devel (and -testing) are disabled Aug 22 14:05:46 i havent enabled them yet and i think HAM is terribly slow Aug 22 14:06:11 is there a nice way to use gpg with the email client? Aug 22 14:08:55 http://www.youtube.com/watch?v=6Zx8qdVs7Ug # how do i get the n900 to show the kernel messages when booting and why have I never seen this REALLY EXCITING VIDEO when booting mine? Aug 22 14:12:33 http://wiki.maemo.org/User:Joerg_rw/tools#replace_shaking-hands_video Aug 22 14:12:55 for the boot messages you need a kernel supporting that Aug 22 14:13:12 power-kernel should be able? Aug 22 14:13:38 I thik so Aug 22 14:13:40 i want to use LUKS soon, i suppose that requires another kernel anyways Aug 22 14:13:41 or can r&d mode with vanilla kernel do that? Aug 22 14:13:43 think* Aug 22 14:13:59 korhojoa: no Aug 22 14:14:07 okay Aug 22 14:14:09 unfortunately, the url the youtube video gives is dead Aug 22 14:14:14 vanilla kernel has no fb Aug 22 14:14:22 ah, okay Aug 22 14:16:03 also iirc you need a patched kernel cmdline to tell kernel to actually do logging to fb console Aug 22 14:16:28 KP comes with cmdline patched afaik Aug 22 14:18:12 IroN900:~# cat /proc/cmdline Aug 22 14:18:13 init=/sbin/preinit ubi.mtd=rootfs root=ubi0:rootfs rootfstype=ubifs rootflags=bulk_read,no_chk_data_crc rw console=ttyMTD,log console=tty0 snd-soc-rx51.hp_lim=42 snd-soc-tlv320aic3x.hp_dac_lim=6 Aug 22 14:18:41 ^^^ this is vtock kernel cmdline and it doesn't have the needed directive to log to console Aug 22 14:19:00 my n900 doesn't find any wifi network :s Aug 22 14:19:18 i think it's because there are so many of those Aug 22 14:19:22 there's awful congestion Aug 22 14:19:53 hard to believe that not a single beacon comes thru Aug 22 14:20:20 i find no wifi network Aug 22 14:20:54 iwlist scan Aug 22 14:21:24 er, wot Aug 22 14:21:25 Network is down Aug 22 14:22:04 http://paste.opensuse.org/64047300 Aug 22 14:22:34 No scan results Aug 22 14:22:44 after an ifconfig wlan0 up Aug 22 14:23:46 ifup ? Aug 22 14:24:55 not ifup Aug 22 14:25:05 ifup is the debiany thing that also runs scripts that don't actually exist in maemo Aug 22 14:25:21 this is silly Aug 22 14:25:25 it finds no networks whatsoever Aug 22 14:25:44 uhuh Aug 22 14:26:28 and ifconfig is the thing that directly talks to hardware then, without letting system know about state change in wlan? Aug 22 14:26:32 yeah Aug 22 14:28:50 Aug 12 07:34:47 IroN900 wlancond[2462]: WLAN Connection Daemon 3.0.20.1+0cssu2 started. Aug 22 14:28:52 Aug 12 07:34:47 IroN900 wlancond[2462]: WLAN flight mode changed to "normal" Aug 22 14:29:38 and I'm pondering to revert to "good old" stock wlancond Aug 22 14:30:30 too many strange thing happening to my WLAN since quite some time Aug 22 14:31:26 that's not a strange thing, but should give kerio sth to think about: Aug 22 14:31:27 Aug 12 07:34:49 IroN900 kernel: [43183.658721] wl1251: 151 tx blocks at 0x3b788, 35 rx blocks at 0x3a780 Aug 22 14:31:29 Aug 12 07:34:49 IroN900 kernel: [43183.674316] wl1251: firmware booted (Rev 4.0.4.3.7) Aug 22 14:31:30 Aug 12 07:34:49 IroN900 wlancond[2462]: Scan issued Aug 22 14:31:42 wait, wl1251? Aug 22 14:31:45 oh, wl12xx Aug 22 14:33:39 this however is kinda strange, though I guess it always been like this: Aug 22 14:33:41 Aug 12 07:34:50 IroN900 wlancond[2462]: Scan results ready -- scan active Aug 22 14:33:42 Aug 12 07:34:50 IroN900 wlancond[2462]: Scan results (6 APs) to :1.81 Aug 22 14:33:44 Aug 12 07:34:50 IroN900 icd2 0.87+fremantle10+0m5[2553]: connecting iap 0x420f8 in state ICD_IAP_STATE_SCRIPT_PRE_UP: interface is '(null)' Aug 22 14:33:45 Aug 12 07:34:50 IroN900 kernel: [43185.088592] wl1251: down Aug 22 14:34:58 Aug 12 07:34:51 IroN900 icd2 0.87+fremantle10+0m5[2553]: connecting iap 0x420f8 in state ICD_IAP_STATE_LINK_UP: interface is 'wlan0' Aug 22 14:35:03 ~cssu Aug 22 14:35:04 cssu is, like, http://wiki.maemo.org/Community_SSU Aug 22 14:35:12 funny bot Aug 22 14:38:09 ~+chaninfo Aug 22 14:38:09 126/126 channels, 5049 users, 3807 unique: #debian/1336, #kde/431, #macosx/337, #gsoc/306, #maemo/233, #asterisk/194, #wowuidev/139, #oe/136, #cubieboard/125, #harmattan/114, #meego/107, #wowace/106, #arm-netbook/89, #utah/82, #webos-internals/75, #openmoko/69, #wowhead/67, #openmoko-cdevel/67, #edev/62, #bzflag/56, #asterisk-dev/50, #maemo-ssu/45, #htc-linux/41, #sc2mapster/40, #curseforge/40, #gllug/36, #n9/35, #uphpu/35, #slug/31, #elinux/29, ... Aug 22 14:38:16 meh, still doesn't work Aug 22 14:38:18 but it never did properly Aug 22 14:38:22 so i don't think it's cssu Aug 22 14:39:19 sounds like you messed up your system's connectivity subsystem. E.G by installing auto-disconnect, or whatever Aug 22 14:39:51 isn't this a reasonable assumption: if i flashed my device just yesterday, then i'm running PR1.3. Aug 22 14:40:00 yes Aug 22 14:40:14 unless you downloaded an obsolete image file Aug 22 14:40:27 ~pr131 Aug 22 14:40:27 hmm... combined is the rootfs fiasco image of maemo. For N900 latest (PR1.3.1) see http://maemo.cloud-7.de/maemo5/fiasco+co/RX-51_2009SE_21.2011.38-1_PR_COMBINED_MR0_ARM.bin or http://nds2.fds-fire.nokia.com/fdp/interface/FiRe/2011/9/--FID--A0A22MVWFVFAM/--LID--FiRe1317015685654/RX-51_2009SE_21.2011.38-1_PR_COMBINED_MR0_ARM.bin Aug 22 14:40:32 i downloaded the newest i could find, but it was still from 2011 Aug 22 14:40:51 ~pr13 Aug 22 14:41:04 mmmpff Aug 22 14:41:10 i downloaded this one RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin (sha1: 54c9e5d6491ea73f9f4131c9f19cc01f9934fcdb) Aug 22 14:41:32 ~listvalues pr_combined Aug 22 14:41:34 Factoid search of 'pr_combined' by value returned no results. Aug 22 14:41:42 ~listvalues *pr_combined Aug 22 14:41:44 Factoid search of '*pr_combined' by value returned no results. Aug 22 14:41:57 ~listvalues PR_COMBINED Aug 22 14:41:58 Factoid search of 'PR_COMBINED' by value returned no results. Aug 22 14:42:05 damn Aug 22 14:42:30 36-2 is PR1.3 Aug 22 14:42:48 MR0 is 2global" Aug 22 14:42:56 s/2/"/ Aug 22 14:42:56 DocScrutinizer05 meant: MR0 is "global" Aug 22 14:43:46 DocScrutinizer05: so i have PR1.3 Aug 22 14:43:50 yup Aug 22 14:43:52 nice Aug 22 14:44:28 also see ~cssu Aug 22 14:44:39 i did, thats why i was wondering if i had 1.3 Aug 22 14:44:50 it elaborates on how to make sure Aug 22 14:45:13 just noticed. :| Aug 22 14:50:33 please foolow instructions faithfully! Aug 22 14:50:38 follow even Aug 22 14:55:29 hmm, had to fix a few flaws in that instruction Aug 22 14:56:07 did you just update the wiki? Aug 22 14:59:55 DocScrutinizer05: It is possible to add pr1.3.1 to http://tablets-dev.nokia.com/ ? Aug 22 15:00:11 I think it should be easy to accomplish Aug 22 15:00:27 benedikt: yup Aug 22 15:00:34 ADDRCONF(NETDEV_UP): wlan0: link is not ready Aug 22 15:00:44 * benedikt reloads Aug 22 15:00:56 DocScrutinizer05: and mirroring downloads.maemo.nokia.com on *maemo.org? Aug 22 15:00:57 what is this D: Aug 22 15:01:04 is there a newer version of python somewhere for maemo? Aug 22 15:01:22 benedikt: look at maemo extras Aug 22 15:01:29 only there can be Aug 22 15:01:30 ~pkg Aug 22 15:01:31 somebody said pkg was http://maemo.org/packages/ Aug 22 15:01:50 kerio: what? Aug 22 15:02:02 Pali: mirroring? why? Aug 22 15:02:06 whenever i try to look for the available wifi networks Aug 22 15:02:23 DocScrutinizer05: backup + resign repository Aug 22 15:02:38 with new or some maemo.org key Aug 22 15:02:40 err wait, what's on downloads.m.n.c? Aug 22 15:02:54 SSU Aug 22 15:02:57 and others Aug 22 15:03:20 aaah, sure we can mirror it, but it's kinda useless unless we find a way to point all devices to the mirror Aug 22 15:03:22 no? Aug 22 15:04:01 DocScrutinizer05: seems like there is only the 2.5 branch Aug 22 15:04:07 kerio: does it have "IPv6:" prefix? that dmesg message? Aug 22 15:04:12 no Aug 22 15:04:50 kerio: reflash stock, test, report Aug 22 15:04:53 DocScrutinizer05: it can be usefull for cssu users or users who read TMO, IRC, RSS, mailinglist... Aug 22 15:04:59 DocScrutinizer05: fair enough Aug 22 15:05:02 but it's going to be hell :c Aug 22 15:05:03 kerio: same problem I have on desktop Aug 22 15:05:06 Pali: :nod: Aug 22 15:05:15 [ 44.811991] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not read Aug 22 15:05:20 but there is IPv6 prefix Aug 22 15:05:29 Pali: prolly we need to go that stony path Aug 22 15:05:48 this has been an intermittent problem for me, fwiw Aug 22 15:05:48 Pali: since Nokia never will fix the friggin gpgkey Aug 22 15:07:44 kerio: look: http://siso.sourceforge.net/wireless-linux.html Aug 22 15:07:56 ADDRCONF(NETDEV_UP): wlan0: link is not ready, merely means what it says: the link is not up yet. The link will be up once wpa_supplicant negotiated the key. Aug 22 15:08:05 there's no key! Aug 22 15:08:08 there's no wifi network! Aug 22 15:08:13 for crying out loud Aug 22 15:08:42 what's happening Aug 22 15:08:52 maemo not using wpa_supplicant but proprietary nokia eapd SW Aug 22 15:14:24 toldya Aug 22 15:14:29 didn't I? Aug 22 15:15:11 [2013-08-22 16:26:28] and ifconfig is the thing that directly talks to hardware then, 04without letting system know about state change in wlan? Aug 22 15:17:26 Pali: why did KP need epoch? Aug 22 15:18:05 Pali: I try to wrap my head around whether we need to support epoch in maemo.org (autobuilder) Aug 22 15:19:05 DocScrutinizer05: I do not remember correctly, but there was problem when we integrated upstream linus diff from 2.6.28 to 2.6.28.10 Aug 22 15:19:30 how's that related to version string? Aug 22 15:19:52 10 was added to version string Aug 22 15:19:57 mhm Aug 22 15:20:11 and there was problem that dpkg told that new version os older Aug 22 15:20:19 but version string of KP is sth like 52, no? Aug 22 15:20:38 so I added epoch to deb package like in other similar problems in debian Aug 22 15:20:54 alas maemo ignores epoch, at best Aug 22 15:21:09 it has suffix 52 Aug 22 15:21:14 maemo.org infra, that is Aug 22 15:21:57 DocScrutinizer05: 1:2.6.28-10power52 Aug 22 15:22:01 there's already an epoch there Aug 22 15:22:05 HAH Aug 22 15:22:24 yes I know there's an epoch now, I asked WHY Aug 22 15:23:52 of course when you break versioning sheme like 2.6.28power49 -> 2.6.28-10power52, that can not work Aug 22 15:24:08 is there a way to run the UI as some other user than "user"? Aug 22 15:24:21 alas epoch isn't a "fix" in this case either, at least on maemo.org autobuilder Aug 22 15:24:30 benedikt: hardly Aug 22 15:24:36 DocScrutinizer05: meh Aug 22 15:24:44 "user" user hardcoded in some locations Aug 22 15:25:32 DocScrutinizer05: was it only extras-devel that you should disable between uses or should i do that with -testing as well? Aug 22 15:25:51 you probably don't want testing at all Aug 22 15:26:06 it had a bunch of packages i wanted.. Aug 22 15:26:16 it's redundant for regular users, when they use -devel anyway Aug 22 15:26:29 i was trying to stay away from -devel Aug 22 15:27:02 well, then you probably can use testing as a halfway-to-"safe" replacement for -devel Aug 22 15:27:21 DocScrutinizer05: yes this was reason: I wanted to have kernel version same as package version Aug 22 15:27:37 again, every package goes extras-devel -> extras-testing -> extras Aug 22 15:27:42 and I really did not know that maemo infra does not support what is supported on debian Aug 22 15:28:04 Pali: no blaming Aug 22 15:28:45 Pali: anyway a better versioning for KP probably is 52power2.6.28-10 Aug 22 15:30:24 particularly since the real kernel version not really applies to KP with all its backports of features from newer kernels Aug 22 15:30:56 1:2.6.28... is still higher than 52po... Aug 22 15:31:15 btw, how was epoch in repos fixed? Aug 22 15:31:39 no, it isn't, on maemo.org. it however is on apt dpkg etc Aug 22 15:31:40 packages were manually imported? Aug 22 15:31:55 I mean in package interface Aug 22 15:31:57 epoch gets ignored completely Aug 22 15:32:02 on maemo.org Aug 22 15:32:10 aiui Aug 22 15:32:39 ok, so if I put new version (with epoch) will it be imported to package interface? Aug 22 15:32:46 you could use 1:52power2.6.28-10 Aug 22 15:33:43 afaik maemo.org simply does a s/[0-0]:// on _some_ locations Aug 22 15:33:51 ok Aug 22 15:33:53 afaik maemo.org simply does a s/[0-9]:// on _some_ locations Aug 22 15:35:23 so make sure your versioning isn't violating epoch rules (for apt's sake) while your version string *without epoch* should still be strictly monotonically growing Aug 22 15:35:59 and if you can avoid by any means, do not use epoch at all Aug 22 15:36:35 for now, according to freemangordon, our policy is: NO changes in epoch! Aug 22 15:37:10 so if you're using 1:, keep it. If you don't use epoch, don't change that either Aug 22 15:38:29 maemo.org doesn't support epoch really. It at best treats epoch like a normal part of version string, worst case simply ignores epoch all together like it wasn't existing. (or swap best and worst here) Aug 22 15:39:39 for handling mess with epoch changes, we simply nuke all packages from a project, then developer may upload whatever she likes Aug 22 15:40:25 I think we got ~5 projects that use epoch, in extras* Aug 22 15:40:28 just... make it work? Aug 22 15:40:45 kerio: it can't be done since db has no epoch Aug 22 15:40:50 huh? Aug 22 15:41:09 epoch is removed from version string in db Aug 22 15:41:24 well that's fucking dumb Aug 22 15:41:37 that is how it been since birth of maemo Aug 22 15:41:38 who threw away information? Aug 22 15:41:43 for fuck's sake Aug 22 15:42:04 so no way we can fix that stuff Aug 22 15:42:07 * benedikt gets the popcorn Aug 22 15:42:24 yes we can, we know where the package versions are Aug 22 15:42:28 in the .debs Aug 22 15:42:32 and in Packages Aug 22 15:42:46 meh, nobody will edit the midgard db Aug 22 15:43:12 since that will likely make our sun go supernova Aug 22 15:43:59 maemo doesn't support epoch, take it Aug 22 15:45:09 we honestly got much more severe issues we need to fix, that are less effort and less demanding to fix, and yet we don't have the manpower to do those Aug 22 15:45:25 epoch, blablablapfffffff Aug 22 15:46:01 who T F needs epoch? Aug 22 15:50:31 me! me! Aug 22 15:50:35 i need epoch for reasons Aug 22 15:50:40 vital ones Aug 22 15:51:06 anyway, ignoring epoch can work i suppose Aug 22 15:51:07 or let me put it this way: you (and everybody interested) are free to port maemo.org incl all the funny custom made scripts to some CMS that's still supported, like e.g. midgard2, and for sure during such migration effort we could fix epoch issue en passant Aug 22 15:51:25 but write it somewhere, or devs might expect to be able to do that Aug 22 15:52:05 yep, agreed Aug 22 15:53:54 we removed the hack xfade implemented to allow pali to upload and promote his KP with epoch. Now it's back to the state where maemo.org completely ignores epoch (aiui, ask fmg) and thus somebody uploading a package with epoch will notice soon enough that it doesn't promote and thus will come here to ask what's wrong Aug 22 15:55:32 but nonetheless I agree we should publish an according note that epoch should not get used to increment version Aug 22 15:55:55 since maemo doesn't watch for changes in epoch Aug 22 15:57:04 and when you can't change epoch (also not from non-using to using it), then it's useless anyway Aug 22 15:57:47 get your versioning sorted then you don't need to resort to epoch ;-) Aug 22 22:11:47 so like. the n900 charger has a different pinout to charge quicker? Aug 22 23:27:14 macer: it has a standard microusb pinout... Aug 23 00:02:23 ecc3g: yeah but i thought the actual n900 had a different thing about it Aug 23 00:02:35 which allows the n900 to charge faster Aug 23 00:24:51 so, guys, the N950 has a proximity sensor on the front, right? Aug 23 00:26:10 It uses an IR(or near-IR) LED? Aug 23 00:26:57 macer: nah it's standard or at least (mostly) follows the USB consortium's guidelines. Aug 23 00:27:28 the only thing that's broken is that it doesn't always switch modes, it's extremely conservative not to fry anything. Aug 23 00:29:13 i've never seen any phone with a visible or near-IR proximity sensor... heh Aug 23 00:30:04 if it was near-IR then I suppose some people could actually see it lit... Aug 23 00:32:16 ecc3g, exactly what I'm getting at Aug 23 00:32:37 Last night, in the dark, I saw a very /very/ dim red light from the front of my N950 Aug 23 00:32:55 just a tiny little spark, from the top area there. definitely glowing Aug 23 00:33:25 I used my N900 to take a look at it with the camera, and it was light purple, which is what IR tends to look like Aug 23 00:33:45 It was also flickering, via the camera Aug 23 00:34:05 naked eye, no flashing(so a very high rate/optical persistance) Aug 23 00:36:29 I didn't know this was possible(never was able to see IR before) Aug 23 00:36:57 which makes me think it might be a close enough frequency, or possibly some reflecting off the plastic layer above it, or something Aug 23 00:44:16 ir is visible, just, at the edges Aug 23 00:44:33 Yeah. Aug 23 00:44:42 it is possibly the contact sensor Aug 23 00:46:49 Yeah, that's what I think **** ENDING LOGGING AT Fri Aug 23 02:59:59 2013