**** BEGIN LOGGING AT Mon Mar 07 02:59:57 2011 Mar 07 03:45:12 Attempting OBS registration for Meego gives: OBS Web Interface Error: Mar 07 03:45:13 Error Details: Mar 07 03:45:13 Errorcode: unknown Mar 07 03:45:13 Message: undefined method `create' for nil:NilClass Mar 07 04:32:52 Guest7035, try again? Mar 07 04:34:07 morning Mar 07 04:34:24 evening fellas Mar 07 04:34:42 i have Orbiter running on MeeGo Core, and the Joggler. http://www.youtube.com/watch?v=ZRKkqxWcytI Mar 07 04:36:36 morning iekku Mar 07 04:47:54 morning Mar 07 04:48:04 morning Mar 07 05:13:06 hello Mar 07 05:13:25 is there anybody who have been used the virtual keyboard? Mar 07 06:26:11 Now always gives: Use only letters a-z and/or digits 0-9 for username (with a legal username) Mar 07 06:31:28 Guest7035: that means that spaces and special chars like !"#¤, etc. are not allowed Mar 07 07:19:50 hi Mar 07 07:20:06 Stskeeps : hello Mar 07 07:20:12 morning wassupnari Mar 07 07:20:26 I finally launch my apps! :D Mar 07 07:20:29 thanx Mar 07 07:20:42 no problem, glad to hear it :) Mar 07 07:20:57 but there is another problem :( Mar 07 07:21:31 i installed inputmethod*.rpm packages which are related with virtual keyboard, Mar 07 07:21:46 and i changed some configurations and scripts Mar 07 07:22:14 if i launch sms or browser and focus on text input area, Mar 07 07:22:58 then virtual keyboard pops up, but it is not working properly Mar 07 07:23:23 ok, so that might be because you aren't using the mcompositor for instance, because of no GLES use Mar 07 07:23:30 so windows doesn't pop up correctly Mar 07 07:24:06 virtual keyboard is one of the places where this is used as it overlays another app Mar 07 07:24:37 yes, it is.. then, do i have to install another packages? Mar 07 07:24:42 Stskeeps: vgrade: http://www.youtube.com/watch?v=ZRKkqxWcytI ;) Mar 07 07:25:51 wassupnari: not sure what exactly needs to be done but for this feature to run properly you need a working 3d acceleration Mar 07 07:26:11 (i think) Mar 07 07:26:27 okay, i'll find out :) thanx, though. Mar 07 07:27:56 TSCHAKeee, nice one Mar 07 07:28:27 still so much to do Mar 07 07:28:51 wassupnari: do you have a screenshot by chance of how bad it looks when it's not "working properly"? Mar 07 07:28:53 but i had to make a gonzo kickstart, cobbled together from meego core, auke's repo for xfwm4, my orbiter packages, matchbox keyboard, and a fe other bits Mar 07 07:28:58 then, do you have 2d/3d hw acceleration graphic library for TI OMAP4? Mar 07 07:29:16 wassupnari: i think it actually exists Mar 07 07:29:25 i'll ask around Mar 07 07:29:31 okay Mar 07 07:30:29 i cannot capture the screen shot, but it shows keyboard and app, repeatedly Mar 07 07:30:53 o Mar 07 07:30:54 k Mar 07 07:31:20 did you already know this symptom? Mar 07 07:31:45 i haven't tried software rendering lately Mar 07 07:32:05 wassupnari: so basically, you need to get the SGX kernel module built for your device and add in the SGX libraries somehow. i think TI calls it Graphics SDK Mar 07 07:32:36 wassupnari: and then make sure libGLESv2.so.2 symlink -> libGLESv2.so and libEGL.so.2 symlink -> libEGL.so Mar 07 07:33:22 unfortunately, we don't have graphic SDK yet. Mar 07 07:33:35 i think it was released recently Mar 07 07:34:02 isn't that for the OMAP3? Mar 07 07:34:16 it is for omap4 too, i believe Mar 07 07:34:42 could you let me know the link? Mar 07 07:34:55 trying to, my laptop is acting up a little bit so can't browse so easily Mar 07 07:35:30 okay, i can wait :) Mar 07 07:36:55 http://www.imgtec.com/powervr/insider/sdkdownloads/index.asp , scroll down to "TI OMAP3430/4430 (inc. Nokia N900 & BeagleBoard and PandaBoard)" Mar 07 07:38:37 good! appreciated it :) Mar 07 07:40:14 good luck with it, haven't tried it myself, but get some simple 3d demos up and it should be easier with meego Mar 07 07:40:55 okay, i'll let you know if i succeeded it! Mar 07 07:41:39 wassupnari, not sure if they have fixed this but I had to add ARCH=arm to the kernel module build line Mar 07 07:42:40 vgrade : thanks for your additional comment :) Mar 07 07:43:13 wassupnari: also, if you get it working and see corrupted fonts in handset UX, let us know as soon as possible :) Mar 07 07:43:17 it's a bug we're chasing Mar 07 07:56:16 okay, i'll let you know :) Mar 07 08:18:43 anyone here know about the meego/qt at http://download.meego.com/live/Trunk/standard/src/ ? Mar 07 08:20:06 is qt-4.7.1-7.6.src.rpm meant for any specific device? Mar 07 08:31:25 girishr: it's generic Mar 07 08:34:46 thiago_home: ok, so the folks who make a device will pick that up and put more patches on them if needed? Mar 07 08:35:03 basically, i tried it on the wetab and mouse event delivery is broken Mar 07 08:35:20 i am wondering what exactly those packages are being tested with Mar 07 08:35:24 they shouldn't have to patch Mar 07 08:36:26 i think there's a problem with some of the X drivers and xinput2 Mar 07 08:36:30 not giving proper labels Mar 07 08:36:31 thiago_home: is the touch support stable across devices? Mar 07 08:36:43 what Stskeeps said Mar 07 08:36:47 * thiago_home is off to the office Mar 07 08:38:06 as i understand, the whole touch thing on X is a moving target. so i assume device makers with add patches on top of the above Qt. did i assume right? Mar 07 08:38:59 girishr: basically it's better to get the X input drivers working properly, as it should work properly then Mar 07 08:39:54 thiago_home: Stskeeps: ok, thanks Mar 07 09:32:16 I guess this might have been here already: http://www.digia.com/C2256FEF0043E9C1/0/405002251 Mar 07 09:32:37 * Stskeeps reads Mar 07 09:33:15 is this a good or a bad thing? Mar 07 09:33:55 I'm going to sai good, but I'm not sure :) Mar 07 09:36:19 * timeless_w7ip sighs Mar 07 09:36:32 ok, so i think mxr.meego.com/meego.gitorious.org/ will now actually update periodically Mar 07 09:36:47 (well, except for the linux-mssf tree which will cause it to get stuck each time it tries to update, but...) Mar 07 09:38:24 stskeeps: having a defined future is more or less a good thing Mar 07 09:38:38 also, it provides for some useful contractual obligations Mar 07 09:38:41 true Mar 07 09:39:13 what's interesting in that announcement is that it seems that nokia will continue to do core Qt work Mar 07 09:39:22 TrollTech has always had 2 groups Mar 07 09:39:27 (well ignoring embedded) Mar 07 09:39:35 Core and Consulting Mar 07 09:39:46 it looks like Digia has bought out the Consulting side Mar 07 09:40:06 (there's a better word than 'consulting', but it's too early for me to come up w/ it) Mar 07 09:41:48 * timeless_w7ip sighs Mar 07 09:41:50 "oops" Mar 07 09:41:57 i manually updated sources Mar 07 09:42:10 then i killed a stuck task in a cron script which was trying to update sources Mar 07 09:42:17 then i manually started a pair of indexing tasks Mar 07 09:42:20 but! Mar 07 09:42:28 ah the whole digia thing already turned up here Mar 07 09:42:34 * dm8tbr tries to understand what this means Mar 07 09:42:52 the cron script has rules to do automatic updates of those indices, which it started performing once it got past the stuck task Mar 07 09:43:28 dm8tbr: "nothing much" ;) Mar 07 09:44:55 Stskeeps: *nod* still the doomsday-nay-sayers will have a fest over this :) Mar 07 09:52:24 sorry for dumb question but how can someone license something if that instance does not own it :) do they pay some royalties to nokia from license purchases or what? Mar 07 09:53:55 jarkko^: possibly? Mar 07 09:53:57 http://qt.nokia.com/merge_requests/agreement/ Mar 07 09:54:32 timeless_w7ip: "for the next year" Mar 07 09:55:54 N needs Qt for another 1-2 years, so it's no surprise they'll support Qt during that time Mar 07 09:57:14 bunk: the question is though: will they be doing new development or merely maintenance? Mar 07 09:57:56 bunk: you seem to be forgetting that Nokia is transitioning _to_ Qt on their lower-end devices Mar 07 09:58:11 s/license something/sell commercial licenses of something/ Mar 07 09:59:13 ahiemstra: if you read labs blog or track gitorious, you'll see that new development is done all the time Mar 07 09:59:31 so answer to that question is easy to dig Mar 07 10:00:10 jarkko^: sure, there is currently still new developments going on Mar 07 10:00:42 considering a lot of that will be projects that have been started before the entire wm7 thing Mar 07 10:01:03 ahiemstra: wp, not wm (different OS) Mar 07 10:01:36 not really Mar 07 10:03:48 Well, wp = wm would be like e.g. maemo = MeeGo Mar 07 10:05:56 jonnor: You are talking about non-Symbian platforms? Mar 07 10:24:57 Stskeeps: indeed, it seems label mapping is not happenning correctly Mar 07 11:16:20 bunk: nokia's volume is S30/S40 Mar 07 11:16:41 presumably the margins are better in symbian, but volume isn't there Mar 07 11:27:21 Has anyone managed to install MeeGo SDK in Debian Squeeze? To be able to build an app in a specific target (i.e netbook ia32) and then with the appropriate runtime to emulate the app? Mar 07 11:29:58 I tried both with amd64 and i386 Debian Squeeze, but i failed. On the other hand, i installed the SDK in Fedora 14 and it went smoothie. Mar 07 11:42:34 Hello all Mar 07 11:42:36 Nokia sells Qt licensing/support business to Digia Mar 07 11:42:40 EHLO wazd Mar 07 11:43:05 Rst38h: orly Mar 07 11:43:12 Rst38h: o/ Mar 07 11:44:43 RST38h: makes sense Mar 07 11:44:45 moo wazd Mar 07 11:45:25 Rst38h: well, not the worst news evah :) Mar 07 11:45:37 Stskeeps: heya Mar 07 11:46:30 I've recently bought fancy "meecast.com" domain :P Mar 07 11:46:34 hehe Mar 07 11:48:18 Stskeeps: we're looking for rebranding opportunities with omweather :) Mar 07 11:48:52 Stskeeps: cause "omweather" is really weird name imo Mar 07 11:48:56 :nod: Mar 07 11:48:58 wazd: Better add more features :) Mar 07 11:49:05 wazd, omweather still needs the "I looked out of the window and it is not snowing" button! Mar 07 11:49:21 * RST38h starts chanting "ommmmm.... Mar 07 11:49:50 lcuk: Simply change your locations from "Yakutsk" in the settings Mar 07 11:49:55 Lcuk: that involves dedicated server, moderator, cms and lots of other stuff :) Mar 07 11:50:29 wazd, do we know anyone who perhaps has written/created/managed a great cms system ? Mar 07 11:50:59 Lcuk: I know one guy, but he's extremely busy :) Mar 07 11:51:08 Lcuk: like totally :) Mar 07 11:51:31 wazd, of course, people are always busy Mar 07 11:52:10 incidentally, I am thinking of bergie, who are you thinking of? Mar 07 11:52:15 Rst38h: havemyou seen my Marina theme for OMW? :) Mar 07 11:52:27 wazd: No, did notevenknow it was themeable Mar 07 11:52:52 Lcuk: non-community dude :) Mar 07 11:53:45 Rst38h: yeah, and that's one of the problems - hard to discover new functionality after update Mar 07 11:54:48 zalbisser, hi how are you ? Mar 07 11:55:02 Rst38h: settings - widget style Mar 07 11:55:12 aha Mar 07 11:56:07 ahh good, bug 13833 is marked blocking 10753 Mar 07 11:56:10 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=13833 nor, Undecided, ---, ilkka.koskinen, NEW, N900 operating at 500mhz, 16% slower than Maemo Mar 07 11:56:21 wazd: Long time no see! (Don't suppose you fancy drawing an icon for the Maemo Community SSU - currently we're using the "update" icon? http://wiki.maemo.org/Community_SSU Mar 07 11:56:21 bug 10753 Mar 07 11:56:23 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=10753 maj, High, ---, huan.zheng, NEW, video playback is not smooth Mar 07 11:57:20 Jaffa: hey there Mar 07 11:57:55 Jaffa: with pleasure :) Mar 07 12:00:42 lcuk: happy to be thought of :-) Mar 07 12:01:06 :) Mar 07 12:07:50 Jaffa: BTW,has the problem with CSSU breaking "apt-get upgrade" been fixed? Mar 07 12:08:34 RST38h: Define "breaking apt-get upgrade" Mar 07 12:08:58 RST38h: There's https://bugs.maemo.org/show_bug.cgi?id=11831 but no-one has been able to provide reproduction instructions Mar 07 12:09:00 Bug 11831: Removal of mp-fremantle-*-pr breaks apt autoremover Mar 07 12:09:03 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=11831 nor, Undecided, 1.2, max.a.yu, RESO FIXED, [Windows] QtCreator can't start Qemu in MeeGo SDK 1.2 preview-1224. Mar 07 12:09:59 wazd: wind direction circles are off in your theme Mar 07 12:10:17 wazd: both vertically and horizontall (a little bit) , I am afraid =( Mar 07 12:10:42 Otherwise, switching to it whenever ready (Digia stuff is too heavy) Mar 07 12:11:02 Rst38h: off? You mean invisible? Mar 07 12:11:19 Jaffa: Once I install the CSSU, apt-get upgrade starts offering to delete base system packages as "not being required by any package in the system" Mar 07 12:11:21 Rst38h: or misplaced?) Mar 07 12:11:26 wazd: Now, offset a little bit Mar 07 12:11:33 misplaced if you wish Mar 07 12:11:52 Jaffa: Yea, 11831 is what I am talking abot Mar 07 12:12:09 RST38h: Reproduction instructions welcome :-) Mar 07 12:13:45 Jaffa: 1) Install CSSU; 2) run apt-get upgrade; 3) Watch apt-get offer you to break stuff, badly Mar 07 12:14:03 isn't this more of a #maemo talk? Mar 07 12:14:04 :P Mar 07 12:14:30 More or less, yes Mar 07 12:14:59 It is a slow day though, so we are probably not disrupting dead things too much :) Mar 07 12:15:25 RST38h: I did that immediately after a clean PR1.3 reflash and on my N900 and cannot reproduce. Neither can MohammadAG. Anyway, as Stskeeps says, this is better for #maemo-ssu Mar 07 12:15:43 Rst38h: ah, yes, we're working on it :) Mar 07 12:18:04 Recent ia32 images for daily-testing have not been including a pinetrail image (most likely because they are not working), however a generic ia32 image is there Mar 07 12:18:30 does anyone know what the differences are? Mar 07 12:18:33 http://download.meego.com/testing-daily/builds/trunk/1.1.90.6.20110304.81/images/ Mar 07 12:23:41 wazd: No plans to expand to other widgets, as we discussed a while ago? Mar 07 12:24:56 http://www.theregister.co.uk/2011/03/07/kirk_discovery/ heh Mar 07 12:26:21 Well, nice, except that it is the last Discovery launch... Mar 07 12:27:34 RST38h, yes but since many equate space travel with star trek think it is nice :) Mar 07 12:28:12 "Alien" feels like a much better approximation to me... Mar 07 12:33:32 RST38h, did you happen to see the piece of 2001 inspired art recently? Mar 07 12:33:47 an 8 foot monolith, actually a huge VHS video tape of 2001 Mar 07 12:33:50 * lcuk finds link Mar 07 12:34:45 http://www.xclusiveworld.com/artist-transforms-2001-vhs-tape-into-a-real-life-black-monolith.html Mar 07 12:36:14 um how can it be a VHS tape if it's 8 foot tall? Mar 07 12:36:24 i'm pretty sure VHS tapes are supposed to fit in VHS players Mar 07 12:37:29 ali1234, they need to be that big for using in a cinema! Mar 07 12:37:41 although using a broomstick to reset the clock in the VHS player is a PITA Mar 07 12:40:51 lcuk: I prefer cassette tapes:) Mar 07 12:41:32 The VHS ones just do not ring any bells (did not have VCR until 20 or so) Mar 07 12:43:43 question: which package contains qemu? Mar 07 12:44:44 i have located the meego-sdk-qemugl-addon, but i lack to find qemu itself on which the addon is used Mar 07 14:08:03 Hi all I got a problem I cannot get gesture recognized with my QGraphicsObject I get this message QGestureManager::deliverEvent: could not find the target for gesture Mar 07 14:32:13 wathek, still stuck on circle gesture? Mar 07 14:32:53 lcuk, I got an other problem with the QGraphicsObject I cannot grab gesture and I'm getting this message : QGestureManager::deliverEvent: could not find the target for gesture Mar 07 14:32:56 it makes me mad Mar 07 14:33:16 lcuk, I need to fix that first and then look at the circle gesture Mar 07 14:34:19 wathek, is it related to: Mar 07 14:34:21 http://bugreports.qt.nokia.com/browse/QTBUG-14224 Gesture target not found when gesturing on parentless widget. Mar 07 14:34:48 which is one of the first things located after an internet search for the error message Mar 07 14:35:06 I believe on Friday I pointed you towards the code this error is raised from Mar 07 14:35:28 which would allow you (as a developer) to step backwards and work out how to work around/cure the issue Mar 07 14:35:56 lcuk, yep but that was with QGraphicsView which is a QWidget and actually I can grab gestures with my QGraphicsView and I don't have the message any more Mar 07 14:36:12 lcuk, so I tried to do the same thing with QGraphicsObject and I'm getting the message Mar 07 14:36:13 :( Mar 07 14:37:52 lcuk, and nope it doesn't seems to be related to that bug Mar 07 14:40:07 wathek, ok, perhaps you could try follow up the searching yourself and maybe try the general #qt channel if no-one here can help Mar 07 14:43:06 ddenis could probably help me Mar 07 14:43:11 but his not here unfortuantly :( Mar 07 14:44:05 all superheroes need downtime. Mar 07 14:44:27 speaking of which Mar 07 14:44:50 lol Mar 07 14:52:21 Nokia just sold Qt ?!?! Mar 07 14:52:42 georgem: only the commercial licensing and support part Mar 07 14:54:15 Wouldn't be surprised to see the rest go eventually. Mar 07 15:06:42 morn qgil :) Mar 07 15:07:56 dd Mar 07 15:07:56 hi there Stskeeps - I'm looking the news about Qt Commercial & Digia. Surprised by the negative reaction from some Qt lovers. I guess Slashdot's false statement isn't helping Mar 07 15:08:19 * TSCHAKeee yawns Mar 07 15:08:35 overall I think this move is good for the Qt project - and indirectly for the MeeGo project too (many Qt Commercial customers are potential MeeGo vendors) Mar 07 15:08:49 qgil: i think people are fairly calm since they read the original sources and understood what is going on.. but it got blown up and misunderstood by media as usual :) Mar 07 15:08:53 yeah, I think so too. Mar 07 15:09:45 * Stskeeps thinks its an expected move that makes sense Mar 07 15:10:20 qgil: Wanted to say, thanks for stopping by the LinuxMCE booth at SCALE. (it's Thom, here.) Mar 07 15:10:24 "people"... my Twitter follows several insiders and as for now they seem to be more victims of that media rather than calm - oh well, as usual time will tell Mar 07 15:10:38 ...heh. Mar 07 15:10:44 TSCHAKeee: hi! It was very good chatting with you Mar 07 15:11:01 likewise Mar 07 15:30:33 does anyone know which package provides the boot splash on Netbook? Mar 07 15:30:58 plymouth-lite Mar 07 15:31:09 oh ok thanks. Mar 07 15:40:09 qgil, \o hope you managed to have yourself a good weekend Mar 07 15:40:54 lcuk: I can't complain :) Mar 07 15:41:09 heh Mar 07 15:42:08 qgil, when you and Nokia moved into the new offices, there was news about how the workforce would be making use of local facilities, has this been working weel in your eyes? Mar 07 15:42:16 well * Mar 07 15:43:53 lcuk: what does "local facilities" mean? Anyway, I believe Sunnyvale downtown is noticing that we are here, yes Mar 07 15:45:02 lcuk: I even got my last haircut in a loca business around the corner: fast and cheap ;) Mar 07 15:45:32 * Stskeeps looks forward to travelling around california for 14 days Mar 07 15:45:37 heh excellent qgil, I am just trying to find the link Mar 07 15:45:43 we were discussing it the other night Mar 07 15:45:45 local business, I mean. "loca business" is actually a good name for a hairdresser in California - more in the drag queen style though Mar 07 15:46:29 lcuk: what was the discussion about? Anyway, not very #meego topic I guess :) Mar 07 15:48:06 Stskeeps: if you come to California you must come to our house - it's a special local tax for certain foreigners Mar 07 15:48:27 qgil, it was actually in maemo, but has parallels throughout industry Mar 07 15:48:29 lcuk mikhas, Nokia building where qgil is are right in the heart of the town and do not include cafe and stuff afaik 00:42 Mar 07 15:48:29 lcuk I remember reading how it helps to integrate the workforce in with the surroundings and obviously also helps local economy 00:42 Mar 07 15:48:35 http://mg.pov.lt/maemo-irclog/%23maemo.2011-03-05.log.html#t2011-03-05T00:42:24 Mar 07 15:48:45 qgil: will certainly consider :) Mar 07 15:48:57 http://mynokiablog.com/2011/02/05/nokias-new-us-hq-in-sunnyvale-california-silicon-valley/ Mar 07 15:50:18 HQ ... Mar 07 15:50:22 lcuk: most people go to restaurants at walking distance or elsewhere in Sunnyvale - some shopping at Target - internal events have been hiring local catering too. Not a lot more... but there is not a lot more in Sunnyvale downtown atm (Macy's) Mar 07 15:50:33 Some people don't half write some crap on the internet. Mar 07 15:51:01 maybe time to adjust my highlight settings :) Mar 07 15:51:12 Sunnyvale, like Palm HQs =) Mar 07 15:52:54 qgil, it just stood out as eing just another think Nokia are doing right :) Mar 07 15:53:00 being * Mar 07 15:53:05 lcuk: what Jay seems to be missing in that post is that the new Sunnyvale site is formed by the unification of several sites Nokia had already scattered in different location in SF Bay Area. Mar 07 15:53:13 * lcuk prefer to look at positive side of things Mar 07 15:53:57 qgil, perhaps could put him right then since you are on the ground there so to speak :) Mar 07 15:54:43 lcuk: that news piece is a month's old now Mar 07 15:55:09 so enough time for people to expand on it and present as a new piece :P Mar 07 16:04:26 lcuk, you still there Mar 07 16:05:16 The person you have requested is not at his seat, please leave a message after the beep and he will try to get back to you. Mar 07 16:05:20 BEEP. Mar 07 16:05:24 lol Mar 07 16:05:56 I need you just to confirm something Mar 07 16:05:58 lcuk, http://qt.gitorious.org/qt/qt/blobs/v4.6.3/src/gui/kernel/qgesturemanager.cpp#line562 Mar 07 16:06:27 that's the function that shows me the message QGestureManager::deliverEvent: could not find the target for gesture Mar 07 16:06:56 lcuk, and it seems to be logic since that my target is a QGraphicsObject and it looks for a QWidget Mar 07 16:07:13 am I wrong ? Mar 07 16:10:09 wathek, the code does certainly indicate it is trying to use QWidget, but I do not know the specific difference and relationship between QWidget and QGraphicsObject, perhaps some of the many Qt experts around could try to investigate further Mar 07 16:10:15 hi there Mar 07 16:10:25 has anyone ever tried meego with n900 Mar 07 16:10:26 ? Mar 07 16:10:38 diazo, I have meego on my N900 right now Mar 07 16:10:46 lcuk: so... how does bug handling work? :) Mar 07 16:10:52 e.g. bug 14221 Mar 07 16:10:54 lcuk, QGraphicsObject is a class that inherits QObject and QGraphicsItem Mar 07 16:10:55 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=14221 nor, Undecided, ---, peter.j.zhu, NEW, keygen_set_std_prefs dangerously wastes 2 bytes per sym due to extra pointer qualifier in size_of ma Mar 07 16:11:17 lcuk: cool Mar 07 16:11:27 lcuk: does the usb otg work? Mar 07 16:11:36 lcuk: I mean, the host mode Mar 07 16:11:58 diazo: no Mar 07 16:12:01 diazo, not sure, if it is something you need to confirm, please check whether there is a FEA marked or a bug related Mar 07 16:12:27 damn.. Mar 07 16:12:31 k Mar 07 16:12:35 thx for the info :) Mar 07 16:13:34 timeless_office, you mean, how do you take your patch and get it integrated? Mar 07 16:13:47 no Mar 07 16:13:59 i mean how does someone ensure that my patch gets sent upstream Mar 07 16:15:00 or perhaps s/how does someone ensure/what ensures/ Mar 07 16:16:38 timeless_office, depending on who or what upstream is in this context, you have to let them know that a patch exists Mar 07 16:18:06 well, s/you/meego/ Mar 07 16:19:42 timeless_office, is your patch integrated on the meego build already, ie so that somebody could even try to progress it further? Mar 07 16:20:31 lcuk: nope, i don't care about meego or gnupg or whatever Mar 07 16:20:42 i'm filing bugs based on a coverity analysis that happens to contain meego code Mar 07 16:20:54 and since i happen to have a public meego xref, i'm demoing how the xref works :) Mar 07 16:21:06 by littering bugs.meego w/ reports that reference it :) Mar 07 16:22:01 heh Mar 07 16:30:15 bug 14222 Mar 07 16:30:17 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=14222 nor, Undecided, ---, peter.j.zhu, NEW, ss_add_request_table dangerously wastes 4 bytes per size count due to missing pointer qualifiers in Mar 07 16:30:51 * timeless_office notes that perhaps someone should add more components to bugzilla Mar 07 16:31:04 this peter guy is going to have a hard time doing anything Mar 07 16:44:17 * timeless_office finds an *awesome* allocation glitch Mar 07 16:44:28 lcuk: so far we have 2<->4, and 4<->8 Mar 07 16:44:51 this next one is um.... 80<->4 Mar 07 16:46:46 linux is awesome Mar 07 16:48:17 y Mar 07 16:48:50 ans5i: that was sarcasm fwiw Mar 07 16:48:54 timeless_office ok, so you have identified and fixed a number of core bugs which need following up on Mar 07 16:49:37 timeless_office, hopefully the guys on the bug itself will advance Mar 07 16:50:02 * lcuk ccs to see how they progress Mar 07 16:50:32 the 80 one is coming up soon once i get the xref's done Mar 07 16:50:48 * timeless_office frowns Mar 07 16:50:59 coverity took a wrong turn, that's lame Mar 07 16:51:10 * timeless_office wonders how it did that Mar 07 16:51:31 it told me it found a size 80, but when i asked it to follow the trail it got me a much smaller critter ;( Mar 07 16:51:36 s/;(/:(/ Mar 07 16:52:40 oh, no, i took the wrong turn! Mar 07 16:52:42 * timeless_office feels much better Mar 07 16:52:45 * timeless_office pats coverity Mar 07 16:57:16 * timeless_office gets confused and moves on Mar 07 16:58:02 timeless_office, does coverity scan and show up odd malloc things? Mar 07 16:58:21 coverity does all sorts of things Mar 07 16:58:45 what it's doing here is saying target type size is , allocation unit is a multiple of Mar 07 16:58:52 when x and y don't fit well, it complains Mar 07 16:59:10 that makes sense. Mar 07 16:59:24 if someone intentionally sticks an extra buffer after a thing, then that'll be flagged because coverity has no way of understanding that it's intentionally evil Mar 07 16:59:38 otherwise, it's generally right when it complains Mar 07 17:00:13 I got told that openssl reads from a random bit of memory, a program that read the openssl code didn't undersand why and flagged an error Mar 07 17:00:23 the program didn't know why it was there so he just commented it out Mar 07 17:00:36 cosmohill: don't blame coverity for dumb users Mar 07 17:00:42 that's debian-ssl Mar 07 17:00:44 it's a classic Mar 07 17:00:45 this meant you had a 1 in 36K chance of guessing the key instead of 1 in billions Mar 07 17:01:07 timeless_office: yeah, that was in our security lecture Mar 07 17:01:34 we had an idiotic "let's clear coverity issues" crusade a couple of years back at nokia Mar 07 17:01:40 i got some wonderful gems from it Mar 07 17:01:45 I'd of imaged there would have been a nearby comment explaing it Mar 07 17:01:57 timeless_office, same for any scanning tool Mar 07 17:02:19 a scanning tool is like GPS, useful but you shouldn't turn your brain off Mar 07 17:02:23 http://timeless.justdave.net/blog/159/ Mar 07 17:02:29 http://timeless.justdave.net/blog/160/ Mar 07 17:02:33 I used to run different scans in codebase for stuff, then manually check against each hit - but that only worked because I knew the code inside out and backwards so to speak Mar 07 17:02:34 http://timeless.justdave.net/blog/161/ Mar 07 17:02:46 those were the results of our coverity crusade Mar 07 17:03:05 cosmohill/lcuk: please read :) Mar 07 17:03:14 * lcuk was doing Mar 07 17:03:29 they aren't advertised in my blog because i didn't necessarily "finish" them Mar 07 17:04:11 grr Mar 07 17:06:44 timeless_office: they look cool even if a little over my head Mar 07 17:07:26 * lcuk hands timeless_office a /facepalm Mar 07 17:08:02 timeless_office: so I think I'm on UK time... Mar 07 17:08:03 lcuk: i'm expecting a couple :) Mar 07 17:08:09 it's not 5pm... it's 7pm Mar 07 17:08:15 lbt: yeah, that's ok Mar 07 17:08:24 which also explains why it's dark, quiet and I'm hungry Mar 07 17:08:26 can we eat @6pm UK time? Mar 07 17:08:29 oh wait Mar 07 17:08:40 if you're hungry now, then "Let's Eat" Mar 07 17:08:43 sure Mar 07 17:08:52 i have one item i'd like to file first, but i can file it while you walk over here :) Mar 07 17:08:57 OK Mar 07 17:09:04 e3 at the back, "i'll leave the light on for you" Mar 07 17:09:05 if(hungry) { eat(); } Mar 07 17:09:37 if(hungry && !diet) { eat(); }; Mar 07 17:10:29 CosmoHill: for(;;) { ++pancakes; } Mar 07 17:10:42 it's pancake day tomorrow \o/ Mar 07 17:10:52 ahem, while(1){ bacon++; } Mar 07 17:11:17 lbt, you need a big watch Mar 07 17:11:24 you'd just have infinate bacon and pancakes Mar 07 17:11:26 try writing the current time on your wrist! Mar 07 17:11:32 ;) Mar 07 17:11:37 lcuk: Doesn't that fail for small values of 1 ? Mar 07 17:11:42 while(1) { bacon++; eat(bacon); }; Mar 07 17:11:57 lbt, having it updating on its own would be cool Mar 07 17:12:45 double fat (double weight) { if (hungry || !hungry) { return weight + eat() + eat(); }} Mar 07 17:13:10 right, I have to install stuff on this n900 Mar 07 17:13:19 and thanks to aard, also on my ideapad Mar 07 17:13:22 bbl \o Mar 07 17:13:24 lcuk: A lcuk I totally for to tell you about ... Mar 07 17:13:47 X-Fade, could you try English please? Mar 07 17:13:47 if ( weight < 70) /*Kg*/ goto(FRIGDE); Mar 07 17:14:11 btw, hi \o Mar 07 17:14:12 lcuk: Bacon dougnut :) Mar 07 17:14:19 mmmm Mar 07 17:14:30 sounds gross Mar 07 17:14:35 lcuk: I had one at Voodoo dougnut in Portland ;) Mar 07 17:15:00 I have a huge slab of pork in the kitchen downstairs Mar 07 17:15:42 lcuk: http://voodoodoughnut.com/menu.php Mar 07 17:15:44 making it yesterday was awesome, we added about 100 baconstrips ontop Mar 07 17:15:46 lcuk: Second one :) Mar 07 17:16:08 mmm candy bacon Mar 07 17:16:08 * timeless_office wonders if lbt got lost Mar 07 17:16:15 X-Fade: 403 Mar 07 17:16:19 timeless_office, internal gps Mar 07 17:16:38 CosmoHill: Works for me. Mar 07 17:16:54 "There is no hotlinking of any kind allowed to or from this website. " :/ Mar 07 17:17:03 Remove the page then. Mar 07 17:17:18 I saw it too Mar 07 17:17:19 I have Mar 07 17:17:20 http://voodoodoughnut.com/ -> Menu Mar 07 17:17:30 still giving me 403 Mar 07 17:17:34 CosmoHill, perhaps your ISP is baconist Mar 07 17:17:49 cisco http://is.gd/fk6zS | Docs: http://is.gd/3EtT | Mailing List: http://is.gd/3Eu Mar 07 17:17:57 CosmoHill: New tab or other browser ;) Mar 07 17:18:04 cisco http://informatiq-help.blogspot.com/2011/03/cisco-reseau-formation.html Mar 07 17:18:42 X-Fade: "要求されたディレクトリへのアクセス権限がありません。 インデックスドキュメントが存在しないか、 ディレクトリの読み込みが許可されていません。" Mar 07 17:18:42 wtf Mar 07 17:18:47 picasse: piss off Mar 07 17:19:12 reminds me of the liqbase graffiti wall :) Mar 07 17:19:16 /me hungers Mar 07 17:22:55 X-Fade: I sent them an email that bounced :( Mar 07 17:23:25 CosmoHill: Corporate firewall? :D Mar 07 17:25:23 cisco http://informatiq-help.blogspot.com/2011/03/cisco-reseau-formation.html Mar 07 17:25:23 * CosmoHill suddenly remember he has chocolate Mar 07 17:25:25 picasse: shoo Mar 07 18:24:57 a reminder for tomorrow, schedule for Tuesday #MeeGo-Meeting is: Mar 07 18:25:01 Tuesday Mar 07 18:25:01 * QA Meetings: Every Tuesday at 07:00 UTC Mar 07 18:25:01 * SDK Meeting (Release): Every Tuesday at 08:00 UTC, 10:00 EEST Mar 07 18:25:01 * MeeGo Netbook Bug Triage Meetings: Every Tuesday at 13:00 UTC Mar 07 18:25:01 * Community Office Meeting: The 1st and 3rd Tuesday, alternating between 20:00 UTC and 15:00 UTC Mar 07 18:25:02 * Events Meeting: Every other Tuesday at 17:00 UTC Mar 07 18:25:10 that looks busy Mar 07 18:25:27 tuesday and wednesay have numerous meetings Mar 07 18:25:28 http://wiki.meego.com/MeeGo-Meeting_IRC_Schedule Mar 07 18:25:40 busiest days Mar 07 18:44:03 but no community office meeting since this is the 2nd Tues :) Mar 07 18:46:56 DawnFoster, since I saw on twitter oyu are preparing to go on a roadtrip Mar 07 18:47:13 who will chair next weeks meeting?! Mar 07 18:47:39 errr s/chair/herd cats/ Mar 07 18:49:52 we actually moved the next CO meeting to the 4th Tues :) Mar 07 18:50:00 http://wiki.meego.com/Community_Office/Meetings Mar 07 18:50:23 I have an internal intel open source event this week in the Oregon wilderness Mar 07 18:50:46 and Friday I fly to Austin, Texas for sxsw Mar 07 18:51:00 that doesn't at all conjure up images of pagan worship or other rituals.. Mar 07 18:51:00 :P Mar 07 18:51:10 :) Mar 07 18:51:21 ha! Mar 07 18:51:31 well, we will have werewolf :) Mar 07 18:51:55 yay doggy Mar 07 18:51:56 DawnFoster, so it is an external internal meeting Mar 07 18:52:12 (re wilderness) Mar 07 18:52:27 lcuk: we hold people captive away from the office and Portland to prevent people from leaving the meeting :) Mar 07 18:52:46 It's about an hour away from portland 1.5 hours away from the Intel office Mar 07 18:52:58 3 days of Intel open source goodness Mar 07 18:53:17 good stuff Mar 07 18:55:19 DawnFoster, do you guys have a no technology rule for things like this, or is it part of the game to see who can get a signal Mar 07 18:55:40 can't have a tech conference without tech :) Mar 07 18:55:51 they do a pretty good job of providing wireless for us Mar 07 18:56:09 that is stretching the definition of wilderness then :P Mar 07 18:56:30 half the fun is seeing how people cope without it! Mar 07 18:57:31 you're evil Mar 07 18:57:47 * lcuk never said otherwise Mar 07 18:59:00 DawnFoster, last time we went camping, nobody could get a phone signal Mar 07 18:59:36 well, this is a really nice lodge sitting in the wilderness :) Mar 07 19:01:24 * gabrbedd can never find a wifi hotspot when backpacking in a national park... need to write congressman Mar 07 19:02:01 I cannot find power outlets Mar 07 19:02:12 they're usually on walls Mar 07 19:02:28 so I should write my representatives so that they put those "wall" things on parks... Mar 07 19:02:30 walls? what is this, the middle ages? Mar 07 19:02:55 the Helsinki airport lounge had these wireless charging stations Mar 07 19:02:58 javispedro: yeah, that too. Not to mention toilets.... Mar 07 19:03:00 I didn't figure out how to use them Mar 07 19:03:10 you want floor sockets that would make a bear trap proud Mar 07 19:04:28 CosmoHill: Hmmm... electric fence to protect bear bags... Mar 07 19:06:52 http://www.youtube.com/watch?v=wJT_ZbIjNfI <- want to run meego on ;) Mar 07 19:07:42 that is hardly world's first transparent lcd, /me has one... (b&w that is...) Mar 07 19:08:51 biggest issue is that you either control illumination, or basically never read anything from it... Mar 07 19:11:19 omg someone is implementing the wp7 metro UI using qtquick =) http://wiki.forum.nokia.com/index.php/Can_QtQuick_create_a_Windows_Phone_Hub Mar 07 19:11:34 javispedro: i actually wondered about that.. Mar 07 19:12:24 Stskeeps: yeah, makes it use as a window to outside home questionable, but It works well for inwards use. Mar 07 19:13:10 (thinking "display" cabinet or the like) Mar 07 19:15:00 Stskeeps: I remember seeing something like it in the movie "Date Night". I remember thinking... "That looks so cool... but the lack of good contrast would get old really quick." :-) Mar 07 19:15:50 is anyone here familiar with soap and complex types? Mar 07 19:17:50 hello Mar 07 19:17:59 has anyone tried installing face recognition on mego 1.1? Mar 07 19:18:24 (according to some other yt videos, it's not transparent lcd but transparent OLED, which would be a more interesting thing) Mar 07 19:19:16 javispedro, everytime I look at wp7 I am reminded of original liqbase Mar 07 19:19:22 it must be the colorscheme Mar 07 19:20:13 lcuk: I think it's mostly because it is very "texty" UI; liqbase also is. And with the Nokia font as in those screenshots... Mar 07 19:20:31 liqbase also has many areas where captions do not fit :D Mar 07 19:21:48 yeah. they quite copied you =) Mar 07 19:21:51 javispedro, standard font now gone, it is all personalised :) Mar 07 19:21:53 http://liqbase.net/liq.meego.r&d.n900.and.ideapad.fullspeed.20110220_011.jpg Mar 07 19:22:39 lcuk: I know but since I first saw liqbase with the Nokia font, I will always remember it that way (even though I guess the "personalised" one was first ;) ) Mar 07 19:22:49 mentioned before to lbt about writing the time on your wrist, a hand written clock on your wrist would be awesome Mar 07 19:23:03 javispedro, yeah, Nokias Sans is great Mar 07 19:23:23 Meego Clock UX ;) Mar 07 19:23:46 (not a bad idea after all =) ) Mar 07 19:24:02 well it is scalable afterall :) Mar 07 19:24:10 from small to large Mar 07 19:26:50 javispedro, most surprising was seeing mine and tracys handwriting together on a page :) Mar 07 19:27:19 hello Mar 07 19:27:49 CosmoHill: hello reading fc a draw this we? Mar 07 19:27:51 niala: Reading 5 - 2 Middlebourgh Mar 07 19:31:16 javispedro, it also helps in a way that it does not depend upon powervr Mar 07 19:31:25 so that tiny hardware could be usable :) Mar 07 19:31:31 yeah, I quite agree with that. Mar 07 19:32:51 lack of a nonGL UI is probably quite a problem for (for ex.) Meego portability Mar 07 19:33:35 * lcuk likes noteslate too :) Mar 07 19:33:47 it appears that is hardware in need of an os Mar 07 19:34:32 oooh, good one. Mar 07 19:35:30 ooooh, gadget lust starting Mar 07 19:36:00 heh Mar 07 19:36:32 I wanted a cheap largish eink ebook reader Mar 07 19:36:43 heh Mar 07 19:38:41 javispedro, i have to wonder how suboptimal rendering must be at the moment Mar 07 19:38:51 especially if the internals of the system are RGB oriented Mar 07 19:39:16 well, X11 has quite good support for monochrome =) Mar 07 19:39:28 actually, X11 probably has better support for monochrome than truecolor =) Mar 07 19:39:29 whether it is used is another issue Mar 07 19:39:59 yeah, like many other ebooks. Mar 07 19:40:10 however, according to the firmware page it does use linux. Mar 07 19:41:41 yeah, most do Mar 07 19:42:18 javispedro: you think? Mar 07 19:42:32 last time i tried to use X11 with a real monochrome display it failed hard Mar 07 19:42:53 looks like nobody has tested monochrome in about 10 years and it's suffering from severe coderot Mar 07 19:43:04 that might be the case... :( Mar 07 19:43:21 * lcuk never expected to get colour working in liqbase o_O Mar 07 19:43:28 could also be a problem with the framebuffer driver i guess Mar 07 19:43:30 I did not think it would be practical Mar 07 19:43:40 ali1234: was going to say the same, fbdev... Mar 07 19:44:10 in fact I remember someone talking about the many problems using some eink controller via fbdev which explained why every other vendors used its own interface Mar 07 19:44:34 eink has problems with frequent updates Mar 07 19:44:36 all the eink drivers in linux use defered framebuffer Mar 07 19:44:45 yes RST38h Mar 07 19:45:01 so what they do is emulate a 8 bit display and then throw away the extra bits Mar 07 19:45:16 sucks. Mar 07 19:45:40 the fact that the controller is more like a vnc consumer instead of framebuffer doesn't help Mar 07 19:45:44 RST38h, I believe einki is actually based on the principle of sucking balls Mar 07 19:45:55 in a way Mar 07 19:46:25 not to tmention that the entire framebuffer abstraction doesn't work that well when clearing pixels is way slower than drawing them... Mar 07 19:46:35 btw, i wasn't using eink, just defered framebuffer on usb Mar 07 19:46:59 are there any alternatives to the balls yet? Mar 07 19:47:12 MEMS lever arms were once shown I believe Mar 07 19:47:28 which give a rainbow effect without backlight Mar 07 19:47:30 there are many, though I do not know how they are from a commercial PoV Mar 07 19:48:40 either way, looking at the relative success of the nook color and all the Amazon for XXX apps, I guess development might be slowing down on eink. Mar 07 19:48:41 they might be good if you could get them bistable Mar 07 19:50:49 javispedro, battery life for eink is one of the real strong points at the moment Mar 07 19:51:16 But that is about it, as far as eink strong points are concerned Mar 07 19:51:26 this is the best i could squeeze from x11 monochrome: http://al.robotfuzz.com/~al/g15/20100809_001.jpg (that blob to the right is the X11 cursor) Mar 07 19:51:27 and not really, readability Mar 07 19:51:32 that and the contrast ratio Mar 07 19:51:33 not having a backlight etc Mar 07 19:52:30 ali1234: well it is working isn't it? ;) Mar 07 19:52:38 more or less :) Mar 07 19:52:46 i had to hack X11 code to even get this far Mar 07 19:53:19 * javispedro likes the eink contrast ratio. very much. and outside usability. Mar 07 19:53:45 in fact the first ebook reader I tried had crappy battery life because they didn't know how to do PM in Linux. Mar 07 19:54:02 that is a trend that has continued Mar 07 19:54:22 not through lack of effort, just pure number of variations Mar 07 19:54:46 * lcuk still finds it odd how often compatability is broken Mar 07 19:54:56 so that what works on one kernel breaks on a later one Mar 07 19:55:33 has anyone tried installing face recognition on meego 1.1? Mar 07 19:56:40 unfortunately this noteslate thing rings my vaporware sense. sad, because the price is more or less right (I still remeber the +$800 eBooks...) Mar 07 19:57:09 javispedro, hence my observation Mar 07 19:57:18 (also note they say the WiFi model is > $99, but don't say exact price) Mar 07 19:58:01 theriex: be more specific. If you are trying to install something, and want help, ask for that Mar 07 20:00:25 I'm trying to install PAM face authentication on meego 1.1 Mar 07 20:00:36 so far i'm struggling on getting the dependencies Mar 07 20:01:01 is there any way to add aptitude repositories to zypper? Mar 07 20:01:10 no Mar 07 20:01:17 javispedro: $800+ ebooks? you mean laptops? Mar 07 20:01:38 CosmoHill: no. one word: iRex Mar 07 20:01:50 these are the dependencies i need Mar 07 20:01:51 build-essential cmake qt4-qmake libx11-dev libcv-dev libcvaux-dev libhighgui4 libhighgui-dev libqt4-dev libpam0g-dev checkinstall Mar 07 20:02:06 so far I managed to get cmake, the equivalent of build-essentials and libx11-devel Mar 07 20:02:16 and i'm compiling opencv atm Mar 07 20:02:24 opencv is in the repositories Mar 07 20:02:28 it's called "opencv" Mar 07 20:03:07 lol...yes it is Mar 07 20:03:10 any idea on the others? Mar 07 20:03:25 libqt4 is there obviously, but with a different name Mar 07 20:03:34 qt-devel Mar 07 20:04:55 libhighgui probably will be the hardest bit, i've never heard of it Mar 07 20:05:44 unless it is part of opencv? Mar 07 20:05:59 is there any way I can see if these are already on the system? Mar 07 20:06:23 zypper search Mar 07 20:06:29 thx Mar 07 20:06:48 "i" means it is installed Mar 07 20:09:45 any idea if opencv comes with libcv-dev? Mar 07 20:17:48 this is nice: Mar 07 20:17:50 http://science.slashdot.org/story/11/03/07/1953213/Scientist-Records-First-5-Years-of-His-Sons-Life-Analyzes-Language-Development Mar 07 20:27:36 * lcuk goes decorating Mar 07 20:31:48 anyone seen this? http://www.digia.com/C2256FEF0043E9C1/0/405002251 Mar 07 20:32:11 yes Mar 07 20:33:01 sounds good for Qt ... but seems Nokia have decided to withdraw itself out of Qt ... Mar 07 20:34:09 not really, please see the complete story in the qt blog Mar 07 20:36:17 I see Nokia tries to soften the dramatic part of this ... but, after Nokia went to WP7, now selling out Qt parts ... I struggle to see that Nokia has long-term plans on anything else than WP7 Mar 07 20:36:43 seems to me like Nokia will now focus on the hardware, get rid of the software part and let Microsoft tackle that with WP7 Mar 07 20:37:58 dazo: think about this: Nokia is not divesting of its Qt R&D Mar 07 20:38:02 there's a reason for that Mar 07 20:38:17 but anyway, that's off-topic. MeeGo uses the open source version of Qt. Mar 07 20:39:11 * timeless_w7ip pops in Mar 07 20:39:22 hey timeless Mar 07 20:39:26 yeah, Nokia has started to sail their own direction ... I *really* hope other mobile vendors will come with MeeGo ... but I'm not going to put money into a Nokia device in a long time again Mar 07 20:40:05 * dazo don't want to go towards Android ... but that's the only real alternative to MeeGo - considering OS openness Mar 07 20:40:10 well, no one else has announced a MeeGo or MeeGo-like phone but Nokia Mar 07 20:40:19 oh, they've announced it now? Mar 07 20:40:28 Aava Mar 07 20:40:50 RST38h: for consumers? I don't think a 2500€ proto is something consumers will buy. Mar 07 20:41:01 hi thiago Mar 07 20:41:17 I doubt they created a company to sell$2500 phones to developers Mar 07 20:41:37 Even though it looks like Intel's pocket company Mar 07 20:41:42 in finland? Mar 07 20:42:05 I'd guess more like a nokia-ecosystem company Mar 07 20:42:13 like tieto, digia, roivio, etc. Mar 07 20:42:42 with moorestown phones? Mhmmm Mar 07 20:43:16 someone's gonna make a medfield phone, eventually Mar 07 20:43:26 I'm just waiting to see who and whether they put MeeGo on it Mar 07 20:44:20 Future will tell Mar 07 20:47:15 After a "successful" installation of meego from usb to my lenovo netbook I get a "Buffer I/O error on device sr0, logical block $BLOCK_COUNT" when booting from disk (disk is definitely OK!) does this ring a bell for somebody? sr0 in meego the standard root dev? Mar 07 20:48:08 no Mar 07 20:48:14 it's usually cdrom Mar 07 21:06:30 lbt: bug filed. Mar 07 21:06:55 Hello Mar 07 21:07:00 timeless_w7ip: ta Mar 07 21:07:07 ta ? Mar 07 21:07:10 "thanks" Mar 07 21:07:14 French ? Mar 07 21:07:18 nah, british Mar 07 21:07:21 Okay ! :) Mar 07 21:07:23 more like "thank ya" Mar 07 21:07:52 interesting Mar 07 21:08:02 * timeless_w7ip only stores translations not proper expansions Mar 07 21:08:35 ta is a contraction Mar 07 21:08:43 :D Mar 07 21:21:36 zomgbie: you can get this error/warning without it being a problem. What is the issue you're having? Mar 07 21:33:12 ~seen sivang Mar 07 22:06:24 pvanhoof, nice mail. But isn't it obvious that EDS only came in because of the decision to drop Buteo? Apparently, there's no interest in making SyncEvolution work with anything else but EDS. Mar 07 22:07:00 Again, another non-technical decision, I might say ;-) Mar 07 22:07:02 I'm not aware of what Buteo is. Is that something for synchronizing PIM data? Mar 07 22:07:22 ah yes, reading that part Mar 07 22:07:33 yeah Mar 07 22:07:40 http://wiki.meego.com/Buteo Mar 07 22:07:52 The problem is that on harmattan quite a few of Nokia's softwares use Tracker and also hope for the PIM data to be there Mar 07 22:08:21 well, it's certainly good to demand concrete numbers, not just handwaving Mar 07 22:08:38 This is among the reasons why the E-mail team although having their own storage, also store into RDF (because some apps do queries for which they need to know about this, and they don't want to merge queries from different sources together themselves. plus they want to join on the metadata from different domains) Mar 07 22:09:04 Well yes, I do wonder about those scalability issues. The point of tracker is of course to scale pretty ok Mar 07 22:09:50 But it's optimized per use-case. It also has quite a few optimization possibilities. And focus on performance has historically been on the query side, not entry side. Data entry optimizations have only recently started Mar 07 22:10:16 So yes, some concrete numbers and how measurements where made would help a lot Mar 07 22:10:20 * timeless_w7ip chuckles Mar 07 22:10:36 data entry by the browser was shown to be abysmal ages ago Mar 07 22:11:24 Browsers don't need to insert thousands of things, so if the use-case is only bookmark management then I can imagine that the priority for that was low Mar 07 22:11:24 on harmattan, things use tracker because someone ordered things to use tracker Mar 07 22:11:47 well, users don't use bookmarks, they use history Mar 07 22:11:49 The reason is applications integrating with each other's metadata and context Mar 07 22:11:53 timeless_w7ip, right, and not perhaps because of a fully integrated search Mar 07 22:11:53 so you're talking about every page ever visited Mar 07 22:12:10 mikhas: some hand waving Mar 07 22:12:20 timeless_w7ip, that's still not a very frequent use-case. And I wonder what the query was, because that should be reasonably fast Mar 07 22:12:32 pvanhoof: insertion, not query Mar 07 22:12:39 unless you consider INSERT to be a QUERY Mar 07 22:12:43 Yes, sparql update is also a query form Mar 07 22:12:45 (SELECT?) Mar 07 22:12:51 * timeless_w7ip shrugs Mar 07 22:12:58 i didn't do it, from memory the way it was done was stupid Mar 07 22:13:04 You insert-where, and the where part is a query like select Mar 07 22:13:05 but that's not my area of interest Mar 07 22:13:16 it was awful Mar 07 22:13:37 and as for whether the browser was considered a priority use case Mar 07 22:13:38 *shrug* Mar 07 22:13:43 the whole design was absolutely insane Mar 07 22:13:46 To insert a history should be straightforward now that I look at the ontology for browser history .. Mar 07 22:14:01 Can you elaborate why? Mar 07 22:14:10 then there was the fun thing about how there were multiple URI fields floating around Mar 07 22:14:19 so when someone tried to dispatch a browser object, it failed Mar 07 22:14:27 because someone was expecting a different URI Mar 07 22:14:30 joy Mar 07 22:14:43 the wonderful thing about standards is that there are so many from which to choose Mar 07 22:15:00 What about multiple uri fields? Mar 07 22:15:07 * pvanhoof doesn't understand this part :) Mar 07 22:15:22 nie:url is the only URL field Mar 07 22:15:40 there was a second one hidden somewhere else Mar 07 22:15:46 Which one? Mar 07 22:15:49 * timeless_w7ip shrugs Mar 07 22:15:59 we're talking about stupidity from last spring or summer Mar 07 22:16:11 Ok, and that stupidity is now fixed? Mar 07 22:16:20 i think at the time someone added a hack Mar 07 22:16:34 since ontologies tend to last forever, i'd assume it wasn't fixed properly Mar 07 22:16:37 since in general that's impossible Mar 07 22:16:55 if you have access to the internal bug tracker, you could probably find the report Mar 07 22:17:00 ah, there's nfo:uri too, that sounds like should be nie:url, but anyway Mar 07 22:17:07 congrats, you found it! Mar 07 22:17:13 see. it doesn't take a browser engineer! Mar 07 22:17:28 But why was there no bug report about this? Mar 07 22:17:33 ""? Mar 07 22:17:34 there was Mar 07 22:17:40 bug#? Mar 07 22:17:45 *shrug* surely you can find it Mar 07 22:17:55 i'd like to pretend i don't have access to that bug traccker Mar 07 22:18:08 my account expired last year, and i only accidentally reenabled it Mar 07 22:18:13 i have better things to do Mar 07 22:18:14 * pvanhoof searches. I wonder what the reason was Mar 07 22:18:28 Anyway, it also doesn't make things more difficult to have that field, nor do you have to use it Mar 07 22:18:38 * timeless_w7ip chuckles Mar 07 22:18:41 my team's dead Mar 07 22:18:45 ah ok Mar 07 22:18:50 we don't have to do *anything* Mar 07 22:18:57 and no one in their right mind would touch this crap Mar 07 22:19:03 so yes, i don't have to use it Mar 07 22:19:14 but please don't tell me you have the perfect product Mar 07 22:19:24 because i've waded through it, and i still remember the stench Mar 07 22:19:54 sorry, i'm not happy right now, someone played ping pong with me Mar 07 22:20:33 and i bet you could find the bugs that talk about insertion performance if you tried ;-) Mar 07 22:20:53 note that i am not claiming the browser code to do the insertion was high quality Mar 07 22:21:00 i make no claims about high quality code Mar 07 22:21:12 all software is crap Mar 07 22:21:19 now that we agreed on that, can we move on? Mar 07 22:21:23 timeless_w7ip, No im not saying tracker is perfect Mar 07 22:21:34 I just have questions about EDS Mar 07 22:21:44 i seem to remember disliking eds :) Mar 07 22:21:45 Legitimate ones Mar 07 22:21:52 * timeless_w7ip can't remember why Mar 07 22:22:03 And tbh isn't Arjan's current reply eum ... community compatible :) Mar 07 22:22:15 ? Mar 07 22:22:35 That's pretty much an "Meego is now an Intel-only shop" reply Mar 07 22:22:47 url? urn? uri? Mar 07 22:22:58 Ah, maybe I only received it yet Mar 07 22:23:05 nie:url? nfo:uri? Mar 07 22:23:14 ah no, it's on the ML Mar 07 22:23:19 there are urls for that Mar 07 22:23:25 minute Mar 07 22:23:45 http://lists.meego.com/pipermail/meego-dev/2011-March/481900.html Mar 07 22:23:56 Harmattan isn't Maemo Mar 07 22:24:18 It should have something to do with MeeGo. Reinventing the wheel yet again is eum a terrible idea Mar 07 22:24:22 it isn't frankenstein either Mar 07 22:25:00 So basically MeeGo is going to make the exact same mistake Nokia did when it went from Fremantle to Harmattan :) Mar 07 22:25:33 Anyway Mar 07 22:25:43 * pvanhoof goes back to coding :) Mar 07 22:26:59 meh... if harmetten is compliant, it's meego. if not, it isn't. up to whoever is developing it. i don't see why nokia should get a special pass on compliance. Mar 07 22:27:22 Maybe because Nokia steered development? Mar 07 22:27:32 Might have something to do with it Mar 07 22:27:43 i don't think that's a valid reason Mar 07 22:27:53 especially if they steered it off a cliff Mar 07 22:27:57 but hey Mar 07 22:28:04 we will see Mar 07 22:28:28 Either way, still no reason for Intel to make the exact same mistakes Mar 07 22:28:57 Nokia's developers didn't steer it off the cliff, upper management made a strange decision recently Mar 07 22:28:58 heh, wasn't it a year ago that tracker authors appeared on irc and ensured us it was "fixed"? Mar 07 22:29:51 A year ago sparql and sparql update were introduced I think Mar 07 22:30:03 That's a bit more than a year ago I think Mar 07 22:30:04 the mistake nokia made was never supporting the old thing after the new thing came out Mar 07 22:30:24 Right, sounds similar to what Intel is doing with MeeGo now isn't it? Mar 07 22:30:29 they could rewrite the firmware every year and nobody would care if they made it work on the old hardware Mar 07 22:30:41 since meego has no hardware yet, this isn't a problem Mar 07 22:30:43 and complaints about performance coming from @intel.con is damn frightening, they've got beefy hw to run it on :) Mar 07 22:30:59 ali: supporting old things wasn't really the problem Mar 07 22:31:11 it was from what i could see Mar 07 22:31:17 sure Mar 07 22:31:23 ShadowJK, Intel gave Tracker patches, the Tracker team at full priority reviewed and reworked the patch and that patch is now in. it's called the UpdateArray API Mar 07 22:31:25 ali: consider it this way Mar 07 22:31:34 if you don't rewrite things, but instead incrementally improve Mar 07 22:31:41 In fact has Intel contributed several patches and all of those pieces of work are integrated in tracker Mar 07 22:31:43 the likelihood of things still working on older hardware is much higher Mar 07 22:32:18 So upstream tracker isn't reluctant to help intel improving things Mar 07 22:33:19 but the real problem is that nokia's business model involves selling units Mar 07 22:33:29 and nokia is run by bean counters Mar 07 22:33:40 The team had to rework the patch because the way intel had it implemented was sending multiple insert queries as one transaction. This means that if one is wrong, all fail. That's not legitimate. If one is wrong, only one can error and the others might have to resume (if they don't depend on each other) Mar 07 22:34:07 if someone had a way to show that keeping users on old hardware gained money for nokia, then there'd be more reason to continue to support old hardware Mar 07 22:34:18 Development on tracker happens 100% upstream, not sure what Nokia's bean counters have to do with that Mar 07 22:34:30 pvanhoof: this is to ali1234 Mar 07 22:34:33 ah ok Mar 07 22:34:34 I'm not saying they're reluctant. I'm just amused that this daemon which has ranked high on list of user frustration is now after being "fixed" also on dev list of frustrations :) Good thing someone is taking notice and trying to come up with a solution though Mar 07 22:34:34 different conversation Mar 07 22:34:34 :) Mar 07 22:34:50 ShadowJK, what is the solution? Mar 07 22:35:09 ali: make sense? Mar 07 22:35:21 yes, it makes sense. i still don't like it though Mar 07 22:35:24 ali: note that the other problem is unit volume Mar 07 22:35:35 Intel approach of one big query makes sense. A single one takes/took like 5-10 secs on maemo5 :) Mar 07 22:35:54 What do you mean one big query? Mar 07 22:35:57 if you make 10 units of device 1. 100 units of device 2. 1000 units of device 3. and plan to make 10,000 units of device 4 Mar 07 22:36:17 ~curse FF on Ubuntu not working with opensource.samsung.com Mar 07 22:36:18 May the fleas of a thousand camels infest your most sensitive regions, FF on Ubuntu not working with opensource.samsung.com ! Mar 07 22:36:24 multiple as one transaction Mar 07 22:36:25 the return on any effort for devices 1..3 is less than the sales for device 4 Mar 07 22:36:30 ShadowJK, And which query takes 5-10 s on maemo5? And what has Fremantle to do with this? Mar 07 22:36:35 which means if you allocate resources you take everything for device 4 Mar 07 22:36:46 ShadowJK, yes and that solution is in Harmattan's Tracker Mar 07 22:36:55 lardman: ff1.0.9? Mar 07 22:37:24 3.6.14 apparently Mar 07 22:37:32 (actually, i don't think i've seen ff1.* in a while, but i'm pretty sure i saw a 2.0.0.8 or something not too long ago) Mar 07 22:37:45 lardman: please try ff4b12 Mar 07 22:37:54 (or if ff4rc1 is available...) Mar 07 22:38:06 timeless_office: I'll see if it's in the Ubuntu repos Mar 07 22:38:15 timeless_w7ip: see above ;) Mar 07 22:38:16 btw, you get points for having .14 instead of .13 Mar 07 22:38:25 iirc there's a ppa Mar 07 22:38:57 https://launchpad.net/~ubuntu-mozilla-daily/+archive/ppa Mar 07 22:39:31 the package is firefox-4.0 :o Mar 07 22:39:50 how nice of them, you can probably have both installed at the same time :o Mar 07 22:40:08 * lardman needs an apt line Mar 07 22:40:24 if you have js enabled there's a link you can click which gives you the apt lines Mar 07 22:40:35 ah ok, will take a look see Mar 07 22:40:55 oh, it's a select, not a link. whatever. Mar 07 22:41:28 oh, no, it's both, you have to click 'Technical details about this PPA ' first Mar 07 22:41:37 then you use the select Mar 07 22:43:33 sudo add-apt-repository ppa:ubuntu-mozilla-daily Mar 07 22:43:38 job done Mar 07 22:43:44 ali: bah, too simple Mar 07 22:43:47 pvanhoof, anything that triggers fsync or fdatasync on ext3 on flash and you can expect all other I/O to block for 2-10s :) Mar 07 22:43:50 we like complicated stuff Mar 07 22:44:09 ShadowJK, you do know that tracker's rdf store only does fsync on its journal, right? Mar 07 22:44:26 ShadowJK, tracker's rdf store's sqlite database uses synchronization=off Mar 07 22:44:26 i think you need to put /ppa on the end so it's alittle more complicated Mar 07 22:44:40 ShadowJK, just saying, read the source my friend Mar 07 22:44:51 But presumably intel isn't running ext3 on flash, so I wonder what it's doing for them to make them sad Mar 07 22:45:16 http://mxr.meego.com/repo.meego.com/source/tracker/ or http://mxr.meego.com/meego.gitorious.org/source/tracker/ fwiw Mar 07 22:45:22 ShadowJK, Also note that tracker doesn't do an fsync automatically: Mar 07 22:45:26 and magically, you can switch from one to the other :o Mar 07 22:45:45 ShadowJK, i'll explain why: it's not expected that the device can be shutdown uncleanly. And fsync is only useful when you can shut down the FS uncleanly Mar 07 22:46:03 * timeless_w7ip wonders why SPARQL is only a link on one of those Mar 07 22:46:05 ShadowJK, there is a DBus call Resources.Sync() that does the fsync on the journal, though. If apps what to be sure Mar 07 22:46:21 But since that's easily readable in the source-code, no need to explain that Mar 07 22:46:43 answer: it's defined as an identifier in libqttracker which apparently hasn't been found by the meego.gitorious.org crawler Mar 07 22:47:07 timeless_w7ip, documentation is here: http://live.gnome.org/Tracker/ Mar 07 22:47:14 not interested :) Mar 07 22:47:21 i'm not an app developer Mar 07 22:47:29 i own mxr.meego.com, so i'm advertising it :) Mar 07 22:47:32 Ah it was just a troll ? :) Mar 07 22:47:43 ok ic :) Mar 07 22:47:53 legitimate on topic advertising! Mar 07 22:48:23 http://mxr.meego.com/repo.meego.com/ident?i=fsync&tree=repo.meego.com&filter=tracker Mar 07 22:48:45 that's not a troll, this is a troll: "hey, isn't tracker that thing that made my ubuntu go really slow?" Mar 07 22:48:59 * timeless_w7ip chuckles Mar 07 22:49:15 yes, those are the correct places for fsync use timeless_w7ip (that doesn't mean that these syscalls are called constantly, they are called when you do Resources.Sync() in DBus from an app) Mar 07 22:49:31 ali1234, which version of Tracker? 0.6.x isn't the same project anymore Mar 07 22:49:39 i dunno, i was just trolling Mar 07 22:49:47 aaargh, even ff 4.0 is beaten by the Samsung site Mar 07 22:50:07 * timeless_w7ip ponders Mar 07 22:50:13 works under Win7 64bit mind you with ff4.x Mar 07 22:50:19 On the other hand, EDS used to be the largest swapstorm trigger on my desktop computer until I figured out how to uninstall it (and then app developers started spawning mysqld instead, *sigh*) :) Mar 07 22:50:29 heh Mar 07 22:50:40 ShadowJK, right. Which is why I have questions about EDS on the ML Mar 07 22:50:41 timeless_w7ip: just locks up trying to download some kernel source code Mar 07 22:50:43 i'm surprised you haven't complained about sqlite3 Mar 07 22:51:09 pvanhoof: hey Mar 07 22:51:15 what's the opposite of g_open()? Mar 07 22:51:22 g_close? Mar 07 22:51:48 if eds uses some fsync happy sqlite3 we'd be back to fremantle or diablo level of pain :) Mar 07 22:51:49 Might be just close() Mar 07 22:52:07 oddly no. Mar 07 22:52:08 http://mxr.meego.com/repo.meego.com/source/glib2/glib/gstdio.h#64 Mar 07 22:52:15 (to g_close?) Mar 07 22:52:24 ShadowJK, I know they are not using synchronous=off in sqlite, so yes they are causing that libsqlite3 does fsyncs Mar 07 22:52:40 boy that's logical Mar 07 22:52:50 it has no opposite? Mar 07 22:52:51 And if they are using it, then they should have a journal themselves, and afaik they don't have that Mar 07 22:52:54 ali: right Mar 07 22:53:00 you just have to use close() Mar 07 22:53:03 gah Mar 07 22:53:09 ali: see my last link Mar 07 22:53:16 oh joy, so that's 5s freeze for every little event that gets stuffed into eds then :D Mar 07 22:53:17 yes, i did Mar 07 22:53:27 although btrfs would hide some of it Mar 07 22:53:47 timeless_w7ip, I think windows's open() isn't the same as posix or something and so the glib ppl wrapped it. Whereas close is the same everywhere, or something like that Mar 07 22:54:07 55 #if defined(G_OS_UNIX) && !defined(G_STDIO_NO_WRAP_ON_UNIX) Mar 07 22:54:12 No idea, sometimes those wraps are for platform portability, sometimes I find them strange too Mar 07 22:54:12 * lardman assumes this is a sign from God telling me to not look at kernel source this evening Mar 07 22:54:43 timeless_w7ip: g_* family of glib functions just wrap those that accept filenames in order to potentially do charset conversion Mar 07 22:55:07 ShadowJK, im not saying eds will cause that behaviour. It might be that eds will be used far less often for writing things to. Tracker on harmattan, however, is being written to constantly by applications (which is the applications doing that, not tracker, mind you) Mar 07 22:55:31 ShadowJK, so that meant that such fsync() hickups where not acceptable, and that's why it uses synchronous=off mode and does it have its own journal Mar 07 22:55:33 javispedro: clever Mar 07 22:55:47 oh right Mar 07 22:55:53 The journal is also used when making backups. Making a backup is just a copy of the journal to your destination. Mar 07 22:56:12 Restore is just placing your backup at the position of the journal, nuking meta.db, and replaying the journal Mar 07 22:56:13 because the _open() call is probably file system encoding whereas _wopen() is probably ucs32 or something Mar 07 22:56:25 err ucs2 Mar 07 22:56:42 pvanhoof, this sounds very sensible :) Mar 07 22:57:23 ShadowJK, tracker developers are sensible. But they do receive a lot of hate unfortunately. Apparently because Fremantle's tracker was bad. But on harmattan it's a completely different project Mar 07 22:57:27 hey, I've been doing sqlite on n900 stuff recently too =) Mar 07 22:58:27 * javispedro also decided to do synchronous=0 but also journal=memory, otherwise it max at some tens of exclusive transactions per second... Mar 07 22:58:38 pvanhoof, seems it has been getting reinvented by different team for each release then? Mar 07 22:59:32 ShadowJK, in fremantle the current team had to deliver something in a very terse timeframe, meanwhile was a group of developers starting on a superior solution. then the team merged and worked only the new project only Mar 07 22:59:50 so, part of the current team had to deliver something in a very terse .. Mar 07 22:59:53 shadowjk: actually on this, i think they did a good job Mar 07 23:00:13 So ill fully accept that fremantle's tracker was piece of shit Mar 07 23:00:14 they managed to split and do best effort and close to ideal impl Mar 07 23:00:39 pvanhoof: it actually did work though, sadly i don't have the proof Mar 07 23:00:41 But I think that by now each and every line of tracker 0.6 has been replaced Mar 07 23:00:47 because the one good demo of it isn't available anymore Mar 07 23:01:03 * timeless_w7ip has a terrible time remembering the nameof the useful search engine Mar 07 23:01:06 veveo? Mar 07 23:01:14 that's closed Mar 07 23:01:32 and it's for text search only, not for relationships between resources Mar 07 23:01:35 yeah, but it was an excellent demo showing that the fremantle tracker actually worked Mar 07 23:01:45 I don't think that was on fremantle.. Mar 07 23:01:50 Maybe they made a demo of course Mar 07 23:01:51 relationships smationships Mar 07 23:02:10 pvanhoof: hypothetically it was available for use and perhaps even offered to he ovi store Mar 07 23:02:22 I know that a team at nokia opens meta.db manually and reads textual data and then makes it searchable on a quite good textsearch solution Mar 07 23:02:23 hypothetically i was happy for many months Mar 07 23:02:29 I think that might be veveo Mar 07 23:02:36 heh Mar 07 23:02:40 But it duplicates the data Mar 07 23:02:53 * timeless_w7ip shrugs Mar 07 23:02:58 Yeh not nice solution Mar 07 23:03:00 that happens anyway Mar 07 23:03:14 Tracker's own FTS is disabled on harmattan (for that reason) btw Mar 07 23:03:15 tracker is already a duplicate of data Mar 07 23:03:26 FTS in tracker is an index on the textual data btw Mar 07 23:03:35 So the textual data is in meta.db anyway, the FTS is an index on that Mar 07 23:03:35 what, so it doesn't clutter the other thing? Mar 07 23:03:54 But that FTS is ifdeffed out of harmattan's tracker Mar 07 23:04:08 Since it's not used, making the index is an unneeded performance impact Mar 07 23:04:37 actually, i think most of the problems w/ tracker on the n900 had to do w/ the crawlers Mar 07 23:04:46 But tbh I don't know how veveo works, and if I'd know I'd be NDAd Mar 07 23:04:51 which were not really tracker, and were just buggy and sloppy in their own right Mar 07 23:05:00 Mostly users are angry at the indexers (of any software) doing one of 3 things: a) indexing. b) not indexing. c) triggering crashes in dsp code when indexing, requiring reboots to recover :) Mar 07 23:05:10 Yes but again, tracker on n900 is completely different than what I'm talking about here Mar 07 23:05:23 all i know is that hypothetically there was this thing from veveo which made me happy on the n900 :) Mar 07 23:05:23 ShadowJK, yes, should all be fixed Mar 07 23:05:35 until i stopped being able to install it :( Mar 07 23:05:36 ShadowJK, there are some know problems about the thumbnailer on video thumbnails atm. Being solved Mar 07 23:05:43 ShadowJK: you forgot 4) using 100% CPU to reindex every file on the device because i uploaded 1 new jpg Mar 07 23:05:50 or on diablo it triggered device reboots because the crawler was so heavy watchdog thought device had hung :) Mar 07 23:06:04 ShadowJK, some formats sometimes cause DSP crashes. Actually the problem is almost always in gstreamer or some DSP code somewhere, not the thumbnailer itself Mar 07 23:06:14 * timeless_w7ip sighs Mar 07 23:06:17 You can trigger the same crash by just playing the movie Mar 07 23:06:21 dsp crashes = scary as hell Mar 07 23:06:32 users are angry at indexers period ;) Mar 07 23:06:33 yes well, if the requirement is: make thumbnails! Mar 07 23:06:35 You know :) Mar 07 23:06:37 "arbitrary code execution + full root exploit" Mar 07 23:06:45 And the video crashes the dsp Mar 07 23:06:54 then you know, what will the first process be that crashes the dsp? Mar 07 23:07:01 So who gets the blame? Mar 07 23:07:05 * timeless_w7ip chuckles Mar 07 23:07:10 first to find Mar 07 23:07:16 :) Mar 07 23:07:17 so tracker :) Mar 07 23:07:20 Not really fair to blame the thumbnailer, right? It just happens to be the first :) Mar 07 23:07:34 fair? who plays fair? Mar 07 23:07:37 need a good fuzzing :) Mar 07 23:07:38 Well tracker's metadata extraction doesn't use the dsp Mar 07 23:07:42 It's the thumbnailer Mar 07 23:07:58 pvanhoof: most people don't know the difference Mar 07 23:08:03 so they'll still blame tracker :) Mar 07 23:08:05 For us thumbnailer and tracker are seperate, for people reporting bugs they are the same :) Mar 07 23:08:36 tomato, potato, who cares? they're both fruit of the ground Mar 07 23:08:56 * timeless_w7ip frowns Mar 07 23:09:11 probably the only person in this channel who could recognize that statement is sivang Mar 07 23:10:17 google + boreh if you want to understand the reference Mar 07 23:11:23 ShadowJK, so on the topic of indexing... Mar 07 23:12:01 It was planned to ship the device with a preinstalled database, so no initial indexing would be needed. That support was planned to be made usable for distributions too Mar 07 23:12:21 Right now with the WP7 announcement I think it's fair to say that priorities have changed Mar 07 23:12:44 oddly, in theory priorities for harmattan haven't changed much Mar 07 23:12:53 You need a way to detect when the user is at least 30 seconds' distance away from device, and a way to undo memory related side-effects of crawling when user starts approaching device :) Mar 07 23:12:59 well, other than actually focusing *more* attn on harmattan Mar 07 23:13:05 As for indexing itself it should have been greatly improved. But there are still improvements possible. For example we have one request to make the signal that tells apps that a file's metadata is deleted happen faster on a unlink of a file Mar 07 23:13:17 shadowjk: well.... Mar 07 23:13:20 Right now that's between 1 and 3s on the proto I think, and that's unacceptable idd Mar 07 23:13:25 you can detect motion and proximity Mar 07 23:13:46 you can also detect cpu usage Mar 07 23:13:54 ShadowJK, well we had plans to integrate with the power management Mar 07 23:14:03 a device that's stationary but using a lot of cpu is playing a movie or recording a video Mar 07 23:14:06 For example, the most battery intensive task isn't indexing but thumbnailing Mar 07 23:14:07 = don't harm Mar 07 23:14:18 kill mass-storage mode to avoid sudden appearance of files :) Mar 07 23:14:20 a device that's stationary with the screen blanked is a good candidate for indexing Mar 07 23:14:32 Thumbnails can be requested on background priority. Right now that uses SCHED_IDLE, but we want to improve that so that it only gets started on wall-power, for example Mar 07 23:14:39 But again, priorities have shifted Mar 07 23:14:48 Else this would be implemented around this time Mar 07 23:15:05 SCHED_IDLE is pretty much what you propose btw Mar 07 23:15:21 wall-power &! usb please Mar 07 23:15:34 ali: we know how m many amps we're getting Mar 07 23:15:39 Do anything with the device, and the processes that you make do something are all higher prioritized than SCHED_IDLE Mar 07 23:15:40 if it's >0.5, then it's a wall Mar 07 23:15:40 nothing worse than pluggin in usb to do something and then the phone starts going nuts and doing everything but what you want Mar 07 23:15:49 ali1234, that's why I said wall-power and not USB-power Mar 07 23:16:08 There are three levels of wattage I think, only the highest is wall-power Mar 07 23:16:14 well, some phones have separate wall power and can't charge on usb... some *nokia* phones that is Mar 07 23:16:16 wall power + no user Mar 07 23:16:17 sadly, i doubt the ui distinguishes between wall and pc Mar 07 23:16:18 Context-Framework can give you those details Mar 07 23:16:20 well, CPU was never much of an issue I think? I/O and memory starvation is the thing making UIs stutter :) Mar 07 23:16:27 Is it possible to run the QEMU emulator in a Windows 7 32-bit VM? I had tried using 64-bit version but it would keep crashing and it also crashes in my VM when I'm trying to use the Netbook Emulator. Any help? Mar 07 23:16:28 therefore these are not mutually exclusive Mar 07 23:16:29 pvanhoof: actually, there's more Mar 07 23:16:37 0.1 <...> 0.5, 1.0 .. 1.2 Mar 07 23:16:45 at least with the n900 those were all possible Mar 07 23:16:52 timeless_w7ip, yes possible. I didn't yet have to look into it because .. priorites Mar 07 23:16:55 you could probably end up w/ 1.1 too if you found the right wall Mar 07 23:17:20 N900 can distinguish 500mA USB from 500mA wallcharger too. Mar 07 23:17:34 shadowjk: does it visibly? Mar 07 23:17:38 no Mar 07 23:17:39 On fremantle was tracker also requesting thumbnails for all images that it saw, it doesn't do that on harmattan Mar 07 23:17:40 well yeah... when you plug in on USB it says "what do you want to do?" Mar 07 23:17:47 apps must request thumbnails themselves, always Mar 07 23:18:04 So that on fremantle was mind-boggling stupid Mar 07 23:18:10 * timeless_w7ip nods Mar 07 23:18:23 And dunno if bme exports it, but it is (or was) in /sys atleast Mar 07 23:18:28 That's something I think somebody should patch out of tracker 0.6 and overwrite on all n900s Mar 07 23:18:37 ali1234: Only on USB host, not on charger. Mar 07 23:18:43 pvanhoof: can you talk to the cssu people about it? Mar 07 23:18:46 SpeedEvil: exactly Mar 07 23:18:47 Some apps will not show thumbnails correctly, but those are wrong to be honest Mar 07 23:18:58 Doing thumbnails in tracker is not stupid. Mar 07 23:19:05 That's the apps that don't use the thumbnail api, but expect thumbnails to be on the FS Mar 07 23:19:06 But it should never, ever impact the user. Mar 07 23:19:16 this is what i am saying: if usb storage, or pc suite, or the other one are active, don't index Mar 07 23:19:29 SpeedEvil, problem is that thumbnailing is very cpu and I/O intensive. So even if it doesn't impact the user, it still drains battery Mar 07 23:19:35 media-whatever Mar 07 23:20:12 actually, usb has perhaps a dozen thingies Mar 07 23:20:13 SpeedEvil, and apps want immediate responsiveness for thumbnails that aren't cached to appear, so they all use high priority. Tracker's request use SCHED_IDLE on fremantle too. But SCHED_IDLE doesn't cause less battery drain Mar 07 23:20:22 the n900 thankfully masks them as two categories Mar 07 23:20:51 And battery time is a bad selling point of the n900. Well, patch auto-thumbnailing out of tracker and it'll improve Mar 07 23:21:27 One line patch Mar 07 23:21:28 pvanhoof: I mean something like only thumbnail if on power, and only if the system is idle for >10s, and only if you can do it without causing memory or IO pressure in case something is needed 'now'. Mar 07 23:21:44 pvanhoof: Unless app actually requests thumb. Mar 07 23:21:48 SpeedEvil, yes, sure, all that isn't implemented as such on the n900 Mar 07 23:21:55 could be worse, i mean, gnome-thumbnailer regularly crashes my computer when it tries to thumbnail 1.8GB tiff files and other such nonsense Mar 07 23:22:01 My operator's packet data has been broken last few days. Battery life is measured in days now when radio isn't used :) Mar 07 23:22:04 And none of that is difficult. Just make the patches and propose. Im sure they'll accept the package Mar 07 23:22:35 ali1234, fremantle's thumbnailer, althuogh a piece of shit software, has a size limit Mar 07 23:23:00 Harmattan's uses quillimagefilter, which should cope with large images Mar 07 23:23:05 i'm sure the gnome one used to but obviously some nutcase with 32GB and 16 cores decided it wasn't needed Mar 07 23:23:11 I think quill uses tiling algorithms to make it work Mar 07 23:23:31 ali: why do you have 1.8gb tiffs? Mar 07 23:23:47 And well, the reason why gnome's thumbnailers crash on this is gdbpixbuf which basically decompresses the 1.8gb tiff file to a RGB ram buffer Mar 07 23:23:48 jpeg can be partially decoded to get a fast low res image :) Mar 07 23:24:02 Which is a bad idea for 1.8Gb tiff files Mar 07 23:24:13 timeless_w7ip: minecraft Mar 07 23:24:17 Hi, I wanted to know if it was possible to run the meego emulator for netbooks on a Windows 7 32-bit virtual machine. I just tried and it keeps hanging and crashing on me Mar 07 23:24:21 "what's that?" :) Mar 07 23:59:34 cyas Mar 08 00:33:04 Stskeeps are you there? Mar 08 00:34:37 uh. Mar 08 00:36:45 do we have someone to fill the void VDVsx left as organizer of GSOC? Mar 08 00:37:23 s/organizer/administrator, ... Mar 08 00:40:36 * javispedro makes mental note to repeat question at at a better time. Mar 08 00:41:01 javispedro, ask on the MeeGo ML? Mar 08 00:41:44 there has been a few questions by prospective students with no answer, so right now I fear the answer is "not". Mar 08 00:42:40 And while a wanted to be a mentor, I'm not sure I will be able to find time to fill in the paperwork required to be an admin... Mar 08 00:43:17 ugh Mar 08 00:59:26 ok, mail fired. please answer! =) Mar 08 00:59:30 anyone here familiar w/ spec files? Mar 08 00:59:34 javispedro? Mar 08 00:59:41 * javispedro is a deb guy Mar 08 00:59:56 if you're asking what the mail is about, see a few lines above. Mar 08 01:00:49 how about the role of someone who speaks .spec? :) Mar 08 01:01:02 ali1234? :) Mar 08 01:01:05 a spec.ialist? Mar 08 01:01:20 michaelg|nok: can you help ? :) Mar 08 01:01:29 what does it mean for a spec file to have: Mar 08 01:01:30 nooo it's been a long time since I've written them Mar 08 01:01:37 Name: helloworld Mar 08 01:01:42 Obsoletes: helloworld Mar 08 01:02:07 umm. that seems a bit confused though Mar 08 01:02:13 really? Mar 08 01:02:43 name of package Mar 08 01:03:15 and list of packages it replaces Mar 08 01:03:16 ? Mar 08 01:03:37 ngharo: see michaelg|nok's observation Mar 08 01:04:04 also surely helloworld is baked into the platform these days... :) Mar 08 01:04:28 in this case, both lines had meego-netbook-settings Mar 08 01:05:30 oh, i see Mar 08 01:09:22 perhaps its supposed to enforce only one of these to be present Mar 08 01:09:34 ? Mar 08 01:09:38 the only reason i can think of to do that is because the package is obsolete and not replaced by anything else Mar 08 01:09:49 so that you can't have 2 versions of the package Mar 08 01:09:52 "don't install me?" Mar 08 01:09:59 "upgrade and remove me?" Mar 08 01:10:03 more than that, also, "if i'm installed, uninstall me" Mar 08 01:10:24 but it's just a theory Mar 08 01:10:30 upgrade-and-remove-old would make sense, a little, but its a guess Mar 08 01:12:17 or maybe its a bug Mar 08 01:18:42 so, anyone have meego and want to find out? :) Mar 08 01:18:49 which version? Mar 08 01:18:52 um Mar 08 01:19:02 0.43-1.1 Mar 08 01:19:04 i think Mar 08 01:19:15 or the other way around Mar 08 01:19:38 nah, the way i wrote it Mar 08 01:19:42 no, which version of meego? Mar 08 01:19:51 meego is versioned too? Mar 08 01:19:56 * timeless_w7ip ponders Mar 08 01:19:56 um... yeah Mar 08 01:20:10 is there a package that'll tell me? :) Mar 08 01:20:13 1.1.0, 1.1.1, 1.1.80, 1.1.90... Mar 08 01:20:34 basically, is this stable, or trunk? Mar 08 01:20:53 no clue sadly Mar 08 01:21:03 well, where did you get the rpm from? Mar 08 01:21:04 the code i have to crawl the repo is capable of crawling anything Mar 08 01:21:29 and since i haven't version controlled it, i can't easily tell which one it crawled Mar 08 01:21:53 so, is there a package that actually would tell you which repo it is? Mar 08 01:22:06 its in the directory structure Mar 08 01:22:26 javispedro, :) you joined twitter Mar 08 01:22:33 evening chaps \o Mar 08 01:22:42 lcuk2: yeap, learning it Mar 08 01:23:01 is there a file that provides lsb_ stuff which would indicate the release? Mar 08 01:24:30 'meego-netbook-settings' isn't a package in netbook 1.1 stable Mar 08 01:26:14 also, version is in /etc/meego-release Mar 08 01:26:54 http://mxr.meego.com/repo.meego.com/source/libzypp/zypp/ZConfig.cc.meego-release is the only meego-release file Mar 08 01:27:50 meego-release-1.1-2.1.noarch provides it Mar 08 01:28:16 1.1.90-1.2 Mar 08 01:28:27 that's presumably the source that generates it... Mar 08 01:29:30 ok so trunk then Mar 08 01:29:33 which i don't currently have installed (nuked it for tablet demo) Mar 08 01:29:33 yeah Mar 08 01:29:45 well, you can grab the (s)rpm Mar 08 01:31:05 * timeless_w7ip wonders why meego-release isn't in mxr Mar 08 01:31:58 whats the name of the src.rpm ? Mar 08 01:32:29 won't mean much to me Mar 08 01:32:50 meego-netbook-settings-0.43-1.1.src.rpm Mar 08 01:33:12 i try to get it in a few mins Mar 08 01:33:24 connman hung up again Mar 08 01:34:49 or my wifi is more hosed than normal, that could be too Mar 08 01:40:40 can I find out accurately within the OS what the image used is? Mar 08 01:40:59 ecause Settings/Product Version is wrong on MeeGo atm Mar 08 01:41:01 how accurate? Mar 08 01:41:07 hheh Mar 08 01:41:18 well more accurate than giving me a Maemo version number from last year would be good. Mar 08 01:41:24 https://bugs.meego.com/show_bug.cgi?id=14234 Mar 08 01:41:25 meego-release Mar 08 01:41:27 <_MeeeGoBot_> Bug 14234 nor, Undecided, ---, michael.leibowitz, NEW, About Product Dialogue shows rootfs (Maemo) product string Mar 08 01:43:18 timeless_w7ip, I was reminded by you digging that I needed to file that bug Mar 08 01:43:26 I noticed it over the weekend and giggled a bit Mar 08 01:44:03 timeless_w7ip, in another thing you would find amusing Mar 08 01:44:06 "TODO: License text needed for meego.com" Mar 08 01:44:10 is on the same page Mar 08 01:44:19 heh Mar 08 01:46:08 timeless_w7ip: it builds it ok here, don't know what that means Mar 08 01:46:39 berndhs: is there a fille named README which doesn't look like a readme? Mar 08 01:47:05 looks like instructions for something in the README Mar 08 01:47:19 it feels like a strange README file Mar 08 01:47:42 it would probably execute as an sh script, but that's probably accidental Mar 08 01:48:25 so, i have a problem... Mar 08 01:48:32 some of the packages i'm dealing with are evil Mar 08 01:48:45 which means that i'm going to need to not use rpmbuild -bp Mar 08 01:48:54 i'd like to leave a file explaining what i've done Mar 08 01:49:04 i was vaguely considering using 'README' Mar 08 01:49:17 but there are a couple of readme files in srpms Mar 08 01:49:31 i would put a huge banner on every html page of that package Mar 08 01:49:49 i'm considering: README, meta-data/, sources/ Mar 08 01:50:02 where the README is an explanation + rpmlint output Mar 08 01:50:21 how about README_why_I_put_it_here Mar 08 01:50:31 and meta-data/ is all the non tarball files from the srpm, with sources/ having all of the expanded tarballs Mar 08 01:50:40 i'm afraid people won;t see that Mar 08 01:50:43 berndhs: the xref has magic handling for README Mar 08 01:50:58 the way it will get used is people will link directly to file/line of code Mar 08 01:51:31 yes but README is a legitimate file for many purposes, so you probably can't assume that its yours Mar 08 01:52:03 http://mxr.meego.com/repo.meego.com/find?string=README Mar 08 01:52:07 is a set of readmes Mar 08 01:52:21 what's important is how the directory list looks, e.g. Mar 08 01:52:22 http://mxr.meego.com/repo.meego.com/source/udisks-helper-extras/ Mar 08 01:52:27 note the text at the top above the list? Mar 08 01:53:04 yeah, most of my website works like that Mar 08 01:53:23 ali: yeah, if people link directly, there's not much i can do Mar 08 01:53:26 that's ok Mar 08 01:53:44 i'm just hoping that anyone who browses into the directory would see it Mar 08 01:53:55 well, you can add more file types than README Mar 08 01:54:02 to that mechanism Mar 08 01:54:09 it actually supports a bunch Mar 08 01:54:43 but if there's a README in some srpms, there's no guarantee whatever i pick won't be in it too Mar 08 01:56:31 unless you pick README_WHY_THIS_PACKAGE_IS_BROKEN Mar 08 01:56:37 heh Mar 08 01:56:38 i doubt you'll find that in any package Mar 08 01:59:39 or README_FOR_XREF, something really specific Mar 08 02:03:49 * ali1234 reopens bug 5266 Mar 08 02:03:52 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=5266 nor, Medium, ---, vivian.zhang, REOP, Changing the default OS in "boot loader operating systems" in anaconda installer has no effect Mar 08 02:10:07 bug 1 Mar 08 02:10:10 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=1 maj, Medium, ---, vivian.zhang, VERI FIXED, [Tracker] tracker-miner-fs should be auto-started after system boot Mar 08 02:10:23 I like ubuntu's #1 better. Mar 08 02:10:49 i don't Mar 08 02:10:55 it shows out of date thinking Mar 08 02:11:06 desktop market share really shouldn't be a priority Mar 08 02:11:14 and when you look at unity, it clearly isn't Mar 08 02:14:58 ok. Mar 08 02:15:15 s/iOS/MeeGo/g Mar 08 02:15:24 erm... Mar 08 02:15:34 :-p Mar 08 02:15:36 s/Windows/iOS/g? Mar 08 02:15:47 yeah... but I totally spoiled it. Mar 08 02:15:53 I'll go stand in the cornere. Mar 08 02:15:57 corner. Mar 08 02:15:59 coroner. Mar 08 02:16:43 did y'all come to a conclusion on the "Obsoletes: meego-netbook-settings" thing? Mar 08 02:16:51 Looks like a goof to me. Mar 08 02:16:57 oh yeah, i forgot with testing the bootloader thing Mar 08 02:18:16 ali1234: thanks for reopening that bug. I think I tried it over the week-end and got the same as you. Mar 08 02:18:32 I filed a couple other related bugs, though. Mar 08 02:18:59 i filed a few for the bootloader already... Mar 08 02:19:16 anything i should take a look at? Mar 08 02:21:12 gabrbedd: perhaps some brave soul can file a big on the Obsoletes: self Mar 08 02:21:15 to find out Mar 08 02:21:19 ok, i have meego-netbook-settings installed, zypper doesn't want to remove it or anything... so i have to agree it looks like a mistake Mar 08 02:21:38 Sure. Bug 14190 and... Mar 08 02:21:41 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=14190 nor, Low, ---, neo.fang, NEW, Partition editor in installer gets confused if primary partition at end of drive Mar 08 02:22:22 ooh, that is a strange one :) Mar 08 02:22:37 recovery partitions are pretty common... Mar 08 02:22:45 Oh yeah, I didn't file one about not touching the MBR... since I don't really care if they fix that one. :-) Mar 08 02:22:52 or "fix" that one. Mar 08 02:23:18 but microsoft has a new trick: win7 uses two primary partitions, then you have two recovery partitions. so you can't repartition without breaking your windows 7 install somehow Mar 08 02:23:21 berndhs: eh, why not? Mar 08 02:23:36 "not touching the mbr" is already filed :) Mar 08 02:23:52 ali1234: ok. Mar 08 02:24:00 bug 12816 Mar 08 02:24:03 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=12816 enh, Undecided, ---, gavin.hindman, NEW, [FEA] No option to install without bootloader step Mar 08 02:24:13 it used to be an option but it got removed somehow Mar 08 02:24:38 ali1234: Ok, I'm on the CC list, now. Mar 08 02:24:48 i think it is important for testers to not be annoyed when installing images to test stuff :) Mar 08 02:25:25 upgrade install would be nice too, so I dont have to rebuild everything Mar 08 02:25:58 chainloading /to/ an extlinux install from the Win7 bootloader was... wierd to set up. Mar 08 02:26:10 there is also bug 12815 and bug 12813 related to install/bootloader stuff Mar 08 02:26:13 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=12815 enh, Low, ---, vivian.zhang, ASSI, Installer does not detect existing Linux installations Mar 08 02:26:15 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=12813 enh, Low, ---, vivian.zhang, ASSI, Bootloader menu is too difficult to access Mar 08 02:26:23 But I actually got the bastard to work. Mar 08 02:27:00 Anybody loaded chain-loaded extlinux from grub? Mar 08 02:27:58 * gabrbedd can't form decent sentances tonight... Mar 08 02:28:06 * gabrbedd can't spell, either. Mar 08 02:28:54 i usually upgrade fedora after i install meego, to get the fedora grub back Mar 08 02:29:28 there is a shorter way of doing it, run grub directly, but i'm too lazy to figure out how to do that right Mar 08 02:30:23 or its grub-install --recheck or something Mar 08 02:30:50 Here's how I had to do it with win7: http://wiki.meego.com/Using_MeeGo_with_the_Windows_7_Boot_Loader Mar 08 02:31:12 extlinux is looking for the partition that's marked bootable... Mar 08 02:31:21 so I bet something similar is required with grub. Mar 08 02:31:56 yes its analogous, you first trash your MBR and then re-make it Mar 08 02:34:19 * gabrbedd files bug 14236 Mar 08 02:34:21 <_MeeeGoBot_> Bug https://bugs.meego.com/show_bug.cgi?id=14236 tri, Undecided, ---, priya.vijayan, NEW, Package meego-netbook-settings declares "Obsoletes: meego-netbook-settings" Mar 08 02:35:22 berndhs: Naw... the trick was that I /needed/ extlinux to write to the MBR. Mar 08 02:35:27 I then saved that MBR... Mar 08 02:35:34 Restored the win7 MBR... Mar 08 02:35:54 And fed the extlinux MBR to the the win7 boot loader. Mar 08 02:39:09 sorry, got disconnected Mar 08 02:39:56 gabrbedd: i don't chainload extlinux, but ubuntu's grub can boot meego with no problems (not tested with btrfs) Mar 08 02:39:59 * javispedro uses grub to boot the meego kernel.. Mar 08 02:40:19 yeah, I had to put the kernel on a separate partition Mar 08 02:40:24 (where grub itself already resided) Mar 08 02:40:25 all it takes is a "sudo update-grub" Mar 08 02:40:27 but that's about it. Mar 08 02:41:09 well, not even MeeGo can boot to btrfs. :-) Mar 08 02:41:14 yeah Mar 08 02:41:19 that's why i use ext3 Mar 08 02:41:31 i already have 8 partitions, don't need another one for /boot Mar 08 02:41:51 I personally get tired of having several os's share the same /boot. Mar 08 02:42:01 So, I've started chainloading where possible. Mar 08 02:42:05 they never cooperate :( Mar 08 02:43:25 I usually just manage the grub partition by myself... most OSes try to kill it in every day funnier ways Mar 08 02:43:43 Meego netbook being an obvious example... Mar 08 02:44:13 if they'd picked a nicer name like butterfly, perhaps Mar 08 02:44:22 um... how do i subscript to bugs without commenting on them? Mar 08 02:44:39 I guess the extlinux choice comes from some seamless bootsplash or the like? Mar 08 02:45:48 ali1234: usually the "add me to CC list" is checked by default. Mar 08 02:45:55 ali1234: So hit "Update" Mar 08 02:46:08 commit? Mar 08 02:46:11 ali1234: It'll send an update to 1 or 2 people, but not the whole list. Mar 08 02:46:39 yeah, commit. Mar 08 02:46:56 heh, it only emailed you :) Mar 08 02:47:44 ha! Mar 08 02:55:28 maybe i'm starting to get the hang of this QML stuff Mar 08 02:55:38 kinda sad actually if thats true Mar 08 02:58:21 berndhs: sad? Mar 08 02:58:50 yeah feels like going back to teh days of open subrouting calls :) Mar 08 02:59:23 s/outing/outine/ Mar 08 02:59:25 berndhs meant: yeah feels like going back to teh days of open subroutine calls :) **** ENDING LOGGING AT Tue Mar 08 02:59:57 2011