IRC log of #harmattan for Sunday, 2011-08-07

fiferboyGBA Mario Kart00:00
javispedrointeresting choice :)00:00
lcukwhats your javis?00:01
fiferboyAnyhoo, got to go for real now00:08
*** fiferboy has quit IRC00:09
hiemanshuah simcity3k is lovely00:09
hiemanshuthough I love to play OpenTTD00:09
GAN900fiferboy, your experience of good games is so limited.00:14
GAN900fiferboy, you should play some Hammerfight00:15
GAN900and Ocarina of Time is clearly the answer.00:15
GAN900Max Payne runs a close second.00:16
hiemanshuGAN900: Hammerfight is lovely00:19
*** Arkenoi has quit IRC00:19
hiemanshuI got with the third HiB00:19
hiemanshu 3 days left, grab it now00:19
hiemanshu7 games, pay as much as you want to00:20
GAN900hiemanshu, infuriating, though.00:21
M4rtinK2wohoo libtiff built, finally :)00:29
M4rtinK2pixbuf is getting closer :)00:29
javispedrolibtiff is on the device00:30
M4rtinK2well it is libtiff4-dev actually00:31
javispedrolibtiff4-dev is on the SDK repo.00:32
* javispedro sighs00:32
javispedroI bet you kept on doing what you're doing on OBS instead of trying scratchbox first?00:32
* RST38h thinks it is time tocall it a day, read a little and go sleep00:32
M4rtinK2I do both actually00:32
javispedrogood :)00:33
M4rtinK2now I first test in scratchbox and then try to build in OBS00:33
RST38hthis qml thing is making me remember a project one German guy has done for Amiga around 199500:33
M4rtinK2still, some packages behave differently00:33
RST38hpretty much the same idea and the same syntax (!), although he obviously did not have JS00:33
M4rtinK2namely gobject-introspection00:33
M4rtinK2it builds unmodified in scratchbox00:34
javispedrowhich is why I say you try this first on sbox, then obs, then note the stuff that is broken on OBS.00:34
M4rtinK2but wont build in OBS complaining that it can't resolve GLIB 2.000:34
M4rtinK2well yeah, that's the logical way to do it :)00:35
M4rtinK2but still, when something builds in OBS on the first try, I spare my self little duplicate work :)00:35
javispedrothere shall NOT be duplicate work00:35
*** piggz has quit IRC00:36
M4rtinK2well, yeah you can actually use the same folders for scratchbox building & as OBS local copy00:37
M4rtinK2try if it builds in scratchbox and then register the files in OBS and commit for build00:38
M4rtinK2BTW, is there a osc command for creating a package from an existing folder ?00:39
M4rtinK2would be faster for this use-case than osc meta pkg + osc up00:40
javispedrono idea00:43
javispedrobut why you don't try _everything_ on sbox first, up to the point where something works00:43
javispedrojust so that you see you're not wasting the time.00:44
M4rtinK2I started with OBS first so I already have something building or almost building there and it would end there anyway00:46
M4rtinK2but I started to first build in sbox recently00:46
M4rtinK2well, I'm quite new to all this packaging stuff anyway :)00:51
* javispedro ponders filling his first bug00:56
*** epage has joined #harmattan00:56
M4rtinK2go for it :)00:59
*** djszapi has joined #harmattan01:03
djszapiehh. fiferboy is gone :(01:04
javispedroyeah, he was here moments ago.01:05
djszapitoo bad01:06
Venemohey again01:13
*** stroughtonsmith has quit IRC01:14
djszapiVenemo hey how is stuff ?01:14
Venemodjszapi, stuff is well :)01:14
Venemodjszapi, if you followed my discussion with fiferboy, you know how well :)01:15
Venemomy main accomplisment for the day:
djszapisorry, I am after a shot, my mind is broken and buggy :)01:16
Venemono problem01:16
Venemoin short, I was fed up with how non-working and useless the SelectionDialog component was01:16
Venemoso I decided to make mine01:16
* djszapi has no idea about components ;)01:17
djszapias for me, it is no go since it is not cross-platform01:17
GAN900If anybody has any requests for topics you think would be usefull in an Improved and Expanded Qt Components Widget Gallery. . . .01:17
VenemoI couldn't be possible compelled to write QML without components.01:17
GAN900Venemo, me neither.01:17
VenemoGAN900 :)01:17
GAN900I want Fremantle Components.01:17
GAN900and GNOME01:17
Venemothere are desktop components, which are claimed to be working fine under Gnome01:18
Venemoas for fremantle, it shouldn't be too hard to implement01:18
lardmannight chaps01:18
*** lardman has quit IRC01:19
* djszapi would like to have a webkit based opensource and free browser on Harmattan...01:19
* djszapi is thinking of starting a qml webkit harmattan project...01:21
GAN900Nokia's built-in browser is so shit.01:21
djszapiit is good and webkit based. However it is unfortunately not opensource01:22
GAN900The w22 one is01:22
djszapiIhave no issue with it01:22
GAN900and the focus on MOBILE browsing doesn't suit me at all.01:22
djszapiand as rzr said, opensource or die :p01:23
javispedrowhy webkit01:24
javispedroport microb01:24
djszapibecause I like wekit(2)01:25
djszapiand apparently Nokia is also fine with that01:25
Venemojavispedro, can't port microb 'cause it's closed01:25
javispedromicrob is open01:25
djszapiporting does not really work for browsers01:25
javispedroit is the UI that is closed.01:25
djszapijust look at the mozilla firefox mobile "version".01:26
djszapiit is almost a new project....01:26
djszapisince you drop everything almost apart from the render engine01:26
javispedroGAN900: thank god there's that Opera around..01:27
javispedroone thing I specially hate is how clicking on a link from RSS seems to randomly select the window where the link will open01:28
javispedrosometimes not opening it altogether01:28
javispedroon Diablo, the formula was something like the "most recent activated window"01:29
javispedroon Fremantle, I really really really enjoyed the always open in a _new_ window behaviour01:30
javispedroon Harmattan, seems that we're back to 0 again.01:30
djszapihehe another complaing ongoing :)01:30
javispedroand here comes another one!01:31
javispedrocombined with the previous one. After the link has opened in the worst window possible, when I try to hit on the address bar in another window to go to the contents that previously was on the "overwriten" window,01:32
djszapiVenemo: do you write ui test codes ?01:32
djszapifor qml based apps ?01:32
javispedrostack overflow :)01:32
SpeedEvilThis 'you're on an old image, don't bother repoting bugs' I can see as being annoying.01:32
djszapiwhy ?01:33
Venemodjszapi, yeah, I'm writing some dummy code until I don't make a backend.01:33
djszapiif you do not like, return your gadget01:33
javispedrorestart: Clicking on a link in the history menu that pops when the address bar has focus sometimes hits the link BEHIND the history menu.01:33
djszapiyou are not forced to use it...01:33
SpeedEvilAt least unless we have someone running the latest that we can poke.01:33
SpeedEvildjszapi: ...01:33
SpeedEvildjszapi: Of course that's not what I meant.01:33
djszapinot sure what you actually meant...01:33
javispedrotherefore, you try to go back to the 4th or 5th element and you suddenly find that you instead hit a link that was on the page....01:34
SpeedEvildjszapi: I don't mean annoying for us - I mean not being able to report bugs that may still be extant in the image, and might in principle be fixable before release, making it a better, more saleable device.01:34
djszapiyou are wrong01:35
SpeedEvilThere will be no bugs in the released device?01:35
djszapiThere are bunch of testers over there with clear concept01:35
djszapiwho are reporting bugs...01:35
djszapithe most common-bugs are known and wip01:36
javispedroYes, I feel betther now.01:36
javispedroHalf of the bugs are caused by bugs in the UI spec, like it was on the n900 times.01:36
SpeedEvilSpecification bugs are more annoying in some aspects than code bugs.01:36
djszapiSpeedEvil: we will not change our timeframe with so less available resource01:36
*** seif has joined #harmattan01:37
djszapiSpeedEvil, sorry but nobody cares about you at nokia01:37
SpeedEvilI know.01:37
djszapi* Nokia01:37
djszapiand I think it is good01:38
djszapisince the planned workflow is not disturbed.01:38
djszapiit would actually be more annoying to mark your bug "won't fix".01:38
djszapifor both us01:38
djszapiboth of us*01:38
SpeedEvilFor that set of bugs that have been found and internally considered.01:39
djszapiI see no point in reporting already known bugs....01:39
SpeedEvilIs the set of bugs that are in the 'would like to fix, but don't know about' category null is another question.01:39
javispedrowell, there's a point: you report a bug to at least know that it is known.01:40
djszapino, we will not investigate into community pebkac or real issues01:40
djszapiwe have enough thing to do, mainly nowadays01:41
djszapiwe trust our paid testers, period.01:41
Venemofrals, is there a way to progmatically check from QML whether the VKB is open or not?01:41
SpeedEvilPaid testers test what you ask them to, which is important.01:42
djszapiyou are seriously wrong01:42
javispedrodjszapi: I know that we are a bunch of crazy nitpickers, but I cannot really believe that you, a KDE contributor, just made that quote.01:42
djszapiit does not really work like that...they are not putting a button and that is...01:42
javispedro"we trust our paid testers, period"01:42
djszapiActually, those are more talented people than few developers01:42
djszapibut it might be that you have no idea about proper testing in industrial environment01:43
GAN900djszapi, we're not morons. :P01:43
djszapianyway, you can report any bugs to the mailing list I guess01:43 also has "device" bugs section01:43
GAN900We have lots of experience with Nokia's testers.01:44
djszapiI see no problem apart from noise01:44
GAN900We know EXACTLY how thorough they are.01:44
SpeedEvilOf course they are not just pressing buttons. But if the test environment is not representative of some subset of users, then they may get different results.01:44
SpeedEvilI'm reminded of the wifi authentication thing.01:44
djszapiI do not have time for this discussion, we have a timeframe, dot.01:44
djszapilet me not repeat myself again, again and again01:44
GAN900djszapi, don't engaged.01:45
djszapiwe do not really wanna investigate into issues which might be pebkac01:45
GAN900You keep complaining about wasting time.01:45
GAN900Well, don't.01:45
Venemoin that case, give us a new image so that we shut up.01:45
djszapiVenemo: funny statement :)01:45
djszapias if I could do anything....01:45
djszapilook, it is full of noise unneccesarily01:46
djszapiyou can report bugs if you want01:46
SpeedEvilThere is also the issue that with more up-to-date images workarounds can be found.01:46
Venemodjszapi, I'm not talking about you personally, I'm talking about the kind of attitude that Nokia is displaying to us01:46
djszapieven though, I would really like to avoid community reports01:46
djszapisince we do not have time with so less human resource for investigating whether it is a pebkac or not01:46
djszapiit would be a bit crazy if we do not do the HIGH priority tasks01:46
djszapibut we start investigating every community report01:46
djszapibut I think you were just be joking.01:46
djszapiVenemo: I seriously disrespect your habit01:47
VenemoI can understand that you can't give us flash/skype/whatever for now. so then until you can, make a bugfix release which only contains the software that is already there.01:47
djszapiNokia gave you a device freely01:47
djszapiand you are complaining about Nokia attitude01:47
djszapiit is really /not/ nice.01:47
VenemoI'm not "complaining", just merely dislike the fact that I'm sitting on a few months old buggy release.01:48
GAN900djszapi, you're awfully condescending.01:48
javispedroNow I'm in a bad mood and I'm going to report not only the QSparql thing but also several other nuisances.01:48
djszapiI hope all of them ignored01:48
djszapi* are01:48
djszapiat least security will, that is for entirely sure :)01:48
GAN900djszapi, I've been involved with Maemo in one way or another since 2005.01:49
GAN900I've put THOUSANDS of hours into the platform and its community.01:49
djszapibut we do not care, period01:49
GAN900I will probably put thousands more.01:49
GAN900The device was hardly free. :)01:49
GAN900djszapi, yes, that's quite clear.01:49
djszapiactually we care the most01:50
djszapisince we are concentrating on the highest priority bugs01:50
GAN900djszapi, what, exactly is your endgame here?01:50
djszapiand not investigating maybe a pebkac issue01:50
*** epage has quit IRC01:50
Venemodjszapi, umm... "I hope all of them ignored"? how can you say that, if you're also part of a FOSS community yourself? how would you feel if KDE just said that they will ignore your bugs?01:50
GAN900That's what triagers are for.01:50
djszapiwell, I am not really happy with the complaining people here.01:50
GAN900We know how bug lifecycles work.01:50
djszapiI am really tired of it01:50
GAN900Have a little respect.01:51
djszapipeople keep complaining here about everything01:51
djszapipeople got a device freely01:51
GAN900It's called venting.01:51
djszapithey actually got a good device01:51
djszapithey are getting proper help from Nokians.01:51
GAN900Complaining often results in exciting new things.01:51
GAN900Well, allow me to kiss your toes, almighty one.01:51
Venemodjszapi, please don't take all of our complaints personally :)01:51
djszapithis is seriously embarrassing01:51
VenemoI am very grateful for all the help I'm getting from the Nokians.01:52
GAN900Our time and expertise is clearly so worthless.01:52
djszapiguys you are keep complaining all the time01:52
djszapiNokia is blabla01:52
djszapiCannot you really express it more politely ?01:52
djszapiit is really no go01:52
GAN900I dunno why you guys deigned to send devices out in the first place.01:52
alteregodjszapi: do you not agree with their concerns/complaints?01:52
djszapiheavily disgree01:52
djszapi* disagree01:52
alteregoWell, denying there's an issue doesn't mean that issue goes away :P01:53
djszapiyou lost me01:53
djszapiseriously, I am just losing all my sake on this channel01:53
djszapijust complaining, complaining and complainging01:53
djszapi* complaining01:53
GAN900djszapi, code is quiet.01:54
alteregoWell, complaints just highlight issues that need to be resolved, in one way or another.01:55
djszapicomplaining is just plain wrong01:55
djszapiyou can give a constructive feedback, but whining is tiresome01:55
djszapilike how the device sucks01:55
djszapihow aegis sucks01:56
djszapihow components sucks01:56
VenemoI never said either of them sucks!01:56
djszapiand so on, I do not read anything else here.01:56
alteregoMeh, venting helps people deal with being pissed.01:56
djszapithere was a guy here few days ago01:56
djszapihe was asking whether this gadget is good or not01:56
djszapisomeone told that, it is worse than N90001:56
alteregoWell then, those are clearly issues that need to be addressed, don't you think?01:56
djszapiso the guy did not even get the sake to apply for a developer device01:56
djszapiwhat is this ?01:56
Venemodepends on your point of view.01:57
javispedrodjszapi: his opinion.01:57
djszapiis this really the encouraging of people ?01:57
alteregoEither through education or code.01:57
djszapiI am really tired about the negative feeling flowing around here.01:57
alteregoWelcome to Maemo :P01:57
djszapiif you do not like it, send your gadget back01:57
djszapiif you like it, deal with it. It is that simple.01:57
alteregoThat'ts bs01:58
javispedroI am going to explain for 78th time why I kept the N810 instead of going back to a Palm.01:58
djszapiI do not care01:58
alteregoYou don't here people complaining about all the good points, and as aegis and qt components are the hottest topics ...01:58
alteregohear ..01:58
*** rcg has quit IRC01:59
javispedrowhen I found out the N810 didn't support PAP, I initially came to to rant.01:59
djszapithere was also a complain about connect button for the internet connection01:59
javispedroThen someone hinted me that I could implement such support on my own.01:59
djszapiwhich makes perfect sense and much better than fremantle01:59
javispedroI did.01:59
djszapibut there was a big whining about that as well01:59
javispedroOn _any_ other platform, I would have got a "Thanks for your suggestion" note like the ones you are giving right out now, djszapi.01:59
djszapiI think this channel has a fairly odd negative feeling about things, most of the time even without real facts.01:59
alteregoThe fact is, aegis has hindered peoples exploration of the platform and it's stopped people from getting things done, people need to either be educated about aegis to resolve their problems or we need to fix aegis ..02:00
javispedro"Thanks for your suggestion but now go away" note actually.02:00
alterego"Fixed in WP7" :P02:00
djszapiI stopped helping today about aegis, sorry.02:01
javispedrodjszapi: you _cannot_ help about aegis.02:01
*** epage has joined #harmattan02:01
GAN900djszapi, when did you start with Maemo?02:01
javispedroyou have said so numerous times.02:01
djszapiyeah, I could not...02:01
fralsVenemo: not sure, at least its possible to close/open it from the TextField components scope02:01
djszapiGAN900: one year ago02:01
Venemofrals, yes, I noticed that, platform*SoftwareInputPanel02:01
Venemofrals, but I see no way of checking whether it's open.02:02
GAN900djszapi, some of us have been around since 2005. We know Nokia's games.02:02
djszapiHowever I have never seen so negative feeling on a channel anywhere yet, mainly when you are getting an expensive device freely, come on...02:02
fralsVenemo: what are you after in the case of open/closed?02:02
GAN900Little respect for your elders. :P02:02
alteregoVenemo: there isn't unfortunately, though you could just force it one way or the other ;)02:02
djszapiGAN900: how cares about 2005 ?02:02
djszapiI am speaking about that, it is no go for me to listen to complaining people02:02
alteregodjszapi: you don't care because you've not bee burned.02:03
djszapi* whining, put it mildly.02:03
Venemoalterego, I don't want to force it, but I want to shrink the size of my stuff when it's visible02:03
fralsVenemo: ideally your whole main window/page is being resized when it appears (iirc)02:03
javispedrodjszapi: it is you who is anchored in the past.02:03
javispedrodjszapi: these days, even HP will send me a free device and gladly listen to my musings.02:03
alteregoYou've only been here a year? Some of us have invested a lot into maemo only to have it dropped.02:03
Venemofrals, yeah, but this mechanism is buggy with dialogs.02:03
alteregoNegativity is understandable ..02:04
fralsVenemo: right... well, I'll go back to my statement "I don't think you are suppose to have text input in a Dialog" :)02:04
djszapialterego: keep whining then, I just ignore you...02:05
fralsVenemo: is it for a filepicker or something along those lines?02:05
alteregoI'm not whining :P02:05
alteregoYou're whining about whiners afaict02:05
Venemofrals, no, it's a channel switcher dialog and a user list dialog (in which you can query a user by clicking on his/her name)02:06
fralsVenemo: aha, but you want a search bar at the top?02:06
alteregonyurgh, I need to sleep. bbl02:06
Venemofrals, yes, a type-to-filter stuff.02:07
fralsVenemo: to be in line with platform UX (afaik) i think it really should be in a separate view or a sheet and not in a dialog :/02:07
fralsVenemo: but best of luck with trying to solve it :)02:07
fralsplease let me know if you do :)02:07
Venemofrals, well, it's not that big of a problem02:08
djszapialterego: seems, it is just another #maemo channel02:08
Venemofrals, thing is that this kind of selector looks the best in a SelectionDialog.02:08
Venemodjszapi, the good point is that the people without Harmattan devices are not on this channel, so we are at least free from their whining :)02:08
fralsVenemo: aye, i understand, but dialogs should not really have a textfield (at least none of the platform apps do afaik) -- this from UX/UI pov (my guess)02:09
Venemofrals, I don't recall seeing a guideline that forbids it.02:10
fralsotoh i havent read the ui/ux guidelines that carefully, i usually get yelled at when doing it the wrong way by my designer :D02:10
djszapifrals: Ui (iphone) designers are a joke02:12
fralsthe ones i work with are quite agreeable, as long as you keep in mind they are forced to certain points to adhere to platform style, which is not defined by them02:13
fralsmore than quite agreeable in fact ;)02:13
djszapithey are replicating iphone from what I can say.02:13
djszapiwhich is not any "agreeable".02:14
djszapibut I am gone for tonight02:14
Venemodjszapi, jó éjt! :)02:15
djszapiyeah, sure...02:16
Venemogood night02:16
Venemofrals, what kind of GUI element would you suggest for font size selection?02:16
*** djszapi has left #harmattan02:16
fralsVenemo: im by no means the right person for that, but most likely a query dialog but without a text input02:17
Venemofrals, QueryDialog is a yes/no kind of dialog02:18
*** epage has quit IRC02:18
fralssorry, the normal dialog thingy02:19
fralswith the list in it02:19
VenemoI thought about it, but it doesn't look good with numbers02:19
fralsVenemo: thats the one used by platform apps, at least02:20
Venemofrals, which app for example?02:20
VenemoI'll check it out02:20
GeneralAntillesfrals, didn't your vacation start today?02:20
fralstap on the "more" icon when having the format toolbar open02:20
fralsGeneralAntilles: it started on friday ;(02:20
GeneralAntillesfrals, go play video games or something. :P02:20
fralsGeneralAntilles: been playing for more or less 9hours straight, need some rest ;-)02:21
GeneralAntillesWe're all so bad.02:21
GeneralAntillesIRC is NOT A BREAK02:21
Venemofrals, ah yes, "X points" looks better than plain "X". I'll do that too, thanks for suggesting :)02:21
fralsVenemo: np :)02:21
fralslooking forward to testing the application when the backend is hooked up, really missing a decent irc client on my n902:22
GeneralAntillesSo, to reiterate my previous question: What do people want in an improved Qt Components Widget Gallery?02:22
Venemofrals thanks for your help with it :)02:22
fralshappy to help when i can :)02:22
* MohammadAG sends a PlayStation 3 copy of an FPS game to frals02:23
* frals sends a nuke in MohammadAGs general direction02:23
fralsheretic playing FPS on console :(((((((02:23
MohammadAGBreaking news, Israel nukes Finland02:23
* frals was going to check what's in qt-comp gallery atm, but apparently got the wrong build in all devices :<02:24
MohammadAGfrals, precisely why I sent it02:24
GeneralAntillesfrals, sadly the only way to play with friends these days. :(02:24
fralsGeneralAntilles: there is a solution... get better friends! ;D02:24
javispedroMaemo Global Thermonuclear War.02:24
fralsbut yeah, i know the feeling, lots of people play on console nowadays for some reason02:25
GeneralAntillesfrals, for instance I'd like to try to include an interactive demo of the ins and outs of PageStacks.02:25
Venemowhat? why would I need a PS3 when I can have Starcraft 2 on my craplaptop?02:25
GeneralAntillesfrals, cheap, easy.02:25
GeneralAntillesIt always works when you put the disc in02:25
fralsi guess they dont feel as bad sucking when playing on console as they do while on a proper machine :-)02:25
GeneralAntillesvoice chat is included out of the box.02:25
GeneralAntilleseverybody else has one.02:25
MohammadAGjavispedro, MeeGo 1.2 Harmattan*02:25
MohammadAGsorry, we're not WMP compatible with maemo02:25
VenemoGeneralAntilles, looking forward to your new QML components gallery02:25
fralslast console i bought was xbox1 ;(02:25
GeneralAntillesMost college students can't afford a decent gaming rig.02:25
javispedrobut as Venemo put these days even a craplaptop runs sc202:26
javispedrowell, not a Meego Netbook...02:26
fralsyeah, had to pull some extra work to keep my rig in decent shape during school02:26
MohammadAGunfortunately most friends play FPSs :P02:26
Venemookay, well, mine is not _that_ crap, but... :P02:26
VenemoI hate FPSs02:26
Venemobut I love RPGs and RTSs02:26
GeneralAntillesjavispedro, well, sorta. :P02:27
Venemojavispedro :)02:27
fralsotoh ive only upgraded the GPU since my last full overhaul of the machine which was in like 200802:28
Venemofrals, forgive me for harassing you, but is there a way to permanently show a ScrollDecorator?02:29
fralsgrrrrr, scrolldecorator, the bane of my war on fps while panning lists02:29
fralsVenemo: __hideTimeout: 0, i would hope02:30
Venemowhy does it begin with __ ?02:30
* Venemo doesn't trust properties whose names begin with __02:30
fralslikely because its not supposed to be changed ;o02:30
fralscrap, it wont work02:30
Venemono, it doesn't work02:31
fralsmaybe if you set it to MAX_INT or something, but that would just make it one long transition :<02:31
Venemo__alwaysShowIndicator doesn't either02:31
fralsyou could probably do something ugly like02:31
fralshmm, transition, bah02:32
fralsprobably cant get to that one with children[] in a nice way02:32
Venemo__hideTimeout: 9999 can do the trick02:32
Venemobut I can't find a way to show it in the first place.02:32
fralsDISCLAIMER: ugly hack inc02:32
frals*maybe* if you do Component.onComplete: { scrolldec.children[(whatever number to get to verticalIndicator)].shouldShow = true; }02:33
frals(not working code but you get the point i guess)02:33
Venemobut this indicator doesn't have a show() method02:35
Venemonor a shouldShow02:35
fralsno but inside the ScrollDecorator is an Item called verticalIndicator which has the property shouldShow02:36
*** antman8969 has joined #harmattan02:36
fralswhich is then set to always true with the hack, which should hopefully always show it02:37
fralsbut yeah, its going to break if they add another children before the verticalIndicator in the implementation02:37
fralsbut shit happens ;-)02:37
javispedrosince this is a QML experts conversation... :)02:38
javispedrohave any of you done anything that resembles a Tracker Live Query in QML?02:38
javispedrothat is, a self updating list of found items02:38
Venemofrals, qrc:/qml/harmattan/WorkingSelectionDialog.qml:83: TypeError: Result of expression 'children[6]' [undefined] is not an object.02:39
fralsVenemo: id suggest a js loop over the children printing the id so you know which one is which, iirc it should 302:39
fralsVenemo: *IIRC* children[] only counts drawable objects02:39
fralsjavispedro: nope, sorry02:39
javispedroah, np.02:40
*** NIN102 has joined #harmattan02:41
Venemofrals, doesn't seem to be working in any way.02:42
fralsjavispedro: throwing the live query results in a qabstractlistmodel or something similar and then exposing it to qml should be doable, but ive managed to stay faaaar away from tracker so far02:42
Venemofrals, can I just copypaste the code of that indicator and tailor it?02:42
lcukfrals, good to see performance still being considered02:42
fralsVenemo: you probably could, but its a slippery slope doing that ;)02:42
Venemofrals, yeah, just like making my own SelectionDialog02:43
javispedrofrals: yes, that's the only way I know, just pondering whether there's a sikret one :)02:43
fralsVenemo: aye02:43
Venemofrals, but I had to, the original one didn't work with my model :(02:43
Venemoor any C++ model for that matter02:43
*** NIN102 has quit IRC02:44
*** NIN101 has quit IRC02:44
fralsyeah, the original taking a ListModel instead of just aliasing the model properly hit me a few days back during work02:44
*** NIN101 has joined #harmattan02:44
Venemobut even the "fixed" one isn't working, because it accesses a non-existant property02:45
VenemoI realized that after I examined it once more.02:45
fralsbecause apparently ListElement properties cant contain scripted values, like for example... qsTrId(your_string) :((02:45
Venemoit calls model.count02:46
Venemowhich works for ListModel, but not for any C++ model.02:46
fralsnot for any c++ model which does not declare count you mean ;)02:47
Venemowell, QAbstract*Model don't have it.02:47
Venemothey have rowCount() but that is not Q_INVOKABLE so can't be called from QML either.02:47
Venemostill, this is an easyfix, since ListView also has a count property, which actually works :)02:48
fralsyeah, would have to subclass it02:48
Venemothat is what I use in my version.02:48
Venemoanyway, I LOVE the newest Qt Creators02:49
antman8969frails, you can work around the ListElement not taking scripts by using the Component.onCompleted in a ListModel02:52
antman8969and adding items through a script instead02:52
antman8969 line 157,02:54
fralsantman8969: aye, same solution i used in the end :)02:55
*** meegoexperts_mob has joined #harmattan02:55
fralsanyways, im off to bed, already 0300 over here and im knackered o/02:56
lcukfrals, do you know how well qml would work on windows phone?02:56
lcuknm, see you later \o02:56
fralslcuk: would depend how well qt would work i assume ;)02:56
Venemofrals, good night :)02:56
lcukhope harmattan leads to some decent apps02:57
fralsi heard some bright minds in the community were looking at qt on wp7 but not heard much about it02:57
Venemolcuk, QML doesn't work on Windows Phone.02:57
lcukVenemo, how is the irc client coming on?02:57
Venemolcuk, nicely. I fought my way through QML this whole day.02:57
javispedrofrals: they are rewriting Qt in C#? ;)02:57
lcukVenemo, is the code published, I would like to try it on my n95002:57
Venemolcuk, sure, but it doesn't have a backend yet.02:58
lcukI may attempt to help if I can understand it02:58
lcukthat is ok :) just curious how an app would take shape02:58
Venemolcuk, it is an "empty shell" :)02:58
lcuk:) what are 80% of liqbase apps?02:58
Venemookay, here it is:
javispedroone think I've pondered once02:58
lcukVenemo, you mould the app02:58
lcukand make the data work with it02:59
Venemolcuk, that is what I'm doing02:59
javispedrois how costly it would be to port a webos JS app to QML02:59
javispedrothey have a pretty good JS IRC client02:59
lcukyou are learning qml and your insights and code methods might prove useful, thanks02:59
Venemolcuk, any time :) if you have questions about my spaghetti code, ask away :)02:59
Venemolcuk, I'm not joking, that code is full of voodoo magic03:00
lcukVenemo, question 1: how do I get this onto my n950?03:00
lcukis there an obs repository of it?03:00
Venemolcuk, either let Qt Creator package it, or
Venemolcuk, I haven't felt the need to package it yet.03:00
lcukVenemo, I do not have much success with qt creator03:01
lcukit is voodoo03:01
lcukif qt creator packages it, could you share the deb somewhere?03:01
Venemolcuk, umm, I directly configured it not to package the app.03:01
lcukyou said you let qt creator package it?03:01
antman8969if it's that big a deal, I'll package it right now03:02
Venemolcuk, I said you _either_ let Qt Creator package it _or_ -> and this latter is what I do03:02
lcukantman8969, !! :D03:02
lcukplease do03:02
Venemothe topic is "Howto: Run/Debug apps with Qt Creator faster"03:02
lcukcommunity in action!03:02
Venemolcuk, you can also uncomment a line and then the app will "fake" a new message to every channel in every second.03:03
*** Rizzer has joined #harmattan03:03
*** wicket64 has quit IRC03:04
Venemolcuk, btw, I plan to implement your feature suggestion too, but not in the first version.03:06
antman8969freshlly downloaded from git, it dosnt seem to startup on my n95003:08
antman8969attempts to chmod something, fails and dies03:08
Venemoantman8969, which Qt Creator are you using?03:09
antman8969just installed with dkpg and it seems fine..  need to try to run03:10
Venemojust asking because the project file is set up to be used from the 6th august nightly, and with skipping the packaging step03:10
Venemoeg. it puts the binary into /home/developer03:10
antman8969I didnt use your pro.user file tho03:10
antman8969permission dendied03:11
antman8969need to su developer03:11
antman8969I think03:11
antman8969permission denied everywhere lol03:11
antman8969why did you decide to put the executable there?03:11
antman8969got it... nice, auto joins meego maemo and harmattan :)03:12
Venemothe .pro.user file is not even committed to git.03:17
Venemoso you couldn't used it either way03:17
antman8969in any case, it's running well now03:17
VenemoI decided that because it can't upload to /usr/bin03:17
antman8969why is that?03:17
antman8969and alternatively, maybe /opt/irc-chatter/bin?03:18
Venemobecause 'developer' doesn't have write permissions to /usr/bin03:18
antman8969do you mind if I put it up for lcuk?03:19
VenemoI don't mind03:19
antman8969probably take it down later tonight.... don't want ppl thinking I made it lol03:19
VenemoI'll probably invent a sane way of doing things, once I finish it and start packaging03:19
antman8969although I should say lcuk, You need to chmod +x /home/developer/irc-chatter as root03:20
antman8969and then su user to run it lol03:20
antman8969platform rules are getting annoying03:20
Venemoand you may need aegis-developer-mode --relaxed-exec too03:20
antman8969I started it without... but I can't actually post or read anything03:20
antman8969or is it just not implemented yet03:21
Venemo[01:58] <Venemo> lcuk, sure, but it doesn't have a backend yet.03:21
Venemo[01:58] <Venemo> lcuk, it is an "empty shell" :)03:21
Venemoantman8969, it is an "empty shell"03:21
antman8969lol aah03:21
Venemojust a GUI, no backend.03:21
Venemono need to fuss about it yet :)03:22
antman8969haha sorry, I came in after that03:22
Venemoas I said, you can uncomment a line and then the app will display fake messages so that I can test the delegate that displays them03:22
Venemoit is also filled with dummy data.03:22
antman8969I like  the gui tho, looking good03:23
*** Rizzer has quit IRC03:30
*** deimos has quit IRC03:33
*** NIN101 has quit IRC04:04
antman8969does anyone have any idea how to get a service to auto start on harmattan04:05
antman8969getting frustrated04:05
Venemoantman8969, there is a thread about this in the maemo-developers mailing list04:06
antman8969what can I google?04:06
Venemosee "Aegis - Upstart script not working in Harmattan"04:06
antman8969thanks venemo04:07
Venemoyou're welcome antman896904:07
javispedrothat thread is fun04:08
javispedromy portmap script on /etc/init/portmap.conf finally worked04:08
javispedroand in the current iteration aegis hasn't triggered self destruction yet04:08
Venemoand I managed to make a nick autocompletion algorythm04:09
javispedroand contradicting that thread I used the "start on started group-mce, stop on stopped dbus" stanzas.04:09
antman8969javispedro, will I find my answer in that thread04:09
antman8969reading right now..04:09
javispedroantman8969: you know what upstart is?04:10
antman8969not yet04:10
javispedrothat thread does not explain04:10
javispedrobut google will04:10
javispedrolet's say they have a specific text file format for describing services04:10
Venemothat thread only explains how to get it working on Harmattan, but it won't explain how upstart works <- antman896904:11
antman8969I was just using a script in init.d and update-rc.d to set it... but now I see that there aren't even any rc#.d directories04:11
antman8969which makes me wonder why there is update-rc.d04:12
antman8969and sitll no cron  / crontab04:12
antman8969I also didn't have a manifest file either04:12
javispedroyou do not necessarily need a manifest file04:13
javispedrobut try your packaged application from the command line before putting in the upstart script, in case you get a reboot loop.04:14
antman8969well I know the daemon works. Right now i set it up so that it just echos "fish" to /home/user/fish.txt ever 5 seconds04:14
antman8969and I can start / stop it with /etc/init.d/script start04:14
antman8969lol I guess it's supposed to go in /etc/init/ and NOT /etc/init.d04:15
antman8969thats part of the problem....04:15
antman8969then why have init.d still  there!?04:15
javispedroto confuse you :)04:16
*** djszapi has joined #harmattan04:16
*** djszapi has left #harmattan04:16
javispedroantman8969: but I meant that you should try running your daemon as installed by the package, just in case there's a problem with the package04:16
antman8969yea that's what I've been trying to do04:16
antman8969but I can't get it to autostart yet04:17
javispedro_before_ setting the init script04:17
javispedrobecause if, say, aegis refuses to execute it from the package, you might get a reboot loop.04:17
antman8969I guess I'm confused as to what you're suggesting... install the script through a deb and try executing it?04:17
SpeedEvilantman8969: Oh my! A fish daemon! Awesome! :)04:17
javispedroantman8969: install the daemon then try executing it.04:18
antman8969yea, thats already done04:18
antman8969works fine04:18
javispedro*install the daemon through a deb04:18
javispedrook then04:18
antman8969I guess I'll try moving it to /etc/init/apps04:18
antman8969and see how that goes04:18
antman8969lol waiting..04:19
javispedrocrash, destruction, and end of the universe.04:19
antman8969WELL... im listening04:19
javispedroI told you to google :)04:19
javispedroeither way, an upstart job looks like this04:19
antman8969as opposed to the debian skeleton file in /etc/init.d I suppose04:20
antman8969i'll read up04:20
*** eman has joined #harmattan04:32
*** meegoexperts_mo1 has joined #harmattan04:40
antman8969javispedro, do you know if the executable location matters? Does it have to be in sbin?04:40
*** meegoexperts_mob has quit IRC04:40
javispedrodon't think so.04:40
*** hardaker2 has joined #harmattan04:49
*** eman has quit IRC04:50
*** seif has quit IRC04:55
*** meegoexperts_mo1 has quit IRC05:03
*** M4rtinK2 has quit IRC05:06
*** hardaker has joined #harmattan05:19
*** hardaker2 has quit IRC05:21
*** epage has joined #harmattan06:06
*** javispedro has quit IRC06:37
*** epage has quit IRC07:18
*** Venemo has quit IRC08:28
ieatlintyou know, the double tap to unlock feature would be a lot more awesome if it were able to use the proximity sensor to see if it's in my pocket08:36
ieatlintbut i guess that would use the battery too much...08:36
antman8969I know what you mean08:44
antman8969I usually face it away from my leg08:44
antman8969and I also have to keep it right side up because the button presses too easy on top08:44
ieatlintyeah, i'd like it even if they disabled that button if the proximity sensor was triggered (or had the option to, anyway)08:47
*** hardaker has quit IRC08:58
*** hiemanshu_N950 has joined #harmattan09:21
*** hiemanshu_N950 has quit IRC09:22
RST38hgood moorning09:34
*** seif has joined #harmattan09:47
*** lardman has joined #harmattan09:48
*** lardman has quit IRC09:48
*** lardman has joined #harmattan09:48
hiemanshumorning lardman09:48
lardmanhi hiemanshu09:48
hiemanshuwhats up09:48
lardmanam cursing packaing09:49
lardmanor packaging even09:49
hiemanshuyeah, .deb packaging is stupid, RPM ftw09:49
lardmanno way!09:49
lardmandeb packaing itself isn't too bad, it's working out package dependencies, etc., for mBarcode-lite which is not fun09:50
lardmanwould far prefer to be writing code than arsing about with things like this09:50
lardmanthough with that said, how does one generate more than one package from a single .pro file, etc.09:56
lardmanfor example I should finally get round to generating both mbarcode-lite and mbarcode-lite-dev from the same project, rather than having to manually update header files09:57
*** seif has quit IRC10:06
*** rcg has joined #harmattan10:07
rcgmorning all10:15
lardmanmorning rcg10:22
lardmananyone happen to know how to get header files to be included in a package?10:22
lardmanI've created the relevant .install file for the package, and the package is packaged up, but it only the automatically included changelog10:23
rcgso you basically want a foo-dev package?10:42
rcgor include the header files in a regular package?10:43
lardmansorry, key got stuck there10:43
lardmanI want to create an mbarcode-lite-core package with the binary, an mbarcode-lite package which has the depends on the plugins and an mbarcode-lite-dev package with the header files10:44
lardmanhmm, not sure what's happened now, my binary is placed in the mbarcode-lite package rather than the -core package, and no headers are placed in the -dev package10:47
lardmanthis is really rather frustrating10:47
*** antman8969 has quit IRC10:49
rcghmm just a second.. trying to figure out how they did that for the system-ui package10:53
rcgi was messing around with this one lately anyhow10:54
lardmanI'm going to try adding some stuff to the rules file now, with a heading for each package10:54
lardman dpkg-genchanges  >../mbarcode-lite-core_0.0.1-0_armel.changes10:59
lardmandpkg-genchanges: error: cannot read files list file: No such file or directory10:59
lardmandpkg-buildpackage: error: dpkg-genchanges gave error exit status 210:59
lardmanI can see what the problem is, but is this step necessary?10:59
rcghmm.. what i saw so far you need to explicitly copy the headers in your project file.. then you need to add the -dev package to the debian/control file and finally need to add debian/foo-dev.dirs and .install files10:59
lardmanI've not added the .dirs file, but have done all the rest11:00
rcgalright.. it build the -dev package and this included the header files i copied for testing11:02
lardmaninterestingly mbarcode-lite-core contains a /usr/bin dir, but it's empty - in the .install file I list that dir and the binary name to be placed there11:02
rcghmm strange thing11:03
lardmanwhat format do your .install files take?11:03
rcgcat debian/meepasswords-dev.install11:03
lardmanhmm, I wonder if mine are wrong like this then: maemobarcodewindow.h usr/include/mbarcode-lite11:04
rcgi just used some of my existing projects and basically tried to mimic what they did in the system-ui package11:04
lardmansure, ok, will give that format a go and also add the .dirs11:04
rcgafaik you need to take care via the project file that your headers are "installed" in the output directory11:05
rcgthen you only need the relative path in *.dirs and *.install to make dpkg pick em up11:05
*** djszapi has joined #harmattan11:06
djszapilbt: hello11:06
lardmanrcg: yeah I was a bit worried about that11:06
djszapiDo you know something ongoing why the distribution does not work on c-obs ? The built packages does not reach the repository.11:06
rcgthats what the relevant part of my project file looks like11:07
rcglardman: maybe go step by step.. do your headers end up in the INSTALL_ROOT as configured in debian/rules?11:08
lardmanrcg: thanks11:08
lardmanno they don't seem to, which looks to be the root of that problem11:08
lardmanthough assuming I can fix that, I'm still ending up with the binary in the wrong package11:09
lardmanlet me add the headers stuff to the pro file and test that quickly11:09
lardmanok, so .dirs works and I get the correct dirs in each package11:12
lardmanand adding the headers bit means the headers get included, but in the wrong package11:12
rcgah well... then i assume you need to configure which file to include in which packge via debian/foo-*.install files11:13
lardmandone that already and it's ignored completely11:14
lardmanseems like the .pro file overrides everything11:14
rcghmmz.. strange thing11:14
rcgwhat do your install files look like?11:14
lardmanmbarcode-core looks like this: mbarcode-lite usr/bin11:15
lardmanbut doesn't get the binary11:15
*** mariob has joined #harmattan11:15
lardmansorry, needs a -lite- in the middle of the name there11:15
rcgah.. could you try: usr/bin/mbarcode-lite11:15
lardmanmbarcode-lite has an empty install file as it is a meta-package11:15
lardmanbut it gets the binary + headers installed in it11:16
rcguh.. ok.. i have no clue about meta-packages11:16
lardmannor me! ;)11:17
lardmanthanks for your help though, I'll do some more tweaking11:17
rcgyw :)11:17
lardmanI think I need to somehow define the other package names in the rules file11:17
rcg <-- just for your reference.. thats how they did it for the system-ui package11:18
rcghmm i do not think you need to tweak the rules file.. what i can see here is that they only added all these packages to the control file11:19
*** Venemo has joined #harmattan11:20
Venemogood morning harmattanites! :)11:21
rcgmorning Venemo :)11:22
lardmanmorning Venemo11:22
lardmanrcg: thanks for those, mine are very similar11:22
lardmando you have a link to the .pro file?11:23
rcguhh.. well.. they use quite a lot of subdirs and includes in that package11:23
lardmanyeah I reckon that's what I will need11:24
lardmanis this a Nokia package?11:24
Venemomorning rcg & lardman11:24
lardmanapt-get source system-ui ?11:24
rcgsrc/systemui/ <-- this should be the relevant project file then11:25
* lardman has a look to see how to enable internet access in his new sb install11:25
lardmanhmm, looks like it uses an extra .pro file for each target11:30
lardmanseems like overkill when my extra packages are just headers and a meta-package11:31
lardmanI guess the debian/rules file might be a better approach in that case.... perhaps....11:31
rcglardman: what i tried here with my meepasswords package was to simply add some -dev package and i only needed to add this "headers.path ... INSTALL += headers" to the project file plus then change debian/control and add -dev.install and dirs11:36
rcgdunno if that fits you use case though11:36
rcgafaik the project file essentially takes care that everything is installed in the INSTALL_ROOT at the right place11:38
lardmanit would, but doesn't work for me11:38
lardmanI just get everything placed in my -core package11:38
rcgdid you add an install file for that -core package?11:38
lardmanthe INSTALL += headers will install them in whichever package the .pro file creates afaiu11:38
rcgafaik the .pro file does not create a "package" it just copies everything to INSTALL_ROOT11:40
rcgthe content of the packages should then be determined by the .install files.. at least as what i can see here11:41
lardmanI'm going to go and grab a shower, perhaps I'll come back refreshed and with a good idea11:42
lardmanotherwise I'll just push the code online and ask for more help :)11:42
lardmanthanks rcg11:42
*** lardman is now known as lardman|afk11:43
rcgalright.. i am just trying to quickly check that11:43
*** seif has joined #harmattan11:45
*** seif has quit IRC11:53
rcglardman|afk: <-- these are my results of the quick test11:54
*** crevetor has quit IRC11:57
rcglardman|afk: one issue that may arise here is when INSTALL_ROOT equals one of your package names... that may actually be a reason why you would have one package containing everything12:04
*** smoku has joined #harmattan12:06
alteregoAny sign of new n950 firmware?12:07
*** mariob has quit IRC12:11
*** lardman|afk is now known as lardman12:14
*** eman has joined #harmattan12:14
*** djszapi has quit IRC12:16
*** andre__ has quit IRC12:16
lardmanrcg: interestingly my builddir doesn't contain the headers, but they end up in a package12:16
RST38hMoo lardman12:22
RST38hMoo all12:22
lardmanhi RST38h12:22
* RST38h has got to passing data from QML back into C++ code. Doable?12:22
lardmanPresumably these are not your own QML components?12:24
rcglardman: hmm that sounds really strange indeed12:25
rcghi RST38h12:26
rcgRST38h: sure, it is possible12:26
*** djszapi has joined #harmattan12:31
lardmanhmm, so looking at the Makefile created by qmake, adding the headers has created an installation rule for them, but copying them from their original location, not from the builddir12:34
lardmanso perhaps this is the problem, both that they are not found by debian/rules and that this installation rule installs them in the wrong package too12:35
RST38hrcg: how?12:38
RST38hwhere should I look?12:38
RST38hlardman: you do not need qmake in the first place12:39
RST38hmakefiles work just fine12:40
lardmannot sure I can be bothered to run moc etc myself12:40
RST38hah it is just 5 minutes to add a .h.moc rule12:40
*** eman has quit IRC12:40
RST38hrcg: thanks =)12:40
lardmanRST38h: hmm, perhaps12:40
rcgyw :)12:40
*** Rizzer has joined #harmattan12:41
lardmanso, the only question now, is how to I tell debian/rules where my files are to be installed?12:43
RST38hyou do it in makefile12:43
lardmanI've added words to the effect of: ../headerfile.h path/to/beall/installed/headerfile.h to the .install file12:44
djszapiyou can do it in the install file.12:44
lardmanRST38h: I don't want to mess about with Makefiles if I can avoid it12:44
RST38hinstall file then12:44
RST38hit did not work for me in maemo though12:44
lardmanRST38h: I'm creating 3 packages from a single dpkg-buildpackage here12:44
djszapior if it is very special, it depends on the way you are using, cdbs or dh_install12:44
lardmanRST38h: I've got the same problem here12:45
lardmandoesn't seem to do what it says it ought to12:45
RST38hmakefile then. moment.12:45
djszapiyou just override the install rule12:45
* RST38h finishing his marshmallow12:45
lardmandjszapi: I need the install rule for one of the other packages though...12:45
lardmanand really, it should be able to do it itself, this isn't very complex12:46
RST38ha moment lardman12:46
lardmanwhat I'd like to know is whether rules is run more than once for each package found in the control file, and if not, what is?12:48
lardmans/more than once/12:48
RST38hlardman: This is a Makefile that resides at the same levelas debian/ dir12:49
RST38hlardman: dpkg will use it to build and "install" your package locally, into debian/12:49
Rizzerwhat's the current state of N950 tethering (USB or bluetooth)?  I've got a USB connection going, I can ping the N950 at from my Debian box. But it doesn't seem to be routing to external IPs.  Should routing tables be automatic (route -n "looks" fine)?12:49
RST38hlardman: So, the idea is to write it as if you were writing a normal install: target in your makefile12:49
lardmanis that going to work with the Makefile generated by qmake too then?12:51
*** M4rtinK2 has joined #harmattan12:57
*** djszapi has quit IRC12:59
*** djszapi has joined #harmattan12:59
lardmandoh! missing dh_install13:02
*** NIN101 has joined #harmattan13:04
RST38hlardman: no idea, not using qmake after finding out it ignores some of my sources13:05
Venemo~seen RzR13:09
infobotrzr <> was last seen on IRC in channel #meego, 1d 2h 16m 42s ago, saying: 'as someone a backup of win7 ?'.13:09
*** Venemo has quit IRC13:12
RST38hShit, I am still unable to pass any data into QML or back13:12
RST38hsetContextProperty() does not get reflected in the qml13:12
*** seif has joined #harmattan13:14
*** mariob has joined #harmattan13:16
RST38hAha, now it is reflected13:18
*** seif has quit IRC13:21
*** Venemo has joined #harmattan13:43
*** djszapi has quit IRC13:48
*** eman has joined #harmattan13:49
RST38hCanI call QCoreApplication::exit() from QML???13:55
lardmanRST38h: thanks13:55
lardmanok, I've sorted it, I was missing dh_make13:55
lardmanconfusingly it still builds with qmake, etc.13:55
lardmanvery odd13:55
RST38hqmake does appear to hide a lot of things13:56
RST38hsome of which are not even possible with make (or nobody nows how to do them)13:56
*** eman has quit IRC14:12
RST38hSo, how the hell do I *exit* the pagestack?????14:14
RST38hAs in, exit the event loop14:14
fralsbtw RST38h, i guess you tried the stuff at for including the qrc properly without qmake?14:14
RST38hno, have not seen it14:15
fralsseems to be possible using rcc, but ive not tried it. hopefully its helpful14:15
RST38hI will try thanks14:16
RST38hfrals: do you have any idea how to exit the QMLapp back into C++ code?14:16
RST38hpopping the last page does not help14:16
fralsuh, exiting as in? removing the view?14:17
RST38hexiting the event loop started with exec()14:17
fralsi wonder if Qt.quit() would quit the application or the event loop14:17
RST38hprints message about handlermissing for quit9014:18
fralsno idea actually, sorry14:19
fralsi mean, besides exporting the eventloop to the qml and calling loop.exit() on it14:19
lardmanhmm, eaten a sandwich (bacon of course) and now trying to apply the same multi-package goodness to another of my apps, and again, not working14:31
* lardman tries to remember all the steps he took before14:31
SpeedEvilOpen bread, apply butter, add filling, close sandwitch around multiple packages of filling.14:33
RST38hno way to get file size or modification date in QML?14:37
*** djszapi has joined #harmattan14:50
MohammadAGRST38h, well, if's a QDeclarativeView it inherits QWidget14:53
MohammadAGQWidget::close() should send you back14:54
*** eman has joined #harmattan14:59
*** eman has quit IRC15:04
*** hardaker has joined #harmattan15:07
*** seif has joined #harmattan15:12
*** cpscotti has joined #harmattan15:14
*** seif has quit IRC15:23
*** mariob has quit IRC15:31
alteregofrals: Qt.quit() sends the quit() signal which should be linked to QApplication::quit()15:32
alteregoSo if you're ctreating your own declarative view you need to connect that signal and slot, otherwise it does nothing.15:33
alteregoRST38h: ^15:34
alteregoIf you do that then it'll cause QApplication::exec() to return15:35
*** mariob_ has joined #harmattan15:35
alteregoRST38h: QObject::connect(view, SIGNAL("quit()", QApplication::instance(), SLOT("quit()"));15:37
*** seif has joined #harmattan15:39
*** deimos has joined #harmattan15:39
*** epage has joined #harmattan15:42
*** mariob_ has quit IRC15:43
*** seif has quit IRC15:45
MohammadAGalterego,  qApp = QApplication::instance() right?15:46
deimosMohammadAG, yes15:51
RST38hOk. File selector kinda done.15:51
RST38hBut how the hell do I EXIT from the damn QML code???15:51
lardmanhmm, all due to a typo - Architecture: Devel should have been Any15:53
lardmanis there such a thing as a paste-o rather than a typo?15:54
Venemomorning people16:04
Venemowhat should I do now? 1) implement a crude IRC backend hardcoded to freenode 2) keep polishing the GUI and begin the backend when it's done16:06
SpeedEvilAs a zeroth order approach, you could tail logs of some other chat program16:07
SpeedEvilProbably not sane though.16:08
Venemo'tail logs'?16:08
SpeedEvilFollow log files of someother IRC program, while it writes them, and pop the messages up in channels16:08
VenemoI've got no problems with GUI testability16:08
VenemoI already have a way to display fake messages.16:09
VenemoI also have no problem with the IRC backend, there are easy-to-use libs for that.16:09
SpeedEvilIf you're at the point of needing actual user-testers I suppose it'd be useful16:09
SpeedEvilAssume this won't run on n900.16:10
Venemoit will run on the N900, but it will need some minimal effort before it can16:10
Venemoand I'm willing to spend on it that effort once it's complete.16:10
lardmananyone compiled ocropus for harmattan yet?16:11
Venemowhat is that?16:12
RST38halterego: thanks =)16:12
lardmanVenemo: OCR16:12
* RST38h copies16:12
* SpeedEvil has not yet found an open source GPS better, or even comparable with the 1995 program that came with his scanner.16:13
RST38hyohoho, with alterego's suggestion I am actually going to make it work inside a C++ app16:13
VenemoMohammadAG, which option are you votion on? 1) or 2) ? :)16:14
*** lardman has quit IRC16:33
MohammadAGVenemo, don't hardocde, have a config file, there's no need for a GUI yet16:34
VenemoMohammadAG, the GUI is already 85% complete16:45
*** lardman has joined #harmattan16:45
*** lardman has joined #harmattan16:45
MohammadAGVenemo, I meant a GUI for settings16:48
MohammadAGit doesn't have to support multiple servers in 0.0.116:48
Venemoand it doesn't yet.16:48
Venemoit doesn't support any servers, not even one, right now :P16:53
*** fiferboy has joined #harmattan16:57
*** lardman has quit IRC17:00
*** epage has quit IRC17:05
*** NIN101 has quit IRC17:10
fiferboyVenemo: The filtering in WorkingSelectionDialog has the same performance issues as my attempts17:10
*** NIN101 has joined #harmattan17:10
fiferboyIt take a long time to initially load the component that call the dialog, althought the filtering itself goes smoothly17:10
fiferboyI'm going to try with a QAbstractSortFilterModel17:10
*** lardman has joined #harmattan17:12
djszapihi fiferboy :)17:13
*** epage has joined #harmattan17:16
djszapififerboy: Could you figure out the way of managing columns from QML ?17:18
djszapidoes anybody have issues on c-obs with built packages not getting into the repository ?17:21
*** NIN101 has quit IRC17:21
fiferboydjszapi: I think OBS is having issues this weekend17:21
fiferboyIt should make in there eventually, or you may have to trigger a rebuild later on17:22
*** NIN101 has joined #harmattan17:22
fiferboyI haven't figured out an _easy_ way to manage columns17:22
fiferboyBut there are ways to do it17:22
djszapififerboy: it is not about this weekend, it has been like that for weeks17:22
djszapififerboy: I guess the proper solution would actually be a TableVoew17:23
djszapi(since it is a table in theory, not a list)17:23
*** NIN101 has quit IRC17:26
fiferboyI'll be back later17:28
*** fiferboy has quit IRC17:29
*** NIN101 has joined #harmattan17:29
*** lbt has quit IRC17:49
*** lbt has joined #harmattan17:50
*** lbt has joined #harmattan17:50
*** djszapi has quit IRC17:51
*** Venemo has quit IRC17:54
*** rm_you has quit IRC17:54
*** rm_you has joined #harmattan17:55
*** rm_you has joined #harmattan17:55
*** Scifig has joined #harmattan17:57
*** javispedro has joined #harmattan18:03
*** Rizzer has quit IRC18:08
*** piggz has joined #harmattan18:16
alteregoIf our browser is html5 why can't I watch youtube vids?18:24
fralsare you being served the html5 video?18:26
GAN900alterego, the w22 browser is failcity.18:26
alteregoNope, is there a special gateway/site?18:26
alteregoAh there's a html5 trial ..18:27
javispedroI can watch youtube vids18:28
javispedroit does something weird18:28
javispedroseemingly it opens them in  fullscreen18:28
alteregohrm, iss working now.18:28
Tronicalterego: You need to go to and enable it.18:32
rcgdid anyone already figure out how to add an entry into the status indicator menu yet?18:32
rcg <- i compiled the profile example and placed the .so and .desktop file in the corresponding directories18:33
rcgi also updated .config/sysuid/ to include the new plugins (i also compiled a custom plugin based on the profile example18:34
hiemanshuI can watch videos, but they start at 30:0018:34
hiemanshuwell the timer shows em like that18:34
rcgunfortunately i seem to miss a little, yet important, bit of information to actually _show_ the new plugins18:35
rcgturning the bluetooth item in the menu on and off shows that the entry for the bluetooth plugin remains in .config/sysuid/ no matter whether its shown or not.. so there has to be another place where the visibility status of those plugin is actually triggered..18:36
rcgsome dbus signal gets send when the bluetooth item status is changed in the control panel.. and also some notifications are send about the status change.. so i assume that one of these notifications goes to sysuid or something to actually change the visibility of the item in the menu18:38
rcgalso.. while the shows up in .config/Trolltech.conf my two plugins do not show up there.. from this i conclude that the plugins are actually never loaded18:41
rcgnote i did restart the device.. and you can also force a restart of sysuid by killing it as user18:50
rcgi feel like being pretty close to the solution though somehow i cannot find it right now...18:53
rcgbut maybe this impression is also wrong and i am still miles away from seeing my custom entry in the status indicator menu ;)18:54
rcghmmm... qdbus /statusindicatormenu19:05
*** wazd has joined #harmattan19:06
rcgthis is at least something i did not look into yet :)19:06
*** lardman has quit IRC19:09
*** smoku has left #harmattan19:13
wazdHi all19:16
rcghi wazd19:17
wazdI've just realized that Harmattan browser needs "web app" option for bookmarks19:18
wazdso it would be opened without the statusbar19:18
*** epage has quit IRC19:22
rcg"..  For example, use QRegExp("$^") to allow no extensions, .."19:37
rcgalright.. that pretty much seems to narrow it down19:38
rcgalright.. obs is not working right?19:55
rcgah wait.. its going on19:58
*** Scifig has quit IRC20:17
*** lardman has joined #harmattan20:21
*** lardman has joined #harmattan20:21
rcgAegis rejecting system-ui_0.20.75_armel.deb: package 'system-ui' already installed from '' -- not replacing it from unknown origin20:22
rcgwell shoulda have seen that one coming.. so patching the code to accept other than the predefined plugins is also not an option... :(20:23
hiemanshurcg: you can purge the package and reinstall it20:23
dm8tbrwould be worth a try to experiment with force purging the package20:23
rcghmm i feel like being on the razors edge to reflashing :)20:24
dm8tbrpossibly :)20:25
*** Arkenoi has joined #harmattan20:27
rcgdpkg -P --force-all system-ui && dpkg -i system-ui_0.20.75_armel.deb worked... though killing sysuid yields in a reboot loop..20:33
rcgso going to reflash :S20:33
rcgRM680-22-6_PR_RM680:~# dpkg -i system-ui_0.20.75_armel.deb20:35
rcgPackage system-ui-screenlock-nokia refers to 'system-ui::/usr/bin/sysuid' which does not exist -- reference removed20:35
rcgSelecting previously deselected package system-ui.20:35
rcg(Reading database ... 35266 files and directories currently installed.)20:35
rcgUnpacking system-ui (from system-ui_0.20.75_armel.deb) ...20:35
rcgaegis-installing system-ui (from '')20:35
rcgPackage system-ui: denied 'usb-moded::USBControl' -- origin '' does not allow it20:35
rcgPackage system-ui: denied 'usb-moded::USBControl' for '/usr/bin/sysuid' -- origin '' does not allow it20:35
rcgSetting up system-ui (0.20.75) ...20:35
rcgfor the sake of completeness.. this was the output of installing the "custom" system-uid after --force-all purging the preinstalled one20:36
hiemanshurcg: and it didn't install?20:37
rcgit did install20:37
hiemanshuwell reboot it?20:37
rcgwell thats what it does since then ;)20:37
rcgRM680-22-6_PR_RM680:~# grep sysuid /var/lib/aegis/refhashlist20:38
rcgS 0  H 40 9d125eaddfd78b5747ac1a6651696147db4cb346 A 0 0 33261 P 9 system-ui R 14 usr/bin/sysuid20:38
javispedro'' origin, so no special privileges20:38
rcgthough it was worth the try..20:38
*** hardaker has quit IRC20:39
dm8tbrrcg: you had the same manifest as the original package?20:39
rcgi used the apt-get source system-ui20:40
rcgand only changed src/systemui/statusindicatormenu/statusindicatormenuverticalview.cpp20:40
rcgkinda sad they do not allow us to add plugins to the status indicator area menu20:41
rcgwell well.. its flashing20:46
dm8tbrwell I hope that soon we'll be able to figure out how to 'neuter' this thing and boot harmattan on top of a custom kernel20:47
javispedroand what will you lose when you do it.20:47
dm8tbrif I groked correctly what was said previously GPS might stop working and 'accounts' like nokia/ovi20:48
dm8tbrstill, worth exploring20:48
rcgwell.. i do understand that they want some level of security or whatsoever.. though it would still be nice if we could at least add menuextensions via the void MApplicationExtensionArea::setOutOfProcessFilter mechanism20:50
hiemanshuwell anything that needs a certificate or security will stop working20:50
javispedrothat would mean all connectivity would stop workin20:51
hiemanshuwell the modems will still working, but no nokia account, so no GPS or drive20:52
dm8tbrMeeGo CE also has to work20:52
javispedrorcg: complain loudly about this20:52
dm8tbrso some things must still work20:52
rcgjavispedro: am already looking up the official harmattan bug tracker20:52
dm8tbrit's on
rcgi assume thats the right place to bug nokia about this20:53
dm8tbrif you mean the public tracker20:53
javispedroMohammadAG: ping20:53
dm8tbryes and post the bug here, so we can vote for it20:53
hiemanshuif nokia looks at it20:54
dm8tbrI think they do20:55
dm8tbrand we can get the right people to look at this too20:55
dm8tbrlike attila or jukka20:55
rcgwell i could see a reason why they want to prohibit running extensions in the same process.. but not extensions being run in their own processes20:56
javispedrohiemanshu: (re no gps or drive) connectivity (wifi, etc) and _all_ account providers use aegis for either certificates or protected storage, so depending on how aegis is hammered, the stuff you will lose is rather impressive.20:56
hiemanshujavispedro: well yeah but it would make a nice music player :P20:57
javispedroor "alarm clock" :)20:57
hiemanshuyeah that too21:00
rcgwell as a nice side effect i now know that the lock code is not reset during a firmware flash21:06
javispedroI wouldn't put a lock code on the device21:06
javispedroaegis takes it very seriously21:06
* lardman wonders if his lockcode is what's killing his battery21:07
javispedro(reading I ponder why people set severity "enhacement" to bugs like fenix crashes or sudden black screens21:08
javispedroare expectations truly that low?21:08
lardmanwhat battery life do you chaps get?21:14
rcgheh.. "enhancement" is the default ;)21:15
lardmanI was getting to the end of the day with ~0.5 left, but these days the device has gone into power saving mode before I've had supper21:15
javispedroon the par with the N900, a bit less.21:16
rcgtheres the bug report21:17
lcuklardman, i am on a replacement battery already21:18
lardmanlcuk: oh, they are that poor are they, or something to do with the charging?21:19
lardmanhow does one replace the battery? Torq screws or did I miss the slide off cover? Also, what type is it?21:21
lcuklardman, for some reason the battery just stopped charging via usb21:22
lcukand i had to replace it21:22
lcukcannot remember the number, but it is same one as n821:23
lardmanhmm, not ideal21:23
hiemanshulcuk: Torx T5 or T4 shoulr work21:23
lcukhiemanshu, already done and fastened back up and working nicely21:24
hiemanshuI used one of those, cant recall which correctly21:24
hiemanshulcuk: yeah, that was meant for lardman, tab fail :P21:24
lcuklardman, are you having your supper later?21:27
lcukor are you using device differently21:27
npmso how does one mount a camera stand to the n950.21:33
* javispedro fills another feature request for djszapi's bugs to ignore list21:34
npmi have this flowering vine growing into second story blinds21:34
npmfigure i should capture png's once a minute and turn it into a movie21:34
npmactually n900... since i'm using the n95021:35
*** antman8969 has joined #harmattan21:36
lcuknpm, afaik timelapse is in maemo extras-*21:37
lardmanlcuk: nah, not sure why it's suddenly fallen off a cliff actually21:41
lardmanI guess it might be that I've been home more rather than at work21:41
lardmansignal strength, etc21:41
* lardman thinks of a random app after Holly asks about National Trust butterfly survey21:42
lardmanthey want photos of butterflies and want a tweeted picture with postcode for their map21:43
lardmanand a hashtag of course21:43
lardmanso while the subject material isn't really my thing, the idea is quite cool21:43
lardman fwiw21:44
* SpeedEvil wonders how to teach mbarcode about butterfly patterns.21:44
lardmanPhotoAnalyser now21:44
npmlcuk ... timelapse is easy to do in a script... infact i have a gstreamer pipeline that'll do it as a one-liner i think21:44
lardmanSpeedEvil: need to write a butterfly identifier plugin, then send dbus message to twitter client21:44
npmbasically you let it run, the ^C it and it writes the last frame of the movie...21:44
lardmanam currently trying to write a text identifier plugin, but wasted a fair bit of time with tesseract21:45
SpeedEvilnpm: you can just take a picture every 60s surely?21:45
lardmannow looking at ocropus21:45
SpeedEvilwith gstreamer21:46
npmtry to not let them get in my teacup though :-)
* lardman worries if that is work safe ;)21:46
lardmanI don't fancy the chances of producing something Google Goggles like with only on-board processing21:47
lardmantime to intercept some of my Tab's comms perhaps21:48
* SpeedEvil puts on some Oasis.21:49
npmlardman: let me know when you get java running :-) (re )21:50
lardmannpm: not going to be soon, and I'd want to compile to binary code anyway21:51
lardmanbut yes will do as and when I do21:51
npmSpeedEvil: yes, but the table on which i can put the camera jiggles a bit if i lean on it... wouldn't work very well for timelapse21:51
SpeedEvilSo don't lean on it?21:52
npmalthough duct-tape and a microphone stand might do it :-)21:52
lardmannpm: what are you taking timelapses of?21:52
lardmanflowers opening?21:52
SpeedEvilI've used a book leaning against it to push it against a window.21:52
npmnot very important, just something poking into my office from outside21:52
*** seif has joined #harmattan21:52
* SpeedEvil pops out to take a pic21:53
npmor i could get back to work :-)21:54
* lardman is now on holiday, and it's nearly 8pm here21:54
lardmanah and it's Sunday too, I forget now that I'm on holiday ;)21:55
*** smoku has joined #harmattan22:00
*** seif has quit IRC22:04
*** harbaum has joined #harmattan22:13
wazdsorry to bother you with my design crap again guys, but check it out :)
wazdKinda final layout, don't think that there's anything wrong with it22:13
lcukwazd, is this qml?22:13
wazdnot yet22:14
SpeedEvilwazd: looks ok22:14
wazdbut it's about to become qml :D22:14
lcukwazd, extra bonus points if you get the wind direction arrow pointing in the right direction!22:14
SpeedEvilAnd the raindrop.22:15
lcukwazd, also, you are missing nighttime temp22:15
javispedrogood idea22:15
lcukfrom the Now view22:15
SpeedEvilIf now is daytime, it can't have nightime temp.22:15
wazdwell, it's "now"22:16
lcukSpeedEvil, the other days have a day/night temp22:16
SpeedEvilI guess I'm assuming now = wednesday22:16
lcukwhat is the temp tonight?22:16
wazdit shows, like, current temp :)22:16
lcukif now is 30deg, what is it tonight?22:16
wazd12 :)22:16
SpeedEvilToday might be better tha now.22:16
lcukie, you are missing a reading22:16
wazdforgot to switch wednesday to today22:16
SpeedEvilAnd leave off 'wednesday'22:16
lcuksince this is all qml, is it possible to have rain and wind particles?22:18
wazdno :)22:18
lcukno because that would frustrate you, or no because it is not feasible?22:18
lardmanlcuk: +1 :)22:18
lardmanwazd: looks good though22:19
wazdwell, first of all AccuWeather is doing that kind of app with fancy animations and stuff :)22:19
* lcuk agrees with lardman22:19
wazdWe just wanted some plain simple application :)22:20
lcukI understand wazd, it looks good22:20
lcukand at some point in future upgrades ca go in22:21
lcukit would make for a nice set of tutorials on qml to make those pretty animations though22:21
javispedrowas that ever released?22:23
wazdIt's gonna be I guess22:23
wazdif not bundled with n922:23
wazdshit, sorry, wrong video :D22:23
wazdthat one :D22:24
javispedrothat was never released22:24
wazdjavispedro: anyway, there's accuweather app for n900 :)22:24
wazdjavispedro: check ovi store :)22:24
javispedroah, ok :)22:27
antman8969hey javispedro, you see
antman8969came up on my google news alerts, pretty cool22:28
javispedrowell, you shouldn't be surprised :)22:30
antman8969of course not lol22:30
*** Ken-Young has joined #harmattan22:37
Ken-YoungI am trying to test a Harmattan app using Qt Creator.   I want to run the app on my N950, not QEMU.  I have configured Qt Creator with a Maemo Device Configuration for the N950, and I have made that configuration the default.   Qt Creator definitely can communicate with my N950, because the "Test" and "Remote Processes..." buttons on the Qt Creator Maemo configuration page gives me the correct results.22:39
Ken-YoungHowever, I cannot get Qt Creator to actually run the app on the N950.22:40
Ken-YoungThe app always gets run in the QEMU window.   Does anyone know how I can get Qt Creator to run the app on the N950, rather than the QEMU emmulator?22:41
alteregobottom left of qt creator you need to select the build target22:52
*** harbaum has quit IRC22:55
*** hardaker has joined #harmattan22:57
Ken-Youngalterego, Thank you.22:58
*** hardaker2 has joined #harmattan23:04
*** hardaker has quit IRC23:05
lardmanhmm, low battery, and I've only turned the phone on once today and sent one sms23:06
*** wicket64 has joined #harmattan23:07
*** hardaker2 has quit IRC23:07
*** hardaker has joined #harmattan23:09
Ken-Youngalterego, I don't see anything at the bottom left of the Qt Creator that allows me to select the build target.   The bottom item on the left side is an icon labelled "Start Maemo Emulator".   Where is what I'm supposed to use to choose the target, relative to that "Start maemo emulator" icon?23:11
*** javispedro has quit IRC23:11
lcuklardman, hm do you have all apps running?23:27
Ken-Youngalterego, Oh, nevermind.   I finally stumbled into the correct menu.23:29
ieatlintyeah, as i recall for qtcreator and switching between the emu and phone, you hit the projects tab, select the "run" tab at the top for the harmattan target, extend the details of the deployment step, and switch from qemu to your n950 -- simple, right? :P23:37
Ken-Youngieatlint, Yeah,  It's nice to be funally running on the phone.   I don't think I'll be using QEMU much, it's so sluggish in comparison.23:39
ieatlintyeah, nokia kinda sucks at emulators23:39
ieatlintthe qemu one at least is a huge improvement from the previous one23:40
Ken-YoungThe one for the N900 was never pixel-perfect.   Lots of little gotchas like slightly different fonts on the device and emulator.23:40
ieatlintoh, yeah, the scratchbox based one was23:41
ieatlintthe one that was thrown in with the qt sdk was the same as for symbian, and equally incorrect23:41
*** [XeN] has joined #harmattan23:49
*** [XeN] has quit IRC23:52
*** [XeN] has joined #harmattan23:52

Generated by 2.13.0 by Marius Gedminas - find it at!