**** BEGIN LOGGING AT Sun Jul 10 02:59:58 2011 Jul 10 07:11:03 thinking whether to flash or not my n900? Jul 10 07:11:14 why is it broken? Jul 10 07:12:11 usb support has an issue Jul 10 07:13:18 unable to connect to pc Jul 10 07:16:47 * RST38h laughs satanically Jul 10 07:17:01 Did you hack Aegis? Jul 10 07:17:20 No, I managed to run my stuff on Android. Jul 10 07:17:46 Only Gingerbead though, Froyo would require at least as much work. What a mess... Jul 10 07:18:32 It's Android. what do you expect? Jul 10 07:19:02 Think Windows 95; a new OS with bits hacked on Jul 10 07:22:29 wasn't win95 just dos with a shinny tacked on? Jul 10 07:25:04 ruskie: Sort of, yes. Jul 10 07:25:12 that was 3.1 you're thinking of Jul 10 07:25:21 but dos was still integral to 95 Jul 10 08:05:40 hah, so much spam in the nitdroid wiki Jul 10 09:42:40 what is the easiest way to setup scratchbox on a 64 bit system? Jul 10 09:43:09 the installer regrets to install as it is "not supported" Jul 10 09:55:18 32bit virtual machine? Jul 10 09:57:03 never tried, but a 32bit chroot might work too? Jul 10 09:59:39 my aim is to avoid a virtual machine resp. chrooted env Jul 10 10:00:17 i will go for the latter if there is no native solution Jul 10 10:04:56 check the harmattan pages, but I think 64bit is not supported Jul 10 10:06:37 BluesLee: there is no easy way to do so Jul 10 10:06:46 BluesLee: it took me a while to make it work Jul 10 10:07:13 BluesLee: there is a switch that makes it ignore 32bit Jul 10 10:07:15 64bit* Jul 10 10:07:19 and you have to disable VDSO Jul 10 10:07:43 ehlo... anynone here from nokia? Jul 10 10:07:45 :) Jul 10 10:08:33 anyone here that participates in the N950 devkit program ? Jul 10 10:08:48 anyone here ? Jul 10 10:09:22 * jonwil is here Jul 10 10:09:27 Progra Jul 10 10:09:37 hello :) Jul 10 10:09:41 Programmatic PDF generation ... Pain in the ass .. Jul 10 10:10:00 do you participate in the N9 devkit prog ? Jul 10 10:10:08 nope, I do not Jul 10 10:10:13 yes, I do Jul 10 10:10:21 did you order your n950 ? Jul 10 10:10:27 Just a happy N900 owner looking for something useful to do :) Jul 10 10:11:09 alterego: did you order the n950? Jul 10 10:11:36 alterego: .. and... in which batch are you in ? Jul 10 10:12:15 I was in the first batch, I ordered my N950 on Friday as there was an issue with Nokia Developer finding my developer account. Jul 10 10:12:24 ok... trick question Jul 10 10:12:33 how many times did you order it until now ? Jul 10 10:12:36 Quim mentioned there were 10 or so people they couldn't identify on Forum Nokia Jul 10 10:12:41 I mean, Nokia Developer. Jul 10 10:12:53 I ordered once, I'm not one of the ones that got multiple emails ;) Jul 10 10:13:00 I actually didn't get a single email. So ... :) Jul 10 10:13:04 oh Jul 10 10:13:07 Caco3: congratz you built some deb ! Jul 10 10:13:10 because somehow, i ordered it twice :) Jul 10 10:13:20 and don't know if it is ok Jul 10 10:13:22 jacekowski: disable vdso as a bot parameter? what do you mean with the switch to ignore 32bit? Jul 10 10:13:30 i got to "a nokia n950 is waiting for you" mails Jul 10 10:13:35 and have placed two different orders Jul 10 10:13:52 perhaps I should look at meego forum Jul 10 10:13:53 I dont think they will send multiple N950s out Jul 10 10:14:05 I guess so Jul 10 10:16:49 BluesLee: on maemo wiki there is whole article that describes it Jul 10 10:16:52 BluesLee: read it Jul 10 10:17:12 * jonwil is going to play with the N900 icd2 daemon Jul 10 10:17:16 find out what it does Jul 10 10:17:52 RzR950: huu? none got built. either it sais failed or excluded :( Jul 10 10:20:28 jacekowski: thanks, you mean http://wiki.maemo.org/Documentation/Maemo_5_Final_SDK_Installation#Installation_of_x86-64_Debian_based_distributions Jul 10 10:28:27 RzR950: would be great if you could give me a hint what I am doing wrong :| Jul 10 10:28:47 sudo flasher-3.5 -F -f -R Jul 10 10:29:00 Where is my firmware image for N900 ? Jul 10 10:29:14 you download it? Jul 10 10:29:21 ~flash Jul 10 10:29:22 rumour has it, flashing is http://wiki.maemo.org/Updating_the_tablet_firmware Jul 10 10:29:31 iirc it has all links there Jul 10 10:30:24 ruskie: is there a latest one ? Jul 10 10:30:36 yes PR1.3 Jul 10 10:30:42 everything is linked to from that page Jul 10 10:30:53 it tells you where to download how to access the downloads etc... Jul 10 10:31:09 should I give the IMEI number ? Jul 10 10:31:18 you'll have to Jul 10 10:31:19 http://tablets-dev.nokia.com/nokia_N900.php Jul 10 10:31:22 yup Jul 10 10:32:03 that way nokia is tracking all those who updated firmware :D Jul 10 10:32:16 Or just use a fake one Jul 10 10:32:20 Or one digit off Jul 10 10:32:28 just because you downloaded it doesn't mean you updated it Jul 10 10:33:20 or you can use other people's IMEI　if you know how to ask google properly Jul 10 10:34:16 lol Jul 10 10:35:43 I'm not joking, I found 2 the other day, one of them was working. The similar technique could be used for obtaining firmwares for other NIT you don't even own Jul 10 10:37:17 you know that you can get somebody elses imei on tmobile website Jul 10 10:37:31 just type his tmobile phone number and you get imei and phone make and model Jul 10 10:37:47 seliously??? Jul 10 10:37:51 lol Jul 10 10:37:53 http://support.t-mobile.co.uk/help-and-support/index?page=home&cat=SOFTWAREDOWNLOADS Jul 10 10:38:22 i typed mine and what i got was Jul 10 10:38:23 Nokia N95 IMEI: 356962017909996 Jul 10 10:38:27 when I once needed an IMEI (as I didnt have my N900 yet) I just googled for an N900 image in google :) Jul 10 10:40:35 I entered a random one, first time was no go, second time I got imei of some sony w95 or something like that lol Jul 10 10:41:03 so that's how those china fakes get valid imeis Jul 10 10:43:30 lol I　got an unused t-mobile number, or at least not registered yet :D Jul 10 10:47:11 how can i get rid of U-Boot? apt-get --reinstall install kernel ? Jul 10 10:47:18 vdv: flash it Jul 10 10:47:37 and maybe then apt-get --reinstall install kernel-flasher Jul 10 10:50:40 jacekowski, won't it work without flashing? Jul 10 10:52:45 apt-get --reinstall install kernel kernel-flasher should work Jul 10 10:59:21 MohammadAG, should i also deinstall U-Boot explicitly? Jul 10 10:59:47 no, it lives in the kernel partition Jul 10 11:00:10 i.e. dpkg or apt-get knows nothing about it? Jul 10 11:00:52 i mean won't package manager come to unstable state? Jul 10 11:01:32 nope Jul 10 11:08:05 MohammadAG: ok, thanks Jul 10 11:39:51 Morning everyone Jul 10 11:43:53 protesters have been stopped in their tracks after Facebook aided Israel in cracking down on the group of activists from the UK, France, and Belgium who planned their event using the popular social networking site. Jul 10 11:47:44 Facebook = evil Jul 10 12:06:08 DocScrutinizer, were we told to keep i2c-tools in -devel? Jul 10 12:07:40 I don't recall such advice, but I think it doesn't matter anyway. We should however make sure it's "hidden" (like a lib or sth) Jul 10 12:08:34 yeah we should make sure i2c-tools doesn't appear in h-a-m or as an option for "normal people" to install Jul 10 12:09:11 it's clearly nothing a "user" wants to install, like "mc" or "puzzlemaster" Jul 10 12:09:51 "user" does not want "mc", "mc" is too "ugly" for a "user" Jul 10 12:11:21 MohammadAG: rationale is: i2ctools isn't more dangerous than messing with sysnodes e.g. for corevoltage directly Jul 10 12:12:39 my yesterday experience shows aegis for example should also get such a big warning on installation, as i2ctools has Jul 10 12:12:51 ;-P Jul 10 12:14:39 iirc it wont show as it doesn't have a maemo-display-name or whatever Jul 10 12:15:21 MohammadAG: (were we told) maybe I myself suggested formerly to keep it in devel - I am allowed to change my mind, am I? Jul 10 12:15:53 DocScrutinizer, robbiethe1st asked me to promote it Jul 10 12:15:58 for Backupmenu(??!) Jul 10 12:16:09 lol what? Jul 10 12:16:11 yeah, he has charging in BM Jul 10 12:16:35 ah Jul 10 12:16:38 if he adds it as dependency, doesn't it get autopromoted if backupmenu is promoted? Jul 10 12:16:43 nope Jul 10 12:16:44 since quite some time, works flawlessly obviously ;-D Jul 10 12:16:59 otherwise things that are not ready for Extras would be promoted Jul 10 12:17:20 ShadowJK, it should Jul 10 12:17:28 I thought libraries and such came along by themselves :P Jul 10 12:17:50 well, I2Ctools *is* tested, stable, working, and ready for extras. Though definitely not a user tool Jul 10 12:18:19 grr Jul 10 12:18:30 do I need a bugs email/url for such a package? Jul 10 12:18:44 how would a library ever gather the 10 thumbs-up? Jul 10 12:19:13 MohammadAG: usb hostmode thread Jul 10 12:19:26 I did same for *# Jul 10 12:21:24 as h-e-n is the "main consumer" of i2ctools, I guess it's fine to point to the tmo thread for reporting bugs Jul 10 12:22:00 of course the pkg maintainer should have subscribed that thread in tmo then ;-) Jul 10 12:22:54 MohammadAG: are you going to repackage h-e-n or what? Jul 10 12:23:19 I'd like it when you'd include my improved boost script then Jul 10 12:24:22 DocScrutinizer, just i2c-tools Jul 10 12:24:30 aah, k Jul 10 12:24:31 USB hostmode really needs an update :/ Jul 10 12:24:40 indeed Jul 10 12:24:47 a *major* update Jul 10 12:25:03 a rewrite Jul 10 12:25:11 with a QML UI and fancy transitions Jul 10 12:25:16 the gui, well, yes maybe Jul 10 12:25:18 * MohammadAG includes sarcasm quotes Jul 10 12:27:31 I think the gui could use a semi-automatic mode, that automatically does the (user adjustable) waiting and button clicking, maybe also some syslog/dmesg scanning for trigger keywords Jul 10 12:27:46 yep, the old code's shitty at best Jul 10 12:28:08 I'll rewrite the old concept in not so shitty code Jul 10 12:28:12 we can improve on that Jul 10 12:28:22 I'd happily help to define the structure (way it works) of a rewrite Jul 10 12:28:24 when I'm done I'll nuke the old git repo Jul 10 12:28:49 I'm not too happy about those system("command &") lines Jul 10 12:28:57 there's a good QProcess library for that Jul 10 12:30:03 hmm, I'm not too worried about *how* the gui calls the backend. I however have a clear notion about what's backend's duty and what's the API Jul 10 12:31:46 I actually forgot how all this stuff works, brb, gonna look for my adapter Jul 10 12:32:09 as this idea is a tiny bit different to how things work now, it'd be nice if you consult me prior to releasing or even putting to much effort into writing the new GUI Jul 10 13:19:20 moo javispedro Jul 10 13:19:30 morning Jul 10 13:23:12 morning javispedro RST38h Jul 10 13:23:41 who said dhl works on weekends? my package is sleeping at leipzig airport for 33 hours now ... Jul 10 13:23:51 dhl works, couriers do not Jul 10 13:23:57 my package moved on saturday afternoon Jul 10 13:24:07 gri: I did, dhl works on weekends, but no deliveries within cities Jul 10 13:24:48 from what I gather, you can call them and they will deliver you on a sunday even, for an extra fee. Jul 10 13:24:58 yup Jul 10 13:25:05 ah ok ... leipzig is still about 600 km from here Jul 10 13:25:50 gri: there is a good chance it might move tonight Jul 10 13:25:59 javispedro: yours at the local DHL yet? Jul 10 13:26:08 * hiemanshu is still waiting for it to be shipped Jul 10 13:26:14 yes Jul 10 13:26:42 javispedro: nice, so you'll have it tomorrow morning :( Jul 10 13:26:44 it's on a warehouse they have near the city airport. Jul 10 13:27:01 ah Jul 10 13:27:15 but I do expect getting it tomorrow morning, I'm pondering how to justify going late to work =) Jul 10 13:27:17 within city means next day Jul 10 13:27:54 javispedro: diarrhea :P Jul 10 13:28:27 hiemanshu: You can check how long it should take for a package when it's on its way here: http://dct.dhl.com/input.jsp?langId=en Jul 10 13:29:25 btw, wouldn't anyone know a nice Qt QIODevice wrapper that allows me to read individual bits? Jul 10 13:30:42 gri: it says if it ships tomorrow, it will be here wednesday, as I expected :/ Jul 10 13:34:01 javispedro: you can only read a byte at a time, then I'd use a mask or something to get individual bits. Jul 10 13:34:39 Or a high 0b10000000 bit bask, then shifting to get each bit. Jul 10 13:34:53 I know how to do that, but it's pain, it's much natural to write whatever->readBit() and let whatever remember which the last bit was for me Jul 10 13:35:25 in fact I have this whatever thing implemented for glib IOChannels, and even Java InputStreams, was thinking maybe I could skip reimplementing for Qt ;) Jul 10 13:35:34 Well, you can just inherit from QIODevice something like IOBitReader and implement readBit ;) Jul 10 13:35:45 Ah, right. Jul 10 13:35:57 sadly, if I inherit from QIODevice means I cannot use e.g. ZipIODevices Jul 10 13:36:06 so it has to be a wrapper. Jul 10 13:36:37 Hrm .. Jul 10 13:36:47 Inherit from ZipIODevice? :) Jul 10 13:36:54 and the complication comes if you try to sync readByte() with readBit(), cause it means there could be unaligned readBytes() Jul 10 13:37:03 Yeah Jul 10 13:38:26 ah well, let's write it. Jul 10 13:38:53 * javispedro inaugurates projects/Qt dir Jul 10 13:47:01 * jonwil never thought he would see so many people so focused on refreshing a package tracking page so frequently Jul 10 13:48:21 javispedro: moo Jul 10 13:48:22 someone here installed Harmattan Platform SDK (scratchbox)? Jul 10 13:48:37 mo. Jul 10 13:49:13 after running the python installer i cant locate the "scratchbox" binary Jul 10 13:49:22 javispedro: may I invite you to temporarily join h-e-n for N9(50) project, on a consulting-occasionally basis? Jul 10 13:49:49 still same channel? =) Jul 10 13:49:57 sure Jul 10 13:50:11 still same shortcut mhd as well Jul 10 13:50:44 not that there's any particular action like huge discussions right now atm Jul 10 13:51:22 * javispedro adds to autojoin list Jul 10 13:51:28 it's just the location where dm8tbr and me (and MohammadAG ) are about to start tossing around ideas and findings Jul 10 13:54:59 DocScrutinizer: can I pop in too? Jul 10 13:55:53 hiemanshu: it's a developer-only no general discussions no support chan, but yes of course, you're welcome to lurk and to contribute when you feel appropriate Jul 10 13:56:24 DocScrutinizer: what chan? I can help with the GUI part, and I know a little kernel hacking too Jul 10 14:00:25 I'm already on the GUI :P Jul 10 14:04:09 MohammadAG: you are? well I am sure I can help :P Jul 10 14:22:28 what is the right way to check for maemo6 in qmake? the examples use "unix:!symbian:!maemo6" but this matches a regular linux desktop too Jul 10 14:22:56 I meant: unix:!symbian:!maemo5 Jul 10 14:24:48 also, I miss the Q_WS_MAEMO_6 define. I've found somewhere in the net that this has been removed intentionally. But I couldn't find the reason why and what to use instead... Jul 10 14:30:56 mzanetti - I had a similar problem to identify scratchbox in qmake, I used exists() to look for a file that only/also exists in the target environment. In my case I used exists(/usr/bin/hildon-home) to identify scratchbox with FREMANTLE targets Jul 10 14:31:54 I didn't know Q_WAS_MAEMO_6 was removed. My code uses still it. Ughh Jul 10 14:32:09 Scifig: ok... that should work, thanks. Still this feels like a workaround and it would be interesting to know why the MAEMO_6 defines are removed. Jul 10 14:36:41 mzanetti - I agree, this is just a workaround. Hopefully they are replacing Q_WS_MAEMO_6 with some MEEGO entry or something Jul 10 14:48:01 mzanetti - On the MeeGo mailing list, it is recommended to use Qt Mobility API (QSystemInfo). More lines of code but i think it works. Jul 10 14:48:24 instead of Q_WS_MAEMO_6 i mean. Jul 10 14:50:12 Scifig: I've seen this. But it doesn't do the job: Jul 10 14:50:18 a) it doesn't work for meego specific includes Jul 10 14:50:40 b) it gives only the hardware platform (e.g. N9 or N950) but not the software platform Jul 10 14:53:16 cancel b)... It does provide the operating system informations... still, the include problem persists Jul 10 14:54:15 What include problem? Jul 10 14:54:32 if I want to include a header thats only available on meego Jul 10 14:55:04 Yes, i got it Jul 10 14:55:22 sure... I could check that in the pro using exists()... But we're at the workaround again then... Jul 10 14:58:19 This seems like a good topic for Meego mailing list discussion. It would also help other devs. Can you post ur question there? Jul 10 14:59:46 yes Jul 10 15:00:49 I don't think that's a work around, it's perfectly reasonable to test your dependencies properly pre build .. Jul 10 15:01:30 Ironically, Q_WS_MAEMO_6 is used in the example given for QSystemInfo --> http://doc.qt.nokia.com/qtmobility-1.1.0-beta/sysinfo-dialog-cpp.html Jul 10 15:03:10 ok, typing seems to work, just barely Jul 10 15:03:23 alterego: I'm talking for example about about qmkeys.h. This is part of the official Qt release for Harmattan and Meego. Jul 10 15:03:56 while I agree that it makes sense to check for third party includes I don't think it is reasonable for platform internal stuff Jul 10 15:06:39 Scifig: also all the harmattan example projects use it. But since it has been removed they define it themselves in the .pro using "unix:!symbian:!maemo5" Jul 10 15:07:05 which, as I said, interferes with basically all other linux systems Jul 10 15:07:59 I use Q_WS_MAEMO_6 Jul 10 15:08:19 then when they roll the new define I could just check for that and add Q_WS_MAEMO_6 Jul 10 15:08:28 My new N900 seems to have another fault :( Jul 10 15:08:38 contains (MEEGO_FLAVOR, "harmattan) DEFINES += Q_WS_MAEMO_6 Jul 10 15:08:51 two glowing lights on the middle of the screen during a white screen Jul 10 15:09:27 * pavin900 loves maemo . debian and even the n900 phone but has to return it !! Jul 10 15:10:45 MohammadAG - Q_WS_MAEMO_6 is not going to supported sooner or later. It is being removed from Qt documentation. Jul 10 15:11:31 Scifig, so add it yourself Jul 10 15:14:05 MohammadAG: what is MEEGO_FLAVOUR? Jul 10 15:14:20 its empty here Jul 10 15:14:26 mzanetti, it'll be added soon Jul 10 15:14:42 ah ok. Great. Jul 10 15:14:42 FLAVOR btw, no U Jul 10 15:15:08 afaik this was discussed on #meego Jul 10 15:17:34 Interesting idea. Guess this MEEGO_FLAVOR also works for WeTab etc then? Jul 10 15:19:30 not sure, ask there Jul 10 15:19:57 Scifig: pretty sure it will. this is also be a good explaination why the MAEMO_6 specific stuff as this is way more flexible Jul 10 15:20:21 ... why the MAEMO_6 specific stuff has been removed ... Jul 10 15:21:18 mzanetti: just define it yourself? Jul 10 15:22:04 hiemanshu: no, the MEEGO_FLAVOR thing Jul 10 15:42:09 MohammadAG: make the new h-e-n GUI in a way it will work for fremantle and harmattan likewise Jul 10 15:42:21 hi, is it possile to run a 2nd instance of maemo from sd? Jul 10 15:43:31 DocScrutinizer: will h-e-n work on n9? Jul 10 15:48:37 I am downloading RX-51_2009SE_20.2010.36-2.004_PR_COMBINED_004_ARM.bin Jul 10 15:48:41 is it the new one ? Jul 10 15:49:28 yes Jul 10 15:49:35 BluesLee: we're working on it Jul 10 15:49:35 DocScrutinizer, won't work, conceptually Jul 10 15:49:58 the concept of h-e-n on Harmattan should be in the status menu Jul 10 15:50:13 MohammadAG: won't work? too bad - why? Jul 10 15:50:23 :nod: Jul 10 15:50:34 can't we do the same for N900 fremantle? Jul 10 15:50:37 way too hard to do with C/Gtk for Fremantle Jul 10 15:50:41 at least for me Jul 10 15:50:46 so unless javispedro wants to step up :p Jul 10 15:50:56 aaah, maybe hiemanshu can help? Jul 10 15:51:07 he's a Qt guy, like me Jul 10 15:51:14 I see Jul 10 15:51:23 ~lart gtk Jul 10 15:51:23 * infobot duct-tapes gtk to the floor and drools on him Jul 10 15:51:28 * javispedro sighs Jul 10 15:51:37 ^ Gtk fan Jul 10 15:52:23 ... doing Qt work. Jul 10 15:54:17 MohammadAG: how about tha generic status applet executable? Jul 10 15:54:45 you can probably load Qt applets in Harmattan, but I wouldn't do it Jul 10 15:54:49 *Fremantle Jul 10 15:57:38 iirc I once suggested a cmdline binary to show an arbitrary png in status systray, and also to add a button to the status menu, and to call a callback function whenever that button gets hit. Did that idea went anywhere? Jul 10 16:00:05 anyway the real h-e-n GUI will be a fullscreen window on both fremantle and harmattan, I guess it should be portable Jul 10 16:00:23 adding somthing in statusmenu on fremantle is pretty trivial Jul 10 16:00:46 and we should change the UI paradigm to a policy not to stop hostmode on closing the main window Jul 10 16:00:46 http://maemo.org/packages/source/view/fremantle_extras_free_source/statusmenu-fmms/1.1/ Jul 10 16:02:22 (not stop hostmode on close of GUI) that's a rationale - one of them - why booston should stay a autonomous process that can live on without the invoking GUI Jul 10 16:03:07 DocScrutinizer: using a .pid file, running it in background, and making the UI aware of the .pid file makes sense :) Jul 10 16:03:59 hmm, yet undecided about the pid file, given the fact the process has a unique name Jul 10 16:04:36 DocScrutinizer: you can have the GUI handle the creation of a .pid file, so only when its started via GUI a .pid file is created Jul 10 16:04:47 my original design planned for a named pipe so a UI frontend instance can talk to the daemonized instance Jul 10 16:05:03 hiemanshu: fair enough Jul 10 16:05:51 so GUI wouldn't tear down a booston process that's not been started by the GUI - sounds good Jul 10 16:05:59 yup Jul 10 16:06:21 and QProcess can handle this easily Jul 10 16:07:34 your domain of expertise - I'll keep my sticky fingers out of h-e-n GUI details mostly ;-) Jul 10 16:08:14 I'm the dude for the dirty lowlevel stuff and general system architecture Jul 10 16:08:51 just sugesting we should try to keep h-e-n GUI in sync for fremantle and harmattan Jul 10 16:09:28 for me this means I need to specify platform agnostic APIs for the low level stuff Jul 10 16:10:13 for you it means you need to think about nuisances like gtk-status-applets vs qt Jul 10 16:11:52 frals, adding it isn't the hardest part Jul 10 16:11:55 making it work is :P Jul 10 16:12:40 DocScrutinizer, technically, I could allow what I did for sociality Jul 10 16:12:53 X button hides the window from the user, clicking the icon makes it show again Jul 10 16:13:31 fair enough, but will it work for harm "close all" on taskswitcher? Jul 10 16:13:33 technically, sociality is still running, but not to the average user Jul 10 16:13:37 DocScrutinizer: /me is getting ready to flash the N900 ! Jul 10 16:13:49 DocScrutinizer, the close button simply sends the window manager's close signal Jul 10 16:14:12 example, in fremantle it sends close to all stacked windows + the main window Jul 10 16:14:30 so just reimplementing closeEvent() and making it ignore the event would not close the app Jul 10 16:15:04 see https://gitorious.org/qtbook/qtbook/blobs/master/mainwindow.cpp#line614 Jul 10 16:15:06 just make sure it won't interfere with harmattan's weird way to close apps Jul 10 16:15:24 it's the standard way all window managers' X button works :) Jul 10 16:15:40 close all is probably just a foreach(app, applist) close Jul 10 16:15:50 harmattan has no classic window manager X-buttons Jul 10 16:16:01 yup, probably Jul 10 16:16:30 it does, they're just hidden by default Jul 10 16:16:35 tap n hold and they're there Jul 10 16:16:47 we easily can test that by getting an app into an unresponsive state, then clicking "close all" Jul 10 16:16:48 now what would be interesting is if it detects that an app crashed Jul 10 16:17:05 that can be easily done Jul 10 16:17:15 QMainWindow::closeEvent(QCloseEvent *e) Jul 10 16:17:22 { e->ignore(); } Jul 10 16:17:35 it won't respond to the window manager's X button press Jul 10 16:18:09 I thought more along the lines of killall -sigstop calendar - then click closeall Jul 10 16:18:20 see what happens to calendar app Jul 10 16:18:33 wouldn't it simply close? Jul 10 16:18:43 see if window manager complains Jul 10 16:18:57 Harmattan does complain if an app stalls Jul 10 16:19:07 does it? Jul 10 16:19:09 noticed that in the SDK Jul 10 16:19:11 yes Jul 10 16:19:13 ok Jul 10 16:19:17 MeeGo's way is way better though Jul 10 16:19:28 App sucks, do you want to close it? Jul 10 16:19:43 and what does harm? Jul 10 16:19:47 Harmattan does it in a boring "App is not responding" Jul 10 16:20:09 hmm, sounds similar Jul 10 16:20:54 DocScrutinizer, http://i52.tinypic.com/1t62iv.jpg Jul 10 16:21:26 LOL Jul 10 16:22:41 dman, my clock shows 18:18 but iut seems to get dark outside. Did the relocate my building? Jul 10 16:22:56 MohammadAG: LOL Jul 10 16:23:08 score one extra point for MeegoCE ;) Jul 10 16:24:02 this reminds me that I used to work on a project where the internal error page had a textbox with the label: "Please fill the following textbox with any kind words you might want to send to developer". Jul 10 16:27:56 that's btw the 'engineering english' locale Jul 10 16:31:48 MohammadAG: using a .pid file would make sense to conserve resources, and stopping someone from killing it from the command line Jul 10 16:37:52 I think the GUI will also implement the status applet, so it might want to stay alive, main window hidden. Nevertheless it must recover if somebody or sth decides to kill it, and user restarts it from applauncher. With pid file it can pick up on prev state about hostmode&booston established, rather than trying to set up things new and thus messing everything up Jul 10 16:39:12 the only thing you'd lose when killing the hidden h-e-n GwindowUI implementing this sheme, would be the status bar applet and status menu button to re-open the main GUI window Jul 10 16:39:42 s/Gwindow/G/ Jul 10 16:39:43 DocScrutinizer meant: the only thing you'd lose when killing the hidden h-e-n GUI implementing this sheme, would be the status bar applet and status menu button to re-open the main GUI window Jul 10 16:40:48 also would implicitly immunize the app against multiple invocations Jul 10 16:41:16 sure we probably don't need that as I guess it's an immanent feature of appstarter anyway Jul 10 16:45:01 OTOH maybe the main GUI wants to spawn a separate lightweight process for the applet, so resource footprint stays low Jul 10 16:46:01 doesn't make much sense to keep the whole GUI data structures and program text in RAM, while actually only the applet is needed (if at all) Jul 10 16:46:36 MohammadAG: using a .pid file would make sense to conserve resources, and stopping someone from killing it from the command line Jul 10 16:46:42 someone who kills it from the command line can rm the pid file Jul 10 16:46:46 (could be a user config option "show statusbar applet when hostmode active") Jul 10 16:47:11 MohammadAG: OOM manager could kill it Jul 10 16:47:21 MohammadAG: it could segfault Jul 10 16:47:26 whatnot else Jul 10 16:48:02 OOM? Jul 10 16:48:06 MohammadAG: also see my rationale about resource footprint some few lines up Jul 10 16:48:15 Out Of Memeory Jul 10 16:48:17 well yeah, agreed about that Jul 10 16:48:19 yay, my n950 cleared customs Jul 10 16:48:26 1GB + OOM?" Jul 10 16:48:30 ieatlint, FU!!! Jul 10 16:48:35 :P Jul 10 16:48:45 on that thought Jul 10 16:48:53 is there a way to dump an app into a file and reload it later? Jul 10 16:49:57 the short answer is no Jul 10 16:50:21 MohammadAG: that's called swap ;-) Jul 10 16:50:23 but a very long answer with some hypotheticals can say yes Jul 10 16:51:11 and generally you don't want to do that Jul 10 16:51:26 MohammadAG: so to get an n950, just go to a meego meetup Jul 10 16:51:34 ieatlint: FU FU FU! Jul 10 16:51:48 ieatlint: we are waiting for ours to get shipped :P Jul 10 16:51:53 look for the scrawniest guy with one (but not me), and grab and run Jul 10 16:51:54 NB aiui program text never gets swapped out as you can reload it from the original binary executable file any time Jul 10 16:52:19 ieatlint: I am gonna grab yours and run Jul 10 16:52:42 i will hunt you down and ask politely that you return it Jul 10 16:54:00 and I'll shoot you with my shotgun, I hate polite people Jul 10 16:54:28 bah, i'm an american.. just imagine how many guns i own Jul 10 16:55:21 none? Jul 10 16:56:10 it'd be better for me if you applied some sort of stereotype than if i answered that question Jul 10 16:56:49 you are a geek, you have computers, not guns? :P Jul 10 16:57:20 phones too, apparently Jul 10 16:57:59 phones have guns? Jul 10 16:58:07 what a wacky world. Jul 10 16:58:49 Device convergence. Jul 10 16:59:09 my phone-gun is awesome Jul 10 16:59:36 yeah, phone has gun, fake ones Jul 10 16:59:37 The lack of a competitive phone-gun is what has killed Nokia in the US. Jul 10 16:59:48 ican shoot someone while videoing it for youtube, and then tweet about it Jul 10 16:59:58 friggin asian HTC guns. Jul 10 17:01:00 * javispedro suggest you to parse the previous sentence in the most P.C. way possible. Jul 10 17:01:42 ok, i clearly need to file a patent for a gun/phone combo Jul 10 17:02:40 Ken-Young: :-D Jul 10 17:03:56 errrr Jul 10 17:04:37 * DocScrutinizer goes editing the http://wiki.meego.com/Community_Office/Community_device_program/Nokia#Completed page Jul 10 17:04:37 i also need to buy a nokla phone sometime Jul 10 17:06:03 Limb Saver needs to make an attatchment that puts a phone at the cheek weld. Jul 10 17:06:10 ...no, not a gun app for N950, but close to it ;-D Jul 10 17:06:44