| *** willer_ has quit IRC | 00:07 | |
| *** Scifig has quit IRC | 00:10 | |
| rcg | mgedmin: well, this /opt/Foo/bin scheme is the default for new Harmattan projects in the new QtSDK | 00:11 |
|---|---|---|
| rcg | it also does not place a symlink in /usr/bin because the Foo.desktop file refers directly to the executable in that path | 00:12 |
| qgil | so... since I really wanted to share source files & package with my Miniature team mates I ended up wit... a bug report containing all the stuff ;) https://bugs.maemo.org/show_bug.cgi?id=12315 | 00:13 |
| povbot | Bug 12315: Placeholder to upload UI source and packages | 00:13 |
| MeeGoBot | Bug https://bugs.meego.com/show_bug.cgi?id=12315 nor, Medium, ---, alexey.kuznetsov, RELE FIXED, libmeegochat should be ported to telepathy-qt4 0.5.x | 00:13 |
| rcg | This whole scheme is configured via the deployment.pri file in your project | 00:13 |
| rcg | the "Meego graphics system destroyed" message seems to be normal afaik.. at least all apps i started via a shell here did this as well | 00:15 |
| qgil | piggz: can you access the Nokia Store? there was a screenshot app there, see e.g. http://forum.meego.com/showthread.php?p=27536#post27536 | 00:15 |
| rcg | actually you take screenshot with some preinstalled app | 00:15 |
| rcg | just a second... | 00:15 |
| rcg | *can take | 00:16 |
| rm_work | screenshot with one of the gallery apps | 00:16 |
| rcg | Widgets Gallery -> Debug Tools ->Take a Screenshot | 00:16 |
| rm_work | yeah that | 00:16 |
| qgil | ah, I believe there was a shortcut for screenshots with the N950 - but of course the N9 needed a different solution :) | 00:16 |
| *** Venemo has joined #harmattan | 00:17 | |
| piggz | qgil: which store? (ive been using the debug tools method, and added another entry in the meego coding comp) | 00:17 |
| qgil | ah, maybe there is no Store in the developer image for the N950 piggz | 00:18 |
| piggz | qgil: yeah, it says 'coming soon' | 00:18 |
| piggz | i heard on twitter a fw update is coming soon? | 00:19 |
| qgil | piggz: URL? | 00:19 |
| rcg | btw just out of curiousity.. is there any way to close an app directly.. instead of the detour via the "task switcher" thing | 00:19 |
| piggz | qgil: url for the twitter talk? i think it was just speculation | 00:20 |
| qgil | I thought speculation also had URLs, this is why I wanted to see the source :) | 00:20 |
| kimju | someone got accidentally a device with an *older* firmware and some people were jumping to wrong conclusions from that | 00:20 |
| Venemo | hmm | 00:21 |
| kimju | it was hot topic #here yesterday | 00:21 |
| rm_work | <rcg> the "Meego graphics system destroyed" message seems to be normal afaik.. at least all apps i started via a shell here did this as well | 00:21 |
| piggz | qgil: http://twitter.com/#!/search/%23n950%20firmware | 00:21 |
| rm_work | yes, that happens every time you "minimize" an app | 00:21 |
| Venemo | Qt Creator doesn't like me still.... | 00:21 |
| Venemo | ":-1: error: Failed to upload file 'C:\Projects\IrcChatter\IrcChatter'." | 00:21 |
| Venemo | what the heck is this? | 00:21 |
| rm_work | rcg: yes there is a way to close app directly... can just do an exit command (whatever the one QT uses is) | 00:22 |
| rm_work | there's a couple apps I have that do it | 00:22 |
| rm_work | that actually have a Quit button | 00:22 |
| *** CaCO3 has joined #harmattan | 00:23 | |
| rm_work | like cuteTube-qml does | 00:23 |
| mgedmin | some apps exit when you press ctrl-q | 00:23 |
| rcg | rm_work: ic.. was just wondering if there was some "magic swipe" that does the same as a quit button ;) | 00:23 |
| mgedmin | like meego-terminal, boo! | 00:23 |
| rm_work | ah no | 00:23 |
| rm_work | not that i'm aware of :/ | 00:23 |
| rm_work | i would like that too :P | 00:23 |
| qgil | piggz Nokia_Fan and colmsmyth are bloggers - perhaps they know but, and they they would knew better than me. I'd rather go back to coding ;) | 00:23 |
| mgedmin | rcg, I heard there's a setting that makes swipe down close the app instead of hiding it | 00:23 |
| rm_work | cuteSoma does it easily enough | 00:23 |
| mgedmin | rcg, but I think it's not present on the firmware we have | 00:23 |
| rm_work | hrm yes i think i would like THAT | 00:23 |
| rm_work | swipe down = kill, not hide | 00:23 |
| piggz | qgil: sure :) | 00:24 |
| rm_work | swipe side = task switcher, like normal | 00:24 |
| rcg | hmm.. yeah that would be quite nice :) | 00:24 |
| qgil | the magic swipe close the apps, yes | 00:24 |
| qgil | I believe it is implemented at framework level = app developers don't need to care about it | 00:25 |
| qgil | For instance, my dummy Miniature UI app gets closed by swiping down | 00:25 |
| rm_work | yeah | 00:25 |
| rm_work | that should be correct | 00:26 |
| rm_work | there IS a "close app" it is just only sent when you go to task manager, hold-click, and hit the X | 00:26 |
| rm_work | so just switch the swipe-down to send that signal instead of minimize? | 00:26 |
| qgil | swipe down = closep app too | 00:26 |
| rm_work | on our devices, swipe-down just minimizes it | 00:27 |
| rm_work | it isstill running in task switcher | 00:27 |
| qgil | in the current Harmattan release swipe down closes the apps | 00:27 |
| qgil | current = latest | 00:27 |
| rm_work | the one we don't have but you do? :P | 00:27 |
| qgil | that one | 00:27 |
| rm_work | lol | 00:27 |
| rm_work | <3 qgil | 00:27 |
| qgil | there is even a youtube video somewhere | 00:27 |
| rm_work | so i assume we get that with the next firmware? | 00:28 |
| piggz | later ppl...tired from moutnain biking :) | 00:28 |
| qgil | of course | 00:28 |
| rm_work | later piggz | 00:28 |
| Venemo | ok | 00:28 |
| *** piggz has quit IRC | 00:28 | |
| Venemo | what do I need to do to allow Aegis to run arbitrary binaries | 00:28 |
| *** crevetor has quit IRC | 00:29 | |
| rm_work | devel-su / develsh should let you | 00:29 |
| Venemo | rm_work, that is not applicable to my situation | 00:29 |
| kimju | package them | 00:29 |
| Venemo | rm_work, Qt Creator connects to the device with the 'developer' account and tries to run it | 00:29 |
| rm_work | your situation of.... needing to run arbitrary binaries? | 00:29 |
| rm_work | err | 00:29 |
| rm_work | ... | 00:29 |
| rm_work | ok | 00:30 |
| Venemo | I want to avoid packaging my app every time I want to run it! | 00:30 |
| rm_work | lol | 00:30 |
| rm_work | then edit on-device | 00:30 |
| mgedmin | aegis is not your friend | 00:30 |
| rm_work | and run your app | 00:30 |
| Venemo | there is some aegis package I need to install. | 00:30 |
| Venemo | but I forgot what it is | 00:30 |
| rm_work | i edit my python apps on the device in vim | 00:30 |
| mgedmin | --relaxed-exec or something | 00:30 |
| rm_work | and run them... | 00:30 |
| rm_work | from commandline | 00:30 |
| rm_work | relexed-exec not necessary | 00:30 |
| rm_work | same with mplayer testing... | 00:30 |
| mgedmin | why does fbreader ignore ~/.FBReader/toolbar.xml? strace shows me it's being opened, but all my edits are ignore | 00:31 |
| mgedmin | if I edit /usr/share/FBReader/default/toolbar.xml instead, my changes are visible | 00:32 |
| *** qgil has quit IRC | 00:32 | |
| Venemo | https://bugreports.qt.nokia.com//browse/QTCREATORBUG-5459 | 00:33 |
| Venemo | There is a package aegis-dev-mode that enables you to run executables that were not installed via a package - otherwise even that is not possible | 00:33 |
| Venemo | OK, I did 'apt-get install aegis-dev-mode' | 00:33 |
| Venemo | still doesn't | 00:33 |
| kimju | I don't think that packaging and installing the app after changes is any more work than copying it over (if you have the packaging stuff already made) | 00:34 |
| mgedmin | Venemo, there's a recipe somewhere on the meego forum | 00:34 |
| Venemo | kimju, but it's faster! a lot! | 00:34 |
| kimju | make or dpkg-buildpackage or something does the work.. | 00:34 |
| mgedmin | you need to devel-su, then develsh, then run aegis-something --something-something, then reboot, and then your device gets bricked | 00:34 |
| mgedmin | or something like that | 00:34 |
| Venemo | hmm | 00:34 |
| mgedmin | http://forum.meego.com/showthread.php?t=3827&page=2 | 00:35 |
| Venemo | mgedmin, mhmm | 00:35 |
| Venemo | thanks, I'll try | 00:36 |
| mgedmin | ok, giving up on the segfault | 00:39 |
| mgedmin | how about I make fbreader intercept volume keys? | 00:39 |
| lardman | night all | 00:39 |
| *** lardman has quit IRC | 00:39 | |
| MohammadAG | mgedmin, there's a binary that interprets keypresses, x something, can't remember atm | 00:40 |
| MohammadAG | you could also string the volume daemon | 00:40 |
| Venemo | aaah, thank you mgedmin | 00:41 |
| Venemo | mgedmin, it works like a charm!!! :) | 00:41 |
| Venemo | MohammadAG, have you got your device yet? | 00:41 |
| MohammadAG | http://forum.meego.com/showthread.php?p=27465#post27465 | 00:42 |
| *** CaCO3 has quit IRC | 00:42 | |
| Venemo | MohammadAG, humm | 00:42 |
| mgedmin | http://www.developer.nokia.com/Community/Discussion/showthread.php?226283-Grab-Volume-Keys-on-Harmattan fails to answer the question completely | 00:42 |
| Venemo | MohammadAG, look at the bright side. once you get it, it'll already have an IRC client | 00:42 |
| mgedmin | which of the standard apps handle volume keys themselves? | 00:43 |
| mgedmin | Camera | 00:44 |
| MohammadAG | string that for an atom | 00:44 |
| MohammadAG | Venemo, I'm looking on the brighter side, by the time I have it, I can trade it for an N9 | 00:44 |
| * mgedmin runs xprop | 00:44 | |
| Venemo | MohammadAG, hehe | 00:44 |
| Venemo | MohammadAG, do you think N9 will be better? | 00:45 |
| * mgedmin runs strings | 00:45 | |
| mgedmin | http://library.developer.nokia.com/index.jsp?topic=/MeeGo_1.2_Harmattan_API/html/qmsystem2/main.html has an example of listening to those key events | 00:51 |
| mgedmin | camera-ui does not appear to be using QmKeys | 00:55 |
| mgedmin | hmm..._ZN15QGraphicsWidget17grabKeyboardEventEP6QEvent (I hate C++) | 00:55 |
| mgedmin | probably irrelevant | 00:56 |
| * mgedmin runs strings on /usr/sbin/qmkeyd2 | 00:58 | |
| Venemo | ookay, does anyone have 5 minutes to test something for me? | 00:58 |
| mgedmin | oh, cool, qmkeyd.cpp is open source | 01:04 |
| mgedmin | I wonder if it differs much from qmkeyd2, which I was unable to find | 01:04 |
| GeneralAntilles | rm_work, ping? | 01:04 |
| mgedmin | I don't think QmKeys is what I want... | 01:06 |
| mgedmin | this is for background apps to be able to react to volume and other system keys | 01:06 |
| mgedmin | so, fbreader sees vol+/- as regular key events and I can map them to scroll forward/backward | 01:08 |
| mgedmin | the only thing is how to prevent the default system action | 01:09 |
| mgedmin | how do I figure out which processes are connected to a unix domain socket? | 01:12 |
| mgedmin | lsof does not help | 01:14 |
| mgedmin | netstat doesn't show pids (-p option not supported) | 01:15 |
| *** infobot has quit IRC | 01:20 | |
| mgedmin | sysuid is one of the four listeners | 01:22 |
| mgedmin | it's the only listener, actually, but it listens multiple times | 01:26 |
| mgedmin | what fun | 01:26 |
| mgedmin | strace -e read $(lsof /usr/lib/libqmsystem2.so.1|awk 'NR>1 {print "-p", $2}') | 01:26 |
| mgedmin | press volume key, see which pids do a read() | 01:26 |
| mgedmin | and receive a 16-byte event | 01:26 |
| *** hardaker has quit IRC | 01:28 | |
| mgedmin | could this be another open-source component? | 01:29 |
| * mgedmin clones meegotouch-systemui | 01:30 | |
| rm_work | GeneralAntilles: pong | 01:32 |
| rm_work | just about to leave tho | 01:32 |
| rm_work | whats up? | 01:32 |
| *** Scifig has joined #harmattan | 01:33 | |
| *** baraujo has quit IRC | 01:33 | |
| mgedmin | so! apparently libresourceqt arbitrates which app controls the volume keys | 01:34 |
| mgedmin | http://harmattan-dev.nokia.com/unstable/beta/api_refs/showdoc.php?pkn=libresourceqt&wb=daily-docs&url=Li94bWwvZGFpbHktZG9jcy9saWJyZXNvdXJjZXF0 | 01:35 |
| *** djszapi_ has joined #harmattan | 01:36 | |
| mgedmin | I need to ask for ScaleButtonResource | 01:36 |
| *** vandenoever has quit IRC | 01:39 | |
| GeneralAntilles | rm_work, later, then. | 01:39 |
| GeneralAntilles | rm_work, was gonna powwow about SnapGo. | 01:39 |
| Venemo | hm, does anyone here have any experience with the QML TextArea item? | 01:43 |
| *** rm_work has quit IRC | 01:47 | |
| *** Arkenoi has quit IRC | 01:53 | |
| *** seif has quit IRC | 01:53 | |
| *** Arkenoi has joined #harmattan | 01:54 | |
| *** rcg has quit IRC | 01:57 | |
| DocScrutinizer | mgedmin: (<mgedmin> the only thing is how to prevent the default system action) in fremantle e.g for xchat it's simply like "unset the F7/8(?) key assignments in xchat's key editor, and they will work as volume keys again" | 02:02 |
| DocScrutinizer | and vice versa | 02:02 |
| DocScrutinizer | not entirely, but basically it's like that | 02:03 |
| *** seif has joined #harmattan | 02:06 | |
| *** wicket64 has quit IRC | 02:13 | |
| MohammadAG | DocScrutinizer, there's some hildon_ call xchat's doing then | 02:13 |
| MohammadAG | you need to set an X11 atom on the window | 02:13 |
| MohammadAG | HILDON_ZOOM_KEYS I think | 02:14 |
| DocScrutinizer | yeah, it's for sure an X11 thing. If I had to guess, I'd say (or design it like) there's a chain of event handlers, a call to X11 to un/register a handler, and on every event X11 is passing the event to each handler in the chain in sequence, until a handler returns "eaten it" | 02:17 |
| *** deimos has quit IRC | 02:27 | |
| *** Scifig has quit IRC | 02:27 | |
| *** Arkenoi has quit IRC | 02:29 | |
| *** meegoexperts_mob has joined #harmattan | 02:38 | |
| *** epage has joined #harmattan | 02:45 | |
| *** meegoexperts_mob has quit IRC | 02:45 | |
| *** infobot has joined #harmattan | 02:51 | |
| *** ChanServ sets mode: +v infobot | 02:51 | |
| infobot | Okay, I'm here. (courtesy of docscrutinizer) | 02:51 |
| Termana | morning | 02:57 |
| Venemo | morning Termana | 02:58 |
| *** jcharpak has joined #harmattan | 03:04 | |
| *** sebas has quit IRC | 03:04 | |
| *** sebas_ has joined #harmattan | 03:04 | |
| npm | http://nielsmayer.com/meego/qml/yberbrowser_0_0_1_armel.deb -- built a new browser | 03:26 |
| SpeedEvil | y? | 03:26 |
| npm | y not | 03:27 |
| npm | to replace the simplistic browser in http://nielsmayer.com/meego/qml/qtzibit_0_0_2_armel.deb | 03:30 |
| *** Venemo has quit IRC | 03:44 | |
| *** _NIN has quit IRC | 03:45 | |
| *** Arkenoi has joined #harmattan | 03:45 | |
| *** Rizzer has joined #harmattan | 03:47 | |
| *** epage has quit IRC | 03:52 | |
| *** meegoexperts has joined #harmattan | 03:53 | |
| *** antman8969 has joined #harmattan | 04:29 | |
| *** antman8969 has quit IRC | 04:33 | |
| *** M4rtinK2 has quit IRC | 04:44 | |
| *** leinir has quit IRC | 04:53 | |
| *** antman8969 has joined #harmattan | 05:11 | |
| *** antman8969 has quit IRC | 05:17 | |
| *** epage has joined #harmattan | 05:33 | |
| *** jcharpak has left #harmattan | 06:01 | |
| *** hardaker has joined #harmattan | 06:02 | |
| *** mzanetti has quit IRC | 06:05 | |
| *** macmaN has quit IRC | 06:07 | |
| *** macmaN has joined #harmattan | 06:14 | |
| *** bugzy has quit IRC | 06:25 | |
| *** mzanetti has joined #harmattan | 06:25 | |
| *** epage has quit IRC | 06:27 | |
| *** seif has quit IRC | 07:22 | |
| *** DocScrutinizer has quit IRC | 07:31 | |
| *** DocScrutinizer has joined #harmattan | 07:31 | |
| *** seif has joined #harmattan | 07:31 | |
| *** antman8969 has joined #harmattan | 07:31 | |
| *** twoboxen has joined #harmattan | 07:58 | |
| twoboxen | question for you folks… in the qt project file, what is maemo6 harmattans make spec? I see maemo5, symbian, unix, etc | 07:58 |
| twoboxen | [ | 07:58 |
| twoboxen | 04:58AM | 07:58 |
| twoboxen | ] twoboxen: | 07:58 |
| twoboxen | meego, maemo6, harmattan, etc don't work | 07:58 |
| twoboxen | this line is generated "unix:!symbian:!maemo5"… i don't care about what it ISN'T | 07:59 |
| *** twoboxen has quit IRC | 08:11 | |
| *** twoboxen has joined #harmattan | 08:12 | |
| *** antman89691 has joined #harmattan | 08:12 | |
| *** antman8969 has quit IRC | 08:15 | |
| *** seif has quit IRC | 08:30 | |
| *** hardaker has quit IRC | 08:34 | |
| hiemanshu | twoboxen: well for now, maemo6 is defined like that | 08:51 |
| *** seif has joined #harmattan | 09:01 | |
| w00t | mgedmin: "MeeGo graphics system" is a bit special | 09:05 |
| w00t | mgedmin: it uses EGL when the application is in the foreground | 09:05 |
| w00t | mgedmin: then it switches to raster (and drops the GL context) when you swipe an application to the background | 09:06 |
| w00t | mgedmin: (this is for resource consumption, holding a GL context on SGX drivers means you lock a large chunk of RAM in memory, not swappable, and not reusable - not nice) | 09:06 |
| w00t | mgedmin: but, QPixmaps should stay valid between the transition iirc | 09:07 |
| *** djszapi_ has left #harmattan | 09:09 | |
| *** Venemo has joined #harmattan | 09:30 | |
| *** Venemo has joined #harmattan | 09:30 | |
| *** seif has quit IRC | 09:31 | |
| *** spenap has joined #harmattan | 09:47 | |
| *** seif has joined #harmattan | 09:53 | |
| *** harbaum has joined #harmattan | 09:55 | |
| *** aleksander_m has joined #harmattan | 10:16 | |
| *** antman89691 has quit IRC | 10:21 | |
| macmaN | mkay some twitter noise going on last night about new fw coming for n950 | 10:39 |
| macmaN | in a few days? | 10:39 |
| macmaN | https://twitter.com/#!/colmsmyth/status/96131700316311552 | 10:39 |
| Venemo | macmaN, "for some"? who? | 10:50 |
| macmaN | frck if i know | 10:50 |
| macmaN | "them" : | 10:50 |
| macmaN | :) | 10:50 |
| Venemo | :P | 10:51 |
| Venemo | hm. | 10:51 |
| Venemo | how would you go about channel switching in a mobile IRC client? | 10:51 |
| Venemo | I was thinking about adding tabs, but that is definiely not manageable if you join more than 4 (or max 5) channels | 10:51 |
| Summeli | I got a snes-emu running at some level: http://www.summeli.fi/?p=2682 | 10:51 |
| RST38h | Venemo: current config of XChat does it well enough for me | 10:52 |
| RST38h | Venemo: tiny little tabs at the bottom of the screen that you can quickly flick through with your finger | 10:53 |
| Venemo | RST38h, xchat (as I saw it on the N900) is not very finger friendly, and this wouldn't work well in portrait mode either | 10:53 |
| Venemo | I'm thinking of simply adding a button which displays a selection dialog in which you can switch channels | 10:54 |
| RST38h | Venemo: for portrait, arrange them vertically along an edge | 10:54 |
| Venemo | my other idea is to add some unique finger gesture for this | 10:54 |
| RST38h | Venemo: and no, I guess these channel-flicking tabs are the most finger friendly feature of the current XChat, otherwise it kinda sucks :) | 10:54 |
| Venemo | if I align them vertically in portrait mode, then there won't be enough comfortable room for the messages | 10:54 |
| RST38h | Forget gestures, they suck moose balls | 10:55 |
| Venemo | not a complicated one | 10:55 |
| RST38h | Venemo: which means I will be able to trigger it accidentally. thanks. | 10:55 |
| Venemo | lol. | 10:55 |
| spenap | Venemo, what about something like the fb top bar | 10:55 |
| Venemo | spenap, 'fb'? | 10:55 |
| spenap | and then some "smart" way to hide/show the least used | 10:55 |
| spenap | 10:56 | |
| RST38h | even better | 10:56 |
| RST38h | the freaking bar will jump out whenever it thinks fit! | 10:56 |
| spenap | xD | 10:56 |
| Venemo | and how should it determine whether it's fit or not? | 10:56 |
| spenap | have you used gtalk in android? | 10:56 |
| RST38h | on its own of course | 10:57 |
| spenap | you could "little-swipe" to change between open channels | 10:57 |
| spenap | (this little swipe I mean is the one used in the music app) | 10:57 |
| spenap | to change between songs | 10:57 |
| *** vandenoever has joined #harmattan | 10:57 | |
| Venemo | my current idea is to "listen to" horizontal swipes on the toolbar or something, and switch between channels according to that. | 10:58 |
| spenap | on the toolbar? | 10:58 |
| RST38h | toolbar? what toolbar? | 10:58 |
| spenap | I don't think (my opinion, though) that the toolbar should be used that way | 10:58 |
| Venemo | maybe not the toolbar then | 10:58 |
| Venemo | maybe the text area | 10:58 |
| Venemo | but then this will interefere with text selection | 10:58 |
| Venemo | RST38h, the one that appears on the bottom of every Harmattan app | 10:58 |
| spenap | Venemo, if you haven't, check the gtalk application in android | 10:59 |
| spenap | because that behavior feels quite natural | 10:59 |
| Venemo | spenap, I don't have neither an android device, nor a gtalk account | 10:59 |
| Venemo | spenap, could you please gimme a video of it or just describe it in detail? | 10:59 |
| spenap | let's go with the description, then | 11:00 |
| spenap | there's this big widget where the conversation is held, and the messages you send and receive appear | 11:00 |
| spenap | this TextArea | 11:00 |
| Venemo | yeah | 11:00 |
| spenap | then you have the text field | 11:00 |
| spenap | to write | 11:00 |
| Venemo | yeah | 11:01 |
| spenap | when you "drag" the text area to the sides | 11:01 |
| Tronic | Venemo: Not very standard but vertical channel list on the left side would be best for me. | 11:01 |
| spenap | it will go to the previous / next conversation | 11:01 |
| Venemo | Tronic, that wouldn't be finger friendly | 11:01 |
| spenap | and there are menu options which would show the currently open conversations | 11:01 |
| Tronic | Venemo: How come, too close to the edge? | 11:01 |
| spenap | that appears in something quite similar to Harmattan sheet's | 11:02 |
| Venemo | Tronic, that too, plus it takes away too much space from the messages | 11:02 |
| spenap | so you can either discard or click on one of this open conversations | 11:02 |
| Venemo | spenap, yeah, I already have a button to show all the current channels | 11:02 |
| Venemo | spenap, but that requires two taps to switch channels, which might not be comfortable to everyone | 11:02 |
| spenap | well: two taps is for selecting one channel "far from the current" | 11:03 |
| spenap | but if it is the previous / next, you can just swipe | 11:03 |
| spenap | (you could add some reordering in this "extended view") | 11:03 |
| Tronic | Venemo: The list could only appear when you put your finger on it, and the browsing could be a scroll gesture. | 11:03 |
| Tronic | Venemo: Solves both problems. | 11:03 |
| Venemo | mhm | 11:03 |
| Venemo | well yeah, but this one still requires two taps | 11:04 |
| Tronic | I was actually thinking just one scroll. | 11:04 |
| Tronic | Press => it appears, release => it switches to whatever is chosen. | 11:04 |
| Venemo | aaah | 11:04 |
| Venemo | hmm. | 11:04 |
| Tronic | Of course it could be horizontal too, since screen space is not actually limited if you have to scroll in any case. | 11:05 |
| Venemo | here is how it looks now: http://i52.tinypic.com/n4ur2d.png | 11:06 |
| *** sebas_ is now known as sebas | 11:06 | |
| *** sebas has joined #harmattan | 11:06 | |
| Venemo | also keep in mind that it needs to look good in portrait mode as well. | 11:06 |
| Tronic | But you'd also want to know window activity, to quickly choose windows where someone has spoken to you (high activity level) or where there has been discussion at all (normal activity). | 11:06 |
| *** lardman|gone has quit IRC | 11:09 | |
| w00t | Venemo: you might want to talk to sput (on #quassel).. he's working on a harmattan version of quassel, i think someone mentioned | 11:09 |
| w00t | so maybe you can steal ideas off each other | 11:09 |
| Venemo | sput? | 11:10 |
| *** Wirta has joined #harmattan | 11:10 | |
| w00t | a person ;) | 11:10 |
| Wirta | O_o | 11:11 |
| Venemo | someone told me that there is such a version of Quassel, made by someone called sgiessl | 11:11 |
| Venemo | but he never responded to my pings | 11:11 |
| Venemo | in any case, if someone beats me to it, I'll gladly forefit this project. I just started it because there is no IRC client yet for Harmattan | 11:13 |
| vandenoever | should a harmattan app have 80x80 or 64x64 icons or both? | 11:13 |
| frals | afaik 80x80 for applauncher and 64x64 for pkgmanager | 11:14 |
| Venemo | mhm | 11:14 |
| harbaum | i am using svg icons and it seems harmattan just renders them nicely | 11:14 |
| harbaum | so no need to think about size | 11:14 |
| vandenoever | harbaum: how do you put that in the control file? | 11:14 |
| vandenoever | i mean the XB-Maemo-Icon-26 entry | 11:15 |
| Venemo | vandenoever, base64 encode it | 11:15 |
| Venemo | this Sput guy doesn't respond to my pings either | 11:16 |
| Venemo | seems that we will have 3 IRC clients then :P | 11:16 |
| Venemo | I'd gladly drop mine if I was assured that the others will be user-friendly enough for my needs | 11:18 |
| *** djszapi has joined #harmattan | 11:20 | |
| *** smoku has joined #harmattan | 11:20 | |
| *** achipa has quit IRC | 11:21 | |
| *** lardman has joined #harmattan | 11:22 | |
| infobot | lardman: please check http://mg.pov.lt/harmattan-irclog/%23harmattan.2011-07-22.log.html#t2011-07-22T18:42:19, cheers DocScr | 11:22 |
| *** slaine has joined #harmattan | 11:22 | |
| lardman | morning | 11:22 |
| lardman | infobot: stop reminding me please | 11:23 |
| * infobot leaps to his feet and stops reminding me please | 11:23 | |
| *** seif has quit IRC | 11:26 | |
| djszapi | haha xD | 11:28 |
| frals | http://twitter.com/#!/dwineman/status/96357433756229632 | 11:28 |
| frals | rofl | 11:28 |
| *** deimos has joined #harmattan | 11:29 | |
| djszapi | Elleo: mmh, liblzma-dev seems to be a bit duplication, the fix seems to be on the way for a while by Nokia. | 11:30 |
| djszapi | and you cannot overwrite their packages without force-all. | 11:30 |
| *** Wirta has quit IRC | 11:30 | |
| *** leinir has joined #harmattan | 11:31 | |
| vandenoever | i'm putting two .desktop files in one project and now the build system does not pick the up | 11:33 |
| vandenoever | at least, i think that's the problem, is it allowed to have two .desktop files for app icons in one package? | 11:33 |
| djszapi | Elleo: I raised the issue at Nokia about the missing packages since it is not nice, and they told me they fix it asap. If it is not really an urgent story, I do not think it is worth doing them. | 11:34 |
| *** smoku has left #harmattan | 11:34 | |
| *** Wirta has joined #harmattan | 11:36 | |
| khertan1 | Morning, | 11:40 |
| djszapi | Funky, OBS lost the files I uploaded yesterday.. | 11:44 |
| *** smoku has joined #harmattan | 11:45 | |
| *** achipa has joined #harmattan | 11:48 | |
| khertan1 | is there a way to fix ssl problem on current developer n950 device ? | 11:48 |
| djszapi | what ssl problem ? | 11:49 |
| djszapi | lbt: I cannot upload files onto OBS via the webinterface. It ends up an empty page after pressing the "Save changes" button. | 11:50 |
| lbt | djszapi: hmm | 11:51 |
| djszapi | it has been like that for 1-2 days actually. | 11:51 |
| lbt | we had some disk space issues... http://stackoverflow.com/questions/3456159/how-to-shrink-purge-ibdata1-file-in-mysql FWIW | 11:52 |
| djszapi | lbt: well, please let me know if it works agian. | 11:53 |
| lbt | djszapi: however, if you've lost stuff I'd like to know about it | 11:53 |
| lbt | AFACT we should be OK | 11:53 |
| djszapi | well, from what I can say, it does not work. | 11:54 |
| khertan1 | djszapi: try to login to the meego forum : ssl handshake failed ... many webmail : ssl handshake failed | 11:59 |
| lbt | khertan1: sorry... are you having ssl problems | 12:01 |
| lbt | djszapi: could you log a bug for me please ... details | 12:01 |
| djszapi | lbt: sure, a bit later. | 12:02 |
| djszapi | khertan1: I am not getting that type of error, but I am getting this: "Couldn't open page" https://meego.com/user/login?destination=forum | 12:03 |
| *** cpscotti has joined #harmattan | 12:04 | |
| *** Arkenoi has quit IRC | 12:05 | |
| *** Arkenoi has joined #harmattan | 12:05 | |
| *** Rizzer has quit IRC | 12:15 | |
| *** rcg1 has joined #harmattan | 12:18 | |
| djszapi | lbt: are you planning to extend the hardware capability of the Harmattan target PC or if not, any collaboration with the maemo.org people who had a hefty machine ? | 12:24 |
| lbt | djszapi: we are .. the new hardware has arrived | 12:25 |
| djszapi | which is ... ? | 12:25 |
| lbt | we actually have 5 machines with 24 cores and 24Gb RAM | 12:25 |
| lbt | and more arrived | 12:25 |
| Stskeeps | nice | 12:25 |
| lbt | sadly they are crap | 12:25 |
| djszapi | well, I have been told it had 6 GB | 12:25 |
| lbt | and the uptime is <6 hrs | 12:25 |
| Arkenoi | Does anybody maintain a list of fremantle apps that run on harmattan without rebuild? | 12:25 |
| Stskeeps | you can send them to me if you don't like them | 12:25 |
| Stskeeps | :P | 12:25 |
| lbt | then the RAID controller (proprietary blob) craps out | 12:25 |
| lbt | the 6GB is per worker VM | 12:26 |
| djszapi | Arkenoi: why not just put it onto c-obs/harmattan target ? | 12:26 |
| lbt | 24/4 | 12:26 |
| djszapi | Stskeeps: haha, me too :p | 12:26 |
| lbt | Stskeeps: yeah ... I'm pissed at this ... why are the core workers just fine | 12:26 |
| djszapi | lbt: ok, but it cannot be a memory issue with kdelibs then... | 12:27 |
| djszapi | if it is not that, seriously no clue.. | 12:27 |
| DocScrutinizer51 | moo | 12:38 |
| Venemo | DocScrutinizer51, moo | 12:40 |
| *** steventroughtons has joined #harmattan | 12:41 | |
| *** steventroughtons is now known as stroughtonsmith | 12:42 | |
| khertan1 | djszapi: i got certificat error on most all things where ssl is used | 12:43 |
| DocScrutinizer51 | djszapi: how would I find out if a file is under aegis aegis, I.E. protected? Usecase: I want to edit a pyrhon script I found on device, and I don't want to risk the "dear customer, aegis nuked your system. Reflash!". So what is the command to 'ls' or 'stat' a file to find out if it may be edited or not? | 12:46 |
| djszapi | DocScrutinizer51: Could you give the exact path of an example ? :) | 12:51 |
| DocScrutinizer51 | not yet | 12:51 |
| *** seif has joined #harmattan | 12:52 | |
| *** smoku has quit IRC | 12:53 | |
| khertan1 | djszapi: http://mg.pov.lt/meego-irclog/%23meego.2011-07-13.log.html#t2011-07-13T18:15:05 | 12:53 |
| DocScrutinizer51 | it's been explained on some blog that you can install python with a manifest to request the needed permissions, but then you MUST NOT edit this file after installation | 12:53 |
| djszapi | ls -lca and stat work here. | 12:54 |
| *** smoku has joined #harmattan | 12:54 | |
| djszapi | khertan1: well, I was trying with N9. Let me try it out with N950 | 12:55 |
| DocScrutinizer51 | djszapi: thanks, will try as soon as I'm back home | 12:55 |
| *** razvanpetru has joined #harmattan | 12:56 | |
| razvanpetru | hi all, what is the easiest way to zoom a WebView? | 12:56 |
| razvanpetru | the PinchArea just swallows all mouse clicks to the input fields... :( | 12:56 |
| DocScrutinizer51 | djszapi: could you tell what a protected file looks like on e.g. ls -lca ? | 12:57 |
| SpeedEvil | won't all system binaries be protected? | 12:57 |
| SpeedEvil | /usr/sbin/aegis - for example | 12:57 |
| DocScrutinizer51 | should, yeah. | 12:57 |
| SpeedEvil | Or /etc/rc.d/whatever | 12:57 |
| DocScrutinizer51 | k, thanks | 12:57 |
| djszapi | SpeedEvil: except that, python scripts can be out of sbin | 12:59 |
| SpeedEvil | Indeed. | 12:59 |
| DocScrutinizer51 | yup, and sbin might have a dir level protection | 12:59 |
| khertan1 | lol now i got strange error on some account connection | 13:00 |
| SpeedEvil | Can you not tell python 'include this script' - without telling aegis about it? | 13:00 |
| khertan1 | please verify hour and date parameter | 13:00 |
| SpeedEvil | If it's not executable - it'll just look like a file read | 13:00 |
| khertan1 | python is protected | 13:00 |
| DocScrutinizer51 | SpeedEvil: obviously you can | 13:00 |
| khertan1 | python script aren't | 13:00 |
| khertan1 | i code on device | 13:00 |
| spenap | razvanpetru, there's a quite-complete example of a WebView inside a Flickable content (if I recall well) | 13:00 |
| spenap | I'll try to find it | 13:00 |
| khertan1 | execute my script | 13:00 |
| khertan1 | without taking care of aegis ;) | 13:01 |
| spenap | razvanpetru, http://doc.qt.nokia.com/4.7-snapshot/demos-declarative-webbrowser.html | 13:01 |
| DocScrutinizer51 | khertan1: this changes as soon as you install a python script with a minifest | 13:01 |
| khertan1 | djszapi: it s probably a bug of image we got on the 22 week firmware | 13:01 |
| spenap | btw, does the webview work for you inside scratchbox? It was segfaulting to me last time I tried | 13:01 |
| khertan1 | DocScrutinizer: of course ... | 13:02 |
| khertan1 | DocScrutinizer but it didn't prevent you to import other module ? | 13:02 |
| DocScrutinizer51 | dunno | 13:02 |
| djszapi | DocScrutinizer51: only /etc/ is integrity protected apart from /etc/init/apps which is the 3rd party interface | 13:03 |
| DocScrutinizer51 | khertan1: the static protection is nonsense anyway | 13:03 |
| djszapi | There is no integrity protection for other things, not even for sbin! | 13:03 |
| razvanpetru | spenap: I use Qt SDK on Win7 and it works | 13:03 |
| DocScrutinizer51 | djszapi: oooh 'only' /rt/ ;-P | 13:03 |
| djszapi | What is done is a hash check defense, thus effectively you can use dpkg -i --force-all, but you will not be getting the whole credentialset since it is not coming from secure origin. | 13:04 |
| razvanpetru | the best thing I can say about webview is that at least it doesn't segfault | 13:04 |
| DocScrutinizer51 | /etc/ | 13:04 |
| spenap | razvanpetru, :P | 13:04 |
| razvanpetru | why the hell can't I have a decent webview that has pinchzoom and all the other crap and instead I have to reinvent the wheel every time | 13:04 |
| djszapi | /etc/ is integrity protected because of the configuration data integrity protection, so for making sure you cannot just easily brick the device with that damaged. | 13:04 |
| SpeedEvil | Does aegis have a 'config file' - or is the different behaviour on internal and external builds built-in? | 13:04 |
| spenap | razvanpetru, yes: it's quite disappointing | 13:05 |
| djszapi | SpeedEvil: sorry ? | 13:05 |
| DocScrutinizer51 | djszapi: when the measure to make sure you can not easily brick is just aegis is bricking it for you, that's rather paradox | 13:05 |
| *** stroughtonsmith has quit IRC | 13:06 | |
| djszapi | DocScrutinizer51: not getting you. | 13:06 |
| SpeedEvil | There has been different behaviour on internal - phones owne by nokia people - and external - community phones. | 13:06 |
| djszapi | SpeedEvil: nope | 13:06 |
| djszapi | maximum because it is an old image on N950, but that is. | 13:06 |
| SpeedEvil | ah | 13:07 |
| DocScrutinizer51 | djszapi: when I edit a protected file (like that hypothetical python script installed with a manifest), the legend says I get instant brick on save-file | 13:07 |
| * spenap bets that 'aegis' is the most used word in this channel, at least since the N950 started arriving to the developers :D | 13:07 | |
| razvanpetru | spenap: I could write an entire article on having to reinvent the wheel | 13:08 |
| Mek | well, and because the n950 is not in open mode and phones owned by nokia people are likely in open mode | 13:08 |
| Mek | that gives quite a difference | 13:08 |
| djszapi | Mek: for me, about nothing. | 13:09 |
| djszapi | spenap: :) | 13:09 |
| Mek | for one in the ones in open mode you can disable aegis, fake origins of packages (AEGIS_FIXED_ORIGIN), and do lots more things like that | 13:09 |
| djszapi | DocScrutinizer51: I am not getting you, but last time when there was "some aegis breakge", I did not really get anything how I could reproduce it. | 13:09 |
| Venemo | what do you guys think about this GUI so far? http://i56.tinypic.com/10pvnup.png http://i53.tinypic.com/bfnog4.png | 13:10 |
| djszapi | Mek: I am not sure it is a public information, I do hope so.. | 13:10 |
| *** NIN101 has joined #harmattan | 13:10 | |
| Stskeeps | isn't there like a irc plugin for telepathy? | 13:10 |
| Venemo | IRC plugin for telepathy can't join channels | 13:10 |
| djszapi | telapthy-idle | 13:10 |
| Stskeeps | well, on n900 it couldn't no as ui didn't support it | 13:10 |
| Mek | djszapi: what, how to get the device on open mode? I haven't seen any information on that so far | 13:10 |
| djszapi | * telepathy-idle | 13:10 |
| vandenoever | Venemo: i like the tabs on the side | 13:11 |
| DocScrutinizer51 | djszapi: see my post #94 with the nice picture. I'm sure somebody else can provide actual URL | 13:11 |
| spenap | Venemo, couple of things: you can hide the status bar, and you could hide and show the channel list on some gestures | 13:11 |
| spenap | and even the same for the navigation | 13:11 |
| Venemo | spenap, I plan on adding such gestures! | 13:11 |
| djszapi | Mek: not sure what your point is. | 13:11 |
| djszapi | I do not see any difference for my workflow. | 13:11 |
| Venemo | spenap, this is how it looks when everything is visible :) | 13:11 |
| Venemo | vandenoever, thank you | 13:11 |
| spenap | Venemo, that's fine :) (but I'd hide the status bar in landscape anyway :D) | 13:11 |
| spenap | always hidden, I mean | 13:12 |
| Venemo | spenap, that'll be switchable from Settings :) | 13:12 |
| spenap | ok | 13:12 |
| Mek | well, for my workflow of wanting to do anything with packages that are normally installed by nokia, AEGIS_FIXED_ORIGIN would be quite nice, but not available if not in open mode | 13:12 |
| *** NIN101 has quit IRC | 13:13 | |
| *** NIN101 has joined #harmattan | 13:15 | |
| djszapi | Mek: you can have open mode with N950 as well | 13:15 |
| razvanpetru | Venemo, the sidetabs won't work on touch, you need something else. | 13:15 |
| DocScrutinizer51 | for F in /etc/*; do echo >$F; done | 13:16 |
| djszapi | not sure why you would do anything with Nokia packages :) | 13:16 |
| djszapi | That is the platform anyway. | 13:16 |
| razvanpetru | Flickable top tabs should be fine... and you could also make the window corners square :) | 13:16 |
| Mek | djszapi: apps provided by nokia; and how do I get open mode? | 13:17 |
| DocScrutinizer51 | I wonder if that will just trow 'permissions missing' errors or brick the device instantly | 13:17 |
| djszapi | Venemo: is it zoomable ? | 13:17 |
| mgedmin | qmlcompass bug: if it starts in landscape, orientationAngle is not set correctly | 13:17 |
| mgedmin | although I'm not sure I have the latest version | 13:17 |
| djszapi | DocScrutinizer51: well, give me steps. I have been asking it from the beginning, never really gotten steps in order to be able to reproduce :) | 13:17 |
| djszapi | until that, it is not really an arguement. | 13:18 |
| djszapi | Mek: why do you wanna change the platform ? That is called Harmattan. | 13:18 |
| djszapi | If you do not wanna harmattan, flash your own kernel and you are in open mode, that is. | 13:18 |
| DocScrutinizer51 | djszapi: the post #94 HAS steps, I'm just not able to find the URL here (mobile on N900) | 13:19 |
| Mek | djszapi: I don't want to change the platform, I want to be able to change apps that run on top of the platform but happen to be made by nokia (or maybe I want to use some of the capabilities that are only allowed for nokia origin packages) | 13:19 |
| djszapi | DocScrutinizer51: take your time, and I am looking forward to getting them. | 13:19 |
| Mek | djszapi: so, where are the step-by-step instructions to get open mode? | 13:19 |
| djszapi | Mek: one step, flash your own kernel | 13:19 |
| razvanpetru | Venemo: also the icons are unclear... :) | 13:19 |
| DocScrutinizer51 | I bet several other users here can help out | 13:19 |
| Mek | djszapi: where do I get that kernel? | 13:20 |
| djszapi | Mek: wherever you want from :) | 13:20 |
| Mek | djszapi: I don't care, as long as I get open mode | 13:20 |
| Mek | djszapi: so, tell me please | 13:20 |
| DocScrutinizer51 | was sth along the line # aegis-exec +p=sgid bash | 13:20 |
| djszapi | "I don't want to change the platform, I want to be able to change apps that run on top of the platform but happen to be made by nokia (or maybe I want to use some of the capabilities that are only allowed for nokia origin packages)" -> No, it is not good. Fork it something else, but you are not more trusted than Nokia on a Nokia device ;-) | 13:21 |
| DocScrutinizer51 | result: instant brick | 13:21 |
| djszapi | Mek: kernel.org ? | 13:21 |
| Mek | djszapi: I can get ready-build kernels for the n950 from there? | 13:21 |
| djszapi | you can build something different than harmattan, that is actually your plan, not having what Nokia provides... | 13:22 |
| djszapi | I am not sure how much sense that makes though, but can be also a community project :) | 13:22 |
| djszapi | I would not do... | 13:22 |
| djszapi | Also, the kernel is open as I have been told anyway, you can just modify it ? | 13:23 |
| Mek | okay, so there is no way to get the device in open mode without actrually changing everything underneath it. that is one major difference with devices nokians have then, as those are in open mode wihtout changing everything underneath | 13:23 |
| Mek | I don't know, where are my step-by-step instructions on getting it in open mode... | 13:23 |
| djszapi | well, I would not like to discuss what happens inside Nokia, please.. | 13:23 |
| djszapi | Also, as I have been saying from the minute first, I am not using that workflow anyway.. | 13:23 |
| mgedmin | djszapi is a nokian? kudos for producing a wonderful device! | 13:24 |
| Mek | well, then please stop telling nonsense like how easy it is to get the device in open mode | 13:24 |
| razvanpetru | :) | 13:24 |
| djszapi | Mek: except that I have never said it is one click or "easy". | 13:25 |
| DocScrutinizer51 | well, achipa said "djszapi is not a Nokian" | 13:25 |
| DocScrutinizer51 | I dunno and I don't care | 13:25 |
| djszapi | mgedmin: nope, I am not a Nokia employee. | 13:25 |
| djszapi | Mek: Also, with documentation things bug the SDK team, and /not/ me. | 13:26 |
| razvanpetru | but will mgedmin retract his kudos now?? | 13:26 |
| mgedmin | contractors that worked on harmattan deserve kudos too | 13:27 |
| DocScrutinizer51 | definitely | 13:27 |
| * mgedmin waits for djszapi to confirm or deny the new allegation :) | 13:27 | |
| razvanpetru | kudos for everybody! | 13:27 |
| DocScrutinizer51 | bbl | 13:27 |
| djszapi | Mek: I am still thinking of it, but for a third party application developer, it is the same process. | 13:29 |
| djszapi | like people here. | 13:29 |
| mgedmin | aah, the cron mails that n950's mail client can't understand all have Content-Type: text/plain; charset=ANSI_X3.4-1968 | 13:30 |
| mgedmin | that's a weird way to spell ASCII | 13:30 |
| djszapi | tricks like that was made for platform developers for testing local packages and so forth, it is really not the use case of third party developers. | 13:30 |
| mgedmin | yep, if I edit the email and replace charset with ASCII, nokia's mail understands it | 13:32 |
| mgedmin | now do I hack cron on 7 servers to produce a different charset, or do I file a bug about the mail app? | 13:32 |
| *** epage has joined #harmattan | 13:34 | |
| mgedmin | what's the right venue to report fenix bugs? | 13:34 |
| djszapi | Mek: if you turn off the security or you misforce the source origin policy, your activity is pretty towards not using the platform or forcing the platform behaving how it is not supposed to be designed and work. | 13:34 |
| *** baraujo has joined #harmattan | 13:39 | |
| Venemo | djszapi, text size will be adjustable | 13:40 |
| Venemo | razvanpetru, top tabs would take away too much space from the messages | 13:41 |
| razvanpetru | then do this iOS trick: make them "pull-down" and hidden by default | 13:42 |
| DocScrutinizer | djszapi: I thought that's the purpose of development ;-D Forcing the platform to do things that it hasn't been developed for originally | 13:42 |
| Venemo | razvanpetru, why don't you like them on the side? | 13:43 |
| DocScrutinizer | s/develped/designed/ | 13:43 |
| spenap | Venemo, I agree with razvanpetru: this pull-down to make them appear would be nice | 13:43 |
| razvanpetru | very hard to read, and the ui becomes unbalanced | 13:43 |
| DocScrutinizer | s/developed/designed/ | 13:43 |
| infobot | DocScrutinizer meant: djszapi: I thought that's the purpose of development ;-D Forcing the platform to do things that it hasn't been designed for originally | 13:43 |
| djszapi | DocScrutinizer: well, it sounds like you wanna have for a different platform with different policies ? :) | 13:43 |
| djszapi | * wanna developer | 13:43 |
| djszapi | * develop, err ... :) | 13:43 |
| Venemo | razvanpetru, ok, I'll think about changing it :) | 13:44 |
| djszapi | Venemo: I guess you plan it as meego terminal then with the text size ? | 13:44 |
| DocScrutinizer | no, just want to teach the platform to do tricks it didn't know about :-) | 13:44 |
| Venemo | djszapi, dunno about meego-terminal, but I'll have a settings entry for text size | 13:44 |
| djszapi | DocScrutinizer: sounds like forcing the linux kernel to be able to do userspace driver development /only/ like in case QNX. Meanwhile it is fairly far away the plaftorm and design and reality :) | 13:45 |
| DocScrutinizer | djszapi: like doing USB hostmode, using a FOSS maemo hotspot rather than a proprietary closed joiku spot, etc | 13:45 |
| Venemo | razvanpetru, problem with your idea is: pulldown will flickscroll the text area, so can't be used for the tabs. | 13:45 |
| djszapi | Venemo: zooming is done nice in meego terminal | 13:46 |
| *** NIN101 has quit IRC | 13:46 | |
| DocScrutinizer | make airodump work | 13:46 |
| Venemo | djszapi, I see, but I don't think it's a setting that you would adjust too often | 13:46 |
| djszapi | DocScrutinizer: not sure I understand why you cannot do a third party hotspot application... | 13:46 |
| DocScrutinizer | get a better smarted brightness applet - wait harmattan seems to come without any applet for brightness | 13:47 |
| DocScrutinizer | djszapi: maybe it needs a "fixed" kernel module driver for WLAN? | 13:47 |
| razvanpetru | Venemo: how about having the tabs on a small row and flicking between them with "the text area"? So if you flick sideways you switch tabs | 13:48 |
| razvanpetru | you could make it full screen and use the status bar spae for the tab display | 13:48 |
| mgedmin | devel-su; develsh; /sbin/iptables -vnL --> yay iptables are there! | 13:48 |
| Venemo | razvanpetru, that would work for a small number of channels, but would fail miserably when you have 10+ | 13:48 |
| djszapi | DocScrutinizer: I think we discussed enough time how aegis-validator actually works with the kernel modules. | 13:49 |
| mgedmin | /sbin/iptables -vnL -t nat -> f%%k, no NAT support | 13:49 |
| Venemo | razvanpetru, but I promise to add a gesture to hide the bar, that's for sure :) | 13:49 |
| DocScrutinizer | djszapi: fact is aegis has a default policy of granting as little permissions as possible, and that pretty much restricts development/messing-with-system to what aegis maintainers could imagine a developer would need to do | 13:49 |
| SpeedEvil | Which is just fine for people doing fart apps. But not so much adding support for unsupported bits of hardware. | 13:50 |
| DocScrutinizer | mgedmin: nice example | 13:50 |
| SpeedEvil | Like some of us are interested in. | 13:50 |
| mgedmin | wasn't an example, was a complaint | 13:50 |
| DocScrutinizer | SpeedEvil: exactly | 13:51 |
| SpeedEvil | Or indeed mgedmin's example that I have wondered about writing a flash debandwithiser. | 13:51 |
| razvanpetru | Venemo: one last shot - use the menu button to show a list of tabs :P | 13:51 |
| *** NIN101 has joined #harmattan | 13:51 | |
| SpeedEvil | Using something like rtmpsuck and ffenc or something to shrink the bandwidth of a flash stream | 13:51 |
| razvanpetru | I think popup menus won't be that popular... | 13:51 |
| Venemo | razvanpetru, the chat bubble button on the toolbar will do that | 13:52 |
| DocScrutinizer | mgedmin's example absolutely fits into " not sure I understand why you cannot do a third party hotspot application..." | 13:52 |
| mgedmin | Venemo, I missed the beginning of your conversation, but are you perhaps talking about adding multiple terminal windows/tab to meego-terminal? | 13:52 |
| Venemo | mgedmin, nope! | 13:52 |
| khertan1 | Does oneclick flasher available at https://www.developer.nokia.com/info/sw.nokia.com/id/db230178-aa63-4c73-ba7f-20930da13cad/Nokia_N950_OneClickFlashers.html flash everything including the /MyDocs partitions ? | 13:52 |
| mgedmin | shame :) | 13:52 |
| Venemo | mgedmin, near miss... | 13:52 |
| razvanpetru | venemo: then you're set... but change the icon :P | 13:52 |
| Venemo | mgedmin, http://i56.tinypic.com/10pvnup.png http://i53.tinypic.com/bfnog4.png | 13:52 |
| *** baraujo has quit IRC | 13:53 | |
| mgedmin | ooh, irc client? | 13:53 |
| mgedmin | nice! | 13:53 |
| DocScrutinizer | khertan1: yes | 13:53 |
| khertan1 | Venemo: i would suggest to unify the both toolbar | 13:53 |
| Venemo | razvanpetru, if you can suggest a PlatformIconId that's better, I'll change it. but I haven't found a better one :( | 13:53 |
| khertan1 | Venemo: and made a contextual one | 13:53 |
| khertan1 | DocScrutinizer: ouch ... | 13:53 |
| * SpeedEvil idly wonders if the above runs on n900. | 13:53 | |
| khertan1 | DocScrutinizer: so need to backup everything before ... | 13:53 |
| DocScrutinizer | khertan1: indeed | 13:54 |
| Venemo | khertan1, now that sounds like a good idea... hmm... :) | 13:54 |
| SpeedEvil | (venemo) | 13:54 |
| djszapi | DocScrutinizer: you can write an adhoc hotspot userspace application anytime | 13:54 |
| Venemo | mgedmin, it's just a GUI so far | 13:54 |
| djszapi | there is no limitation in the kernel for that. | 13:54 |
| khertan1 | DocScrutinizer: hum ... does we can choose to what to restore of the backup ? | 13:54 |
| SpeedEvil | djszapi: How do you overcome the lack of iptables in kernel? | 13:54 |
| DocScrutinizer | khertan1: I don't see any option for it | 13:54 |
| khertan1 | :( | 13:55 |
| djszapi | 7:ip_tables 8122 1 iptable_filter | 13:55 |
| mgedmin | HotSpot on the N900 came with a custom kernel module for doing NAT | 13:55 |
| djszapi | SpeedEvil: lsmod | grep -rni iptable | 13:55 |
| DocScrutinizer | khertan1: you might find ways not to flash MyDocs though - maybe | 13:55 |
| djszapi | 6:iptable_filter 1854 1 | 13:55 |
| khertan1 | DocScrutinizer but i'm on windows currently : | 13:55 |
| khertan1 | :) | 13:55 |
| DocScrutinizer | :-/ | 13:56 |
| *** baraujo has joined #harmattan | 13:56 | |
| djszapi | khertan1: did you port windows8 to N950 ? :p | 13:56 |
| SpeedEvil | Or rather NAT support. | 13:56 |
| Venemo | SpeedEvil, well, someone ported the MeeGo components to Maemo 5, so with some tweaking, it might work | 13:56 |
| *** razvanpetru has left #harmattan | 13:56 | |
| mgedmin | djszapi, now try /sbin/modprobe iptable_nat :) | 13:56 |
| SpeedEvil | (11:49:03 AM) mgedmin: /sbin/iptables -vnL -t nat -> f%%k, no NAT support | 13:56 |
| DocScrutinizer | djszapi: [2011-07-28 12:49:03] <mgedmin> /sbin/iptables -vnL -t nat -> f%k, no NAT support | 13:56 |
| djszapi | do not spam ! :) | 13:56 |
| khertan1 | djszapi: certainly not ... i m just saying that desktop pc use windows | 13:56 |
| lcuk | morning \o | 13:56 |
| djszapi | khertan1: cannot handle jokes ? :) | 13:56 |
| *** divan_ has joined #harmattan | 13:57 | |
| djszapi | lcuk o/ | 13:57 |
| *** divan_ is now known as divan|work | 13:57 | |
| Venemo | lcuk, http://i56.tinypic.com/10pvnup.png http://i53.tinypic.com/bfnog4.png whaddoyouthink? | 13:57 |
| khertan1 | djszapi: not when talking about forced use of windows at works which already crash 2 time this morning | 13:57 |
| DocScrutinizer | lcuk: hi | 13:57 |
| mgedmin | lcuk, you are a nokian, right? | 13:57 |
| mgedmin | any clues where I should complain about bugs in fenix? | 13:57 |
| DocScrutinizer | no | 13:57 |
| khertan1 | what is fenix ? | 13:58 |
| mgedmin | the mail client on the n950 | 13:58 |
| MohammadAG | email client | 13:58 |
| khertan1 | oh ... i have complain too to post :) | 13:58 |
| Venemo | mgedmin, complain to frals | 13:58 |
| DocScrutinizer | or achipa | 13:58 |
| khertan1 | no signature with mail for exchange | 13:58 |
| khertan1 | exchange only works one time with gmail | 13:58 |
| khertan1 | :) | 13:58 |
| mgedmin | signature shmignature, I can't read emails that were sent by cron | 13:58 |
| mgedmin | because of a silly charset name not being recognized as ASCII | 13:59 |
| MohammadAG | how well is fenix compared to modest? | 13:59 |
| mgedmin | MohammadAG, feels faster, but that's probably due to better hardware :) | 13:59 |
| khertan1 | anyway, better than modest :) | 13:59 |
| Venemo | MohammadAG, fenix is a LOT better | 13:59 |
| Venemo | MohammadAG, it has "mark all as read", and it hasn't segfaulted for me yet | 13:59 |
| w00t | fenix is short for fantastix | 13:59 |
| mgedmin | lets me access IMAP folders without having to subscribe to them from a desktop | 13:59 |
| Venemo | w00t :) | 13:59 |
| lcuk | mgedmin, not a Nokian (yet?) but the same place you file any bug for harmattan | 14:00 |
| MohammadAG | w00t, got your device? | 14:00 |
| mgedmin | lcuk, which was "if it's meego core component, go to bugs.meego.com, otherwise go to developer.nokia.com" | 14:00 |
| khertan1 | Fenix -> Xinef -> Xchange is not entirely functionnal | 14:00 |
| mgedmin | and I've no clue what is a "meego core component" and what is not! | 14:00 |
| w00t | MohammadAG: yes | 14:00 |
| lcuk | mgedmin, put it on developer.nokia and let it be passed back if required | 14:00 |
| w00t | just the other day | 14:01 |
| MohammadAG | update the wiki then :P | 14:01 |
| w00t | yes, yes :P | 14:02 |
| Venemo | w00t, you liking? | 14:02 |
| w00t | i've had other things happening | 14:02 |
| djszapi | SpeedEvil: well, again this is harmattan with fixed kernel capability. I am not sure what your point is :) | 14:02 |
| spenap | Khertan, accounts -> MfE -> Mail -> signature | 14:02 |
| w00t | Venemo: yes, it's sexy sauce | 14:02 |
| djszapi | internal tricks will not help you with shipping it with the production image anyway | 14:02 |
| w00t | MohammadAG: Venemo: you guys? | 14:03 |
| MohammadAG | I'm still on New Order, yay | 14:03 |
| Venemo | w00t, I have, MohammadAG not yet | 14:03 |
| spenap | and while it fails quite often sync'ing (gmail on MfE), it's usable | 14:03 |
| Venemo | yeah, and it doesn't display HTML e-mails over MfE correctly | 14:03 |
| DocScrutinizer | djszapi: that's exactly the point where I think N9 harmattan is *useless* in the end, thanks aegis | 14:03 |
| Venemo | and sometimes fails on my hotmail as well just as spenap says | 14:03 |
| lcuk | Venemo, do this and I will love you longtime: | 14:04 |
| lcuk | http://liqbase.net/liq.20091201_191242._maemo.scr.png | 14:04 |
| djszapi | DocScrutinizer: again, if you do not wanna Harmattan, send your device back :) | 14:04 |
| MohammadAG | IRC's working, but anything else related to the interwebz isn't, any ideas? | 14:04 |
| djszapi | DocScrutinizer: It seems to me the same story with Android overall anyway | 14:04 |
| Venemo | lcuk, I'm not familiar with liq enough for this :( plus, I wanted to learn QML :) | 14:05 |
| lcuk | Venemo, do it in qml | 14:05 |
| djszapi | DocScrutinizer: except that, they do not even have too much security. | 14:05 |
| lcuk | idontcarehhowyouwriteit | 14:05 |
| DocScrutinizer | djszapi: yes, exactly. And same story as iPhone | 14:05 |
| Venemo | lcuk, I'm doing it in QML | 14:05 |
| djszapi | DocScrutinizer: Also, I have been told you need customer kernel for nat support in case of N900 as well anyway | 14:05 |
| Venemo | lcuk, then what were you referring to when you said "do this"? | 14:06 |
| djszapi | maybe the guy who told me was not right, but that is I have been told about | 14:06 |
| DocScrutinizer | he was right | 14:06 |
| djszapi | so what is your point ? | 14:06 |
| djszapi | you can have your own kernel here as well ? | 14:06 |
| DocScrutinizer | no I can't | 14:06 |
| djszapi | fail to see why not.. | 14:07 |
| DocScrutinizer | I can have my own kernel with arbitrary NON-HARMATTAN system running on it, while on N900 fremantle happily runs on custom kernels | 14:07 |
| lcuk | Venemo, the pictures | 14:08 |
| DocScrutinizer | (and even the mere suggestion of custom kernel seems on the test bench still) | 14:08 |
| djszapi | DocScrutinizer: This is just plain wrong | 14:08 |
| djszapi | DocScrutinizer: You can get the source of the Harmattan kernel, and modify it a bit with your superninja module ? | 14:08 |
| DocScrutinizer | then do what? | 14:08 |
| DocScrutinizer | boot harmattan with it? | 14:08 |
| djszapi | I do not see the difference between N900 and Harmattan regarding this. I do not know the N900 way though, I have just been told about that. | 14:09 |
| DocScrutinizer | we still miss a word of Nokia how many of the core apps will fail to work on such a aegis-free system. As you can't sugest a custom kernel will happily run aegis in "normal mode" | 14:10 |
| djszapi | DocScrutinizer: You are missing the point. You do *not* disable aegis | 14:10 |
| djszapi | You *add* your own superninja module extending the aegis whitelist, come on. | 14:10 |
| khertan1 | hum ... backup didn't works | 14:10 |
| khertan1 | lol | 14:10 |
| DocScrutinizer | no, the unsigned kernel will disable aegis | 14:10 |
| djszapi | DocScrutinizer: fail to see why you could not modify that bit of aegis, if it is that way... | 14:11 |
| DocScrutinizer | toherwise aegis would be a hoax | 14:11 |
| djszapi | no, it is pretty simple kernel module. | 14:11 |
| DocScrutinizer | interacting with a TPM, yes | 14:12 |
| DocScrutinizer | ~aegis | 14:12 |
| infobot | [aegis] http://www.developer.nokia.com/Community/Wiki/Harmattan:Developer_Library/Developing_for_Harmattan/Harmattan_security/Security_guide , or "The purpose of this framework is: ... to make sure that the platform meets the requirements set by third party software that requires a safe execution environment.", or http://en.wikipedia.org/wiki/Trusted_Computing#Criticism | 14:12 |
| DocScrutinizer | for that | 14:12 |
| DocScrutinizer | >>>"The purpose of this framework is: ... to make sure that the platform meets the requirements set by third party software that requires a safe execution environment."<<< ! | 14:12 |
| Stskeeps | DocScrutinizer: you'll be happy to know i've booted my own stuff with moslo, but you'll have to wait a bit as we had a bug in the fw images you guys have | 14:12 |
| Stskeeps | (work in progress) | 14:12 |
| DocScrutinizer | Stskeeps: pleased to hear | 14:12 |
| DocScrutinizer | Stskeeps: I gather you weren't able to start closed secured Nokia blobs on your "own stuff" | 14:13 |
| djszapi | DocScrutinizer: so it seems to me it is not any different to N900 and you can do it, period. :) | 14:14 |
| DocScrutinizer | even less you should be able to start >>third party software that requires a safe execution environment<< on your "own stuff" | 14:14 |
| MohammadAG | tbh, if I had an N950 I'd just try it | 14:15 |
| DocScrutinizer | if you were able to do that, then aegis implementation is a hoax | 14:15 |
| MohammadAG | why not do the same? | 14:15 |
| MohammadAG | get the kernel source, recompile, flash | 14:15 |
| MohammadAG | see what happens, experiment, it's a dev device after all | 14:15 |
| djszapi | DocScrutinizer: few lines... | 14:15 |
| djszapi | does not require too much kernel experience either. | 14:15 |
| DocScrutinizer | MohammadAG: >><Stskeeps> DocScrutinizer: you'll be happy to know i've booted my own stuff with moslo, but you'll have to wait a bit as we had a bug in the fw images you guys have<< | 14:15 |
| MohammadAG | saw it | 14:16 |
| MohammadAG | but why not try it with flasher-3.5 -k image -f -R | 14:16 |
| khertan1 | someone have already tryed backup tool ? does it s works ? | 14:16 |
| MohammadAG | see what happens, it won't "brick" anyway | 14:16 |
| khertan1 | seems freezing at first stage | 14:16 |
| Stskeeps | MohammadAG: been there, that's the bug i spoke of ;) | 14:17 |
| * mgedmin tries to create a backup with a blank title | 14:18 | |
| DocScrutinizer | MohammadAG: won't even boot, javispedro did it | 14:18 |
| mgedmin | khertan1, worked for me | 14:18 |
| Stskeeps | so, patience.. :P | 14:19 |
| MohammadAG | DocScrutinizer, a hammer should make it work | 14:19 |
| djszapi | :D | 14:19 |
| DocScrutinizer | MohammadAG: the kernel has a signature, this signature changes when even a single byte (like build time) changes -> kernel and modules don't match anymore | 14:20 |
| djszapi | updating a hash come on... :) | 14:20 |
| djszapi | really big taaaask :) | 14:21 |
| DocScrutinizer | device freezes before linux even starts aiui | 14:21 |
| Stskeeps | can i propose someone to build kernel, different version, so different /lib/modules/<path> ? | 14:21 |
| DocScrutinizer | djszapi: the hash is against a cert that we don't have access to | 14:21 |
| DocScrutinizer | djszapi: NOLO notices the kernel isn't signed by Nokia rootcert, so it immediately disables TPM (in best case, worst case it simply refuses to load kernel all together) | 14:23 |
| djszapi | DocScrutinizer: you are not right | 14:25 |
| DocScrutinizer | even best case, the harmattan system as is will most likely blow chunks on bootup, when one of the several aegis components (kernel modules, executables) tries to access TPM. This is what "open mode" basically means, and it's been frequently stated here harmattan isn't able to run in open mode on systems without R&D cert | 14:25 |
| djszapi | Thing is the following: | 14:26 |
| dm8tbr | nolo should put the device in open mode if the kernel is not signed AIFAUI | 14:26 |
| DocScrutinizer | dm8tbr: exactly | 14:27 |
| DocScrutinizer | but harmattan isn't supposed to work in open mode | 14:27 |
| djszapi | If you have your own kernel module, and you add your MySuperNinjaModuleEver, you actually are in open mode. What certificate you were talking about is t he libbb5 five and the calibration area because you actually turned off your device into open mode with that. You do not have access to the libbb5, calibration area any longer. You do not actually even need to modify the validator anymore, you do not need to check it, it will work out-of-the-box! Also, t | 14:27 |
| Stskeeps | who says that? | 14:27 |
| Stskeeps | :P | 14:27 |
| DocScrutinizer | that's what a number of Nokians clearly stated here | 14:27 |
| djszapi | 14:27 < DocScrutinizer> but harmattan isn't supposed to work in open mode -> funky, I told gazillion times the opposite.... | 14:28 |
| DocScrutinizer | achipa (iirc) said "open mode is not possible on devices without R&D cert, I.E. the CE devices you got" | 14:29 |
| djszapi | It is really not any different to N900 and even easier than ever... | 14:30 |
| DocScrutinizer | frals said similar things, though more specifically | 14:30 |
| *** baraujo has quit IRC | 14:30 | |
| *** baraujo_ has joined #harmattan | 14:30 | |
| DocScrutinizer | djszapi: ok, so which of the nokia blobs will refuse to work in open mode - assuming it is possible to enter open mode by flashing my own custom kernel? | 14:31 |
| *** baraujo_ is now known as baraujo | 14:31 | |
| djszapi | well, what they were talking about it in my opinion is the R&D image, not a customer open mode kernel image... | 14:32 |
| djszapi | DocScrutinizer: why should it stop to work, any reason for this guesswork ? | 14:33 |
| DocScrutinizer | the reason and rationale is: "The purpose of this framework is: ... to make sure that the platform meets the requirements set by third party software that requires a safe execution environment." - plus the fact e.g bme is pretty much "protected" in a normal harmattan in "normal mode" | 14:34 |
| djszapi | what is "bme" ? | 14:34 |
| DocScrutinizer | I'd not think Nokia can refrain from protecting own "IP" same way it offers for "third party" | 14:35 |
| DocScrutinizer | ~bme | 14:35 |
| infobot | extra, extra, read all about it, bme is http://wiki.maemo.org/N900_Software_BME | 14:35 |
| djszapi | At any rate...Your custom kernel is not a safe execution environment | 14:35 |
| DocScrutinizer | battery management entity | 14:35 |
| djszapi | anybody using it should be aware of this... | 14:35 |
| djszapi | That is a completely different question whether it still works, imo should. | 14:35 |
| Elleo | djszapi: ah cool, I'll hang fire on liblzma-dev then, thanks | 14:35 |
| djszapi | even if they work, everybody should be really aware of this | 14:35 |
| djszapi | Elleo: it is the fault of the SDK team | 14:35 |
| Venemo | lcuk, hmm! | 14:36 |
| djszapi | I am trying to complain about it to them for 2-3 weeks | 14:36 |
| djszapi | desperate... | 14:36 |
| Venemo | lcuk, where does it get the pictures from? | 14:36 |
| DocScrutinizer | djszapi: the promised link to post with instructions how to brick your device: http://forum.meego.com/showpost.php?p=25529&postcount=94 - the command is # aegis-exec -a CAP::setgid -a CAP::setuid bash | 14:38 |
| DocScrutinizer | where bash can probably get replaced by arbitrary binary | 14:38 |
| DocScrutinizer | that tries to do a setgid | 14:38 |
| lcuk | Venemo, from the irc comment | 14:39 |
| lcuk | it parses what people say | 14:39 |
| lcuk | and any pictures or links or info are turned into buttons below each comment | 14:39 |
| lcuk | showing thumbnails | 14:39 |
| Venemo | lcuk, mhm | 14:39 |
| djszapi | DocScrutinizer: am I your friend for a beer if I cannot brick it ? ;) | 14:39 |
| DocScrutinizer | sure thing | 14:39 |
| djszapi | :p | 14:40 |
| Venemo | lcuk, so you basically need say your avatar image every time you post a message? | 14:40 |
| lcuk | Venemo, no? | 14:40 |
| lcuk | avatar is different | 14:40 |
| Venemo | lcuk, then where does it get the avatars from? | 14:40 |
| DocScrutinizer | only if you can give me the bits needed so it won't brick my device as well | 14:40 |
| lcuk | avatars actually come from maemo.org | 14:40 |
| lcuk | :D | 14:40 |
| Venemo | mhmm | 14:40 |
| djszapi | DocScrutinizer: well, I need to write some new Qt classes proposal into Qt5, but I will try it out when I get there. | 14:40 |
| DocScrutinizer | take your time | 14:41 |
| djszapi | Also, wouldn't a bugreport be a better occasion for this ? | 14:41 |
| Stecchino | djszapi: new classes? Curious minds want to know :) | 14:41 |
| DocScrutinizer | javispedro had an even more evolved method to brick it same way, was along installing rpcbind and then waiting for bricking at random time | 14:41 |
| Venemo | lcuk, I think I could implement these ideas, but I'd need a better way for avatars... if IRC could tell me everyone's e-mail addresses, I could just use Gravatar, but it doesn't | 14:41 |
| DocScrutinizer | djszapi: do you think this is a bug? | 14:42 |
| DocScrutinizer | looks pretty much intentional | 14:42 |
| lcuk | Venemo, do the best you can with what you have to hand | 14:42 |
| djszapi | DocScrutinizer: Look, developers do not follow the forums, but they are assigned to reports. | 14:42 |
| djszapi | * not every developer | 14:43 |
| Venemo | lcuk, getting 'em from maemo.org might work for us Maemo users, but won't for anyone else | 14:43 |
| DocScrutinizer | look, I don't even have a manpage for aegis-exec | 14:43 |
| lcuk | Venemo, other people are not so important obviously :P and it would need updating to be meego.com user accounts | 14:44 |
| lcuk | afterall, it will be a meego irc client? | 14:44 |
| lcuk | question: | 14:44 |
| DocScrutinizer | so how to base a ticket on a random incantation of aegis-exec that we _imagined_ might do what we need, but actually does something pretty much different? | 14:44 |
| Venemo | yeah | 14:44 |
| lcuk | given this version: Version: 1:0.4.0-258+0m6 | 14:44 |
| lcuk | is the -258 BUILD only | 14:44 |
| lcuk | and no code changes from a (for instance) -240 | 14:44 |
| DocScrutinizer | djszapi: I'd agree though that bricking device is always a bug - in a *normal* system | 14:45 |
| djszapi | DocScrutinizer: why not bug against the SDK team ? | 14:45 |
| DocScrutinizer | I got pretty much flamed for "doing a `rm -rf /` alike nonsense with this aegis-exec command" | 14:46 |
| djszapi | Stecchino: You are a KDE contributor iirc, so you might know this use case. Pure Qt solution for archiving :p | 14:47 |
| djszapi | something like K{Archive,Ar,Zip,Tar,etc} | 14:47 |
| DocScrutinizer | and honestly I'm not exactly eager to help make aegis a success story. If I could do anything to abolish it, I'll do | 14:47 |
| Stecchino | djszapi: arK like? | 14:47 |
| djszapi | Stecchino: Qt classes. | 14:47 |
| djszapi | just the functionality in the platform. | 14:48 |
| Mek | KArchive, Ar, Zip, Tar etc are all already pure-qt with no other kde dependncies... | 14:48 |
| Stecchino | djszapi: couldn't they also be K-classes without kdelibs deps? | 14:48 |
| djszapi | not really, no | 14:48 |
| Venemo | lcuk, also, what do you think of my channel switching solution? | 14:48 |
| djszapi | Mek: You are wrong. They depend on KSaveFile | 14:48 |
| djszapi | They depend on other things, like kde_file.h, kde_file_win.h and so forth | 14:48 |
| djszapi | I made a pure Qt version for our projects, but it seems a common requirement. | 14:48 |
| djszapi | They depend on KFilterDev, KFilterBase and so much other things. | 14:49 |
| djszapi | KMimeType. | 14:49 |
| djszapi | and I could enumerate... | 14:49 |
| djszapi | Stecchino: from what I see is that, in my past at least 3-4 Qt projects required it and I also heard it from others as well | 14:49 |
| Stecchino | yup, it does seem to tie into a lot of kdelibs' conveinience classes, but nothing runtime right? | 14:50 |
| lcuk | The following NEW packages will be installed: | 14:50 |
| lcuk | autoconf automake autotools-dev coreutils debianutils m4 mktemp | 14:50 |
| lcuk | 0 upgraded, 7 newly installed, 461 to remove and 0 not upgraded. | 14:50 |
| lcuk | Need to get 6,621kB of archives. | 14:50 |
| lcuk | After this operation, 310MB disk space will be freed. | 14:50 |
| lcuk | Do you want to continue [Y/n]? y | 14:50 |
| Stecchino | what worries me more is kdelibs runtime components | 14:50 |
| Stecchino | speaking of which | 14:50 |
| djszapi | Stecchino: that is the point, they do not have runtime dependencies and that is why their implementation is a bit longer than invoking libarchive and so on :) | 14:50 |
| Stecchino | Mek, djszapi: have you seen the progress status API in Harmattan? | 14:51 |
| * lcuk thinks removing the mp-harmattan-rm680-pr meta package might cause some things to stop working :$ | 14:52 | |
| DocScrutinizer | lcuk: \o/ | 14:55 |
| DocScrutinizer | lol, you bet removing 461 will stop sth to work, no? | 14:55 |
| DocScrutinizer | >>After this operation, 310MB disk space will be *freed*(!!!) | 14:56 |
| djszapi | DocScrutinizer: you did not set up any link to /bin/sh, did you ? | 14:56 |
| *** Rizzer has joined #harmattan | 14:56 | |
| Stecchino | Mek, djszapi: https://www.developer.nokia.com/swipe/ux/pages/Transfer_manager.html | 14:57 |
| Stecchino | BTW, why are you guy not on #kde-mobile ATM? | 14:57 |
| DocScrutinizer | I'd be concerned about dis space wasted if that doesn't result in anything ceasing to work | 14:57 |
| DocScrutinizer | djszapi: nope | 14:57 |
| djszapi | Stecchino: I am sorry, I have no experience with this :( | 14:57 |
| djszapi | Stecchino: btw, if you want to use kdelibs, I can give it to you. | 14:57 |
| lcuk | DocScrutinizer, :D | 14:58 |
| djszapi | I have been trying to make it work on c-obs for 3 weeks, no luck yet :p | 14:58 |
| lcuk | amusing, just rebooting now to see what is left | 14:58 |
| Stecchino | djszapi: could be good to have that be used by KJob ui-servier | 14:58 |
| lcuk | after it removed ssh | 14:58 |
| lcuk | I was viewing the app grid | 14:58 |
| lcuk | and icons were vanishing | 14:58 |
| djszapi | Stecchino: I ported KJob also to pure Qt in Gluon :) | 14:58 |
| * lcuk was slightly amused | 14:58 | |
| lcuk | the store icon was the last official icon to vanish! | 14:58 |
| DocScrutinizer | lcuk: reflash time ;-D | 14:58 |
| Stecchino | djszapi: I have some experience with OBS, ping me in #kde-mobile and I'll have a look | 14:58 |
| djszapi | Stecchino: thing is that it gets stalled. | 14:59 |
| djszapi | :( | 14:59 |
| lcuk | DocScrutinizer, of course | 14:59 |
| djszapi | or rpc timeout, that are the two ccasions. | 14:59 |
| lcuk | I was only trying to build sgx library on device | 14:59 |
| lcuk | where is the image download? | 14:59 |
| djszapi | Stecchino: https://build.pub.meego.com/package/live_build_log?arch=armv7el&package=kdelibs&project=home%3Adjszapi&repository=MeeGo_1.2_Harmattan | 14:59 |
| DocScrutinizer | lcuk: http://wiki.meego.com/N950_landing_page | 15:00 |
| *** hardaker has joined #harmattan | 15:00 | |
| djszapi | DocScrutinizer: well, it should /not/ brick it then :) | 15:00 |
| Stecchino | djszapi: have you reported a bugs.meego.com about this? | 15:00 |
| djszapi | Stecchino: I keep repeating it almost every day to lbt | 15:00 |
| Stecchino | I doubt you have broken the base system, but could have triggered OOM killer | 15:00 |
| DocScrutinizer | djszapi: it doesn't just "brick", it does an emergency boot and a *lock* | 15:01 |
| Stecchino | djszapi: bugs is still the best way | 15:01 |
| djszapi | Stecchino: They have 24 GB memory | 15:01 |
| djszapi | Stecchino: Actually, I reported it few weeks ago | 15:01 |
| *** trx has quit IRC | 15:01 | |
| Stecchino | individual processes can still go over some limit | 15:01 |
| djszapi | lbt ^ ? | 15:01 |
| DocScrutinizer | djszapi: punishment for "stupid" devels | 15:01 |
| djszapi | DocScrutinizer: Thanks, kindly. | 15:01 |
| DocScrutinizer | djszapi: I hope you don't misunderstand that statement | 15:02 |
| *** epage has quit IRC | 15:02 | |
| Stecchino | DocScrutinizer: keep in mind you are talking to the devs of a 5*10^6 loc codebase here ;) | 15:02 |
| DocScrutinizer | I think the "aegis has detected problems and locked your device" is perceived by developers as a punishment for them doing "sth stupid" | 15:03 |
| djszapi | Stecchino: It was you arraging the table football competition at Randa ? :) | 15:03 |
| Stecchino | yes | 15:03 |
| Stecchino | and winning.... | 15:03 |
| Stecchino | :) | 15:03 |
| djszapi | heh :p | 15:04 |
| DocScrutinizer | Stecchino: djszapi: wasn't my intention to offend any of the developers of harmattan | 15:04 |
| Stecchino | should check to see if there will be any tables around in Berlin :) | 15:04 |
| Stecchino | DocScrutinizer: don't worry, we're not offended :) | 15:04 |
| djszapi | Stecchino: I did not even use the KDE_ENABLE_FINAL option in the rules file :) | 15:05 |
| lcuk | book reading: this is ona a large format device: http://liqbase.net/liq.20110725_163156.liqbookreader.scr.png would you expect page by page reading | 15:05 |
| djszapi | I tried to lower the memusage :p | 15:05 |
| Stecchino | but i was talking about the KDE codebase, it used to be close to 5 million lines of code | 15:05 |
| lcuk | or would kinetics work well do you think | 15:05 |
| DocScrutinizer | how should you, I wasn't even targeting the "offending" post at Nokia | 15:05 |
| mgedmin | but but but books are black on white, not white on black! | 15:05 |
| Stecchino | djszapi: what worries me is that it seems to be building kjs, which contains some automatically generated code if I'm not mistaken | 15:05 |
| djszapi | Stecchino: It stalls at random stages :( | 15:06 |
| Stecchino | oh, that is worse | 15:06 |
| DocScrutinizer | Stecchino: but your and djszapi's reaction made me think you took it as targeted at harmattan developers, while it was a completely different statement | 15:06 |
| hiemanshu | mgedmin: white on black is better for the eyes, and the books dont understand that sadly | 15:06 |
| djszapi | There are situations when it stalls after the complete build, but prior to the installation. | 15:06 |
| mgedmin | [citation needed] | 15:06 |
| djszapi | DocScrutinizer: no worries ;) | 15:06 |
| Stecchino | DocScrutinizer: to be honest, I also did some Harmattan work, but not enough to call me a harmattan dev. | 15:06 |
| *** trx has joined #harmattan | 15:07 | |
| djszapi | Stecchino: I was proposing to a direct repository update from my local PC few days ago to lbt, I did not get any answer | 15:07 |
| djszapi | it just blocks the kde-mobile project :( | 15:07 |
| DocScrutinizer | I was not , repeat NOT, talking about "stupid harmattan devels" - no way. I tried to characterize what the locking of device by aegis implies to say to the developer that caused that locking | 15:08 |
| Stecchino | djszapi: I wonder if we can try splitting the build into separate packages | 15:08 |
| djszapi | DocScrutinizer: Could you send me the bash package, please ? | 15:08 |
| Mek | if it is any consolidation, the diablo buildserver never managed to compile kdelibs either, and with fremantle I think it took a while too before it managed to build kdelibs | 15:08 |
| DocScrutinizer | djszapi: it's from nokia repo | 15:08 |
| djszapi | Stecchino: That is a good idea for the time being. | 15:08 |
| Mek | (in those cases it was mostly bugs in qemu that were problematic) | 15:08 |
| DocScrutinizer | djszapi: the URL for wget is in post #94 | 15:09 |
| lbt | djszapi: does a local build work? | 15:09 |
| djszapi | lbt: yep, I told you that bit ;) | 15:09 |
| Stecchino | with one master dep meta package that fetches the source, and the other just leeching of it so they don't have to download kdelibs | 15:09 |
| DocScrutinizer | djszapi: btw I think I've sent you a verbouse mail about it | 15:09 |
| djszapi | lbt: my temporary proposal is to push the deb packages to my repo directly... | 15:09 |
| DocScrutinizer | verbose* | 15:09 |
| djszapi | lbt: and get replaced when OBS first can build it up. | 15:10 |
| djszapi | lbt: it worked locally 2-3 weeks ago. | 15:10 |
| djszapi | I have been just waiting for OBS since that. | 15:10 |
| lbt | clarify local... osc build ? | 15:10 |
| djszapi | yes. | 15:10 |
| lbt | ISTR asking you to check the mem peak | 15:11 |
| lbt | you didn't want to wait 5h ... :) | 15:11 |
| djszapi | lbt: what do you mean ? | 15:11 |
| lbt | peak memory usage during a local osc build | 15:11 |
| djszapi | well, we can debug it meanwhile it is available in my repository by direct copy. | 15:12 |
| DocScrutinizer | djszapi: just in case you missed the mail I sent: http://paste.debian.net/124359/ | 15:12 |
| djszapi | Stecchino: the problem is that you still need to build kdelibs, so the modularization would not avoid the build stage. | 15:16 |
| *** lbt has quit IRC | 15:16 | |
| *** lbt_ has joined #harmattan | 15:16 | |
| *** lbt_ is now known as lbt | 15:16 | |
| djszapi | Mek: mmm, maybe qemu bug here as well :p | 15:17 |
| Stecchino | djszapi: but the memory will be cleared for each package. So if anything is leaking memory and causing a OOM it's less likely to happen with shorter build times | 15:17 |
| djszapi | Stecchino: but the last reference I showed got stuck during the build. | 15:18 |
| Stecchino | djszapi: well, if it's qemu getting stuck there is not much that can be done either way | 15:19 |
| djszapi | lbt: is it not possible to copy it directly to my repository from local build in this special case ? | 15:19 |
| Stecchino | since the same qemu session should live for the whole dep tree. | 15:19 |
| *** smoku has quit IRC | 15:19 | |
| djszapi | Stecchino: yep, I do not have any idea than pushing it directly to my repository, for the time being :/ | 15:20 |
| *** smoku has joined #harmattan | 15:20 | |
| Stecchino | djszapi: it's quite complicated since the repo's are auto generated. In any case it's better to fix the problem then to circumvent | 15:20 |
| *** smoku has quit IRC | 15:20 | |
| Stecchino | tough I remember packages being imported in OBS repo's before | 15:20 |
| Stecchino | not meego's not c-obs though | 15:20 |
| djszapi | yes, it is better, but without a workaround, it can take ages. | 15:20 |
| djszapi | until it can be used, so kde-mobile is pretty much blocked till that :/ | 15:21 |
| Stecchino | djszapi: will you be at the DS intel-appup event? | 15:22 |
| DocScrutinizer | djszapi: btw I've sent this mail to lpapp kde, nine days ago, after yiu asked me to do so | 15:22 |
| djszapi | Stecchino: theoritically yes, practically not sure :) I have already had one in Tampere :p | 15:22 |
| Stecchino | ah, ok | 15:22 |
| djszapi | Stecchino: so as far as you know, it is not possible to explicitely touch the repository of mine ? | 15:23 |
| Stecchino | it's always possible, but i worry that the manual changes could get overwritten very quickly | 15:23 |
| djszapi | can it get overwritten if there is no successful build for that package ? | 15:24 |
| Stecchino | djszapi: other package builds | 15:25 |
| Stecchino | also not sure it will be picked up by the builds | 15:25 |
| Stecchino | normally it should just install from the packages | 15:25 |
| djszapi | https://build.pub.meego.com/package/live_build_log?arch=armv7el&package=kdelibs&project=home%3Adjszapi&repository=MeeGo_1.2_Harmattan -> Let us see in 3 hours =) | 15:27 |
| djszapi | maybe it requires some round number to suceed, like 100th try :p | 15:28 |
| DocScrutinizer | djszapi: (bash) meanwhile I'm running this very bash binary by # devel-su -c bash | 15:33 |
| DocScrutinizer | djszapi: it works | 15:33 |
| DocScrutinizer | djszapi: but doesn't xplain the core issue of aegis-exec bricking the device | 15:34 |
| DocScrutinizer | djszapi: FYI as I learned it actually matters: I'm logging in to device via ssh root@<IP_of_N950> | 15:36 |
| DocScrutinizer | I meanwhile learned this doesn't yield same permissions as a devel-su | 15:37 |
| djszapi | Stecchino: this is also ++: http://lists.qt.nokia.com/pipermail/qt5-feedback/2011-June/000386.html | 15:38 |
| Stecchino | djszapi: lot's of really valid stuff in that thread | 15:46 |
| *** hardaker has quit IRC | 15:50 | |
| achipa | DocScrutinizer: no. You need a special cert only to flash an R&D firmware (which you wouldn't have outside of Nokia anyway). Open mode should be unrelated to this (i.e. if it's not Nokia signed, we don't care about the cert) | 16:03 |
| MohammadAG | so even if the R&D firmware was leaked it can't be flashed? | 16:05 |
| achipa | correct | 16:06 |
| MohammadAG | and if the cert was leaked? | 16:06 |
| MohammadAG | or is that somewhere in the HW? | 16:06 |
| *** fiferboy has joined #harmattan | 16:07 | |
| Mek | even inside nokia very few people have access to the cert (you can download the cert form a device, but then it is tied to that specific device) | 16:07 |
| djszapi | MohammadAG: yes, it is inside the cal area, hardware and nokia specific | 16:08 |
| MohammadAG | cal can be backed up with flasher | 16:09 |
| MohammadAG | and restored | 16:09 |
| djszapi | libbb5 thingy | 16:09 |
| Mek | but at least for the cert part only on the device you backed it up from | 16:09 |
| achipa | MohammadAG: as said, you can download, but cannot flash it back to a different device | 16:09 |
| *** Venemo has quit IRC | 16:09 | |
| MohammadAG | achipa, I'm sure it can be circumvented :p | 16:09 |
| MohammadAG | write the cal backup of a Nokia device and flash the firmware | 16:10 |
| achipa | no, IIUC this is actually tied to the hardware | 16:10 |
| achipa | i.e. even if you managed to flash it, it just wouldn't boot | 16:11 |
| MohammadAG | anyone feel like risking their job and future employment to provide me with both? :P | 16:12 |
| achipa | (actually, that's a surefire way to permanently brick a device, as in has to go back to the factory type of brick) | 16:12 |
| frals | MohammadAG: job, future employment AND most likely not to bright economic prospects :p | 16:12 |
| MohammadAG | frals, cool, so that's a yes? | 16:13 |
| MohammadAG | :P | 16:13 |
| frals | :p | 16:13 |
| mgedmin | oh, hi frals! someone said I should bug you about fenix not understanding that "Content-Type: text/plain; charset=ANSI_X3.4-1968" means it's ASCII | 16:14 |
| mgedmin | because of this errors from cron scripts show up with an empty body | 16:15 |
| fiferboy | No hi mgedmin, now I'm going to bug you about the fbreader swipe issue :D | 16:15 |
| frals | i know who to bug about it, ill pass it along mgedmin | 16:15 |
| djszapi | DocScrutinizer: send the mail to my archlinux.us address then, I need to leave, will be back later. | 16:15 |
| mgedmin | thanks, frals! | 16:15 |
| mgedmin | fiferboy, I spent half the night yesterday attacking it | 16:15 |
| mgedmin | and the rest trying to make volume keys turn pages | 16:16 |
| mgedmin | I think I know how to achieve the latter | 16:16 |
| fiferboy | mgedmin: And how did that go? | 16:16 |
| *** djszapi has left #harmattan | 16:16 | |
| mgedmin | fbreader gets regular key events, so those can be mapped | 16:16 |
| mgedmin | the trick is to prevent sysuid from also increasing/decreasing the volume | 16:16 |
| mgedmin | and you can do that by acquiring the ScaleButtons resource via libresourceqt | 16:16 |
| fiferboy | Excellent | 16:16 |
| mgedmin | this is just theory, I haven't written any code yet | 16:17 |
| fiferboy | And did you manage to get anywhere with the swipe issue? | 16:17 |
| mgedmin | incidentally, if qmkeyd2 and system-ui weren't open source packages, I'd never have gotten this far | 16:17 |
| mgedmin | re: swipe -- fbreader creates a QPixmap, and a QPainter(on that pixmap), and keeps it for a long time | 16:18 |
| mgedmin | when you swipe out, that QPainter becomes invalid somehow | 16:18 |
| mgedmin | the pixmap gets deleted on the server side maybe? | 16:18 |
| mgedmin | is it backed by an X pixmap, or by a GL texture? | 16:18 |
| MohammadAG | former afaik | 16:18 |
| mgedmin | dunno, lack of Qt debug symbols makes gdb hard | 16:19 |
| fiferboy | It is strange that the pixmap would become invalid, since it still paints in the task manager | 16:19 |
| mgedmin | can I get /usr/lib/debug/lib/armel-linux-gnu/libQtGui.so.4 from somewhere? | 16:19 |
| mgedmin | fiferboy, that's a different pixmap | 16:19 |
| MohammadAG | Try setting Qt::WA_OpaquePaintEvent | 16:19 |
| * mgedmin googles | 16:20 | |
| MohammadAG | setAttribute(Qt::WA_OpaquePaintEvent), no need to google | 16:21 |
| mgedmin | fiferboy, did you see http://pastie.org/2281247 ? | 16:21 |
| mgedmin | MohammadAG, I need to google to figure out what it means, and where I should stick it | 16:21 |
| fiferboy | mgedmin: Yeah, I saw that. Too much assembly for me :) | 16:22 |
| MohammadAG | constructor, reduces repaints, which seems to be something you're doing manually iiuc | 16:22 |
| fiferboy | MohammadAG: Yeah, fbreader uses zlibrary which likes to control everything, painting included | 16:23 |
| Mek | when the app is active, QPixmap is an opengl texture, when it gets de-activated, the app switches to the raster paint engine, so no longer opengl allowed | 16:24 |
| Mek | active app has opengl, inactive app has raster | 16:24 |
| mgedmin | Mek, thank you! | 16:24 |
| mgedmin | I suspected something like that | 16:24 |
| mgedmin | ~QMeegoGraphicsSystem() destroys the shared widget that holds all the GL textures, AFAIU | 16:25 |
| fiferboy | Ah, I am beginning to see | 16:25 |
| mgedmin | so I need to intercept some signal and re-create the QPixmap() | 16:25 |
| frals | mgedmin: check your query | 16:25 |
| w00t | mgedmin: the QPixmap will remain valid, the QPainter probably won't | 16:25 |
| Mek | or maybe use a QImage instead of a QPixmap... | 16:25 |
| mgedmin | https://github.com/geometer/FBReader/blob/master/zlibrary/ui/src/qt4/view/ZLQtViewWidget.cpp#L97 is how fbreader paints | 16:26 |
| mgedmin | https://github.com/geometer/FBReader/blob/master/zlibrary/ui/src/qt4/view/ZLQtPaintContext.cpp#L47 is where QPixmap gets created | 16:26 |
| mgedmin | w00t, thank you for the clue! | 16:26 |
| mgedmin | QPainter is also long-lived | 16:26 |
| mgedmin | more specifically, QPainter's engine | 16:27 |
| mgedmin | so, a cheap fix would be to fix ZLQtPaintContext to call myPainter->begin() and myPainter->end() every time I get a paintEvent | 16:33 |
| fiferboy | mgedmin: Are you going to give that a try? | 16:35 |
| mgedmin | yes | 16:37 |
| mgedmin | technically, I'm at work now... :) | 16:37 |
| fiferboy | techincally so am I... but let me know if you need a tester :) | 16:38 |
| khertan1 | flashing the device on a windows pc host is completely unreliable ! | 16:41 |
| *** achipa has quit IRC | 16:45 | |
| khertan1 | how time does it take for you to do the flash ? | 16:47 |
| *** rm_work has joined #harmattan | 16:47 | |
| *** achipa has joined #harmattan | 16:48 | |
| *** twoboxen has left #harmattan | 16:49 | |
| * fiferboy grumbles about rebooting | 16:54 | |
| fiferboy | BRB | 16:54 |
| *** fiferboy has quit IRC | 16:54 | |
| mgedmin | what happens if I invoke QPainter::end() twice, i.e. the second call has no matching begin()? | 16:57 |
| *** crevetor has joined #harmattan | 16:57 | |
| Mek | you'll get a qWarning("QPainter::end: Painter not active"); | 16:59 |
| *** Wirta has quit IRC | 17:01 | |
| khertan1 | mgedmin: and aegis will ask you to reflash your device | 17:02 |
| khertan1 | :) | 17:02 |
| lcuk | lol | 17:03 |
| lcuk | khertan1, why are you flashing yours? | 17:03 |
| khertan1 | aegis :) | 17:03 |
| lcuk | i reflashed quite easily from windows | 17:04 |
| lcuk | after innocently removing most packages from my build | 17:04 |
| *** harbaum has quit IRC | 17:04 | |
| * mgedmin tries http://pastie.org/2284660 | 17:05 | |
| lcuk | aegis never stopped that one! | 17:05 |
| khertan1 | yep but replacing ssl libs does :) | 17:05 |
| *** qgil has joined #harmattan | 17:05 | |
| lcuk | lol | 17:05 |
| khertan1 | + many other things | 17:05 |
| khertan1 | i ve play trying to understand why exchange for email sync never end | 17:07 |
| khertan1 | oh ... nokia ovi account seems working now | 17:07 |
| mgedmin | hm, fbreader still segfaults | 17:09 |
| mgedmin | but this time in a different place | 17:10 |
| *** fcrochik has joined #harmattan | 17:10 | |
| SpeedEvil | Progress! | 17:10 |
| lcuk | mgedmin, I thought somebody made a qml reader interface? | 17:10 |
| lcuk | not fbreader though | 17:10 |
| mgedmin | dunno, lcuk | 17:10 |
| lcuk | but something with fancy pants page turning | 17:10 |
| mgedmin | link me if you find it please | 17:10 |
| lcuk | hold on | 17:10 |
| hiemanshu | mgedmin: google reader? | 17:11 |
| SpeedEvil | Fbreader core has some useful features. | 17:11 |
| mgedmin | I've seen blog posts about an ebook reader on planet maemo, I think... | 17:11 |
| SpeedEvil | It can read a slew of book formats. | 17:11 |
| hiemanshu | ah ebook reader | 17:11 |
| SpeedEvil | Do library managment, ... | 17:11 |
| mgedmin | fbreader's library management is crap | 17:11 |
| mgedmin | vi /path/to/book.html to fix a typo -> fbreader loses all metadata | 17:11 |
| lcuk | [MeeGo-community] Starting a tablet book reader project http://lists.meego.com/pipermail/meego-community/2011-March/003653.html | 17:11 |
| mgedmin | because the file size or mtime changed or something | 17:12 |
| lcuk | and there was a ui created | 17:12 |
| *** fcrochik has quit IRC | 17:12 | |
| *** fcrochik has joined #harmattan | 17:12 | |
| mgedmin | ooh, another list I ought to maybe subscribe to... | 17:12 |
| * lcuk is happiest with plaintext books and simplicity | 17:12 | |
| * SpeedEvil reads a _lot_ on the n900 | 17:13 | |
| lcuk | i read more on the n810 | 17:13 |
| lcuk | i read a lot on my ideapad meego instance | 17:13 |
| mgedmin | problem with plaintext books: no italics, no place to put <!-- ??? is this really a word --> comments | 17:13 |
| SpeedEvil | Nokia-N900:/home/user/MyDocs/ebooks# du -s | 17:13 |
| SpeedEvil | 186784. | 17:13 |
| SpeedEvil | Grep. | 17:14 |
| SpeedEvil | Being unable to search dead trees annoys me. | 17:14 |
| SpeedEvil | Especially with casts of thousands. | 17:14 |
| lcuk | SpeedEvil, how many .txt? | 17:14 |
| lcuk | or translatable to using various means | 17:15 |
| SpeedEvil | Being able to click a char that is reintroduced 200 pages and have you taken back to the introduction is awesome. | 17:15 |
| SpeedEvil | Most html or lit or epub | 17:15 |
| * mgedmin has a fancy shmancy python script to convert most formats to .txt or .html | 17:15 | |
| lcuk | how did you buy so many ebooks? | 17:16 |
| mgedmin | because baen free library had weirdly broken epubs | 17:16 |
| *** fiferboy has joined #harmattan | 17:16 | |
| SpeedEvil | lcuk: baen free library too. | 17:16 |
| mgedmin | fiferboy, bad news: the fix was incomplete, I still get a segfault in a different place | 17:16 |
| SpeedEvil | lcuk: But also fictionwise. | 17:16 |
| khertan1 | lol on usb fat transfer, take same time to transfer 5 Mb made of 1400 elements than 2Gb made of 40 elements | 17:16 |
| khertan1 | :) | 17:16 |
| mgedmin | iops kill | 17:17 |
| SpeedEvil | lcuk: And some are bookz versions of titles I own in paper, but can't actually find. | 17:17 |
| SpeedEvil | I think I've mentioned I'm slowly churning through ~6K books to list them on amazon. | 17:17 |
| mgedmin | fiferboy, new segfault: http://pastie.org/2284708 | 17:18 |
| mgedmin | I suppose it's not enough to call QPainter::end(), I must create a new QPainter() every time? | 17:18 |
| mgedmin | Mek:^ ? | 17:19 |
| fiferboy | Hmm, still a painter issue | 17:19 |
| Mek | I'm not sure, but yeah, that's what I suspect... | 17:19 |
| mgedmin | QPaintDevice is the QPixmap, I suspect | 17:20 |
| *** fcrochik has quit IRC | 17:20 | |
| rcg1 | mgedmin: ".. Construct and customize (e.g. set the pen or the brush) the painter. Then draw. .." | 17:20 |
| rcg1 | thats what the api docs say.. so i'd say: yes you need to create one each time | 17:21 |
| rcg1 | thats at least what i do for my custom widgets.. | 17:21 |
| mgedmin | will try that then | 17:21 |
| rcg1 | dunno actually where i got that info from to that time but there had to be some good reason to do so ;) | 17:21 |
| qgil | hi, does anybody know how to setup a font.family to be used by default in all the stext strings of my app, unless I define something different for a specific item? Basically, I want to make sure Nokia Pure is used always | 17:21 |
| qgil | this is for a Qt Quick app | 17:22 |
| vandenoever | qgil: you get set the font in the qApp | 17:22 |
| vandenoever | that's on the c++ side | 17:22 |
| vandenoever | qApp->setFont(comicsans); | 17:23 |
| qgil | vandenoever: you mean this main.cpp_ | 17:23 |
| vandenoever | qgil: yes | 17:23 |
| qgil | ah, makes sense. uuuhhh first time ever I touch a .cpp <( | 17:23 |
| qgil | <( | 17:23 |
| vandenoever | qgil: they dont bite too much if you stay away from the macros | 17:23 |
| qgil | ah, for some reason my ES keyboard behaves like EN / thank you freshly updates Ubuntu | 17:24 |
| vandenoever | qApp->setFont(QFont("comicsans")); | 17:24 |
| qgil | ;) now | 17:24 |
| qgil | thanks vandenoever | 17:24 |
| vandenoever | qgil: i just did the same yesterday to get a larger font in warptree, so it was fresh | 17:25 |
| Arkenoi | ha, reflashing n950 erases all mass storage partitions :-) | 17:28 |
| mgedmin | valgrind of fbreader: http://pastie.org/2284745 | 17:28 |
| Stskeeps | if you use OCF | 17:28 |
| mgedmin | Arkenoi, I was just wondering about that | 17:28 |
| qgil | vandenoever: do I need to put this setFont inside the int function or just out of it, at the end? | 17:30 |
| npm | re: qgil Nokia Pure font question... so what font name do i use in "qApp->setFont(QFont("comicsans"));" to get those fonts? | 17:31 |
| qgil | npm: "get" you mean use in your app or really get? | 17:31 |
| qgil | npm: the Nokia pure font is supposed to be already in the device, availableç | 17:32 |
| Tronic | Btw, I tried Nokia Pure on N900 and desktop Linux but the kerning doesn't seem right. | 17:33 |
| npm | qgil: use in my app | 17:33 |
| Tronic | Aren't combinations such as "co" supposed to go very closely together, yet I get rather large spacing there. | 17:33 |
| lcuk | Tronic, how did you install the Pure font on random other devices? | 17:33 |
| Tronic | Haven't paid attention to that on N950. | 17:33 |
| vandenoever | qgil: you can do it on line 2 of main() | 17:33 |
| npm | so i just meant the actual name of some of those fonts that you find nice and useful for harmattan devel | 17:34 |
| Tronic | lcuk: They provided a TTF file download at some point. | 17:34 |
| lcuk | Tronic, under what license? | 17:34 |
| Tronic | No idea. | 17:34 |
| lcuk | because Jukka and I both actively pondered it | 17:34 |
| lcuk | for meego and n900-ce use | 17:34 |
| lcuk | afaik the license said use only after written license request | 17:35 |
| lcuk | so I half heartedly pondered a hand written request letter to nokia for it | 17:35 |
| npm | speaking of fonts, the "Installed Fonts" demo in qtzibit wins my contest for javascript hackery.... http://qtzibit.googlecode.com/svn/trunk/exhibit/src/webapp/examples/InstalledFonts/installed-fonts.html | 17:35 |
| npm | http://sites.google.com/site/exhibitecmanaut/installed-fonts.html | 17:35 |
| vandenoever | qgil: actually it's (slightly) cleaner to do "QApplication app(argc, argv);\n app.setFont();" | 17:35 |
| qgil | vandenoever: right, will do | 17:36 |
| Tronic | lcuk: http://taskumuro.com/nokia-uusii-brandiaan-puhelimiin-uusi-fontti-ohjeet-kayttoonottoon | 17:36 |
| qgil | btw, to get a list of fonts available what I'm doing is to use this Qt Quick toolbar, then whenever you type font.family: "" in Qt Creator a nice drop down menu shows you what is available e.g. font.family: "Nokia Pure Headline" | 17:36 |
| *** meegoexperts has quit IRC | 17:37 | |
| lcuk | for scratchbox, does the harmattan target sit happily alongside fremantle ? | 17:38 |
| lcuk | Tronic, thanks but I was thinking a more general MeeGo usecase | 17:38 |
| qgil | mmm... there might be a different name for these fonts for QML and C++ since "Nokia Pure Text" appears in green in the QML file but in plain black in the function vandenoever posted here... | 17:39 |
| npm | qgil: ah... sounds like you have a newer version of qtreator than i | 17:40 |
| rcg1 | lcuk: it does | 17:40 |
| qgil | ... but in green at setFont(QFont("Nokia Pure Text")); :) | 17:40 |
| lcuk | rcg1, awesome | 17:40 |
| rcg1 | lcuk: though i set up the harmattan targets manually from inside scratchbox | 17:40 |
| qgil | npm: if you are targetting Harmattan then getting an SDK up to date with all the Experimental stuff is in order | 17:40 |
| rcg1 | lemme see if i find the page with the instructions | 17:40 |
| lcuk | thx | 17:41 |
| vandenoever | QFontDatabase::applicationFontFamilies is helpful on the c++ side, perhaps there's qml equivalent | 17:41 |
| * npm i'll have whatever font quim's having.... | 17:41 | |
| *** djszapi has joined #harmattan | 17:42 | |
| rcg1 | lcuk: http://www.developer.nokia.com/Community/Wiki/Harmattan:Platform_Guide/Troubleshooting_Harmattan_Platform_SDK/Installing_Harmattan_Platform_SDK_manually | 17:42 |
| khertan1 | too bad can't select what restore from backup | 17:43 |
| rcg1 | lcuk: though, as i am on gentoo i installed/updated the scratchbox stuff via portage and then started at http://www.developer.nokia.com/Community/Wiki/Harmattan:Platform_Guide/Troubleshooting_Harmattan_Platform_SDK/Installing_Harmattan_Platform_SDK_manually#Creating_and_configuring_targets | 17:43 |
| khertan1 | as it s just simple zip file ... i m asking what will happen if i remove some from the backup file | 17:44 |
| khertan1 | :) | 17:44 |
| *** andrunko has joined #harmattan | 17:45 | |
| hiemanshu | gah, I'll have to setup everything again :( | 17:46 |
| rcg1 | lcuk: if you already have scratchbox installed it should be safe to skip the user and group creation/configuration steps | 17:48 |
| Stskeeps | hiemanshu: i accidentially flashed my rootfs earlier, that was painful | 17:48 |
| Stskeeps | :P | 17:48 |
| hiemanshu | Stskeeps: I rm -rf'ed / on my desktop by mistake :/ | 17:49 |
| hiemanshu | Stskeeps: rm -rf / home/hiemanshu/* :P | 17:49 |
| Stskeeps | heh | 17:49 |
| Stskeeps | that's why i backup :P | 17:49 |
| hiemanshu | I have backups | 17:49 |
| hiemanshu | but I'll have to QtSDK, scratchbox and what not | 17:49 |
| hiemanshu | setup* | 17:49 |
| khertan1 | yeah stupid * is too near than / on keyboard | 17:49 |
| lcuk | rcg1, just going to try the default sdk installer | 17:49 |
| hiemanshu | khertan1: naah, I added a space after / | 17:49 |
| lcuk | for fremantle at least on ubuntu it worked cleanly | 17:50 |
| *** willer_ has joined #harmattan | 17:50 | |
| lcuk | hope it does again | 17:50 |
| qgil | vandenoever: your command right below the #includes produce this error: | 17:50 |
| lcuk | but not drastic if it doesnt | 17:50 |
| rcg1 | lcuk: hehe ic.. someone told me this should also work.. ;) | 17:50 |
| khertan1 | lol removing something from the backup when restoring didn't mean you will keep current settings | 17:50 |
| qgil | error: expected constructor, destructor, or type conversion before '(' token | 17:50 |
| khertan1 | no ... it s just reset and restore nothing | 17:50 |
| vandenoever | qgil: yes, the command should go in the function "int main(int argc, char** argv) { ..." | 17:51 |
| qgil | ah ok, sorry I misunderstood | 17:51 |
| khertan1 | grrrr connecting to nokia account works well after flash | 17:52 |
| khertan1 | didn't after restore | 17:52 |
| vandenoever | qgil: http://paste.debian.net/124378/ | 17:52 |
| vandenoever | qgil: that's from an app where i wanted a different font size, but keep the font family | 17:53 |
| hiemanshu | my FB feed stopped updating on the 21st, the call dialog vanished a day later, and other random stuff happening for me :P | 17:53 |
| lcuk | haha hm | 17:53 |
| lcuk | the clock lost a digit | 17:53 |
| hiemanshu | yeah that too | 17:53 |
| lcuk | "15:52 wa"s showing as "15:5" | 17:53 |
| hiemanshu | I had the time show up as __:__ once | 17:54 |
| hiemanshu | it was all 1 :P | 17:54 |
| fiferboy | hiemanshu: Am I correct in my recollection that you have run a plain Qt program on the N950? | 17:55 |
| hiemanshu | fiferboy: yes | 17:55 |
| khertan1 | lcuk: at 11:11 this morning i got 1 : | 17:55 |
| fiferboy | hiemanshu: Did you have any issues with the program when you swiped it ou? | 17:55 |
| fiferboy | out? Did it show up in the task view? | 17:56 |
| hiemanshu | fiferboy: nope, runs just fine | 17:56 |
| hiemanshu | yes it id | 17:56 |
| hiemanshu | it did* | 17:56 |
| khertan1 | lcuk: sometime also specially when trying to connect to account i see clock going back 1 min ... then going to right time and going back ... | 17:56 |
| lcuk | bug filing time | 17:56 |
| fiferboy | hiemanshu: Thanks for the info, fbreader qt4 is having issues with that | 17:56 |
| *** mikhas has joined #harmattan | 17:56 | |
| mikhas | o/ | 17:57 |
| hiemanshu | fiferboy: everything works just fine, except for the part that it looks ugly | 17:57 |
| khertan1 | arg carrefull backup didn't backup contact image !! | 17:57 |
| khertan1 | s/image/picture | 17:57 |
| qgil | vandenoever: still an error, a different one: | 17:57 |
| qgil | error: 'helvetica' was not declared in this scope (it doesn't matter what font name I put, the same error comes up) | 17:58 |
| Stskeeps | is it in quotes? | 17:58 |
| vandenoever | qgil: you need double quotes around strings in C++ | 17:59 |
| mikhas | First QML developer, now C++ developer. What will be next!? | 18:00 |
| vandenoever | qgil: you could also use QFont::Helvetica like this: | 18:01 |
| hiemanshu | mikhas: android developer? | 18:01 |
| mikhas | weeeeeell | 18:01 |
| Stskeeps | html5 ; | 18:01 |
| Stskeeps | ) | 18:01 |
| qgil | now I get other errors but I must join a call :) | 18:01 |
| mikhas | vandenoever, wont help - there is no enum yet for Nokia Pure font | 18:01 |
| vandenoever | QFont helvetica; helvetica.setStyleHint(QFont::Helvetica); app.setFont(helvetica); | 18:01 |
| vandenoever | mikhas: well, qgil had 'helvetica' in his err message | 18:02 |
| mikhas | yes, he was trying to mislead you ;-) | 18:02 |
| vandenoever | using QStyleHint will choose the preferred font in that style | 18:02 |
| qgil | because I wanted to make sure the problem was not with Nokia Pure Text :) | 18:02 |
| qgil | ok, the experts meet, I can just comment that piece of code and move forward :) | 18:03 |
| vandenoever | choosing a style is a good idea, because the fallback on a system with no nokia pure is nicer | 18:03 |
| djszapi | Elleo: heh, we also made a duplication again. I also packaged vim yesterday :) | 18:13 |
| djszapi | damn duplication all the time... | 18:13 |
| Elleo | djszapi: ah, strange I search obs for vim before starting that package and it didn't show anything | 18:13 |
| Elleo | searched* | 18:14 |
| djszapi | Stecchino: https://build.pub.meego.com/package/live_build_log?arch=armv7el&package=kdelibs&project=home%3Adjszapi&repository=MeeGo_1.2_Harmattan | 18:14 |
| Elleo | oh well, I'll just switch to using your package | 18:14 |
| djszapi | Elleo: Yeah, I could not upload because OBS was broken here. | 18:14 |
| djszapi | and then I did not try later. | 18:14 |
| Elleo | ah | 18:14 |
| Elleo | is your package online anywhere? | 18:14 |
| djszapi | I have just wanted to upload | 18:15 |
| djszapi | I basically took the internal harmattan packaging thing. I see you used more like the upstream debian way. | 18:15 |
| Elleo | yeah | 18:16 |
| Elleo | although it seems obs has a bug in its dependency checking | 18:16 |
| Elleo | since it didn't realise that libgpmg1-dev wasn't available prior to trying to build | 18:16 |
| Elleo | presumably because it uses the [linux-any] versioning tag | 18:16 |
| djszapi | that is scary for OBS yeah | 18:17 |
| djszapi | do not use things like that | 18:17 |
| Elleo | I should have removed that as a dep anyway though | 18:17 |
| Elleo | since it's just needed for some of the gui packages (which I disabled) | 18:17 |
| Elleo | oh no, that's needed for mouse support (also disabled) | 18:18 |
| djszapi | I upload my package since it has been like that for ages internally, so it is proven | 18:19 |
| Elleo | okay, cool | 18:19 |
| djszapi | but I am fine if it is not used :) | 18:19 |
| Elleo | I don't have any particular desire to maintain a vim package, I just want vim available ;) | 18:19 |
| djszapi | heh, OBS still fails by uploading.. | 18:19 |
| djszapi | yeah, it would be nice if someone could maintain the packages I need.. | 18:20 |
| Elleo | djszapi: have you tried via osc? | 18:20 |
| Elleo | I'm not having any trouble uploading with osc | 18:20 |
| *** andrunko has quit IRC | 18:22 | |
| djszapi | Elleo: ok, it worked. | 18:24 |
| djszapi | we should really make a shared repository btw | 18:25 |
| *** andrunko has joined #harmattan | 18:25 | |
| mgedmin | Elleo, use packrat to find packages | 18:28 |
| mgedmin | ~packrat | 18:28 |
| infobot | extra, extra, read all about it, packrat is a cute little rodent that keeps everything it finds. http://www.ageofikon.com/prh/ | 18:28 |
| djszapi | mgedmin: mmh, it is nice page, thanks ! | 18:30 |
| lcuk | am I right in thinking aegis will prevent re-installation of: | 18:30 |
| lcuk | xserver-xorg-video-fbdev-0.4.0 | 18:30 |
| *** slaine has quit IRC | 18:30 | |
| Elleo | mgedmin: neat | 18:30 |
| djszapi | lcuk: nope | 18:31 |
| djszapi | lcuk: if you /really/ have a serious reason for that, you can use "--force-all". | 18:31 |
| lcuk | djszapi, ahh right so I can do some tinkering with that package and reinstall? | 18:31 |
| lcuk | I thought system packages would not be updatable | 18:32 |
| djszapi | lcuk: I do not recommend it... | 18:32 |
| lcuk | i am of course assuming I can build it at all | 18:32 |
| lcuk | djszapi, well i have to | 18:32 |
| lcuk | bug 13084 is assigned to me | 18:32 |
| MeeGoBot | Bug https://bugs.meego.com/show_bug.cgi?id=13084 maj, Medium, ---, liquid, ASSI, [n900] Horizontal tearing with xvimagesink | 18:32 |
| lcuk | and it impacts harmattan also | 18:32 |
| lcuk | so I have to do some debugging and testing | 18:32 |
| * lcuk hasn't a clue how else to try and fix it | 18:33 | |
| djszapi | mmh, there is no "irssi" package yet. | 18:34 |
| mgedmin | is there screen? | 18:34 |
| mikhas | lcuk, uses xrender, right? | 18:34 |
| mgedmin | or should I use tmux instead? | 18:34 |
| djszapi | there is screen | 18:35 |
| lcuk | mikhas, not sure, problem exhibits itself with xvideo mode | 18:35 |
| lcuk | i can only speak for liqbase because that is where I encounter it, the n900-ce has same problem with default movie player | 18:35 |
| djszapi | mgedmin: http://repo.pub.meego.com/home:/djszapi/MeeGo_1.2_Harmattan/armel/screen_4.0.3-14_armel.deb | 18:35 |
| mikhas | well, try to find it out whether it does use xrender | 18:36 |
| mikhas | that's most likely the culprit | 18:36 |
| * mgedmin should add fbreader's icon to debian/control | 18:36 | |
| lcuk | mikhas, liqbase does not directly use it | 18:36 |
| mikhas | no, but the libs you use | 18:36 |
| mikhas | it all goes through the same graphics pipeline, eventually | 18:37 |
| lcuk | libX11-devel libXv-devel | 18:37 |
| djszapi | Elleo: I need a package actually if you have time for packaging :p | 18:38 |
| lcuk | this bug is relating to the yuv overlay | 18:38 |
| lcuk | not a normal x11 window | 18:38 |
| djszapi | http://www.ageofikon.com/prh/index.php?Action=list&System=.%2F&Arg=kde&Section=&Repo=0 -> This is interesting. It seems I have kde packages, but I am not aware of that fact.. | 18:40 |
| Elleo | djszapi: not got much time right now, but I could maybe take a look later, what're you after? | 18:41 |
| djszapi | polkit-qt | 18:42 |
| Elleo | okay, I'll see what I can do this evening | 18:42 |
| lardman | are harmattan apps supposed to be optified? | 18:48 |
| mgedmin | no, afaiu | 18:49 |
| mgedmin | let me look at df | 18:49 |
| Elleo | lardman: nope | 18:49 |
| Elleo | is one big 4gb partition | 18:49 |
| Elleo | so sticking stuff in opt won't make any difference | 18:50 |
| mgedmin | 4 GB /, 2 GB /home, 8.8 GB /home/MyDocs | 18:50 |
| lardman | just that I have some optification stuff in my .pro file and I thought it had been created for harmattan | 18:50 |
| *** M4rtinK2 has joined #harmattan | 18:50 | |
| mgedmin | /opt lives on the root partition, so optification is not going to hurt anything much | 18:50 |
| lcuk | Elleo, just wait until Gb range games are installed | 18:50 |
| *** achipa has quit IRC | 18:50 | |
| lardman | np and thanks, I guess I must have selected a maemo5 target at some point too then | 18:50 |
| mgedmin | if you plan to backport your app to maemo5, maybe that's useful | 18:50 |
| lcuk | then tell us optification isn't needed! | 18:50 |
| lardman | mgedmin: perhaps, should be done automatically thouhh | 18:51 |
| mgedmin | someone said Qt Creator places binaries in /opt/AppName/bin | 18:51 |
| mgedmin | but doesn't create symlinks in /usr/bin | 18:51 |
| lcuk | are qml apps bigger or smaller than similar native apps? | 18:51 |
| mgedmin | which makes terminal lovers like me unhappy -- can't gdb appname, must gdb /full/path/to/bloody/binary/appname | 18:52 |
| DocScrutinizer | lardman: no optification on harmattan | 18:52 |
| mgedmin | lcuk, I thought qml apps _were_ the native apps for harmattan | 18:52 |
| DocScrutinizer | lardman: there's not even /opt iirc | 18:52 |
| lcuk | mgedmin, elf loader currently does not support directly running *.qml | 18:52 |
| lcuk | therefore not a native app yet :P | 18:53 |
| mgedmin | DocScrutinizer, the /opt directory exists, but there's no separate /opt partititon | 18:53 |
| DocScrutinizer | though optified pkgs will install just fine, adding cruft to the system | 18:53 |
| mgedmin | lcuk, Qt Creator compiles those .qmls into the main binary somehow... | 18:53 |
| *** khertan1 has left #harmattan | 18:53 | |
| DocScrutinizer | mgedmin: I can't check anymore as I already installed optified pkg from fremantle :-) | 18:53 |
| mgedmin | I'd hoped I could vi *.qml of an installed package to play with javascript expressions, but no... | 18:53 |
| lcuk | mgedmin, therefore the main native app is still binary | 18:54 |
| mgedmin | yes | 18:54 |
| lcuk | just because you used qml to describe the ui | 18:54 |
| djszapi | lbt: How can I delete a package from c-obs, harmattan target ? | 18:54 |
| lcuk | ok, are qml files created with whizbang flashy graphics generally bigger than their oldschool counterpart? | 18:54 |
| hiemanshu | mgedmin: then dont use the .qrc file, instead call the qmls directly? | 18:54 |
| mgedmin | they what? | 18:55 |
| hiemanshu | well the qmls are stored in the .qrc file, which compiles into the binary, instead call em directly | 18:55 |
| hiemanshu | I remember reading a post about this somewhere | 18:55 |
| mgedmin | I think you're right, hiemanshu | 18:56 |
| mgedmin | strings /opt/QMLCompass/bin/QMLCompass shows me the contents of the QML files | 18:56 |
| mgedmin | and the SVG files | 18:56 |
| *** andrunko has quit IRC | 18:56 | |
| hiemanshu | mgedmin: http://wiki.meego.com/Porting_Fremantle_Applications_to_Harmattan#Store_QML_in_Filesystem_instead_of_Resources | 18:57 |
| mgedmin | ooh, thanks! | 18:58 |
| *** crevetor has quit IRC | 18:59 | |
| *** timoph has joined #harmattan | 18:59 | |
| *** crevetor has joined #harmattan | 18:59 | |
| lardman | bbl, power going off in the office soon | 19:05 |
| *** lardman has quit IRC | 19:05 | |
| djszapi | Elleo: How about putting stuff to this repository ? https://build.pub.meego.com/project/show?project=home%3Arzr | 19:18 |
| djszapi | I am thinking of that I would discuss it with that guy. | 19:18 |
| djszapi | He also has a lot of Harmattan packages. | 19:18 |
| *** meegoexperts has joined #harmattan | 19:18 | |
| Elleo | djszapi: I'd have thought it'd be better to have a separate project to push to, so that rzr still has the freedom to do experimental stuff in his repo without worrying about breaking everyone's system | 19:20 |
| djszapi | he can actually still do that | 19:21 |
| DocScrutinizer | djszapi: you asked me to resend the mail to another addr? | 19:21 |
| djszapi | the thing is that we did not get community account | 19:21 |
| djszapi | so there is no other way than doing at someone's place. | 19:21 |
| DocScrutinizer | (community repo) MohammadAG said he'd plan to create that community repo. Wonder what's up with that | 19:23 |
| DocScrutinizer | last thing I heard was he can't log in to OBS | 19:23 |
| DocScrutinizer | think that's been 2 days ago | 19:24 |
| *** achipa has joined #harmattan | 19:25 | |
| *** achipa has quit IRC | 19:25 | |
| *** achipa has joined #harmattan | 19:25 | |
| DocScrutinizer | http://mg.pov.lt/harmattan-irclog/%23harmattan.2011-07-25.log.html#t2011-07-25T23:34:47 | 19:26 |
| *** seif has quit IRC | 19:39 | |
| *** divan|work has quit IRC | 19:39 | |
| *** andrunko has joined #harmattan | 19:39 | |
| *** aleksander_m has quit IRC | 19:43 | |
| *** piggz has joined #harmattan | 19:46 | |
| *** divan|work has joined #harmattan | 19:46 | |
| *** cpscotti has quit IRC | 19:51 | |
| *** hardaker has joined #harmattan | 19:51 | |
| *** divan|work has quit IRC | 19:59 | |
| *** Ronksu has joined #harmattan | 20:01 | |
| *** crevetor has quit IRC | 20:20 | |
| mgedmin | so, n950 usb tethering works like a charm | 20:20 |
| macmaN | mgedmin: yeah i did bt dun | 20:21 |
| macmaN | while on 900, bt PAN worked very well | 20:21 |
| mgedmin | oh, cool | 20:21 |
| mgedmin | haven't tried bt dun, don't even know if network manager in ubuntu natty supports it | 20:21 |
| macmaN | i guess now that i have ppp installed, dun would work too | 20:21 |
| mgedmin | bluetooth is like the least reliable tech *ever* | 20:21 |
| macmaN | im sure it would | 20:21 |
| macmaN | youre saying you did bt pan on n950? | 20:22 |
| macmaN | with bnep0 and all? | 20:22 |
| *** crevetor has joined #harmattan | 20:22 | |
| alterego | works for me | 20:22 |
| mgedmin | "USB tethering" means no BT :) | 20:22 |
| macmaN | ahhahhaha | 20:22 |
| macmaN | lol | 20:22 |
| macmaN | ewps | 20:22 |
| rm_work | i'm curious about BT tethering though | 20:22 |
| rm_work | in ubuntu | 20:22 |
| rm_work | if it works that'd be awesome | 20:22 |
| mgedmin | I think I saw a howto once | 20:22 |
| rm_work | it never really worked for me on N900 | 20:22 |
| rm_work | i tried :/ | 20:22 |
| rm_work | usb is fine tho | 20:23 |
| mgedmin | there's no UI for it, you have to install some random 3rd party package with crappy UI | 20:23 |
| macmaN | rm_work: works fine for me | 20:23 |
| rm_work | since i want to charge it while raping the battery for 3G constantly anyway :P | 20:23 |
| mgedmin | only then network manager sees your BT PAN/DUN device and lets you connect | 20:23 |
| mgedmin | I wish my n950 acted as a wifi access point | 20:23 |
| rm_work | lol yeah, WAP would be sweet | 20:23 |
| macmaN | mgedmin: yes, there's a btpan setup script involved on n900 | 20:23 |
| rm_work | did the wireless chipset change to something with WAP capability?> | 20:24 |
| rm_work | or not | 20:24 |
| macmaN | mgedmin: n9 has joiku's custom hotspot solution preinstalled | 20:24 |
| macmaN | im sure its there in next fw for n950 | 20:24 |
| rm_work | hrm | 20:24 |
| rm_work | so could we NOT use joiku, but just make our own network? | 20:24 |
| mgedmin | joikuspot on a n900 does a real AP, or ad-hoc wifi? | 20:24 |
| rm_work | IIRC joiku requires accounts and such? | 20:24 |
| *** zehjotkah has joined #harmattan | 20:24 | |
| macmaN | mgedmin: hm, that i dont | 20:25 |
| mgedmin | I'd a shell script on my n900 that set up adhoc wifi with dhcp and everything; the only thing I didn't have was NAT | 20:25 |
| macmaN | rm_work: want the bt setup script im using? | 20:25 |
| mgedmin | I was forced to use ssh as a proxy server | 20:25 |
| rm_work | oh damn i need to get lunch bbl | 20:25 |
| rm_work | nah don't use N900 anymore anyway | 20:25 |
| macmaN | ok | 20:25 |
| macmaN | its off tmo really | 20:25 |
| macmaN | i guess lmk whenever | 20:25 |
| rm_work | k | 20:25 |
| rm_work | ack, missing lunch | 20:25 |
| rm_work | bbl | 20:25 |
| zehjotkah | hi everyone | 20:31 |
| Stskeeps | lo zehjotkah | 20:33 |
| zehjotkah | how are you doing | 20:33 |
| Stskeeps | good good | 20:33 |
| *** fcrochik has joined #harmattan | 20:37 | |
| *** harbaum has joined #harmattan | 20:38 | |
| *** djszapi has left #harmattan | 20:40 | |
| *** Scifig has joined #harmattan | 20:45 | |
| leinir | yo zehjotkah :) you gonna be at c-base next Friday for the pre-desktopsummit-start-party thing? :) | 20:45 |
| zehjotkah | maybe | 20:45 |
| zehjotkah | next week I'll be in a different city | 20:46 |
| zehjotkah | but I'll be back on friday | 20:46 |
| zehjotkah | but we'll do a MeeGo Hack Sonntach | 20:46 |
| zehjotkah | (MeeGo Hacking Sunday) | 20:46 |
| *** djszapi has joined #harmattan | 20:46 | |
| zehjotkah | at the 14th of august | 20:47 |
| zehjotkah | which will be announced at the desktop summit by intel | 20:47 |
| *** [XeN] has joined #harmattan | 20:47 | |
| leinir | Nicely :) | 20:47 |
| [XeN] | blubs | 20:47 |
| leinir | In particular because there's the plasma active guys who, i'm pretty sure, would enjoy joining that :) | 20:47 |
| zehjotkah | okay | 20:48 |
| *** mikhas has quit IRC | 20:48 | |
| Mek | although anything that requires people to stay extra days but isn't pre-announced is probably not likely to attract many people | 20:48 |
| zehjotkah | let me check the time when the pre summit thingi will take place | 20:48 |
| zehjotkah | Mek, yes, that's a problem | 20:49 |
| zehjotkah | leinir, when on friday will this eveent be at the c-base? | 20:50 |
| zehjotkah | (btw. I'm @ the c-base atm... ;)) | 20:50 |
| djszapi | DocScrutinizer: Did you mean this bash ? http://harmattan-dev.nokia.com/pool/harmattan-beta/free/b/bash/ | 20:50 |
| leinir | zehjotkah: Hehe, yeah, that's why i thought to ask ;) You're always there ;) | 20:50 |
| DocScrutinizer | djszapi: err, on first glance yes | 20:51 |
| zehjotkah | at which tiome the event will be? | 20:53 |
| zehjotkah | there is no info at the cbase site | 20:53 |
| DocScrutinizer | djszapi: http://paste.debian.net/124359/ | 20:53 |
| djszapi | zehjotkah: which event, next Friday, pre-registration ? | 20:54 |
| zehjotkah | yes | 20:54 |
| leinir | zehjotkah: https://desktopsummit.org/program/pre-registration :) | 20:55 |
| zehjotkah | thanks | 20:55 |
| zehjotkah | yes, I think I could be there... | 20:55 |
| DocScrutinizer | zehjotkah: 14.8. collides with camp | 20:56 |
| zehjotkah | yes, ... | 20:56 |
| zehjotkah | intel asked me to do the event | 20:56 |
| zehjotkah | to help all the folks who will receive an ExoPC | 20:57 |
| zehjotkah | during the desktop summit | 20:57 |
| DocScrutinizer | just mentioning | 20:57 |
| zehjotkah | to install the latest MeeGo version and install the SDK and so on... | 20:57 |
| macmaN | djszapi: umm, that looks new | 20:57 |
| djszapi | sorry ? | 20:57 |
| macmaN | or what exactly have i not been able to apt-get it | 20:58 |
| *** Rizzer has quit IRC | 20:58 | |
| macmaN | why* | 20:59 |
| mgedmin | macmaN, bash? | 21:07 |
| macmaN | yes | 21:07 |
| mgedmin | did you add the harmattan-dev repo to your sources.list? | 21:07 |
| macmaN | that could be a good reason :> | 21:07 |
| mgedmin | also, be aware that DocScrutinizer encountered a happy fun aegis device lock while playing with bash | 21:07 |
| DocScrutinizer | it was on playing with aegis-exec actually | 21:08 |
| DocScrutinizer | bash is just a random binary I tried to start with augmented permissions | 21:08 |
| DocScrutinizer | using aegis-exec | 21:08 |
| macmaN | yeah | 21:09 |
| macmaN | bash itself should be ok | 21:09 |
| DocScrutinizer | I'm running devel-su -c bash now and that works reasonably good | 21:09 |
| *** smoku has joined #harmattan | 21:09 | |
| DocScrutinizer | even executes >> random.py | 21:10 |
| DocScrutinizer | if random.py is set +x | 21:10 |
| DocScrutinizer | i.e. no need to >> python random.py | 21:11 |
| macmaN | mhm | 21:11 |
| *** [XeN] has quit IRC | 21:11 | |
| macmaN | hmm how do you choose what meego-terminal executes for shell | 21:13 |
| zehjotkah | anyone knows why alien is not saving the created .deb file? | 21:15 |
| DocScrutinizer | I'd bet it has a hardcoded /bin/sh | 21:15 |
| zehjotkah | I'm trying to convert Opera Mobile for MeeGo ARM from .rpm to .deb | 21:15 |
| djszapi | DocScrutinizer: why do you need for aegis-exec for that bash execution ? | 21:15 |
| Stskeeps | zehjotkah: rpm2cpio *.rpm | cpio -id and dpkg-deb -c is what you need | 21:17 |
| Stskeeps | i think | 21:17 |
| leinir | Hmm... anybody know how to just fire off a sound effect, as in a system sound effect? | 21:18 |
| DocScrutinizer | djszapi: as explained in http://paste.debian.net/124359/ javispedro and me tried to start it, and didn't know about root != root != devel-su != develsh, so javispedro concluded the problem is bash is trying to do a setgid and we suspected aegis-exec could help to allow bash doing that | 21:18 |
| zehjotkah | http://www.howtoforge.com/converting_rpm_to_deb_with_alien | 21:18 |
| zehjotkah | atm I'm trying to do it with alien... | 21:19 |
| djszapi | DocScrutinizer: hah misusage ok | 21:19 |
| DocScrutinizer | how can it be misusage? probably that's the secret command to brick devices, and has no further use that exactly that | 21:20 |
| DocScrutinizer | than exactly that* | 21:20 |
| DocScrutinizer | note we have no manpages for any of this aegis cruft | 21:21 |
| djszapi | DocScrutinizer: that commands works just fine | 21:21 |
| djszapi | * command | 21:21 |
| djszapi | I dared to try it out =p | 21:22 |
| macmaN | mgedmin: how do you add harmattan-dev to the sources.list, it has now Packages.gz files anywhere | 21:23 |
| macmaN | no* | 21:23 |
| zehjotkah | ah, alien deletes the .deb file directly after creating it... because it is a different platform than the host system | 21:24 |
| zehjotkah | trying in qemu now | 21:24 |
| mgedmin | macmaN, I would copy it from my scratchbox sources.list | 21:25 |
| mgedmin | if I wanted to do that | 21:25 |
| mgedmin | in case you're curious, my scratchbox has 'deb http://harmattan-dev.nokia.com/ harmattan/sdk free non-free' | 21:26 |
| zehjotkah | oh wait... qemu is not necessary... we've an toshiba ac100 here... :D | 21:26 |
| *** [XeN] has joined #harmattan | 21:27 | |
| macmaN | mgedmin: sonnuvabitch, youre right | 21:28 |
| macmaN | it does have it (looked in my sbox too) | 21:28 |
| djszapi | DocScrutinizer: what I do not prefer in aegis is that, the dpkg is written in perl. | 21:29 |
| djszapi | I always rejected it :) | 21:30 |
| DocScrutinizer | djszapi: the command bricked my device. I logged in via ssh root@N950. | 21:31 |
| DocScrutinizer | djszapi: I believe ther are probably modes (devel-su, develsh, --relaced-exec, whatever) where it does _not_ brick the device | 21:32 |
| djszapi | DocScrutinizer: hard to help if I cannot reproduce it.. | 21:33 |
| mgedmin | oh crap? | 21:33 |
| DocScrutinizer | djszapi: you suggest I do it again to confirm? | 21:33 |
| mgedmin | did I really patch zlibrary/ui/src/ezx/view/ZLQtPaintContext.cpp instead of the qt4 version? | 21:33 |
| mgedmin | what is ezx anyway? | 21:33 |
| DocScrutinizer | djszapi: I doubt I will do that | 21:34 |
| *** hardaker has quit IRC | 21:34 | |
| djszapi | DocScrutinizer: well, at least, a log would be nice | 21:35 |
| DocScrutinizer | djszapi: if you can't eproduce, then either A) you did some detail differently to me (I guess you didn't log in via ssh root@n950); B) your device is different, C) this is Heissenbug | 21:35 |
| GeneralAntilles | People who dropped the DEVELOPER EDITION pngs, did you end up with a blotchy background? | 21:35 |
| fiferboy | GeneralAntilles: I typically notice it on the light coloured apps (unless the app paints a rectangle over the background) | 21:36 |
| GeneralAntilles | fiferboy, yes, but you don't have any answers about how to fix it. :P | 21:36 |
| djszapi | DocScrutinizer: 4) The device likes me :p | 21:36 |
| DocScrutinizer | djszapi: I don't see how to get a log better than what's in http://paste.debian.net/124359/, on a instant booting and bricking command | 21:37 |
| zehjotkah | Opera Mobile for MeeGo ARM is coded with Hard Float... | 21:38 |
| zehjotkah | :/ | 21:38 |
| DocScrutinizer | well, I could try to redirect syslog to a logging host :-P | 21:38 |
| fiferboy | GeneralAntilles: I must have missed you asking for a solution instead of just confirmation of the problem :P | 21:38 |
| Stskeeps | of course, so's harmattan | 21:38 |
| zehjotkah | so alien complains about that... | 21:38 |
| djszapi | Well, me and my colleague did not change credp_kcheck recently, and it failed by you. It should not, and it does not by me. | 21:38 |
| *** antman8969 has joined #harmattan | 21:38 | |
| GeneralAntilles | fiferboy, implied solution request. *g* | 21:38 |
| fiferboy | GeneralAntilles: WONTFIX | 21:38 |
| GeneralAntilles | I mean, since there's no place to report bugs. | 21:38 |
| zehjotkah | probably i have to do the alien thing directly on the N950^^ | 21:38 |
| zehjotkah | anyone would like to build alien for MeeGo Harmattan? ;) | 21:39 |
| Stskeeps | zehjotkah: honestly, just try to untar it on harmattan | 21:39 |
| zehjotkah | okay, will try it | 21:40 |
| Stskeeps | and see if it runs | 21:40 |
| spenap | is there any documentation on how to integrate an application with the system settings (so its options appear under settings -> applications)? | 21:45 |
| djszapi | zehjotkah: I have packaged alien, fwiw | 21:48 |
| zehjotkah | great | 21:49 |
| zehjotkah | mind to send it ;) | 21:49 |
| djszapi | well, c-obs is currently (still?) full of bugs :) | 21:49 |
| djszapi | send me your email in private | 21:50 |
| zehjotkah | zehjotkah at googlemail dot com | 21:50 |
| zehjotkah | btw here is the opera rpm file if you want to try it on your own: http://labs.opera.com/news/2011/05/24/ | 21:51 |
| djszapi | sent | 21:52 |
| zehjotkah | got it, thanks! | 21:53 |
| Scifig | Anybody familiar with MAFW? Can you pls take a look at this post --> http://permalink.gmane.org/gmane.comp.handhelds.maemo.devel/28406 | 21:56 |
| *** djszapi has left #harmattan | 21:56 | |
| zehjotkah | djszapi, too much dependency errors... | 21:56 |
| * alterego starts watching season 2 of "The Wire" .. | 21:57 | |
| DocScrutinizer51 | Stskeeps: you relized we successfully installed and run .deb of fremantle on harmattan? (regarding hardfp vs softfp) | 21:59 |
| w00t | it will work, some of the time | 21:59 |
| *** wicket64 has joined #harmattan | 22:01 | |
| mgedmin | editing patches by hand FTW | 22:02 |
| Stskeeps | DocScrutinizer51: yes, it will work except for procedure prototypes involving float | 22:07 |
| mgedmin | I see harmattan doesn't search/domain to /etc/resolv.conf either | 22:07 |
| *** zehjotkah has quit IRC | 22:07 | |
| macmaN | i think everything goes through local dnsmasq | 22:07 |
| mgedmin | which doesn't add the domain either | 22:08 |
| lcuk | hey | 22:08 |
| lcuk | when swipe is in effect | 22:08 |
| lcuk | how come when you swipe from the left | 22:08 |
| lcuk | the app grid appears | 22:08 |
| lcuk | and when you swipe from the right | 22:08 |
| lcuk | the app grid appears | 22:08 |
| lcuk | isn't the active window in between the launcher and the live grid? | 22:08 |
| alterego | lcuk: it's always the last screen | 22:08 |
| lcuk | ie swipe left for launcher | 22:08 |
| lcuk | swipe right for grid | 22:08 |
| lcuk | ahh well alterego | 22:09 |
| alterego | left, right, up, down, all go to the last screen. | 22:09 |
| * lcuk is used to other places already :O | 22:09 | |
| lcuk | the annoying mis-swipes are frustrating | 22:09 |
| mgedmin | swipe is not as well designed as I'd like | 22:10 |
| lcuk | mgedmin, I have an implementation of swipeish on liqbase | 22:10 |
| mgedmin | I'd even say it's crap, but I want to give it a fair chance | 22:10 |
| lcuk | it swipes between all the running apps | 22:10 |
| lcuk | and cycles round | 22:10 |
| lcuk | like super alt-tab | 22:10 |
| lcuk | works really well on the ideapad too | 22:10 |
| lcuk | like how it works on maemo5 desktops | 22:11 |
| mgedmin | wtf am I offline? n950 shows connected to my lan | 22:11 |
| mgedmin | can't ping router.lan | 22:11 |
| mgedmin | router can't ping it either | 22:12 |
| * mgedmin curses, disconnects, reconnects manually | 22:12 | |
| mgedmin | now it works | 22:13 |
| *** qgil has quit IRC | 22:19 | |
| *** Arkenoi has quit IRC | 22:19 | |
| hiemanshu | mgedmin: yeah happens a lot with my 3G connection too | 22:22 |
| *** fcrochik has quit IRC | 22:22 | |
| hiemanshu | I know its the network, because I use a sim from the same operator on my N900 too which works just fine | 22:22 |
| hiemanshu | s/the/not the/ | 22:22 |
| infobot | hiemanshu meant: I know its not the network, because I use a sim from the same operator on my N900 too which works just fine | 22:22 |
| *** piggz has quit IRC | 22:25 | |
| *** harbaum has quit IRC | 22:26 | |
| mgedmin | building fbreader debs with a new release number, uploading to my repo, and upgrading via apt-get after each change is a colossal waste of time and bandwidth | 22:28 |
| *** Arkenoi has joined #harmattan | 22:28 | |
| mgedmin | and flash rewrite cycles, I suppose | 22:28 |
| mgedmin | fiferboy, woohoo fbreader bug is fixed! | 22:28 |
| mgedmin | get the mg5 version from my repo (it's the latest one) | 22:29 |
| mgedmin | no more segfault on swipe | 22:29 |
| fiferboy | mgedmin: !! | 22:29 |
| fiferboy | mgedmin: Woo hoo! | 22:30 |
| mgedmin | duh! | 22:30 |
| mgedmin | wondering why fbreader doesn't remember book, position or custom keybindings -- I was running it as root | 22:30 |
| *** piggz has joined #harmattan | 22:33 | |
| *** hardaker has joined #harmattan | 22:36 | |
| fiferboy | mgedmin: Installing from your repo isn't working | 22:38 |
| fiferboy | I apt-get update and I can see all your versions | 22:38 |
| fiferboy | But I am getting 404's on install | 22:38 |
| mgedmin | oh? | 22:40 |
| fiferboy | But I can download them directly out of your repo without problems | 22:40 |
| fiferboy | Doesn't look like a fault at your end | 22:41 |
| fiferboy | Something must be messed up with my N950 state | 22:41 |
| mgedmin | I see your 404s in my access.log | 22:41 |
| mgedmin | GET /770/dists/harmattan/experimental/binary-armel/dists/harmattan/experimental/binary-armel/libzlcore-data_0.12.10dfsg-1ubuntu2mg5_all.deb | 22:41 |
| mgedmin | yeah, that's not the right URL | 22:41 |
| *** zehjotkah has joined #harmattan | 22:42 | |
| zehjotkah | opera mobile for MeeGo ARM works on the N950 now! :) | 22:42 |
| fiferboy | mgedmin: Oh? | 22:42 |
| zehjotkah | http://twitter.com/XenGi_ is writing a tutorial atm | 22:42 |
| fiferboy | mgedmin: I added "deb http://mg.pov.lt/770/dists/harmattan/experimental/binary-armel/ ./" to my sources | 22:43 |
| mgedmin | no! | 22:43 |
| fiferboy | Ah, screwed up the line | 22:43 |
| mgedmin | "deb http://mg.pov.lt/770 harmattan user other experimental" | 22:43 |
| mgedmin | 'other' is empty, actually, right now | 22:44 |
| fiferboy | mgedmin: Got it, thanks | 22:45 |
| mgedmin | next: volume keys | 22:46 |
| mgedmin | or a toolbar button for toggleFullscreen | 22:46 |
| mgedmin | to hide the uglyness ASAP after startup, without having to open the hw keyboard | 22:47 |
| mgedmin | :) | 22:47 |
| fiferboy | Hehe, mine starts up in fullscreen :) | 22:51 |
| mgedmin | ooh, how did you make it? | 22:51 |
| mgedmin | oh, it remembers from last time | 22:52 |
| mgedmin | all I had to do was close it (nicely) while it was fullscreened | 22:52 |
| fiferboy | mgedmin: Yeah, it never used to do that (in old versions) | 22:52 |
| fiferboy | mgedmin: I bind ctrl+q to quit to get that | 22:52 |
| fiferboy | exit | 22:52 |
| mgedmin | I've copied my options.xml from the n900 with w/a/s/d keybindings | 22:54 |
| mgedmin | the UI doesn't allow me to make new bindings to alphanumerics | 22:54 |
| mgedmin | something about the way meego-inputmethod works interferes with what fbreader wants | 22:54 |
| mgedmin | so, libresourceqt | 22:58 |
| vandenoever | is there an equivalent to qsizepolicy in qml? | 23:00 |
| fiferboy | mgedmin: Thanks for fixing up fbreader! | 23:02 |
| fiferboy | I tend to use tap scrolling, so keybindings and volume scrolling aren't too big for me | 23:03 |
| spenap | do you know if the package extensions for Fremantle (such as XB-Bugtracker, XB-Homepage or XB-Maemo-Upgrade-Description) are still valid in Harmattan? | 23:05 |
| *** baraujo has quit IRC | 23:10 | |
| *** baraujo has joined #harmattan | 23:11 | |
| mgedmin | spenap, Maemo-Upgrade-Description yes, the rest aren't mentioned in the docs | 23:12 |
| mgedmin | Homepage is a standard field, not an extension, I think | 23:13 |
| mgedmin | dunno about bugtracker | 23:13 |
| spenap | mgedmin, thanks: can you point me to those docs? | 23:13 |
| *** djszapi has joined #harmattan | 23:13 | |
| GeneralAntilles | Can't install from mgedmin's repo | 23:13 |
| GeneralAntilles | 404 | 23:13 |
| mgedmin | http://www.developer.nokia.com/Community/Wiki/Harmattan:Developer_Library/Reference_documentation/Harmattan_Appendix_to_the_Debian_Policy_Manual | 23:13 |
| djszapi | Elleo: https://build.pub.meego.com/project/show?project=home%3Arzr%3Aharmattan -> This will be our shared repository, wanna join ? | 23:13 |
| mgedmin | GeneralAntilles, "deb http://mg.pov.lt/770 harmattan user other experimental" in sources.list? | 23:14 |
| mgedmin | or something insane like what fiferboy had? | 23:14 |
| spenap | mgedmin, thanks | 23:14 |
| GeneralAntilles | Something insane, clearly. | 23:14 |
| Arkenoi | does flash plugin from n900 work on 950? | 23:15 |
| Tronic | Almost certainly not. | 23:15 |
| Tronic | But Youtube at least works in HTML5 mode on N950. | 23:16 |
| * mgedmin updates instructions at http://mg.pov.lt/770 | 23:17 | |
| mgedmin | not that I want people using my repo... | 23:17 |
| GeneralAntilles | mgedmin, COBS account? Your repo is slow. :P | 23:17 |
| mgedmin | C-OBS is in the TODO list | 23:17 |
| mgedmin | a link to a "these are the exact steps you need to do" kind of HOWTO would speed it up | 23:18 |
| GeneralAntilles | mgedmin, there is one, but it never shows up on Google. | 23:19 |
| mgedmin | now, what's this thing about invoker in the .desktop file to enforce "run once"... | 23:19 |
| mgedmin | I saw that in some document somewhre | 23:20 |
| mgedmin | Exec=invoker --single-instance --type=d /path/to/app | 23:20 |
| mgedmin | what's --type=d do? | 23:20 |
| mgedmin | cool, invoker --help works! | 23:21 |
| fiferboy | mgedmin: The better(?) way seems to be Exec=/usr/bin/single-instance /path/to/app | 23:23 |
| fiferboy | mgedmin: At least, the way that seems to make more sense to me | 23:23 |
| mgedmin | oh my poor widdle head, what's the difference? | 23:23 |
| fiferboy | mgedmin: Nothing (porbably), except that the way I listed is the way that first worked for me | 23:24 |
| fiferboy | Same functionality anyway, but I have an easier time remembering my way | 23:24 |
| mgedmin | invoker also lets me use --splash /path/to/image.png if I wish | 23:25 |
| mgedmin | on the other hand invoker --single-instance --type=e fbreader doesn't actually enforce the --single-instance bit :) | 23:26 |
| mgedmin | you win, fiferboy | 23:27 |
| *** rlinfati has joined #harmattan | 23:30 | |
| fiferboy | mgedmin: I have found new installs with the invoker command don't actually launch the application until a reboot | 23:32 |
| fiferboy | But I haven't been able to confirm a link between invoker and the lack of launching | 23:32 |
| mgedmin | I had that problem with your original .desktop file that didn't use invoker | 23:32 |
| mgedmin | remember? | 23:33 |
| fiferboy | Are you sure it didn't use invoker? | 23:33 |
| fiferboy | I remember you having that trouble | 23:33 |
| rlinfati | Hi, i have problem trying to use c-obs for harmattan :( can someone helpme? the build log is http://pastebin.com/1Ga4a4XX | 23:35 |
| rlinfati | g++ -Wl,-O1 -Wl,-rpath,/usr/lib -o LatitudeUpdaterDaemon latitudedaemon.o main-daemon.o gpscontrol.o googlelatitude.o kqoauthauthreplyserver.o kqoauthrequest.o kqoauthmanager.o kqoauthrequest_xauth.o kqoauthrequest_1.o kqoauthutils.o moc_latitudedaemon.o moc_gpscontrol.o moc_googlelatitude.o moc_kqoauthmanager.o moc_kqoauthrequest.o moc_kqoauthrequest_xauth.o moc_kqoauthauthreplyserver.o moc_kqoauthauthreplyserver | 23:36 |
| rlinfati | _p.o -L/usr/lib -L/usr/lib -lQtLocation -lQtNetwork -lQtCore -lpthread | 23:36 |
| rlinfati | moc_kqoauthauthreplyserver_p.o: file not recognized: File truncated | 23:36 |
| fiferboy | mgedmin: You are right, my desktop file didn't use invoker *or* single-instance | 23:36 |
| fiferboy | Must be something with fbreader, because I am having the same trouble with your package :) | 23:36 |
| zehjotkah | Opera Mobile 11 on the N950 or N9 (MeeGo Harmattan): http://meetmeego.org/2011/07/28/opera-mobile-11-furs-nokia-n950-und-n9-harmattan/ | 23:38 |
| * djszapi would like to see rekonq on his mobile. | 23:39 | |
| hiemanshu | zehjotkah: change ssh root@localhost to devel-su | 23:39 |
| zehjotkah | does it make any difference? | 23:40 |
| mgedmin | fiferboy, for the record, my latest package didn't use invoker or anything else | 23:42 |
| fiferboy | mgedmin: Huh, works after a reboot though | 23:48 |
| mgedmin | I didn't even need to reboot, just modify the .desktop file and change it back | 23:48 |
| mgedmin | wish I knew how to reproduce this | 23:48 |
| mgedmin | then maybe I could debug it | 23:48 |
| fiferboy | Maybe the single-instance would help? | 23:49 |
| *** trx has quit IRC | 23:49 | |
| fiferboy | mgedmin: Thanks again for fixing up that bug! | 23:51 |
| *** fiferboy has quit IRC | 23:51 | |
| mgedmin | where does fbreader define its standard keybindings? | 23:54 |
| *** trx has joined #harmattan | 23:55 | |
| djszapi | anybody writing an offline dictionary software, like QStarDict did the trick on N900 ? | 23:55 |
| mgedmin | found it, now how does it parse the key names? | 23:56 |
| *** pancake has joined #harmattan | 23:59 | |
| pancake | hi | 23:59 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!
