IRC log of #harmattan for Tuesday, 2011-08-02

*** antman8969 has quit IRC00:00
*** dcarr_home has joined #harmattan00:03
*** Arkenoi has quit IRC00:22
lardmanany ideas on how to get a pro file to build a binary that is not under /usr/local/ ?00:23
lardmanI was assuming qmake PREFIX=/usr00:24
djszapiINSTALLPREFIX=/usr ?00:26
kimju\o/ - working menu entry in meego-terminal to force it to stay on landscape/stay on portrait/follow device orientation00:26
djszapimmh, PREFIX should do the trick.00:28
kimjunext up, enabling/disabling the special key input bar.00:28
mikhaskimju, grep for MInputMethodState00:29
djszapilardman: use cmake and -DCMAKE_INSTALL_PREFIX=/usr :p00:29
lardmanfor Qt stuff djszapi ?00:29
lardmanhmm, ok00:30
lardmanit is possible with qmake though, as I've done it before00:30
* lardman doesn't like being defeated00:31
lardmanah, needed to edit the pro file00:33
djszapi DESTDIR = $$(PWD)00:34
djszapi message(The project will be installed in $$DESTDIR)00:34
djszapiI think qmake is one serious weakness of Qt.00:36
*** deimos has quit IRC00:36
mikhaswow, Madde really tries hard to be stupid00:40
mikhasthis is where it puts the prf file00:40
mikhasas if that would match the directory structure on the N95000:40
djszapiit reminds for scratchbox :)00:41
djszapi* me00:41
djszapiThere are also things like that over there.00:41
mikhasit is00:41
mikhasthat is the insane part about it00:41
mikhasthey wanted to get rid of sbox … by wrapping it with a dumb tool00:41
djszapiI could never really managed in scratchbox to find the right FindFooBar.cmake module since it puts into some heuritics place inside the debian folder :/00:42
mikhaswell, I actually get along with sbox00:42
mikhasthat's why Madde's stupidity is extra harsh for me00:42
mikhasproductivity killer00:42
mikhasand I only try to get it to work because it comes with that Qt SDK00:43
djszapiohh it is just one small example of the scratchbox things, but already annoying00:43
mikhasso that others (not even me!) can compile my project00:43
djszapithe perl situation is even much worse.00:43
djszapibut it is also true for the whole qemu emulation and others.00:45
djszapiI could never build a package with few or more test cases inside scratchbox without force disable of the test runs, which is ... not nice :)00:47
mikhasand of course, debian install files dont allow for regex00:47
mikhashate hate hate00:47
djszapithey do ?00:48
mikhas*.prf did not work00:48
mikhasto match above00:48
djszapithat is not because of regexp is forbidden00:48
djszapiwell, I always use wild cards for version things for instance.00:48
djszapiso versions. Always worked.00:48
mikhasthe debian manual doesn't think it is important to tell me what kind of regex it uses00:49
*** deimos has joined #harmattan00:50
djszapiwell, I use what is used by bash things, like grep, find and so forth00:52
*** antman8969 has joined #harmattan00:54
lardmanso how do I add extra files to a package with dpkg-buildpackage?00:55
*** NIN101 has quit IRC00:55
lardmane.g. headers00:55
lardmaneven better, how does one create a -dev package?00:56
djszapilardman: add an entry in the control file and create the relevant install file00:56
lardmanah, control file was the missing link00:56
lardmanthanks djszapi00:56
djszapiyou can also play with symbols file, if there is a need for that, but I would avoid that, if you do not have that much time :)00:57
lardmannah, just need to package up some headers into a -dev package01:00
djszapilardman: are you familiar with rpm packaging, btw ?01:01
*** baraujo has quit IRC01:02
lardmanoh I do hate packaing01:06
lardmanI hate packaging01:07
lardmanadded entry to control file, added .install file matching that name, no package01:07
*** wazd has quit IRC01:07
djszapiwhat did you run then ?01:08
djszapidpkg-buildpackage -rfakeroot ?01:08
lardmandpkg-buildpackage -rfakeroot01:08
djszapishow the control file ?01:09
lardmanhmm, I may be wrong, I have a package but no temp dir created under debian01:09
lardmanI may hve spoken too soon01:10
lardmannope, just the docs included in the -dev package01:11
*** seif has joined #harmattan01:14
leinirgah... QSoundEffect(pulseaudio): Error decoding source01:16
kimjuhmm, I don't see any way to hide the toolbar in docs, only unregister it completely.01:16
mikhaslardman, does your qmake project file INSTALLS += target ?01:21
lardmanyes mikhas it does01:22
lardmando I need to add another target?01:23
*** epage has joined #harmattan01:24
*** rcg has quit IRC01:25
lardmanWhat I never understood is how one goes from the *.install file paths to the actual build tree paths01:26
lardmanor is it just vey intelligent?01:26
*** antman8969 has quit IRC01:27
lardmanlaptop battery about to run out, seems like a good time to head for bed01:29
lardmannight all01:29
djszapilardman: it is much of difficult mapping..You are basically just deciding what to really install. You can overwrite the paths and such, but normally it is really straight-forward01:29
djszapi* not much of ..01:29
lardmandjszapi: will have a chat to you tomorrow if you're about01:29
lardmanas it's not working for me atm01:29
djszapisleep well :)01:30
lardmananyway must head off before battery dies completely01:30
*** lardman has quit IRC01:30
*** wicket64 has quit IRC01:49
*** lcuk has quit IRC01:51
*** seif has quit IRC01:54
*** andre__ has quit IRC02:00
*** seif has joined #harmattan02:07
seifstill no new on the next updatE?02:09
djszapiseif: maybe the end of september02:10
seifdjszapi, so notwitter support coming out ?02:10
djszapior october, but it is just a guess work02:10
djszapisorry ?02:10
seifmy harmattan on n950 does not support twitter accounts yet02:12
djszapiI am not sure I can talk about what it will support, sorry ...02:13
djszapimmm, since I can use my N9 outside with PR image..well, it certainly has twitter.02:13
djszapiseif: I think it is not a big deal to get it working though.02:14
seifyeah my pr image is for n95002:15
seifi wanted to know if there is an image update soon02:15
djszapiI think they will ship one with the release. Someone linked a page today here with the date, September for that02:18
djszapibut it is just guess-work :)02:21
*** lbt has quit IRC02:34
rm_youso rebooting didn't fix my blank call window problem02:36
*** smoku has left #harmattan02:37
*** vandenoever has quit IRC02:44
*** mikhas has quit IRC02:44
*** eman has joined #harmattan02:47
*** Rizzer has joined #harmattan02:58
DocScrutinizerMohammadAG: >>Departed Facility in TEL AVIV - ISRAEL  Monday, August 01, 2011  19:03 << \o/03:06
DocScrutinizerMohammadAG: seems they transport from tel aviv to LOD with bike, or taking a walk03:08
* GeneralAntilles yawns.03:10
MohammadAGDocScrutinizer, yeah, testing Sports Tracker03:12
DocScrutinizerMohammadAG: so seems your deadline 08-03 will be the date. Dunno if it's the last day to pick it up, or the first day you're not at home03:12
DocScrutinizerGeneralAntilles: hush! you're several TZ behind, you're not allowed to yawn right now ;-D03:14
GeneralAntillesLong day03:14
DocScrutinizerhow do I get the fqn of a file? except from concatenating CWD with the relative filename03:20
DocScrutinizerdirname and readlink all are not helpful, even for "see also:" in manpages03:22
*** meegoexperts_mob has joined #harmattan03:23
*** deimos has quit IRC03:23
DocScrutinizerstat -c %N x  #also not what I need03:24
*** trx has quit IRC03:25
SpeedEvilIs that a valid question?03:28
SpeedEvilI point at various files in /sys with multiple paths03:28
DocScrutinizerdang, isn't there any function/executabe that, on ># themagicls ~/../B/x; replies >>  /home/B/x03:28
*** trx has joined #harmattan03:29
DocScrutinizermmmpff! I probably could compare inode number of the file in question, and a given FQN path I want to know if it's same file or not03:30
DocScrutinizeris there any tool that lists all hardlinks pointing to a file?03:32
SpeedEvilI don't think it's possible to do that.03:32
SpeedEvilOther than with find, or something like tracker.03:33
DocScrutinizerprobably a dang heavy tool, it had to ls -lr / and compare the inode numbers03:33
SpeedEvilThere is no 'stat' that spits out all of the directories the file is in, or its names, as that's not held in the file inode.03:33
DocScrutinizeryeah find03:33
DocScrutinizerfind / -inum 76463403:34
DocScrutinizeris probably what I look for03:34
DocScrutinizeror -samefile :-D03:35
*** Scifig has joined #harmattan03:42
DocScrutinizeranyway unless I find out what's "A 0 0 33061" in  (grep -v "A 0 0 33" )  /var/lib/aegis/refhashlist03:49
DocScrutinizer... :-S03:50
SpeedEvilinode 33061?03:50
DocScrutinizernope, thought that as well first, but it's either a group ID, or a permissions bitset03:51
DocScrutinizerS 15 H 40 a68426a5b1586cccd918bbb63798c43ca26b7a8d A 0 0 35309 P 22 energy-profiler-server R 30 usr/bin/energy-profiler-server03:51
DocScrutinizerS 15 H 40 871ed0d45e694039478e651295861fb3ec1080fd A 0 0 35309 P 8 libc-bin R 16 usr/lib/pt_chown03:51
SpeedEvilWell - the long string is obviously a hash03:51
*** epage has quit IRC03:53
DocScrutinizerSring length:15 Hash length:40 871ed0d45e694039478e651295861fb3ec1080fd A??? 0 0 35309 Package length:8 libc-bin  Reference length:16 usr/lib/pt_chown03:53
DocScrutinizerin the whole 700some kB refhashlist there are EIGHT lines not matching "A 0 0 33"03:55
SpeedEvilflags indeed03:56
rm_youso rebooting didn't fix my blank call window problem, any thoughts?03:56
SpeedEvilThere should possibly be a packagename in there too for eavh file03:58
DocScrutinizeractually it seems like 33261 and 33188 are the only sets of flags(?) used03:58
SpeedEvilDifferent signers?03:58
* DocScrutinizer checks /etc/groups03:58
DocScrutinizermeh, I guess that's aegis permissions03:59
DocScrutinizerthe whole hashlist was useless if it wasn't for telling aegis WHAT to do with those files04:00
DocScrutinizerthe 3 ints suggest U G O04:00
SpeedEvilSeems logical.04:01
DocScrutinizerthere's exactly ONE line with G!=004:01
DocScrutinizererr 204:01
DocScrutinizerS 15 H 40 838157fee92ca7a8df6c093dea9079145a3ed2f3 A 0 30021 34285 P 14 openssh-client R 17 usr/bin/ssh-agent04:02
DocScrutinizerS 15 H 40 9465b53d341467bfe43b7975b488dd61cc2dde32 A 0 30018 35181 P 21 sdk-connectivity-tool R 18 usr/sbin/devrootsh04:02
DocScrutinizersetgid() ?04:02
DocScrutinizerjavispedro once deduced about missing setgid perm for bash, from a aegis diagnostic printout in syslog04:04
DocScrutinizerwish I could find where from he got the bit meanings04:04
DocScrutinizerhmmm - Aegis: credp_kcheck failed 9990004 bash - is 7 digits04:06
DocScrutinizerwhile hashlist flags(?) are 504:07
DocScrutinizerman 5 refhashlist !!!1!!1!!!!@#%&&!!104:08
DocScrutinizerRM680-22-6_PR_RM680:~# cat /var/lib/aegis/refhashlist|wc -l04:11
DocScrutinizerS 0  H 40 0b2879714e842125406fcee56f7c8881475dbc05 A 0 0 33261 P 2 mc R 30 opt/mc/libexec/mc/extfs.d/uace04:14
DocScrutinizerS 0  H 40 c31fe5dda65c9e711364a26c10a72bd53069d9bc A 0 0 33188 P 2 mc R 28 opt/mc/etc/mc/edit.indent.rc04:14
DocScrutinizerS 0  H 40 e4bbe0cd6120fa1287fcd3ed8a298fe81ce6e89f A 0 0 33261 P 2 mc R 29 opt/mc/libexec/mc/fish/unlink04:14
DocScrutinizerS 0  H 40 d62804c56054aa03ce38d09ea4d308d983f288db A 0 0 33261 P 2 mc R 30 opt/mc/libexec/mc/extfs.d/ulha04:14
DocScrutinizerfun bit: one file got other flags(?)04:14
*** eman has quit IRC04:18
DocScrutinizerhmmmmm, edited opt/mc/etc/mc/edit.indent.rc, started mc, no issues at all, not even any logs in syslog04:19
*** willer_ has quit IRC04:20
DocScrutinizerrefhashlist record for the file didn't change though, ergo it can't be correct now04:21
SpeedEvilmaybe it will now blow up if something tries to exec() the file when it wouldn't before.04:21
DocScrutinizerlet's try it :-D04:22
*** meegoexperts_mob has quit IRC04:28
DocScrutinizernah, I always get "permission denied" when I try to run those scripts as root04:29
DocScrutinizerRM680-22-6_PR_RM680:~# su - user04:30
DocScrutinizersu: can't set groups: Operation not permitted04:30
DocScrutinizeras user as well04:31
DocScrutinizer you can't execute shellscripts. GREAT!04:36
DocScrutinizererr, while in develsh04:37
SpeedEvilI wonder how often they poke the RDA devices to see if anyones bricked them04:39
MohammadAGor every other day04:40
MohammadAGI broke an N97 mini once04:40
DocScrutinizerthere should probably a button "reflash" in the GUI ;-D04:40
MohammadAGAegis, protecting kids from internet porn since 201104:40
MohammadAGsh: ./xxx: Operation not permitted04:40
DocScrutinizeryou've been told not to play doctor games! operations are always not permitted04:42
DocScrutinizerand xxx operations even more04:42
DocScrutinizerthis friggin aegis crap gives me headaches04:44
MohammadAGi'm going to sleep04:44
DocScrutinizere.g. what is develsh, what does it do and why can it do what it does?04:44
MohammadAGmy N950 better be here when I wake up04:45
rm_youso rebooting didn't fix my blank call window problem, any thoughts? :P04:45
DocScrutinizerrestful dreams!04:45
DocScrutinizeryet none04:45
DocScrutinizerrm_you: I guess it's simply a bug04:46
rm_youyeah, but how do I FIX it04:46
DocScrutinizermaybe even a race04:46
rm_youit worked at first04:47
rm_youfor like the first week04:47
DocScrutinizerthe REFLASH!04:47
rm_youthat's not a good solution04:47
rm_youbecause if i don't figure out the problem, it will probably keep happening04:47
DocScrutinizerdon't see any alternative04:47
rm_younot reasonable to reflash every week04:47
rm_youi don't have a problem reflashing04:47
rm_youbut you see the problem, no?04:48
GeneralAntillesKill yourself.04:48
* GeneralAntilles coughs.04:48
rm_youlol GA04:48
GeneralAntillesIt requires blood sacrifice.04:48
DocScrutinizernah, you must reflash and then be alert about the steps causing the problem again04:48
GeneralAntillesIt's how Elop is powered.04:48
rm_youDocScrutinizer: but i know every single action i took before it started04:48
rm_youand none of them seem related04:48
DocScrutinizerthen reflash and redo04:48
rm_youand see if it breaks again?04:49
DocScrutinizerirrelevant what seems related04:49
DocScrutinizereither it's reproducable by your steps, or it's heissenbug04:49
rm_youn950 has quantum processor04:50
DocScrutinizerrm_you: look, it might be a random flipped bit in flash. it might be proximity sensor broken after one week. or it *might* be actually caused by your unrelated steps done during last week04:52
DocScrutinizernobody but you can find out04:52
rm_youi prolly will04:52
DocScrutinizerand your only chance is reflash04:52
rm_youi'm going to investigate04:52
DocScrutinizeresp when you claim your steps are "unrelated"04:53
rm_youwhat was that utility package04:53
DocScrutinizernot even you seem to have an idea where to look and what for04:53
rm_youthere was a util package someone mentioned earlier04:53
rm_youuseful for debugging these sorts of issues04:53
DocScrutinizerprobably script (1)  - make typescript of terminal session04:56
*** M4rtinK2 has quit IRC04:57
DocScrutinizerbisect down to the one instruction/action that triggers the bug, then we'll be wiser04:58
DocScrutinizerbut first: reflash and see if the bug vanishes04:58
DocScrutinizerto ruke out it's been always there or a hw defect04:59
SpeedEvilDoes anyone know where to get large ferrite bars cheaply? I'm wondering about ones similar to the ones in my induction cooker for containing the field on the underside.05:00
DocScrutinizeror ignore the whole issue - if your app you're developing isn't related to making calls, you just as well can ignore the whole thing05:00
rm_youno, it was a util to actually see the inner workings of an active Qt window05:00
DocScrutinizerthat's called gdb ;-P05:01
*** epage has joined #harmattan05:06
ieatlintanyone looked much into changing the standby/lock screen?05:09
ieatlintor more specifically, where the image file is located/configured05:09
*** eman has joined #harmattan05:27
*** Scifig has quit IRC05:49
*** twoboxen has joined #harmattan05:59
*** leinir has quit IRC06:05
*** willer_ has joined #harmattan06:10
hiemanshuDocScrutinizer, ieatlint, dm8tbr: you could always use a universal battery charger around when such things happen -
DocScrutinizerbut... does it come with TX4? ;-D06:24
DocScrutinizerseriously, this MUST NOT happen06:25
infobotThe key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.06:25
DocScrutinizerthe less on a device with "no user servicable parts inside"06:25
hiemanshuyes it MUST NOT happen, and hasn't happened for me either when my battery drained like 5-6 times now, though my N900 has gone crazy, but thats a different story06:26
DocScrutinizerimagine this happening on N906:26
hiemanshu'hey darling, I think I broke my N9' 'Naah the battery just died and nokia didn't care enough to fix the issue, let me charge it up for you'06:27
DocScrutinizerthe service points will need automatic coffee machines and little paper cups. And a corner with armchairs for the waiting customers ;-P06:30
DocScrutinizerhammocks, drinks, and a hula band06:32
*** willer_ has quit IRC06:32
*** eman has quit IRC06:32
DocScrutinizerI bet no developer ever let his N9 sit for 4 weeks with a depleted battery, then tried to start charging it. Odds are the cell (self)depleted enough during those 4 weeks to deadlock the device06:34
*** crevetor has quit IRC06:34
DocScrutinizerbut hey, why am I telling this now, rather than publishing it crypted on my blog, and publish the decrypt key with a big HAHA, wehn the first 10.000 N9 returned to service for charging?06:35
DocScrutinizerI offered free hw review, Nokia wasn't interested. A bit of sarcasm may be ok, no?06:37
twoboxenugh… this security policy makes it a huge pain in the ace to dev for the n95006:38
twoboxenwtf root doesn't have permission to delete a file with user/users ownership06:38
hiemanshuheh well no, they wont be interested in anything, because they uhmm, dont care much about it06:38
DocScrutinizertwoboxen: welcome to the wonderful world of AEGIS06:38
infobot[aegis] , 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
twoboxenwhere is my sudo gain root that actually works06:39
twoboxeni'm about to say f. this… it's difficult EVERY step of the way06:39
DocScrutinizertwoboxen: welcome to the club, I already got tired with ranting :-D06:40
twoboxenif they are going to abandon this platform, at least give us a way to get FULL dev mode06:40
DocScrutinizertwoboxen: you missed the best yet [2011-07-31 21:28:21] <DocScrutinizer> I'm going to eventually document on wiki: "aegis is protecting arbitrary files from editing/deleting/renaming/chmod. You have no means to know if a file is protected that way. Touching such a file will possibly result in bricking and need for a reflash of complete system. That's the situation and you either deal with it or forget about harmattan, as nobody is going06:42
DocScrutinizerto do anything about it to fix this permanent threat" Maybe NOT - CBA06:42
twoboxenso true06:43
twoboxenDocScrutinizer: I know I was counting on qole's easy debian.  and he is pissed as well06:43
DocScrutinizerthe evil notorious aegis selfdestruction function06:44
DocScrutinizerso you're about to "say f. this" - I'm a tad further over the edge06:46
*** antman8969 has joined #harmattan06:48
twoboxenDocScrutinizer: meaning, you already said 'f this'06:50
DocScrutinizer>>After installing the package, do NOT modify the installed files if they request a token. Security FW will discover an unexpected change in the file and lock the device (ops! reflash).<<06:51
DocScrutinizers/files/python scripts/06:52
infobotDocScrutinizer meant: >>After installing the package, do NOT modify the installed python scripts if they request a token. Security FW will discover an unexpected change in the file and lock the device (ops! reflash).<<06:52
twoboxenare there any petitions or forums where people are staging a protest?  WTF does Nokia care what we do?06:52
twoboxenthey didn't for the n90006:52
* DocScrutinizer hands over the banner to twoboxen, sighs, yawns, and waves06:53
twoboxenWebOS is a lot nicer… webos<somedate> and you get ACTUAL dev access… openssh, root access, etc.  See?  The users who do that know what they are getting into06:53
twoboxenDocScrutinizer: So what platform are you moving to now? :)06:53
DocScrutinizerstaying at fremantle for now (and next few years)06:54
*** Arkenoi has joined #harmattan06:54
DocScrutinizernever been interested in N9 anyway06:54
DocScrutinizereventually Samsung might come up with sth similar to maemo, which is *not* bada06:55
hiemanshuDocScrutinizer: do you know what bada means?06:55
DocScrutinizerwave, or sea, or sth06:55
hiemanshubada is big in hindi06:56
hiemanshuand in korean it means ocean or sea06:56
DocScrutinizerseems bada is "made in korea" though06:56
twoboxenDocScrutinizer: what about intel's actual meego releases?  They can't be that far off.06:57
DocScrutinizerno handheld devices06:57
hiemanshuwell yeah, nothing as of yet, but you can run it on the N9{,00,50} apparently06:58
*** epage has quit IRC06:58
DocScrutinizersupport for N9 from Nokia is more than questionable06:58
DocScrutinizerregarding "real" meego06:58
DocScrutinizerand honestly I don't care, as - like I already mentioned - I'm not particularly interested in N9. And it seems there won't be any "meego" device from Nokia after N906:59
DocScrutinizerso no matter how you look at it, it looks like a dead end07:00
hiemanshupretty much yes, and I am not moving to android, so its most likely iOS or webos for me07:01
DocScrutinizernota bene N9 never got attributed as NIT by Nokia07:06
twoboxenhiemanshu: you mean the me ego CE release?  Does it work on n950 yet?07:08
twoboxenon  so does me ego use aegis?07:11
twoboxenwhy the hell didn't they just follow android's security model07:17
twoboxenit meets all their goals on
dm8tbrhiemanshu: yes, such things go unsaid. I just mentioned the hacky approaches07:19
*** twoboxen has quit IRC07:29
*** DocScrutinizer has quit IRC07:31
*** DocScrutinizer has joined #harmattan07:31
*** antman8969 has quit IRC07:31
*** Arkenoi has quit IRC07:33
*** antman8969 has joined #harmattan07:39
*** hardaker2 has quit IRC07:43
*** antman8969 has quit IRC07:47
*** antman8969 has joined #harmattan07:47
*** antman89691 has joined #harmattan07:54
*** antman8969 has quit IRC07:57
*** npm has quit IRC08:13
*** npm has joined #harmattan08:13
*** djszapi_ has joined #harmattan08:40
*** Scifig has joined #harmattan08:43
*** antman8969 has joined #harmattan08:45
*** antman89691 has quit IRC08:48
*** djszapi_ has quit IRC08:52
*** rcg has joined #harmattan09:01
*** npm has quit IRC09:02
*** npm has joined #harmattan09:03
*** arfoll has quit IRC09:09
*** djszapi has quit IRC09:19
*** smoku has joined #harmattan09:34
*** smoku has left #harmattan09:39
MohammadAGhiemanshu, iOS? seriously?09:40
hiemanshuMohammadAG: I have owned an iPhone, so yeah, iOS is better than android, much much better09:40
MohammadAGin what ways?09:41
hiemanshuevery way?09:41
MohammadAGnot really?09:42
Stskeepsit's 8:36am and we're discussing platform wars09:42
MohammadAGStskeeps, I'm actually holding a shotgun for the courrier09:42
Stskeepstrust me, in the future, platforms and their APIs matter less, they'll just be a delivery mechanism to differentiate with the ecosystem being the free web09:43
*** divan has quit IRC09:43
MohammadAGForwarded for delivery, meh09:44
hiemanshuMohammadAG: the main thing about iOS is loads of games \09:45
hiemanshuand I have developed for iOS09:45
MohammadAGhiemanshu, meh, I've been with maemo for a year, never cared about games09:46
MohammadAGdevelopment needs a mac, I don't intend on buying one09:46
hiemanshuMohammadAG: I travel enough to care about games09:46
MohammadAGplus it's Obj-C09:46
hiemanshuMohammadAG: I have a mac, and getting another one some time soon09:46
hiemanshuyeah, Obj-C is beautiful code, and prettier API cals09:46
MohammadAGso why aren't you on an iPhone?09:47
hiemanshuI lost mine09:47
hiemanshuabout a month or so ago, and not buying one right now, waiting for iPhone 509:47
hiemanshuonly reason I am not full time into iPhone dev is no Qt09:47
*** aleksander_m has joined #harmattan09:49
MohammadAGwell, the platform war didn't work well09:50
MohammadAGlet's talk rpm vs deb09:50
* MohammadAG runs09:50
Stskeepsi've used debs, i've used rpm, and i like rpm better, allows me a much nicer workflow09:50
MohammadAGbut RPM sucks!!!1!109:51
Stskeepsnot a productive argument09:52
MohammadAGObviously, I wasn't being serious...09:53
hiemanshuMohammadAG: I have used rpm all my life, deb is just meh10:02
*** harbaum has joined #harmattan10:03
*** antman8969 has quit IRC10:03
*** antman8969 has joined #harmattan10:07
*** andre__ has joined #harmattan10:11
*** andre__ has joined #harmattan10:11
DocScrutinizer<<in the future, platforms and their APIs matter less>>  seems I've heard this since end of last millenium, still it didn't exactly come true, despite chromeOS, thin clients, clouds and all10:13
*** lbt has joined #harmattan10:19
antman8969does anyone know who owns the site? (also
hiemanshuantman8969: see who owns the domain. use a whois service10:21
antman8969I was hoping for a user name lol10:21
antman8969micheal jirz10:21
*** eman has joined #harmattan10:23
*** Wirta has joined #harmattan10:28
*** Scifig has quit IRC10:28
lbtFYI discuss in #meego. Also see and email threads10:48
*** rcg has quit IRC10:55
*** wazd has joined #harmattan10:59
*** Arkenoi has joined #harmattan10:59
harbaumPretty enteraining read ...11:02
* Arkenoi thinks easydebian must be hell to use on capacitive touchscreen11:02
*** antman8969 has quit IRC11:06
*** djszapi has joined #harmattan11:06
*** slaine has joined #harmattan11:06
DocScrutinizeryou bet11:07
*** mikhas has joined #harmattan11:08
DocScrutinizeresp without mouse and kbd support, due to missing hostmode (and alas this doesn't seem to change any time soon)11:08
ArkenoiDocScrutinizer, well, you may use bluetooth ones..11:08
RST38hI am creating a repo on OBS. What repo type should I choose?11:09
DocScrutinizerIf I had, I'd test it to confirm :-D11:10
DocScrutinizerRST38h: strawberry11:10
djszapiRST38h: pick one via advanced interface.11:11
djszapiand armv7el there.11:11
*** leinir has joined #harmattan11:18
*** cpscotti has joined #harmattan11:24
*** eman has quit IRC11:47
*** vandenoever has joined #harmattan11:50
djszapiDocScrutinizer: we are gonna have a documentation review meeting with the SDK team prolly. Let me know what you miss from the aegis documentation (without screaming, and in a mature mood please) and I try to get into this.11:58
DocScrutinizerdjszapi: noticed, thanks11:58
SpeedEvil'man aegis'12:00
DocScrutinizermaybe ask twoboxen and javispedro as well, if they appear from the underbush12:00
DocScrutinizerman 5 refhashlist12:01
djszapiyeah, please tell it to everybody12:02
djszapiwe try to make it work by Friday.12:02
djszapiI would appreciate a wikipage for it with short notes.12:03
Arkenoidamn they call *this* "affordable". $50 for fscking IR diode with simple converter! would be useful as we do not have CIR anymore if not the price12:03
DocScrutinizeraaaah, of course! >>what is a practicable way to make aegis "hush", I.E. get a set of tokens for root that lets root feel like root?"12:03
*** Rizzer has quit IRC12:05
* SpeedEvil writes page overview on wiki.12:05
DocScrutinizerdang, I bet I could do better, by feeding left headset channel with max dBm 20kHz and rectify/multiply it for VCC, while right channel outputs signal12:06
Arkenoiroot cannot even view dmesg, that sucks!12:06
djszapiArkenoi: no, it does not.12:06
djszapiyou have no idea what a user device means.12:06
Arkenoiuser device means you cannot view dmesg?12:06
DocScrutinizerok, this comment was it for me12:06
Arkenoihell even iPhone can do12:06
Arkenoidoesn't it qualify as user device?12:07
djszapiArkenoi: in my opinion, it is much worse approach if you have a unix workstation design for a consumer phone.12:07
*** DocScrutinizer has left #harmattan12:07
djszapiArkenoi: frankly, you should just learn instead of whining, how to do things.12:08
*** wazd has quit IRC12:08
SpeedEvilI'm writing an overview of the issues as I see them on the wiki. Comments or alterations welcome when I have it up - maybe a few mins I'll post the link.12:09
Arkenoidjszapi, i run selinux enforced mode on my desktop, it does not suck that much12:09
djszapiArkenoi: hahahahahaha12:09
djszapimade my day =)12:09
djszapiSpeedEvil: about aegis ?12:09
SpeedEvildjszapi: yes.12:09
djszapicool, thanks really.12:10
SpeedEvilThe broad themes, about why it causes issues, an the narrower things that we want to know.12:10
djszapiWe got an opportunity to meet the SDK team and fix the documentation by Friday12:10
djszapiI would like to see everybody's feedback12:10
djszapiif it does not now happen, it cannot be easily amended in the near future...12:10
SpeedEvilDocScrutinizer can be a little bad-tempered.12:10
djszapiso please take it seriously as an opportunity to improve this field.12:11
SpeedEvilI am.12:11
Arkenoiactually it would be more useful to have thing like that on the desktop as it has potentially bigger malware and remote exploits threat exposure (potentially means who the hell will hack linux desktops when there are zillions of windows machines with careless users)12:13
*** divan has joined #harmattan12:13
Arkenoiwe had vulnerable flash with remote code execution bug on the n900 from the very beginning and no one gives a fsck12:13
djszapiArkenoi: yeah, we know you know better and all the principal security engineers are just silly at Nokia and Intel :)12:16
djszapiincluding Linus Torvalds who is heavily rejecting the selinux kernel model and accepted an upstream approach which is pretty much aegis alike.12:16
djszapi* than all12:16
SpeedEvilThe issues Aegis was intended to solve are pretty much those solely around a mass-market series of handsets.12:17
SpeedEvilIntended to take over much of Nokias line.12:17
Arkenoiselinux is a good thing and its poor adoption is shame on humanity12:17
SpeedEvilThe reality - when it's implemented on one niche handset - is more questionable.12:17
djszapi12:17 < SpeedEvil> The issues Aegis was intended to solve are pretty much those solely around a mass-market series of handsets. -> no, it is not true.12:18
djszapibut nice to see you are confident enough to know better than us :)12:18
SpeedEvilBack when Aegis was designed, this was the vision.12:19
ArkenoiSpeedEvil, i just think a futile attempt to make DRM that does not suck is so futile.. though Aegis security model is worth something even beyond DRM12:19
SpeedEvilThat was the (public) plan.12:19
Stskeepsi still kind of want something like aegis in a world where devices are as promiscious as a US teenage girl studying in europe12:20
djszapiSpeedEvil: never really been a plan at the architecture design. I think you should just learn more about this field before saying so untrue explanations. Just start at the smack kernel implementation.12:20
SpeedEvilThere were security platform announcements around n900 time which I assume were about aegis, which ever made it into the n900.12:20
Arkenoihaving a trusted environment on a consumer device is great as long as the trust relationship are under the full control of the owner12:20
SpeedEvildjszapi: By issues, I mean security issues for a phone maker.12:20
SpeedEvilNot security issues as a general kernel or unix thing.12:20
djszapi12:20 < SpeedEvil> There were security platform announcements around n900 time which I assume were about aegis, which ever made it into the n900. -> so untrue again.12:21
SpeedEvilNo, it's not untrue.12:21
SpeedEvilI assume that.12:21
djszapiok, if you know better, I give up. :)12:21
SpeedEvilIf my assumption is incorrect, fine.12:21
SpeedEvilI have no data, as as far as I know, this has not been published.12:22
Arkenoiphone makers are anal. nokia is double so. see, HTC lets you take HD2 and run android on it and nothing terrible happens. Nokia guys thing that the sky will fall if you are allowed to change OS on a consumer device (n900 is really not an example)12:22
djszapiwhat bugs me is that, I have been trying to help as much as possible here since I feel it my obligation. Nobody does it apart from me in my team, and people keep saying they know better. Don't you think it is a bit soul-destroying ?12:23
djszapiSpeedEvil ^12:23
SpeedEvildjszapi: What aspect of that are you disagreeing about. That there were announcements and talk of a security platform on the n900?12:23
*** smoku has joined #harmattan12:23
djszapiSpeedEvil: aegis has nothing to do with N90012:23
SpeedEvilI know.12:24
SpeedEvilHowever, statements were made at that time which seem to the outside world to be very related to what aegis has turned out to be on the n950. It would seem logical to outside observers that there was an internal development project running at that time to make aegis.12:25
djszapiplease stop speculating.12:25
SpeedEvilAnd at that time, the vision was (publically at least) that maemo was going to take over a large slice of nokias handsets.12:25
Arkenoino one expects the spanish inquisition^W^WElop12:26
SpeedEvilIf you can't publically address these points - say so - fine.12:26
djszapiSpeedEvil: publicly address 100 times ? :)12:26
djszapisorry, I find it really tiresome.12:27
djszapiI have never seen any aegis work on N900, period. (at gazillion times).12:27
SpeedEvilI never said it was working on the n900.12:27
djszapisorry, I find it time wasting, afk :)12:28
SpeedEvilI said it seems from the outside like at least design work was going on at that time.12:28
SpeedEvilAnyway - reading docs and writing overview - please critisise that. Sorry we're talking past each other appparantly.12:28
djszapiSpeedEvil: seriously the last time, 1000005th. Aegis was designed for Harmattan and then a very alike approach was proposed against the Linux kernel as an upstream version. Linus accepted it since he just simply dislikes selinux like almost everybody else in the linux security domain apart from the selinux developers. really period.12:30
SpeedEvilI'm not disagreeing with you.12:31
Arkenoido you mean i am a pervert or work for NSA?12:32
Arkenoi(or both)12:32
SpeedEvilI mean - the above seems to be what I understand too - it's just I expressed it in different ways.12:32
*** NIN101 has joined #harmattan12:32
ArkenoiSELinux is disliked by people who are too lazy to profile their software12:33
SpeedEvilI'm not saying that aegis was ever intended for n900, just that there were (unclear) announcements made at that time.12:33
Arkenoiunfortunately it includes almost all "enterprise software vendors"12:33
Arkenoibut it does not mean SELinux is bad12:34
SpeedEvilSELinux is a good idea in principle. Most software has way too much access.12:34
Arkenoiredhat team made hell a lot of work to write selinux policies for most packages12:34
djszapiSELinux does not really work in the practice, but let us not fight it again. There are enough mailing list threads about it amongst the linux security experts. Since selinux was born there were just new upcoming linux security models. The reason is very simple. It does not work in the practice, hard to configure, not modularized and I could enumerate, really overdiscussed matter.12:35
Arkenoicould you please provide a proof it does not work?12:36
djszapiArkenoi: try to configure selinux for a really simple task and then smack.12:36
djszapiYou will see the difference in no time.12:36
kimjumy wish, apart from core documentation (like description of those /var/lib/aegis/* files) would be documenting all the various capabilities in use, what they (dis)allow and which of those can not be set in "unofficial" packages.12:37
djszapikimju: that has nothing to do with aegis12:37
djszapithose are really application things...12:37
djszapiif an application defined a token, like tracker, tracker should document it, next! :)12:37
Arkenoii have done that, quite simple and there are comprehencive tutorials, actually i am going to do that for all my packages and i do not consider it to be waste of time12:38
kimjudjszapi, technically you're of course correct, but for user, that all seems "aegis".12:38
SpeedEvildjszapi: There - seems - to be no public documentation of these things you are saying. Which is why people are asking questions that are silly.12:38
SpeedEvildjszapi: And getting upset as you're saying they're silly, because we don't know what's in scope or not.12:39
SpeedEvilSorry - this probably isn't productive,  I'll get back to writing coherent thoughts for wiki.12:40
djszapikimju: we can put one line into documentation, it has nothing to do with aegis, ok12:40
Arkenoiwell, people may dislike SElinux or not, but I am yet to see SElinux causing a fraction of PITA aegis happily provides12:40
djszapiSpeedEvil: I told you gazillian times to look for documentations and videos12:40
djszapido I need to make the google for you ?12:41
djszapiok, let us give one example....12:41
SpeedEvildjszapi: I'm looking at them at the moment.12:41
SpeedEvilSorry - I haven't dug as deep into the docs as I should.12:41
SpeedEvilreading hrouhg the released docs.12:41
kimjudjszapi, please make it two lines, to other pointing the blame to the apps :)12:41
SpeedEvilPart of the confuseion may be as I'm basicng my comments of others that may not have read the docs.12:42
SpeedEvilIs there more than at ?12:42
kimjudjszapi, but still, I'd like to see such documentation. even if it's not your (personally nor the teams) responsibility to make such.12:43
Arkenoiand i guess aegis is one of primary reason we are not likely to see cheap chinese phones and tablets with harmattan12:44
Arkenoiwell, the diversity of maemo platform is not a good thing for wide adoption12:44
kimjuand I actually do like some things that aegis makes possible, like secure storage in the device.12:45
Arkenoithere is TPM thing, like "embedded smartcard", it is much more simple and widely available12:47
SpeedEvildjszapi: However - whatever - thanks for the willingness to discuss things, and sorry there seems to be so much drama.12:47
djszapithe drama is only that, people are not constructive if there is some issue :)12:48
djszapithey tend to be smarter than a bunch of principal engineers in full time :)12:48
SpeedEvilI don't think anyone is claiming that.12:48
kimjuI wonder what is wrong with my mg-terminal package, it gets build but then obs gets stuck when it should move the resulting .deb to repo.12:49
SpeedEvilDisagreeing with policy is not disagreeing with the team of engineers, it's disagreeing with their managers.12:49
djszapiSpeedEvil: also, do not forget, meego tends to use the aegis-alike upstream security model.12:49
djszapiy does not have to be complicated.12:49
djszapialso, just good feedbacks in comparison with android, windows security model...12:50
kimjuother packages seem to get build and pushed to repos, mine gets stuck each and every time.12:50
djszapiso yeah, there might be undocumented things that can be improved, but whining all the time when we are still better than everybody else ... :)12:50
djszapiwell, that is not really constructive.12:50
*** eman has joined #harmattan12:50
SpeedEvilThe problem is that at times it's not been clear what the release configuration of the n9 is.12:51
djszapiit is like saying kdevelop's cmake system sucks :)12:51
djszapiit is not perfect, there are minor issue, but still the one available and the best ever for that purpose.12:51
SpeedEvilAnd - for example - from the perspective of most open-source developers - there is the worry about what happens if nokia stops signing stuff 12 months down the line, and drops all support.12:51
djszapiand I think that is something people forget to remember.12:51
SpeedEvil'best ever' - for purposes of nokia.12:52
Arkenoiso can anyone answer me a simple question: how do i look at dmesg?12:52
SpeedEvilAyway - it's probably best if I clarify this in document form.12:53
Arkenoithat reminds me symbian certificate uglyness12:53
djszapi12:52 < SpeedEvil> 'best ever' - for purposes of nokia. -> again untrue :)12:53
SpeedEvilRather than trying to put points at random in off-the-cuff responses that may seem confrontationa.12:53
djszapisuch a security model has nothing to do with, smack and meego....12:53
djszapiand actually upstream linux kernel, so anybody can use it...12:53
SpeedEvildjszapi: Nokia = manufacturer12:54
SpeedEvildjszapi: I'm not saying specifically nokia, sorry.12:54
Stskeepskimju: osc wipebinaries doesn't help?12:54
kimjuStskeeps, there is no binaries12:54
Stskeepskimju: hmm12:55
kimjuif someone wants to have a look12:56
*** infobot has quit IRC13:04
*** M4rtinK2 has joined #harmattan13:04
*** achipa has quit IRC13:05
*** cataska has quit IRC13:05
*** cataska has joined #harmattan13:06
*** achipa has joined #harmattan13:07
*** achipa has joined #harmattan13:07
*** eg81 has quit IRC13:08
*** lcuk has joined #harmattan13:12
Arkenoiis sip account plugin opensource?13:14
w00t_I don't think any of the UI stuff (apart from the framework) is13:16
Arkenoidamn that sucks13:16
* Arkenoi did hope harmattan is more open than fremantle13:17
Stskeepsapps/ui level was kinda announced way ahead of time not to be13:17
*** eman has quit IRC13:18
*** Rizzer has joined #harmattan13:21
*** lcuk has quit IRC13:30
SpeedEvil -1th cut on questions about aegis.13:32
SpeedEvilThis is based on at best a skim over the available aegis docs, I need to do more indepth reading.13:35
*** epage has joined #harmattan13:36
*** NIN101 has quit IRC13:37
*** baraujo has joined #harmattan13:37
*** lcuk has joined #harmattan13:38
*** lcuk has quit IRC13:38
*** lcuk has joined #harmattan13:38
MohammadAGno copy/pasting in terminal eh?13:40
SpeedEvilComments on the above MohammadAG?13:41
kimjuMohammadAG, not yet, at least.13:41
fralsMohammadAG: its OSS so patches welcome im sure ;)13:42
djszapikimju: tokens are documented btw, we did instead of the relevant teams internally....13:42
djszapithat will be fixed :)13:42
djszapiSpeedEvil: we already overdiscussed the first question of yours here :)13:47
djszapiplease make notes about things we discuss, if it is not evident.13:47
MohammadAGSpeedEvil, still haven't hit anything from aegis13:47
MohammadAGit's been with me for an hour :P13:47
MohammadAGoh and I set relaxed exec on, that was the first thing I did13:47
SpeedEvildjszapi: Part of the problem is I'm trying to work out state from others comments, available docs, without an actual device.13:48
djszapihehe :)13:49
SpeedEvilOther than RDA - which as I understand it is not running a similar image.13:49
djszapiRDA ?13:49
SpeedEvilThe above is intended as a framework others can jam stuff into.13:50
fralsremote device access13:50
SpeedEvilRemote Device Access13:50
djszapibtw, it is possible to work with qemu as well to emulate images in the future13:50
djszapiI guess it will fit to the general QtCreator approach.13:51
SpeedEvilSome of the development stuff I'm interested in will never be supported by qemu13:51
SpeedEvilFor example - USB host.13:51
djszapiI am just saying that will also help some people and some things.13:51
*** lardman has joined #harmattan13:52
*** lardman has joined #harmattan13:52
lardmanmorning all13:52
SpeedEvilQemu is a useful tool, though I assume aegis can't work under qemu?13:52
SpeedEvilOr is there a seperate certificate.13:52
djszapiSpeedEvil: From what I see, almost all the questions were discussed here.13:53
SpeedEvil lardman - somewhere to stick questions. Some of them may be bad questions, which can be linked to docs.13:53
djszapiSpeedEvil: I think it would be better to send to the mailing list than highlighting everybody :)13:54
djszapiif any.13:54
djszapiif there is no mailing list, then forum I guess.13:55
SpeedEvilThe questions were discussed - I don't recall answers - I may have misinterpreted the answers, or not been present. However, IRC is a volatile medium, and not the best place for docs.13:55
djszapithat is why I have been saying for a while for everybody to document the hard bits discussed here :)13:57
djszapiI always note the hard parts on the wikipage.13:57
SpeedEvilWhich is why I created the above - to have somewhere to scribble on.13:57
SpeedEvilWhat wikipage?13:58
Arkenoibtw iirc all early aegis presentations mentioned that it *can* be turned off, losing access to DRM-protected content if there is any, but everything other should work13:58
Arkenoiit is quite surprising that it turned out it cannot13:58
StskeepsArkenoi: that's my understanding too, with a self-made kernel, everything else is a bug13:58
Stskeeps(provided no sim lock, device lock)13:58
lardmanSpeedEvil: thanks13:59
SpeedEvilHas it been determined if any core binaries will fall over if aegis is not present?13:59
djszapiArkenoi: you can turn it off....13:59
djszapiopen mode...own kernel...13:59
djszapigazillion times repeated.13:59
SpeedEvilFor example - as addressed in the above page - what happens. Do you just lose angry birds, or does the dialer, maps, ... stop working.14:00
lardmanWhat we could do with is e.g. qgil asking someone from inside the Nokia security side to comment on the page14:00
lardmanthough I'm sure that's pretty clear to everyone14:01
djszapiSpeedEvil: what happens if you do not use the platform ? Mmmh, probably no guarantee for anything by nokia ?14:01
djszapilardman: accept that we are aware of these issues, already and the SDK team was boosted gazillion times.14:01
SpeedEvildjszapi: Which essentially means you can't do anything requiring any kernel mods for users.14:01
SpeedEvilIt's a devs-only hack.14:01
djszapiSpeedEvil: 1) It is very good in my opinion 2) It is not any different to N90014:02
SpeedEvilFor example.14:02
lardmandjszapi: we need it spelled out to us though ;)14:02
SpeedEvilThat is a fairly widely installed package, that cannot be duplicated on the n9.14:03
SpeedEvilThat's a fair difference.14:03
djszapiI am not getting you.14:03
SpeedEvil The custom kernel contains additional modules for IPv6, packet  filtering, QoS, NAT, tunneling, kernel configuration, Wifi mesh  networking, builtin ext3 for booting from other media, ext4, XFS,  reiserfs, NTFS read support, ISO9660, UDF, CIFS, automounter, squashfs,  unionfs, device mapper and dm-loop, cryptography, cryptoloop, EFI  partitions, UTF8 codepages, mouse+joystick input, PPP, PPTP, serial  support, US14:05
djszapiso ?14:05
SpeedEvilThat package cannot be used by users on the n9.14:06
SpeedEvilWhich is a fairly large difference to the n900.14:06
Arkenoi"It is very good thing"14:06
Arkenoinot really owning your own device is "very good thing"14:07
Arkenoithe major advantage of n900 was that you do not need to "jailbreak" it for useful things manufacturer did not keep in mind14:08
Arkenoithat's what "open" means14:08
djszapiSpeedEvil: I answered the questions on the wikipage.14:13
SpeedEvilWithout actually answering the question.14:14
lardmandjszapi: can you give some more details?14:14
SpeedEvildjszapi: SpeedEvil: what happens if you do not use the platform ? Mmmh, probably no guarantee for anything by nokia ?14:14
lardmane.g. qualify ""You can still use it with own kernel image to a certain extent.14:14
SpeedEvilAnd seeming to contradict your earlier comment14:14
djszapiSpeedEvil: do you expect from Nokia to take the responsibility for your kernel ? :)14:15
SpeedEvilIf you can run your own kernel, but it requires replacing large chunks of the software that most users rely on, then it's not an option for users.14:15
SpeedEvilNo, it's a factual question.14:15
SpeedEvilIf you run your own kernel, what stops working.14:16
Arkenoiif "open mode" renders device half-unusable it is practically equivalent of "you cannot actually use open mode"14:16
SpeedEvil'open mode' - for example could mean that the dialer, camera, ... don't work.14:16
SpeedEvilHell - mce and bme too.14:17
dm8tbrDocScrutinizer51: which voltage would you consider low enough to test the recovery? I have one bl4d down to 2.7xV now14:19
SpeedEvildm8tbr: That should probably be low enough. I think that's around the hard shutoff in the n90014:21
SpeedEvilChecking now.14:22
*** epage has quit IRC14:22
lardmanif I've got the meego sdk installed, where does one find the sdk maintenance tool to update it to support Harmattan targets too?14:22
dm8tbrI'm draining it a bit more at around 100\Omega to make sure it stabilizes below 2.7514:23
djszapi14:16 < SpeedEvil> 'open mode' - for example could mean that the dialer, camera, ... don't work. -> nope, works just fine14:23
SpeedEvilIdeally it'd be 0V - the internal protection circuitry kicks in - providing no power.14:23
SpeedEvildjszapi: This will be the case on the release firmware?14:23
djszapiSpeedEvil: I think I will change my help policy, I will document more and talk less here :)14:24
dm8tbrSpeedEvil: *nod* prolly somewhere around 2.3? I'm currently seeing 2.414:25
SpeedEvilI wouln't go lower than that14:25
SpeedEvilThat's in the range of possibly damaging.14:25
SpeedEvilI'm surprised the battery let you do that.14:25
SpeedEvil2724 the n900 shut off at - the lowest voltage I have in logs.14:25
SpeedEvildm8tbr: If that's the case - many of the concerns about the possible impacts of the security model go away.14:26
dm8tbrok, the battery disabled itself. let's see when I see voltage again14:27
SpeedEvilIn principle - if everything goes right, and it always shuts off properly, there should be some months before the battery voltage gets low enough to not allow charging.14:29
SpeedEvilIf it does in fact require power in the battery to charge.14:29
dm8tbrI know someone who had a very low charge problem, the -Owait-charging fixed it14:30
SpeedEvilIt's great to have a recovery method that we know works from absolutely positiverly flat though.14:31
SpeedEvil(Or to know it won't)14:31
djszapiSpeedEvil: this is myth for me only: I could never reproduce it on any devices I had.14:36
dm8tbrok, this battery stays at 0V for now. I'm cooking another one that I'll try to hover around 2.714:36
SpeedEvildjszapi: The images internal nokia people have differ from released images.14:36
SpeedEvildjszapi: At least so I understand.14:37
SpeedEvilOf course - the above could be a bug which has been fixed.14:38
SpeedEvilRather than intentional policy.14:38
djszapiSpeedEvil: I cannot reproduce it neither with N9 nor with the published N95014:39
SpeedEvilIt may have ben a result of earlier poking around, not directly related to the error that seemed to cause it.14:40
SpeedEvilUnsure - can't reprouduce here, for obvious reasons.14:40
SpeedEvil'published n950' is the same image as released to devs?14:41
djszapiyes, of course.14:43
*** lcuk has quit IRC14:44
SpeedEvilIf this is representative of the released n9 - great. If we can run our own kernels, and no critical stock software falls over, many of the concerns dissapear.  Thanks.14:45
kimjulbt, is there anything I can do to try to find out why the obs doesn't like my package? (still the same problem, hangs in building state after succesfully compiling it)14:45
djszapikimju: I have been having the same issue for 3 weeks :p14:47
*** wazd has joined #harmattan14:49
kimjudjszapi, nice to know that I'm not the only one :)14:49
djszapiAlso, SpeeEvil:
djszapiThat is a good starting point about security and linux...14:50
dm8tbrSpeedEvil: device with dead battery now hooked up to a wall wart. the led in the corner came up immediately. let's see.14:50
SpeedEvilGood sign.14:50
*** eman has joined #harmattan14:53
* RST38h writes off the list of places to seek help14:55
RST38hKiddies, *again*.14:56
kimjuit's a web forum, what did you expect?14:57
djszapiSpeedEvil: I think I found a commit relevant to that. It seems that issue was addressed, but I cannot still reproduce the issue :p14:57
djszapiso if any, it is probably an image bug.14:58
*** suy_ has joined #harmattan14:59
djszapiDocScrutinizer51, to you too ^14:59
*** eman has quit IRC15:01
*** lcuk has joined #harmattan15:07
dm8tbrdo we have the led behaviour pre-boot documented somewhere?15:11
dm8tbrcurrently it went from solid to blinking15:11
*** suy_ has quit IRC15:11
SpeedEvilIf it's like the n900.15:12
SpeedEvilSolid is hardware.15:12
dm8tbrwell there you go, device boots15:12
StskeepsSpeedEvil: heh, no15:13
SpeedEvilBlinking is software - it boots into ACT_DEAD state15:13
Stskeepsif you look at serial output it's obvious it's nolo :)15:13
dm8tbrI had to keep the power button pressed for 10s or so15:13
SpeedEvilbelieve him.15:13
dm8tbrI'd need a jig to verify that ;)15:13
* SpeedEvil is feeling a fair bit more positive about the device now.15:13
dm8tbrany chance to get one? :D15:13
SpeedEvil(well - modulo the whole non-having thing)15:14
dm8tbrthe device just recovered from a truly and utterly dead battery15:14
SpeedEvilAnother interesting test would be to try that with a dead battery with .7R in series with it.15:14
SpeedEvil(two year old battery equivalent)15:15
SpeedEvilBut that's not really important yet. :)15:15
dm8tbrSpeedEvil: mmmhhh, I'd need to rip an bl4d apart for that...15:15
SpeedEvilYes - or make a patch lead.15:16
* SpeedEvil looks at his patch lead.15:16
*** hardaker2 has joined #harmattan15:18
djszapiSpeedEvil: one of the Linus comments about selinux: I must say, I agree with him about it.15:18
SpeedEvildjszapi: I'm not advocating it.15:26
SpeedEvilIn fact - I'd be totally in favour of Aegis - and it would be a positive selling point - if I could sign my own keys.15:27
SpeedEvilAs I understand it, the signature has to be done online by the SDK calling a nokia server somewhere.15:28
SpeedEvilIf there isn't verification of what's signed - inspection of source - a major reason it's a good idea falls over. It's all in the implementation.15:30
djszapiaegis-origin is actually written by me :p15:31
djszapithat is an important bit of the design.15:31
SpeedEvilI've no arguments with the concept of aegis - it's all in the implementation as to how likely I am at some point in the future to purchase a n9.15:33
SpeedEvilFrom a platform security point of view - something like the apple store where everything is verified for sanity would be a great thing.15:33
SpeedEvilBut this is some way in the future, and raises obvious issues. Aegis of course can do this.15:33
SpeedEvil(implementation = configuration largely)15:34
*** eman has joined #harmattan15:34
*** rm_work has joined #harmattan15:36
*** rm_work has joined #harmattan15:36
*** NIN101 has joined #harmattan16:04
lcuklardman, ping for later.  when you add items onto the eventwall, are they persisted,  as in when you reboot will the items remain there?16:05
lardmanhey lcuk16:05
lardmanI guess so, it looks like it's a DBus call, so the state is presumably held by the event wall app itself16:06
GAN900OK, I give up on using that thing daily.16:06
djszapilcuk o/16:06
lcukhence me asking, thp do you have any input on this16:06
lcukhiya djszapi16:06
lcukGAN900, which thing?16:06
* lcuk waiting for new battery for his n95016:07
*** fiferboy has joined #harmattan16:08
lcukmorning fiferboy \o16:08
SpeedEvillcuk: Why?16:10
lcukSpeedEvil, just wondering whether an app would need to repopulate it after a reboot16:17
lardmanfar too many twitter posts on mine16:18
lardmanwould be good to be able to determine what appears, and how long for, etc16:19
lardmanI also don't like the way the notifications vanish so quickly, I'd prefer to have to manually clear them a la Android16:19
thplcuk: what?16:20
SpeedEvillcuk: battery - why16:20
djszapilardman: You can actually eliminate manually them since they appear on the other screen anyway16:20
RST38hI have neither Facebook nor Twitter account. Can you imaging how happy my life is? :)16:20
lardmandjszapi: eliminate twitter posts?16:20
lardmanor facebook perhaps they are16:21
lardmanRST38h: :)16:21
djszapilardman: yup16:21
lardmanI'd prefer things like RSS news items to appear there16:21
lardmandjszapi: that's alot of manual work to do though16:21
kimjuany pointer how to prevent vkb from showing up?16:21
RST38hGot a + account though and it kinda sucks16:22
djszapilardman: one swipe more actually than that what you asked for :D16:22
*** apol has joined #harmattan16:23
RST38hWondering when Google geniuses finally figure out the difference between "have X in my circle" and "want to read X's posts"16:23
apolhow do we run qml apps in harmattan? there's no qml viewer?16:23
RST38hthere is16:23
RST38hyou can install it from the harmattan-dev repo16:24
RST38hbut it does not run properly, lots of errors16:24
lardmandjszapi: Is it possible to choose what data sources are accepted?16:25
lardmane.g. so I could completely disable facebook16:25
lardmanand also only enable some twitter sources, etc16:25
apolRST38h: so what is the preferred way?16:25
djszapilardman: sorry ?16:26
lardmanon the events wall thingie16:26
RST38hapol: Apparently code your own app with QDeclarativeView16:26
djszapilardman: there is a feed settings, for instance.16:26
fralslardman: settings-> notifications->16:26
fiferboylardman: You have twitter posts on the event wall?16:26
lardmanno facebook16:27
RST38hlardman: Actually, got a better question: How do I import my Google Reader RSS feed into Harmattan?16:27
frals-> feed on homescreen*16:27
lardmansome things are republished from twitter on facebook, hence my confusion16:27
djszapiapol: you need this one ? qt4-declarative-qmlviewer16:28
lardmanyay, finally I can see RSS feed info :)16:28
lardmanthanks djszapi frals16:29
apoldjszapi: well, I don't need it if I can't use it to distribute my application16:29
djszapiCan we update the image on the N950 device ? For instance I can use N9 out of Nokia with newer Pr images. Hence why asking.16:29
djszapiapol: qmlviewer is just for debugging purposes imho.16:30
lardmandjszapi: we have no source of newer images afaik16:30
djszapilardman: I have, trust me :p16:31
lardmandjszapi: shame we all don't too! ;)16:31
apoldjszapi: well, makes little sense to have everything in a scripting language and force developer to create a dummy c++ application just to load the scripts16:31
GAN900lcuk, what thing do you think?16:32
lardmanGAN900: what's up with it?16:32
lardmanOther than the lack of a notification light16:32
djszapiapol: I am not getting you. What I told is only that qmlviewer can be only good for development purposes.16:32
apoldjszapi: the alternative is to create a c++ app16:33
djszapiapol: yes, I know these things :)16:34
djszapihowever makes little sense for some simple Ui change test in some cases.16:35
djszapibut since you need to package it anyway, it is not much of an overhead.16:35
*** hardaker2 has quit IRC16:36
apol... yay?16:37
*** hardaker has joined #harmattan16:37
djszapilardman: Could you manage the packaging issue from the last night ?16:37
djszapisome dev package installationg.16:38
hiemanshudjszapi: if you dont put the qml in a resource file, you can make small changes without having to recompile or repackage16:38
lardmandjszapi: no, but it's not that important atm, so I'll revisit it later when I need to get it sorted - would prefer to get some coding done now rather than wasting time16:38
djszapihiemanshu: except that you either need to update the hash manually for every change or using some aegis tool which is even worse imho16:39
hiemanshudjszapi: well I did with a C++ backend and qml for the UI, worked just fine even after changes16:39
*** eman has quit IRC16:40
djszapihiemanshu: with some aegis tool ? Point is that I think it is better to follow the platform recommendation and for instance, what QtCreator does. On the other hand, I do not like tray files on my system.16:40
djszapiwriting a hash updating tool for every bit change is technically possible, not much worth doing it though in my opinion.16:41
hiemanshudjszapi: nope, no tools, changed file, restarted app, and it worked16:41
djszapiyeah, in some local area :D16:42
thpwazd: ping16:43
RST38hthp: Look what I found:
djszapihiemanshu: actually, for easier things, qemu might even be better.16:44
RST38hthp: And this little fragment is enabled on Harmattan, if I to believe debian/rules16:45
djszapiprolly as an embedded thing inside the QtCreator like it happened with other mobile platforms.16:45
djszapi* qemu-alike16:45
hiemanshudjszapi: qemu is very very slow :(16:46
djszapihieamnshu: not for me16:47
djszapiI do not feel any difference between swiping in an emulated qemu image and the device. It is almost the same speed here.16:48
djszapiNote, I am not a QtCreator expert, they might even have better technology and background experience with it to speed up inside that.16:48
djszapiand I think it is way very nice you can just run the application from QtCreator without all the transfer things if you make a trivial Ui alignment, for instance.16:49
hiemanshuor you could subsitute with, and use the symbian simulator16:50
RST38hMeanwhile: Nokia ditches letters for all-number names16:50
hiemanshuyeah Nokia 50016:50
RST38hdjszapi: You may be able to, but I can't16:50
RST38hdjszapi: None of the dh tools appear to work in QtCreator on Windows.16:51
vandenoevermore portable than Amiga 50016:51
RST38hSo, applications compile but fail at packaging16:51
vandenoeverand comes with batteries16:51
RST38hNokia 100 was a WinCE-based GPS unit afaik16:51
djszapiRST38h: actually it confirms me that the emulator I said nice...16:51
slaineSo the N9 is now called ?16:52
RST38hdjszapi: well, as none of the apps run for me, it is pretty useless16:52
RST38hslaine: 9? =)16:52
Stskeepsslaine: gruntmaster 300016:52
djszapiRST38h: nothing works for you since I have been here seriously :D16:52
RST38hdjszapi: Well, Maemo5 tools work for me16:53
djszapiRST38h: At least, it is something on your end since my colleague uses Windows and works oob16:53
RST38hdjszapi: Basic Linux GCC works for me. Visual Studio and even Symbian SDK work for me16:53
RST38hdjszapi: Android SDK and NDK both work for me16:53
RST38hdjszapi: Even SB-based platform SDK for Harmattan works for me, especially after some advice from mgedmin16:54
RST38hBut QtCreator does not, sorry =(16:54
djszapiwell, from what I can say, it works here with my colleague's PC.16:54
RST38hYea, it works in some galaxy far away16:55
fralsonly ever used qtcreator for coding, never got it to work with my workflow for development16:55
djszapiAlso, generally "does not work" is good for nothing, steps steps steps16:55
fralsso basically i edit the code, then i switch to terminal and compile and deploy ;(16:55
djszapiand try to report a bug against QtCreator if you think they did something wrong after you make sure.16:55
RST38hdjszapi: Steps: 1) Create an empty harmattan app 2) Run it 3) Creator tries compiling ,packaging, fails with an error message16:55
RST38hEnd of story16:55
djszapiWe cannot reproduce it here...16:56
djszapiIt is a bit slower on Windows, but that is...16:56
RST38hdh_make missing16:56
djszapisounds like an "invalid" bug :)16:56
RST38hyea, that is why I am not going to file it16:56
RST38hWill just say that QtCreator does not work when asked if anyone should use it16:56
djszapiRST38h: do you use msvc or mingw ?16:57
RST38hInstalled mingw when offered by QtCreator install script16:57
djszapiare you sure the debian tools are installed properly ?16:57
djszapiare they also in the path ?16:58
tommaRST38h, use remote compiler =)16:59
RST38hLook, I do not know16:59
RST38hI followed the instructions given to me by the QtSDK install script17:00
RST38hAny sane install script should take care of installing all necessary pieces into the right places17:00
RST38hAnd it certianly did not show any error messages during installation17:00
GAN900lardman, doesn't do any of the things I use my N900 for.17:02
GAN900lardman, excepting: make phone calls and send SMS17:02
RST38hwhat are the other things?17:02
djszapiRST38h: you never did debugging when something is not working ? :)17:02
djszapialso, was madde installed properly ? It does not actually have the "real" debian tools, it is just some short cut for those.17:03
GAN900RST38h, FBReader, IRC, web, GTalk.17:04
RST38hGAN: Ok. GTalk actually works, nicely17:05
*** Wirta has quit IRC17:05
divanRST38h, how did make gtalk work?17:06
RST38hGAN: then try clicking at the top icons to access other google services in mobile form. There will be a link to the mobile gtalk web applet, send it to your application menu17:06
divanRST38h, ahh, in browser?17:07
RST38hGAN: May want to book mark a few other of these mobile pages (Reader, News, Plus)17:07
RST38hSome of them show up in a very plain :generic" style, ignore these17:07
RST38hAlso, mgedmin compiled FBReader and MohammadAG compiled communi IRC client. Both work, albeit not optimally.17:08
hiemanshucommuni is usable but butt ugly17:08
RST38hsame with fbreader17:09
RST38hthey both need a few changes17:09
GAN900RST38h, FBReader segfaults loading any book.17:10
GAN900The web browser crashes on 2/3rds of the site I attempt to load.17:10
RST38hweb browser seems to work for me17:11
djszapihiemanshu: it would be nice to have something working on N9 as well. Someone being aware of Venemo's IRC client ?17:11
RST38hbut I am only lvisiting a very limited number of sites17:11
GAN900So, back to the N900. Will use it for testing until we get a retail image.17:11
hiemanshudjszapi: well I haven't spoken to venemo in a while so no idea really, but it should be too tough to write an irc client using libircclient-qt17:12
* RST38h will try QApplication::setFont() when he gets home tonight. In theory this should automagically make most standard Qt apps usable17:12
hiemanshuRST38h: or fix meegotouch-qt-style17:12
RST38htoo much work17:12
RST38hit is not even being used at the moment, afaik17:13
hiemanshuwell its usable, so fixing it would make all fremantle qt apps just work, well most of it17:13
djszapihiemanshu: I do not think it makes sense to reinvent the wheel since that has actually been what Venemo nicely did.17:13
djszapiHowever I would have used telepathy-idle, prolly.17:13
hiemanshudjszapi: well we'll have to wait till he is around then17:14
RST38hhiemanshu: crashes and burns on me17:14
hiemanshuRST38h: well it worked for my app on meego-ce17:14
hiemanshuand it showed up fine on SB too17:14
RST38hnot on the n950 though17:14
RST38hdunno why17:14
djszapinothing works for RST38h :D17:14
*** cpscotti has quit IRC17:14
lardmanSeems to me that if we had an extras-devel repo we might know about more of these apps ;)17:15
RST38hsee above for the list of things that work for me17:15
RST38hlardman: whip up your own17:15
djszapilardman: there is something like that, actually.17:15
djszapiwith no QA :p17:16
RST38hlardman: if you do, I will upload some stuff there =)17:16
lardmanRST38h: sure17:16
lardmandon't tell me I need to use obs17:16
hiemanshuwell I could set up a repo on my server if anyone wants one17:16
lardmanRST38h: ah don't worry about my how, I thought you meant one already existed17:17
djszapihiemanshu: there is already a short term non QA solution and one QA solution for the long run17:17
lardmanmust learn to not mix lines up while reading :)17:17
djszapiwhat is the sense of sharing it further on ? :)17:17
*** crevetor has joined #harmattan17:18
*** lbt has quit IRC17:18
djszapi* separating17:18
hiemanshudjszapi: is the long run one Ovi Store?17:18
*** seif has quit IRC17:24
*** djszapi has left #harmattan17:24
*** seif has joined #harmattan17:26
*** willer_ has joined #harmattan17:30
*** lbt has joined #harmattan17:39
*** lbt has quit IRC17:39
*** lbt has joined #harmattan17:39
*** mikhas has quit IRC17:42
GAN900Man, the LF is so acutely useless.17:42
GAN900Worse, they're a giant roadblock.17:43
* SpeedEvil passes GAN900 a fork.17:43
*** suy_ has joined #harmattan17:44
GAN900Why did it take until now for the LF to nix apps.meego?17:52
rm_worki assume there's an article or something related?17:56
*** leinir_ has joined #harmattan17:58
lardmanGAN900: so the lawyers could be paid for the whole time?17:59
*** leinir has quit IRC18:01
*** leinir_ has quit IRC18:03 sounds a bit crap imho18:03
*** rcg1 has joined #harmattan18:11
*** javispedro has joined #harmattan18:12
* javispedro comes back from holidays to find about the formeego fiasco18:13
Stskeepswell, patents do suck18:14
M4rtinK2lbt: any idea why I'm getting "ERROR: can't resolve libraries to shared libraries: glib-2.0" ?18:16
M4rtinK2while building Gobject-Introspection18:16
lbtM4rtinK2: sorry ... bit busy atm18:17
M4rtinK2lbt: np :)18:17
GAN900Stskeeps, patents don't suck outdated and subverted patent laws suck18:23
rm_workheard a good NPR segment on software parents last week18:24
GAN900and incompetent government agences granting and enforcing them.18:24
rm_workseems like maybe mainstream is picking up on the issue18:24
GAN900(but I'm being redundant)18:24
javispedroYet another "please do not ask harmattan questions on this ML"18:25
GAN900They should just send an autoreply from the mailserver to emails containing Harmattan.18:31
GAN900"Die! Die, die die! Productive scum!"18:31
*** slaine has quit IRC18:31
javispedrofunny and sad.18:32
*** rcg1 has quit IRC18:32
*** rcg1 has joined #harmattan18:32
* javispedro sighs at the idea of having to follow three forums to keep with harmattan news: tmo (where most n900ers are), fmc (where most previously n900, now n950 users are), and (where most previously s60, now n950 users are)18:37
javispedrofortunately, we have this channel...18:37
*** willer_ has quit IRC18:38
*** harbaum has quit IRC18:41
hiemanshu has anyone converted mp4 video to play on the N950?18:42
lardmanjavispedro: yeah, I've pretty much given up on fora until someone posts a link here18:42
lardmanjavispedro: how was your holiday, go anywhere nice?18:42
SpeedEvillardman: It's depressing when stuff that's useful gets left behind.18:43
*** npm has quit IRC18:43
lardmanSpeedEvil: as in information?18:43
SpeedEvilUsenet put discussion in one place. Supported a wide variety of clients, and had a distributed server architecture.18:43
*** leinir_ has joined #harmattan18:43
SpeedEvilEven global logging.18:43
SpeedEvilNow we have a billion forums all with shitty interfaces.18:43
lardmanI also find the myriad Meego mls too much to follow18:43
macmaNjavispedro: good point there18:44
GAN900javispedro, mwkn!18:45
javispedrolardman: Tuscany (florence, etc.) -- nice place, but too many peopl18:46
javispedroGAN900: indeed!18:46
lardmanjavispedro: that's the problem with going on holiday to sunny places in the summer ;)18:46
javispedrolardman: yep, there was also some kind of public transportation crisis resulting in huge traffic. didn't help :)18:47
*** npm has joined #harmattan18:47
suy_so, there is no specific mailing list for harmattan? only forums or this channel? :-(18:48
* lardman is currently enjoying the Mediterranean summer weather in the UK today18:48
javispedrotechnically, there's no specific Harmattan fora at all.18:48
*** willer_ has joined #harmattan18:48
alteregolardman: is nice isn't it.18:49
* javispedro is still waiting for someone (arjan? ;) ) to post that "Harmattan questions should better not be asked in the Application Developer Support" fmc forum.18:49
*** leinir_ has quit IRC18:50
lardmanalterego: quite amazing really, though I'm sure there's rain not far off!18:50
alteregoYeah, got very cloudy here now.18:51
alteregoLooks like it may rain at some point.18:51
lardmanhas been sunny all day here, which is nice, but ~27C in the shade now18:51
javispedroit rained heavily here last week, plants ruined -- looks like I chose a good week to leave18:51
javispedro*(here at barcelona)18:52
*** Arkenoi has quit IRC18:54
rm_workhiemanshu: all my mp4 play natively, without conversion18:55
rm_workthough i am working on figuring out the perfect profile for video18:55
rm_workwhen I get it figured out i'll let you know18:55
hiemanshurm_work: well the 720p mp4 I tested was very very laggy18:56
rm_work720p doesn't play18:56
rm_workthe screen isn't even 720p tho, so18:56
rm_workbasically you need to re-encode to native18:56
rm_workso it doesn't have to attempt to scale18:56
rm_worki think the scaling is what kills it18:57
hiemanshuI am using total video converter, and trying to figure out the best profile too, so far selecting PSP Video and giving a custom res seems to work18:57
rm_worki use mencoder with a script18:57
hiemanshuwith a very decent video output18:57
hiemanshuI have all my videos on my HTPC which runs windows18:57
rm_workwooo re-encoding ON DEVICE with mencoder :P18:58
* rm_work kries18:58
lardmanhiemanshu: Windows?! oh dear18:58
rm_workwhat size is our screen again?18:59
rm_work856x480 or something random like that?18:59
lardmansorry rm_work, not sure19:00
* lardman calls it quits for the afternoon, catch you all later on19:00
*** lardman has quit IRC19:00
*** leinir_ has joined #harmattan19:04
hiemanshurm_work: 854 48019:04
hiemanshulardman|gone: yes, it works perfectly19:05
alteregoWow, that's the quickeset I've seen ice melt in aaages.19:06
*** npm has quit IRC19:06
*** djszapi_ has joined #harmattan19:08
*** djszapi_ has left #harmattan19:08
*** npm has joined #harmattan19:09
rm_workalterego: lolwut19:10
rm_workmmm, mencoder is getting ~8.3 FPS encoding video on my n95019:14
rm_workx264 to xvid19:15
hiemanshuI have a 30 min video, that will take about 75 mins to encode19:15
rm_workencoding my 4 minute video in 10 minutes19:16
rm_workso yeah19:16
rm_workabout right19:16
rm_work2.5x length19:16
hiemanshuwhich means, time to upgrade your PC!19:16
rm_workmy PC would encode this in ike 30 seconds19:16
rm_workbut since i don't have a USB cable on me and zero wireless connectivity at the office19:17
rm_worki get to do it on the phone19:17
rm_workat 8.3 FPS :P19:17
M4rtinK2what about using the DSP ? :)19:17
rm_workdunno if there is a driver for the DSP in mencoder for encoding lol19:18
javispedroit's a good idea.19:18
hiemanshurm_work: I am doing it on my PC :/19:18
javispedrothere should be in gst19:18
npmrm_work: do you have mencoder .deb? howbout ffmpeg?19:18
npmi think the gstreamer plugin idea woud be better19:18
rm_workyes i have mencoder deb19:19
*** Rizzer has quit IRC19:19
rm_workwhere's infobot19:19
* npm wonders about libvideo-renderer-gst-playbin219:20
* hiemanshu wonders if aegis did something to infobot19:20
rm_workmplayer deb contains mencoder19:20
javispedrooh, packrat is useful again.19:20
rm_workjavispedro: did you miss me fixing it last week? :P19:20
rm_workit REALLY should be in the topic now :/19:21
javispedroyes, I was on holiday.19:21
rm_workah lol19:21
javispedrowonder if we are going BACKWARDS regarding repos =)19:21
rm_workwell... i fixed it to work with harmattan19:21
rm_worknow you are caught up :P19:21
rm_worklol yes, it seems that way19:21
rm_workwow this poor thing is actually heating up a bit encoding19:22
rm_workhooo man and it's eating the battery like a mofo19:22
javispedroWas going to ask for mine to be added but either way I have nothing useful until I get closed-source headers there.19:22
javispedrothere being COBS19:22
javispedroI'll probably have to blow the dust off reprepro for the time being. It also goes well with this entire "going backwards" theme.19:23
npmlooks like is what i was looking for19:23
M4rtinK2nice domain name :)19:24
wazdwell, maybe I'll sound stupid, but is it hard to just port fapman to harmattan and use xtras as a defaault repository?19:25
wazdCause i really don't want to get back to those times with packrat :)19:25
*** leinir_ has quit IRC19:26
wazdsorry for my ignorance anyway :D19:28
javispedrowazd: the idea is to eventually have a common meego repo19:31
javispedrobut this seems more like a obstacle race than anything else.19:31
divanHow to get default theme background color in QML?19:33
rm_worknpm: yes that's my repo, lol19:34
rm_worknpm: what were you looking for? :P19:34 is johnx's domain19:34
wazdjavispedro: yeah, but we still don't have any packag manager except that extravagant built-in updater :)19:34
rm_worki pay for half of the VM so i can use it :P19:35
fiferboydivan: Do you mean set the colour scheme so all the widgets have the correct highlight colours?19:38
wazdfiferboy: ! o/19:39
divanfiferboy, Not exactly. Currently I don't want to change the theme. And now I have Rectangle with default color is set to 'white'. But all other widgets seems to use some kind of light grey color and I'd like Rectangle to be of the same color.19:40
fiferboydivan: I'm not sure how to get the colour of the default widgets19:41
fiferboyI don't know the answer to my intrepetation of your question either :)19:41
fiferboyI was hoping someone would, but it seems not possible in QML ATM19:42
fiferboyHi wazd19:42
wazdfiferboy: how's it going? Had crazy birthday celebration? :)19:45
fiferboywazd: It's going good.  Birthday was nice and quiet (the way I like it)19:47
wazdfiferboy: yeeeah, we all know how you like "quiet" :D19:49
wazdfiferboy: especially when it comes to songs from 80's-90's :D19:49
wazdfiferboy: anyway, do you have any apps to test? :)19:52
fiferboywazd: "Lexicon" is ready to test from my repo19:55
fiferboyYou can find it on Pack Rat19:55
*** smoku has quit IRC19:55
tommafiferboy, you mean themes window background color or similiar with Qt Components?19:56
* javispedro ponders when will be the right time to introduce the issue of nokia-binaries in suposedly OSS COBS19:56
fiferboytomma: IIUI in meego-touch-framework apps you can choose from a bunch of different colour themse that will make the widgets highlights follow the scheme19:57
fiferboyWith Qt Components I haven't found a way to have anything other than blue19:57
Stskeepsjavispedro: been there already19:58
Stskeepsjavispedro: in practice, the harmattan binaries aren't on the actual cobs, except as caching, AFAIK19:58
javispedroStskeeps: is this last week news? Last I remember where some long ago discussions about having to sign an EULA to use those in COBS19:59
Stskeepsjavispedro: no idea how it was solved eventually but sometimes there's easy solutions to things, ie, too trivial issues19:59
Stskeepsdo remember the cobs is driven by volunteers and they have more than enough in their hands after the patent fiasco20:00
javispedroyes, that's what "the right time" part was about ;)20:01
javispedrothey hadn't been imported at all initially due to the EULA thing, but will try now again and see if it's been solved with this rate technical solution to a legal problem ;)20:02
tommafiferboy, there is theme object but i have no idea what you can get from it20:02
*** aleksander_m has quit IRC20:03
tommafiferboy, ok UIConstants.js contains some color defs20:05
javispedrooh, funnilty enough, COBS dependency resolver now thinks it has nokia-binaries but builder VM sources.list doesn't.20:07
* javispedro retries, just in case.20:07
Stskeepsnot sure sources.list matters20:07
Stskeepsisn't pulled in that way20:08
*** antman8969 has joined #harmattan20:08
rm_workmy favorite response to most of those sort of legal issues is "i'm doing it anyway, you can go fsck yourself. sue me." just in this case it's a lot more likely someone will take that suggestion to heart20:09
javispedrothis is as usual probably because I'm doing weird things20:09
javispedroI have this on build-depends: "opengles-sgx-img-common-dev [armel]"20:10
Stskeepsright, there's something about obs not being fond of those20:12
*** leinir_ has joined #harmattan20:13
fiferboytomma: How do you access the theme object?20:16
fiferboyI have seen reference to it before20:16
wazd <- what do you guys think? :)20:23
fiferboywazd: Mockup or actual code?20:24
wazdfiferboy: mockup but it won't take too long to make such an app I guess20:24
fiferboyLooks nice20:25
wazdAlso I'm thinking about color-coding temperatures20:26
wazdlike orange for hot and blue for cold20:26
npmrm_work: thanks for the repo ... i wanted mplayer ... was also looking for ffmpeg.20:26
wazdbut I'm not sure bout it20:26
* npm an hour late20:27
GeneralAntillesrm_work, yeah, but what if they do? :P20:27
rm_worklol dunno :P i go broke?20:29
rm_worknpm: i'll look at compiling ffmeg20:29
javispedro"The setup of the repository is broken, build not possible"?20:31
npmwhile you're at it :-) ...
npm^^ rm_work20:31
* javispedro attacks OBS with Delta Force .... critical damage!20:31
fiferboyjavispedro: Do you have multiple source code versions for that pakcage?20:31
npmthanks for ffmpeg. rm_work.... it's actually a dependency for some functionality in
javispedrofiferboy: you mean multiple source packages? no, I delete the previous one before commiting a new one.20:33
npmas the "download to audio" makes use of it to extract audio from videos20:33
rm_workyou can do that with mencoder20:33
rm_workbut yeah20:33
fiferboyjavispedro: That's what I meant.  I ran into trouble until I deleted the old ones20:33
rcg1anyone knows how to add entries to the status indicator menu drop down list?20:33
wazdfiferboy: <- what do you think bout color coding? :)20:34
npm(and if someone knows the gstreamer equivalent that would invoke the DSPs, that would be cool too)20:34
rcg1afaik you need but i can't find it anywhere20:34
npmalthough that may not be necc. for audio extraction20:34
javispedrofiferboy: thanks for the hint I am to blame though ;)20:34
rcg1another source of information is:
rcg1are these the right places to look at?20:34
fiferboyjavispedro: With OBS it is _always_ user error ;)20:34
fiferboywazd: That looks pretty cool!20:35
divanwazd, colors are pretty intuitive20:36
rcg1wazd: nice work.. and yeah colors are naturally understandable20:36
rm_worknpm: what version is spekle on?20:37
rm_workdon't see it anywhere20:37
lcukwazd, ++ on the subtle color hinting20:39
lcukvery nice effect :)20:39
wazdmaybe we can go even further and slightly change "today" background :)
RST38hmoooo wazd20:40
wazdRST38h: o/20:40
javispedroSeemingly I broke OBS =)20:51
javispedroah well20:51
javispedrolooking at the queue I see I'm not the only one who did20:51
M4rtinK2so it was you :)20:53
RST38hjavispedro: evening!20:54
M4rtinK2looks pretty dead20:54
javispedroevening RST38h, got the device?20:54
RST38hjavis: yea20:54
javispedroM4rtinK2: no wonder, with four packages stuck there for hours20:54
RST38hjavis: currently hacking QMainWindow, trying to get access to that bar at the bottom20:54
RST38hI have found a place where Qt creates it in the source code20:55
javispedroyou are using old school Qt?20:55
javispedroare you sure it's not the bar drawn by mcompositor?20:55
*** leinir__ has joined #harmattan20:56
javispedroand COBS is down!20:56
* javispedro pats himself on the bac20:56
rm_worknpm: wow, lot of depends here :/20:56
npmyes, if you speak of spekle, yes.20:58
*** leinir_ has quit IRC20:58
npmi think if i start working on that again, my head will explode :-)20:58
rm_worklol yes20:58
rm_workand ffmpeg20:58
javispedrowhen you upload ffmpeg, do you enable the restricted codecs or not?20:59
javispedroyou are talking about x264 so I bet yes?20:59
rm_worki'm working on ffmpeg20:59
rm_workgetting as many codecs in as possible20:59
rm_workbut it takes a lot of efforrt20:59
rm_worki'm better with mplayer20:59
* javispedro ignores the legal part of that and congratulates rm_work21:00
npmrm_work: sweet21:00
fiferboyDammit javispedro, I was trying to build something on OBS when you brought it down >:|21:00
RST38hjavispedro: anyway what did you do to obs? =)21:00
wazdRST38h: he used it, obviously :D21:01
*** leinir__ has quit IRC21:01
javispedroI hit "rebuild" on a package that had "repository configuration errors"21:01
javispedroit's now back up and it is building21:02
javispedroso I guess it worked!21:02
fiferboyI can't view the log of my application building, so I have to assume it is still going21:03
javispedroyeah me neither, and it's been building for 38 minutes already21:03
*** harbaum has joined #harmattan21:03
javispedrothat's impossible21:03
javispedroeither way someone's terminal package has been on the queue for 9 hours21:04
*** leinir__ has joined #harmattan21:06
RST38hOk, good news:21:07
RST38hthe QApplication::setFont() thing works21:07
hiemanshuRST38h: it has to :P21:07
RST38hSo, we can now have a decently working Communi IRC client, if MOhammad cares to make the change21:07
hiemanshuwell the buttons will still be small21:08
javispedrofiferboy: this time I think it's been you the one who crashed the COBS frontend =)21:09
javispedroyou job's also stuck.21:09
RST38hhiemanshu: No21:09
hiemanshuRST38h: fixing font size fixes that too?21:09
RST38hhiemanshu: The buttons will have to resize with the font21:09
RST38hat least they seem to do here21:10
M4rtinK2my GI build seems to be stuck too...21:10
lcukwazd, where is the omweather app running?  is there possibility to add to the lock screen?21:10
*** achipa has quit IRC21:10
M4rtinK2and I can't get to the overview page to cancel the build :)21:11
*** hiemanshu_N950 has joined #harmattan21:11
*** hiemanshu_N950 has quit IRC21:11
hiemanshuMohammadAG: around?21:12
javispedroM4rtinK2: according to this, it just went through21:12
*** divan has quit IRC21:13
hiemanshuRST38h: I cant remember the correct XLFD for the dejavu font to test21:14
hiemanshuusing the -font param21:14
RST38hhidemanshu: setPointSize(24)21:15
*** leinir__ has quit IRC21:15
RST38hhidemanshu: forget the fontfamily, although you can set it to "Nokia Pure"21:15
M4rtinK2ah, I see21:15
hiemanshuRST38h: well I want to test it without actually messing with the code, so wondering21:15
*** achipa has joined #harmattan21:16
*** achipa has joined #harmattan21:16
M4rtinK2yay for rpc timeout21:16
* javispedro does not understand what is going on behind COBS queue. Probably human intervention. Thanks then.21:16
RST38hhiemanshu: -font will not work. Dunno why21:17
*** djszapi|windows has joined #harmattan21:17
hiemanshuRST38h: how did you try to use -font?21:17
hiemanshujust normal name or size wont work really21:18
hiemanshuyou need to use XLFD for the font21:18
hiemanshualso IIRC Qt compiled with fontconfig will not accept the -font argument21:18
hiemanshuRST38h: 'The font should be specified using an X logical font description. Note that this option is ignored when Qt is built with fontconfig support enabled.' from the docs21:20
*** antman8969 has quit IRC21:20
RST38hhiemanshu: <sigh> It works when you do QApplication::setFont().21:21
RST38hhiemanshu: I do not care why it does not work otherwise.21:22
hiemanshuRST38h: yeah, because that it how you would normally do it too21:22
javispedrothe communi client looks decent21:23
javispedroin fact it might even require less paintwork than xchat to make it pass as a native app (if anyone wanted) =)21:23
M4rtinK2I hope "No AEGIS_HASH_FDS environment" is fixed in the near future - it is making the logs LOOONG :)21:23
djszapi|windowsM4rtinK2: why ? They do not shop up after the real build, just during the very beginning, initialization stage21:24
fiferboyThere, my build went smoothly that time21:24
djszapi|windowsit is not a biggie since it does not pollute the real happenings21:24
hiemanshudjszapi|windows: lol @ |windows :P21:24
djszapi|windows* shop = show :)21:24
javispedrofiferboy: yeah, my resubmitted build also went fine this time (and I didn't change a letter, so it wasn't my fault after all ;P )21:25
djszapi|windowshiemanshu: well, yeah I made an NSIS installer for our cross-platform project by using cmake.21:25
RST38hjavispedro: XChat is gonna be hopeless, custom Gtk display widget and all21:25
M4rtinK2djszapi|windows: but is is still part of the live log so the page with is like a kilometer long21:26
javispedroGtk apps are still NOT entirely out of the question..21:26
djszapi|windowsM4rtinK2: not sure why it matters.21:26
M4rtinK2javispedro: I'm actually trying to compile GTK and Clutter :)21:26
hiemanshujavispedro: but GTK is meh, and will never get a native theme, so well if its but ugly, meh21:27
javispedroCompiling is easy21:27
javispedrothe hard part is integrating21:27
M4rtinK2no i'ts not :)21:27
javispedroI already did it ..21:27
lcukM4rtinK2, ...21:27
M4rtinK2I'm working on it third day in row not getting any closer :)21:27
djszapi|windowsevening lcuk :)21:27
javispedroM4rtinK2: what problem are you hitting?21:28
M4rtinK2javispedro: oh ? :)21:28
lcukevening djszapi|windows21:28
javispedrohiemanshu: who needs a native theme when you have gtkqt ;) ?21:28
lcukdo you want to tinker with a little windows toy?21:28
hiemanshujavispedro: even that wont work until you fix meegotouch-qt-style :P21:28
javispedroit probably does not work with gtk3, but I'm more interested in the older Maemo versions of Gtk+...21:28
M4rtinK2javispedro: mostly with Gobject-Introspection21:28
M4rtinK2and also pixman does not want to build21:29
javispedroM4rtinK2: try your stuff on scratchbox first21:29
M4rtinK2well yeah, should be faster21:30
djszapi|windowslcuk: I am porting libliqbase to windows8 :)21:30
javispedroand, considering COBS seems to be introducing random failures, easier.21:30
M4rtinK2but it should still end on OBS in the end21:30
javispedroM4rtinK2: in the _very_ end.21:30
M4rtinK2javispedro: okok :)21:30
rm_worki've nearly given up on COBS for now21:30
lcukdjszapi|windows, porting?21:30
rm_workjust using SB to get anything actually compiled21:31
lcukif so cool!21:31
lcukI wasn't sure there was a yuv overlay to use21:31
lcukis that your new project?21:31
djszapi|windows=p :)21:31
RST38hjavispedro: take a look here:21:32
M4rtinK2javispedro: so basically I should just build binary packages in sb21:32
M4rtinK2install them and then compile packages that depend on them ?21:32
RST38hjavispedro: search for QMainWindowLayoutPrivate::init21:33
djszapi|windowsM4rtinK2: it cannot be that ultimate rule, try to use your sanity21:33
djszapi|windowsif a dependency cannot be achived, do not try to force it :)21:33
djszapi|windowsbut yeah, that straight-forward in general21:33
M4rtinK2well, some of the dependencies HAVE to be bogus :)21:34
M4rtinK2like espeak -> portaudio19 -> libjack-dev -> libraw1394-dev -> lynx-cur -> openssh-client :)21:35
djszapi|windowswell, I packaged portaudio1921:35
javispedroRST38h: you mean QMainWindowPrivate::init?21:35
djszapi|windowsI also left out the libjack part21:35
djszapi|windowsthat is reasonable, yeah ;)21:35
M4rtinK2I think I tried your package but espeak didn't like it21:35
javispedroportaudio on the n950?21:36
javispedroah, ok, I confused it.21:36
M4rtinK2djszapi|windows: I'll try again and report results :)21:37
djszapi|windowsI think my package is actually now broken21:37
M4rtinK2djszapi|windows: any idea which espeak version should I use ? :)21:37
djszapi|windowssince we made a shared community repository21:37
djszapi|windowsbut I think there are people who cloned it from me when it worked21:37
djszapi|windowsI always use the debian testing, wheezy21:37
javispedrodjszapi|windows: (shared community repo) WHERE? =)21:38
djszapi|windowsread the wikipage21:38
M4rtinK2djszapi|windows: this one seems to build:
javispedroso it works in that you pull packages instead of packagers pushing them21:40
* fiferboy sighs21:41
fiferboyIs there *really* no date picker in QML for Harmattan?21:41
javispedrofiferboy: there seems to be one in the qt components sample application21:42
djszapi|windowsjavispedro: packagers push packages to a shared repository.21:42
fiferboyjavispedro: Yeah, I am just looking there now21:42
fiferboyBut there is nothing mentioned in the documentation, so I have a feeling they implemented their own21:42
fiferboyAha!  It is in
fiferboyI never knew about this before21:44
fiferboyThis is what the "extras" gallery means!  I feel a whole new level of understanding21:45
javispedrodo not worry, the understanding will eventually dissipate.21:45
fiferboyjavispedro: I hope so, this is a very unfamiliar sensation21:46
rm_workoh man, so many depends21:47
rm_worknpm, you better be happy once i get ffmpeg :P21:47
rm_workwas wondering why i had only like 10 basic crap widgets to choose from21:48
rm_workand QML seemed kinda lame as i couldn't figure out where to get more21:49
javispedrowhat, are you using a VISUAL interface builder?21:49
rm_workQtCreator? :P21:49
Stskeepsjust write it in hand, you imbecile21:49
Stskeepseven my wife can write qt components and qml21:49
rm_worki do21:49
harbaumMine, too!21:49
rm_workbut i used the visual thing to see what they were called21:49
rm_workand to see what was available21:50
rm_workhow else would i know21:50
lcukrm_work, apparantly whinging in #harmattan21:50
javispedrothis visual thing sounds like wp721:50
fiferboylcuk: Hey!  My whinging led to a self-discovery!21:50
lcukof course21:51
lcukwho in here are also publishing their builds to rpm meego x86 for instance?21:51
*** Arkenoi has joined #harmattan21:51
lcukor is the obs here purely used for one kind of app?21:52
javispedrolcuk: I was just going to, wondering whether I can push both rpm spec and debian dsc to the same obs package or I need to create both21:52
Stskeepsobs can do both types of packaging in same21:52
Stskeeps'obs package'21:52
lcukjavispedro, would expect if your project contains both it would work, and Stskeeps confirmed that I tihnk21:53
javispedronice to know21:53
Stskeepsi think it's something about debian.tar.gz and stuff21:53
javispedroI do not know how that part works21:53
djszapi|windowsjavispedro: you can, there are gazillion examples ;)21:53
javispedroI read a bit about the debian part, and found that I can use old style 1.0 packaging21:53
djszapi|windowsyep, but does not make sense to use that21:54
javispedrodepends on where your interest is -- if I want to share source package with fremantle, it does :)21:54
M4rtinK2lcuk: I think there should be a wiki page/howto about this dual packaging :)21:55
djszapi|windowsjavispedro: please try to do quality packages.21:55
rcg1did anybody compile a vnc client yet?21:55
*** MohammadAG has quit IRC21:55
lcukM4rtinK2, for obs it should be a case of just putting the right info into the root folder21:55
Stskeepsi think harmattan should just man up and take rpm packages21:55
djszapi|windowseven vnc gui actually21:55
Stskeepsie, able to install meego compliant ones21:56
rcg1djszapi|windows: sweet.. got a link at hand?21:56
Stskeeps(yes, that's restricted to meego api)21:56
djszapi|windowsStskeeps: what do you mean ?21:56
javispedroStskeeps: if you believe the biggest problem with compliance is just that, well, that can probably be fixed...21:56
*** MohammadAG has joined #harmattan21:56
djszapi|windowsautomatic spec -> debian converter ?21:56
rcg1else i'll just google a little more later21:56
djszapi|windowsthat is not really harmattan specific stuff :p21:56
M4rtinK2lcuk: so I'll have a package that has both spec & dsc inside ?21:57
Stskeepsjavispedro: people have repacked opera for meego arm to harmattan..21:57
javispedrojust ship a rpm.deb with a preloaded db including the few key packages and call it a day21:57
Stskeepsjavispedro: yes21:57
Stskeepsin fact, i'm putting out 100 eur for those who will post such a solution :P21:58
djszapi|windowsM4rtinK2: may I have a question, why do you duplicate packages ?21:58
djszapi|windowsinstead of using the shared repository ?21:58
Stskeepsjust to show how easy it could have been to solve this crap once and for all long ago21:58
M4rtinK2djszapi|windows: what do you mean ?21:59
djszapi|windowsM4rtinK2: portaudio19 is already available at more places.21:59
javispedroupdate manager would be a problem21:59
M4rtinK2djszapi|windows: I just branched the package from you repository22:00
javispedroaegis on the other side can probably be fixed as there was at least work started to make it work with rpm22:00
M4rtinK2or is there another way ?22:00
djszapi|windowsuse the shared repository please22:00
djszapi|windowsjavispedro, no it was not ?22:01
M4rtinK2djszapi|windows: how does it work ?22:01
djszapi|windowsrcg1: this one ?
rcg1djszapi|windows: great! thanks22:01
M4rtinK2I'm learning as I go so I might have missed something important22:01
javispedrodjszapi|windows: there's this, no idea how complete it is:
djszapi|windowsif you miss a package feel free to propose and we can set you maintainer22:02
djszapi|windowsjavispedro: you are wrong, it has nothing to do with aegis22:02
djszapi|windowsit was for meego.22:02
javispedroand it is very, very, very, very similar to aegis.22:03
djszapi|windowsaegis cannot fix it....22:03
djszapi|windowssince it is not aegis project.22:03
djszapi|windowstwo different projects.22:03
M4rtinK2djszapi|windows: well I thought this is a repository for "working" packages22:03
djszapi|windowsM4rtinK2: yep, portaudio19 is working in sandst1's branch22:03
fiferboyBack to square one: " 1.1" doesn't work on N95022:03
djszapi|windowswe should just copy paste it :p22:04
M4rtinK2like I first make sure it works in my harmattan repository and then send a push request in the community one22:04
fiferboyjavispedro: You were right, understanding is fleeting22:04
djszapi|windowsM4rtinK2: it works in sandst1's branch22:04
djszapi|windowswe should just grab it for now22:04
javispedrodjszapi|windows: not saying it will work out of the box (but it even might), but the idea is very much the same (passing hashes to a fd)22:04
djszapi|windowsjavispedro: then again :) aegis cannot fix it.22:04
fiferboyAlthough, /usr/lib/qt4/imports/com/nokia/extras exists22:05
fiferboyMaybe just a version issue?22:05
djszapi|windowsaegis has nothing to do with rpm22:05
djszapi|windowsthe "Software Distribution Area" of aegis deals with dpkg wrapper, that is22:05
javispedrodjszapi|windows: what do you mean by "fix"?22:05
javispedrosign? let it run?22:05
djszapi|windows[22:00] <javispedro> aegis on the other side can probably be fixed as there was at least work started to make it work with rpm22:05
javispedroah, ok =)22:05
djszapi|windowsno aegis cannot be fixed by design, period22:06
djszapi|windowswhat you mean just do not know is called smack :)22:06
M4rtinK2djszapi|windows: isnt the one in sandst1's branch .spec based ?22:06
djszapi|windowsM4rtinK2: nop22:06
javispedroso there I meant "aegis can be hit with a large enough hammer so that it considers binaries installed by RPM as legit as those installed by dpkg"22:07
M4rtinK2this one ?
npmrm_work: i'll be happy even w/o ffmpeg.. but i'll be even happier with... thanks22:07
rm_worknpm: NEARLY there22:07
fiferboyFYI, 1.0 works on the N95022:08
rm_workhad to actually fix some source code to get it to compile, lol22:08
* javispedro ponders why do I always eventually end up talking about aegis22:08
* javispedro decides to call it a day and go rest22:08
rm_worknpm: vorbis and theora are my last two depends, compiling them now22:09
djszapi|windowsjavispedro: to be honest, I do not understand what is the purpose of your saying with rpm22:09
djszapi|windowsrpm is not the guaranteed stuff there, period22:09
M4rtinK2djszapi|windows: so when I see a working package in some repository, what is the right way of using it ? branching it to my repository or some other way ?22:11
javispedrodjszapi|windows: see above discussion with Stskeeps22:11
javispedrohave to really go now, cya22:11
*** javispedro has quit IRC22:11
djszapi|windowsStskeeps: I did not get the purpose of using rpm for fake purposes on Harmattan22:12
djszapi|windowsCould you ellaborate please ?22:12
djszapi|windowsI actually tried this Well I tried to change the fiasco image and I think validator init is protected as well as bb5 validator-init is checked by the kernel and bb5 is doing some stuff with the PA so even if you manage to replace the binaries you won't disable security22:12
djszapi|windowsM4rtinK2: well, if it works it is better to put into the shared repository22:13
djszapi|windowsso that others can have access to it easily.22:13
djszapi|windowsnot from 20-30 scrach personal repository :p22:13
M4rtinK2djszapi|windows: of course, this is the ultimate goal :)22:14
M4rtinK2but I'd really like to see that it works before dumping non-building package + half a ton of its dependencies that are not actually needed in there22:15
djszapi|windowsall the dependencies are available in the shared repository22:15
djszapi|windowsthere is some issue with portaudio19 build in the shared repository, but unfortunately I am not having time to investigate.22:16
M4rtinK2so should I just create a espeak package and submit it in the shared repository ?22:17
*** djszapi|windows has quit IRC22:20
*** wicket64 has joined #harmattan22:22
RST38hAny QML gurus here? (and I mean people who know what they are doing)22:25
*** hardaker2 has joined #harmattan22:25
fralsRST38h: whats up?22:26
fralsnot saying im a guru, but ive done a bit of qml :p22:26
*** djszapi|windows has joined #harmattan22:26
*** harbaum has quit IRC22:27
*** hardaker has quit IRC22:27
djszapi|windowsM4rtinK2: dodgy mobile internet :) Back to the question, I am just making local osc test with the shared repository usage for the dependencies and if it works I push it. Unfortunately, c-obs is tiresome for me to go through the same process twice.22:27
GAN900RST38h, no.22:27
RST38hanyway, I would like to run QDeclarativeView event loop without calling QApplication:exec(). Is there a way to do tht?22:27
GAN900We lack those people.22:27
RST38hIdeally, I want to start it and continue with my own execution when QDeclarativeView closes22:28
fralsthat, i dont know i'm afraid. i guess #qt-qml or #qt might know22:28
*** harbaum has joined #harmattan22:29
RST38hauke: Not familiar with Qt's event loops, are you? =)22:30
M4rtinK2djszapi|windows: I have send a submit request to the share repository22:30
M4rtinK2for Espeak 1.44 from Natty22:31
RST38hactually, found it already22:31
RST38hQEventLoop Loop;22:31
RST38hLoop.exec() (and Loop.exit() inside the loop)22:31
*** leinir__ has joined #harmattan22:32
djszapi|windowsM4rtinK2: you are a maintainer, please use it directly, try to make your best and feel yourself responsible if you break something :p22:33
rm_workwhat version of arm are we?22:34
rm_work--enable-neon --enable-armvfp --enable-armv5te --enable-armv6 --enable-armv6t2 --extra-cflags="-mcpu=cortex-a8 -mfloat-abi=softfp -mfpu=neon"22:34
rm_worki should change that to...22:34
M4rtinK2djszapi|windows: hh ok :)22:36
rm_workhrm don't see one for armv722:36
rm_workwe still cortex-a8 ?22:38
*** harbaum has quit IRC22:39
*** leinir___ has joined #harmattan22:39
Stskeepsno armv5te22:41
Stskeepsit blows up with mfloat-abi=hard22:41
*** leinir__ has quit IRC22:42
*** leinir has joined #harmattan22:45
*** leinir___ has quit IRC22:46
*** leinir has quit IRC22:50
*** leinir has joined #harmattan22:54
*** leinir has joined #harmattan22:54
RST38hBTW, can I have both PageStackWindow and some pages in a single qml file?22:58
fiferboyRST38h: Yes22:58
fiferboyCreate your pages inline and set the page stack initialPage to one of them22:59
*** achipa has quit IRC22:59
*** wicket64 has quit IRC22:59
RST38hfiferboy: You mean, I can't have them underneath pagestack and refer to them by names? Need to have them inside pagestack { } ?23:01
*** wicket64 has joined #harmattan23:02
*** smoku has joined #harmattan23:06
*** mikhas has joined #harmattan23:08
*** javispedro has joined #harmattan23:15
fiferboyRST38h: Yes, they have to go inside pagestack, but you can refer to them by id23:18
RST38hah, ok23:18
fiferboyA QML file is only allowed to have one top-level element23:18
RST38ha small change then23:18
*** meegoexperts has quit IRC23:18
RST38hand I guess for the message boxes one has to use MeegoTouch?23:19
javispedrotechnically if you use MeegoTouch you're doomed.23:22
RST38hIDs cannot start with an uppercase letter23:22
javispedroYou have to use the* components23:22
RST38hjavispedro: That is officially, not technically. Technically, mostof the Nokia's own UI uses meegotouch extensively.23:22
rm_workdamn, ffmpeg is defeating me23:22
javispedroI know, only Facebook and Maps use QML23:22
rm_work"gcc is unable to create an executable file."23:23
RST38hAnd, I only need a freaking message box, for Tentacled's sake!23:23
GAN900fiferboy, did you split the qml files yet?23:23
fiferboyRST38h: Yes, if the ID starts uppercase it is refering to a component file AFAICT23:23
rm_workhow is it failing this hardcore T_T23:23
fiferboyGAN900: Of course now, that would be productive :P23:23
RST38hbtw when MQL screws up it screws up really well, messes up the desktop23:23
GAN900fiferboy, weak sauce.23:23
fiferboyGAN900: Have you made a good icon yet?23:24
GAN900Yes, but I'm not giving it to you until you split the files.23:24
* GAN900 me looks shiftily around the room.23:24
* GAN900 coughs.23:24
djszapi|windowsjavispedro: you cannot bypass the security by using rm23:24
javispedroah, let's continue :)23:25
djszapi|windows"I actually tried this Well I tried to change the fiasco image and I think validator init is protected as well as bb5 validator-init is checked by the kernel and bb5 is doing some stuff with the PA so even if you manage to replace the binaries you won't disable security"23:25
javispedroah, new topic23:25
fiferboyGAN900: Show me a grainy blurry-cam shot of it and I will split the files23:25
javispedrovalidator-init is checked by the kernel? hm..23:25
djszapi|windowsof course23:26
Stskeepsrm_work: don't have mfloat-abi=hard23:26
Stskeepserr, =softfp23:26
rm_worklol k23:26
rm_worki was like "yeah, not using hardfp" :P23:26
rm_workworking i think23:28
javispedroah yes, bootloader passes the hash of validator-init23:29
javispedrovia "vhash=stuffstuffstuff" cmdline23:29
javispedroIf there is R&D certificate present in the device then allow ... validator-init to run ... even if the SHA1 hash did not match."23:30
djszapi|windowsR&D certificate is not PR.23:31
javispedroI bet.23:31
djszapi|windowsand even if you could hack validator-init which is not the case, there is still a low-level bb5 defense23:32
javispedroeither way, before I forget, the RPM stuff was not about bypassing but rather allowing it to do the same dpkg does (for "" origin only)23:32
javispedroso that Meego compliant OSS RPMs can be installed as is.23:33
javispedrobut bb5 is used mostly for secure storage23:34
djszapi|windowsat least I like how debian cares about the API breakages. That is awesome that no distribution cares about.23:34
djszapi|windowsI would not really like to proceed with that workflow, I think the debian way is good enough.23:34
djszapi|windowsrpm packaging.... :)23:34
*** baraujo has quit IRC23:35
djszapi|windowsyes and no23:35
*** willer_ has quit IRC23:35
djszapi|windowsthere are other good bits there as well23:35
*** willer_ has joined #harmattan23:37
javispedrobut if you manage to control validator-init, you control the interesting bits -- you can disable enforcing. what you lose afterwards depends on whether DocScrutinizer's predictions are true or false ;)23:37
javispedrowill you lose cellmo? nokia account? maps? gps? /me has no idea.23:38
djszapi|windowsnot getting you23:38
djszapi|windowsyou cannot turn off validator-init with the pr images.23:38
RST38hOk, what do I do about this one: file:///opt/speccy/Config.qml:2:1: module "Qt.labs.components" version 1.1 is not installed   ???23:39
djszapi|windowsit is of course simply not possible.23:39
javispedro"for now" :)23:39
djszapi|windowshuh ?23:39
djszapi|windowsI think the guess itself really is very wrong23:39
rm_workthey say "it is simple not possible" for just about every security framework on every cellphone OS <_< and they keep getting hacked. i'll agree with javispedro "for now"23:39
djszapi|windowssince that would seriously mean you turned off the kernel part of the security23:40
fiferboyRST38h: Only version 1.0 is available on the device at this time, I think23:40
RST38hok, replacing23:40
djszapi|windowsrm_work: sorry, but this is really not an arguement23:40
djszapi|windowsif it cannot be broken, it is good enough period23:41
djszapi|windowseven if "anybody" says for now, yes, for now until it is not broken :)23:41
rm_worki think he's saying it will be broken23:41
RST38hfiferboy: any idea how I can bring the QML page to front, blocking my QMainWindow?23:41
djszapi|windows"for now" has not much technical meaning23:41
djszapi|windowsrm_work: no, it is technically impossible.23:42
rm_workand now back to what i saifd23:42
fiferboyRST38h: You are loading a QML component from C++?23:42
rm_workthey say "it is simple not possible" for just about every security framework on every cellphone OS <_< and they keep getting hacked.23:42
javispedronot really saying how, Nokia might even eventually relent and provide R&D cert.23:42
djszapi|windowsrm_work: give me the steps23:42
rm_workwe don't know YET. that's why we he said "for now"23:42
djszapi|windowsso what ?23:42
rm_worki believe someone will figure it out23:42
djszapi|windowswhat is the consequence of "for now" ?23:43
djszapi|windowswhat do we benefit from this statement ?23:43
rm_workyou keep saying it is never possible23:43
djszapi|windowsseems quite useless to me23:43
rm_workwe're just saying "it probably will be possible eventually"23:43
RST38hfiferboy: Yes23:43
RST38hfiferboy: Furthermore, I have got a normal QMainWindow and not running QApplication::exec()23:43
djszapi|windowsrm_work: sorry, but I do not have time for non-technical chit-chat. Come back, when you have the steps, period23:44
fiferboyRST38h: So what happens when it loads the QML component?23:44
rm_worki wish i had /ignore for IRC trolls :/23:44
RST38hInstead, I am processing events with QCoreApplication::processEvents()23:44
*** suy_ has quit IRC23:44
RST38hfiferboy: immediately covers it with my QMainWindow, does not let me swipe any more23:44
*** mikhas has quit IRC23:44
RST38hrm_you: You have /ignore23:44
RST38hrm_you: also have /kick and /mode +b, if you wish so23:45
fiferboyRST38h: You can _try_ setting the z property on your pagestack23:45
rm_workah xchat does have ignore somehow23:45
fiferboyRST38h: But it sounds like the QMainWindow should not be showing up at all?23:45
djszapi|windows"for now" :D:D:D23:45
RST38hfiferboy: Well, I would guess it should go to the bottom23:45
RST38hfiferboy: called hide() on it now, let us see how it works...23:46
*** mikhas has joined #harmattan23:46
rm_worksomeone tell me next time he says something23:46
rm_worki think i got it working23:46
javispedrojust forget about the entire thing23:47
rm_workit's not that really23:47
fiferboyRST38h: How is the pagestack being inserted?  As it's own window, or /INTO/ the main window?23:47
rm_worki *will* systematically ignore anyone on this channel who says *anything* "just can't be done"23:47
RST38hfiferboy: as its own. just added existing wqmainwindow as parent, trsting23:47
fiferboyrm_work: You just can't ignore everyone like that.  Can't be done.23:47
RST38hrm_work: add "you are not supposed to do it" morons23:48
RST38hrm_work: I have just ignored the entire for this reason23:48
rm_workthe reason we're all here is usually because we enjoy doing things that other people say can't be done :P23:48
* djszapi|windows is not here because of that23:49
fiferboyRST38h: I take it you are injecting some QML in FBReader?23:49
rm_workcellphone security frameworks is a silly example though, because it happens with just about every phone that comes to market, lol23:49
RST38hfiferboy: nah, much simpler for now23:49
RST38hfiferboy: trying to add a QML-based config page to my emulators23:50
rm_workso unless Nokia can do what Google and Apple can't, it should be *possible* to break23:50
javispedrothey have something23:50
javispedrothey were deadly serious, for some reason more than google or apple =)23:50
rm_workwell, i've been toying with ideas about how to break aegis23:51
fiferboyRST38h: Ah.  Sounds useful for FBReader in the future, though :)23:51
SpeedEvilDesolder the SOC, solder one on which doesn't believe in such things.23:51
djszapi|windowsjavispedro: yes, we got very positive feedbacks compared to android or apple23:51
djszapi|windowsthat is true.23:51
rm_workhaven't had time to try anything yet, and it's definitely very deeply embedded :) but i believe it can be done by someone dedicated enough23:51
RST38hfiferboy: QML is so orthogonal to the traditional Qt that I am afraid it will not be possible23:52
rm_worknpm: do you mind if i don't include x264 as an ENCODER in ffmpeg? it should still be able to decode23:52
fiferboyRST38h: The interaction between QML and Qt code is actually pretty useful, though23:52
*** baraujo has joined #harmattan23:52
RST38hfiferboy: Even for tis simple use case, I am seriously considering just going with the traditional Qt with a bit of Meegotouch where Qt does not23:53
RST38hfiferboy: I am trying to get this interaction working as we speek and failing miserably23:53
javispedrorm_work: btw, I have something for you :)23:53
fiferboyI have some complicated database models that are simple to use in QML, although model interaction is one of the better apects of this23:53
javispedrorm_work: those are the ffmpeg configure flags the pre1&2 used23:53
RST38hjust doesn't work outside of a single documented scenario =(23:53
rm_workpre1 / pre2?23:53
rm_workah lol23:53
*** leinir has quit IRC23:54
rm_worki'm using the flags from Maemo523:54
fiferboyRST38h: I guess a dialog to contain the QML doesn't help any?23:54
lcukfiferboy, in a similar way that visual basic worked well with activex,  qml does have its positive aspects and things it is good at23:54
*** djszapi|windows has left #harmattan23:54
javispedrorm_work: mentioning it because it is a rather complete set with all the popular commercial codecs23:54
fiferboylcuk: I am exceedingly impressed with the way QML list views work with my large database model23:54
RST38hfiferboy: On the contrary, a QML-based dialog would be very very nice to have23:55
RST38hWould solve my problem23:55
fiferboyUsing thread with signals and slots so as not to block the QML UI is quite easy too (after some trial and error)23:55
fiferboyRST38h: What about a Qt dialog with the QML loaded in it?23:55
lcukfiferboy, i have an app rolling around head recently23:55
RST38hnah, head enough threads with android, don't want to bring them here unless absolutely necessary23:55
lcukfor making a garden management system23:55
RST38hfiferboy: how do I achieve that?23:55
lcuksomething to require a plants database23:56
fiferboyRST38h: Create a QT Dialog and use a declarative view and create.component to populate it with the pagestack?23:56
fiferboyRST38h: Should be possible, but I haven't tried that yet23:56
javispedrothere's also the* qml dialogs... ;)23:56
fiferboylcuk: A plant spotting app? :)23:57
javispedrothey are clones of the meegotouch ones, written in qml23:57
fiferboyjavispedro: I can tell you right now, the datepickerdialog is not ready in 1.023:57
fiferboyNeither is the tumbler dialog23:57
lcukfiferboy, no23:57
javispedroWHAT? a plain slot machine date picker and is not even working?23:57
javispedrobut then why it's on the demo application?23:58
lcuka garden system, where I can using gps and climate23:58
fiferboyjavispedro: The powerful drop-in replacement isn't working23:58
lcukknow when to sow seeds etc23:58
lcukfor a blossoming garden at a specified time23:58
fiferboyjavispedro: For the demo application they pretty much write their own from scratch23:58
lcukcoordination system really23:58
fiferboylcuk: Horticulture made easy?23:58
javispedrofiferboy: if you have the time and the qml expertise, you could write your own and publish it23:58
lcuki suppose the same algorithm will be used already in cooking recipe programs23:58
lcukfiferboy, just idle thoughts23:59
javispedromake it calendar-ish instead of slot machine23:59
lcukspending time in the garden recently23:59
fiferboyjavispedro: That is pretty much what I am looking at now23:59
RST38hfiferboy: I tried embeddeding it into a dialog and filed23:59
RST38h(shit, I ate this keyboard)23:59
javispedrowow, you must have been really hungry.23:59
fiferboyRST38h: Is it doing the same thing? Showing the main window in front?23:59

Generated by 2.15.1 by Marius Gedminas - find it at!