luf | Can anybody help me with C? I really have no idea what's wrong. | 00:16 |
---|---|---|
luf | uuid_filter[len] = '\0'; g_warning("get_dict_property: uuid_filter[len] = %c", uuid_filter[len]); | 00:17 |
luf | I see in log: "get_dict_property: uuid_filter[len] = *" | 00:17 |
DocScrutinizer05 | what can I say, '\0' ? | 00:30 |
DocScrutinizer05 | you're sure this works as intended? | 00:30 |
luf | DocScrutinizer05: yes. It usually works ;) I need to make the string shorter ;) | 00:31 |
DocScrutinizer05 | also printing a char(0) is errr | 00:31 |
DocScrutinizer05 | nfc how g_warning is defined, but i'd guess passing 0x00 to a %c parameter is no good idea | 00:32 |
luf | It also works in e.g. bluez: devname[15] = '\0'; | 00:32 |
luf | DocScrutinizer05: please don't elaborate on g_warning. The problem is that uuid_filter[len] = '\0'; does nothing. | 00:33 |
DocScrutinizer05 | what do you mean by "does nothing"? | 00:34 |
luf | It doesn't assign end of "string" to that position. | 00:35 |
DocScrutinizer05 | how do you know? | 00:35 |
DocScrutinizer05 | you're not printing a string, you're printing a %c har | 00:36 |
luf | because g_warning I posted above and also next g_warning: | 00:36 |
luf | g_warning("get_dict_property: UUIDs filter = %s, len = %d, char = %c, full = %d", uuid_filter, len, uuid_filter[len], full); | 00:36 |
luf | get_dict_property: UUIDs filter = 00001106-*, len = 9, char = *, full = 0 | 00:36 |
DocScrutinizer05 | I just can say printf("%c", 0) is a bad idea | 00:37 |
luf | DocScrutinizer05: I don't want to have the first g_warning. I added it because the code doesn't work ... | 00:37 |
DocScrutinizer05 | then add something better than that | 00:37 |
luf | What? I don't understand. | 00:38 |
DocScrutinizer05 | e.g g_warning("get_dict_property: uuid_filter[len] = %s", uuid_filter) | 00:38 |
DocScrutinizer05 | or g_warning("get_dict_property: uuid_filter[len] = %x", uuid_filter[len]); | 00:38 |
luf | DocScrutinizer05: can you skip the %c? It doesn't work even without all g_warnings ... It doesn't work with your suggested g_warning("get_dict_property: uuid_filter[len] = %s", uuid_filter). | 00:41 |
DocScrutinizer05 | you pasted some code and asked what's wrong with it, I answered you shouldn't print char(0) | 00:42 |
luf | The problem is that it doesn't assign \0 to the uuid_filter at position len. | 00:43 |
DocScrutinizer05 | dunno what else I could say about that line of code | 00:43 |
luf | Is there some "mode" when the char *uuid_filter = "00001106-*" can't be set using the way uuid_filter[0] = 'l' ? | 00:45 |
merlin1991 | luf: is uuid_filer filled by something or by a static string? | 00:45 |
luf | static string | 00:46 |
merlin1991 | because char* whatever = "something"; whatever[5] = 'x'; does not work | 00:46 |
*** dhbiker has quit IRC | 00:46 | |
merlin1991 | because something is allocated in a read only block | 00:46 |
DocScrutinizer05 | yep | 00:46 |
DocScrutinizer05 | points to const area | 00:46 |
merlin1991 | what I meant :D | 00:46 |
luf | if (!get_device_properties (conn, (const char*) remote_devname, devprops, "00001106-*")) { | 00:47 |
luf | So this is const initialization? | 00:47 |
luf | static gboolean get_device_properties (Connection *conn, const char *dev, DeviceProperties *devprops, char *uuid_filter) | 00:47 |
merlin1991 | uuid_filter inside the function is a pointer to const memory if you cal lthe function like you do | 00:49 |
DocScrutinizer05 | anyway I'd thing you should get compile time warning | 00:49 |
DocScrutinizer05 | think* | 00:49 |
luf | DocScrutinizer05: no warning :) | 00:52 |
luf | Thanks you (both of you). | 00:52 |
DocScrutinizer05 | yeah, compiler flags can do amazing things ;-D | 00:52 |
DocScrutinizer05 | -Wall ? | 00:53 |
*** M13 has quit IRC | 00:53 | |
luf | yes | 00:53 |
DocScrutinizer05 | I really dunno shit about those flags | 00:53 |
merlin1991 | luf: did you fix it? | 00:56 |
luf | I'm rewriting it a little bit. But I'm now sure what's the problem. | 00:56 |
merlin1991 | you could make the function axcept a const char* and just copy the data inside | 00:57 |
luf | I can't change const char* (do you mean strdup?) | 00:57 |
luf | I recognized if it's prefix or whole string (based on ending *). So I rewrite it that I add parameter like gboolean is_prefix. | 00:59 |
merlin1991 | yep strdup | 01:01 |
*** dhbiker has joined #maemo-ssu | 01:06 | |
luf | It's working now :) | 01:37 |
*** LaoLang_cool has joined #maemo-ssu | 01:53 | |
*** arcean has quit IRC | 02:04 | |
*** jonwil has joined #maemo-ssu | 02:05 | |
*** LaoLang_cool has quit IRC | 02:07 | |
*** _rd has quit IRC | 02:18 | |
*** jon_y has quit IRC | 03:01 | |
*** jon_y has joined #maemo-ssu | 03:17 | |
*** NIN101 has quit IRC | 03:39 | |
*** Martix has quit IRC | 03:58 | |
*** Oposum has quit IRC | 04:11 | |
*** Oposum has joined #maemo-ssu | 04:12 | |
*** luf has quit IRC | 04:32 | |
*** M4rtinK has quit IRC | 04:33 | |
*** nox- has quit IRC | 05:02 | |
*** kolp has quit IRC | 05:11 | |
*** amiconn_ has joined #maemo-ssu | 05:46 | |
*** amiconn has quit IRC | 05:46 | |
*** amiconn_ is now known as amiconn | 05:47 | |
*** DocScrutinizer05 has quit IRC | 06:04 | |
*** DocScrutinizer05 has joined #maemo-ssu | 06:04 | |
*** dhbiker has quit IRC | 06:33 | |
*** dhbiker has joined #maemo-ssu | 06:53 | |
*** jonwil has quit IRC | 08:23 | |
*** Pali has joined #maemo-ssu | 09:51 | |
*** _rd has joined #maemo-ssu | 10:29 | |
*** _rd has quit IRC | 10:37 | |
*** Milhouse has quit IRC | 11:07 | |
*** Milhouse has joined #maemo-ssu | 11:08 | |
*** Milhouse has quit IRC | 11:15 | |
*** Milhouse has joined #maemo-ssu | 11:15 | |
*** jonwil has joined #maemo-ssu | 11:33 | |
Pali | kerio, now new profiled is in cssu-devel | 11:35 |
kerio | ^_^ | 11:36 |
*** Milhouse has quit IRC | 11:39 | |
*** Milhouse has joined #maemo-ssu | 11:40 | |
*** _rd has joined #maemo-ssu | 11:40 | |
kerio | Pali: not only that, i see | 11:48 |
kerio | OMG dsmetool poweroff | 11:48 |
kerio | i love you <3 | 11:48 |
Pali | sudo poweroff should work now too :-) | 11:49 |
kerio | yep, i see | 11:49 |
kerio | why both upstart and system-services are updated? | 11:49 |
Pali | sudo reboot working too | 11:49 |
kerio | same source package? | 11:49 |
Pali | kerio, because both deb packages comes from upstart src | 11:49 |
*** Milhouse has quit IRC | 11:50 | |
*** Milhouse has joined #maemo-ssu | 11:51 | |
Pali | merlin1991: In cssu-devel and cssu-testing are different deb packages rtcom-messaging-ui-portrait with same version | 11:52 |
Pali | and my apt-get still trying to upgrade something... | 11:52 |
Pali | we should consider to *not* rebuild Architecture independent packages when moving cssu-devel to cssu-* | 11:53 |
Pali | or always increase number | 11:53 |
kerio | the one from cssu-testing takes priority for me | 11:56 |
kerio | or maybe it's just newer | 11:56 |
*** luf has joined #maemo-ssu | 11:57 | |
*** Milhouse has quit IRC | 12:05 | |
*** Milhouse has joined #maemo-ssu | 12:07 | |
merlin1991 | Pali: well who's the culprit who uploaded to -deve? | 12:10 |
merlin1991 | *-devel* | 12:10 |
merlin1991 | btw which version has the better css? | 12:11 |
Pali | both are same | 12:11 |
Pali | I updated first version to devel | 12:11 |
Pali | and (propably you) updated recompiled (same version) from git to testing | 12:12 |
merlin1991 | and what's the difference? | 12:12 |
Pali | no difference | 12:12 |
Pali | only apt-get still show me "update" | 12:12 |
merlin1991 | "are different deb packages" | 12:12 |
Pali | and cycle between that two versions | 12:12 |
merlin1991 | hm weird | 12:13 |
Pali | deb packages has different md5sum but files are same | 12:13 |
Pali | control files has different timestamp, so tar part in deb is different | 12:13 |
Pali | so deb packages must be different | 12:13 |
*** M4rtinK has joined #maemo-ssu | 12:14 | |
Pali | btw, operator name widget from devel doing cyclic reboot :-( | 12:14 |
Pali | not I enabled R&D mode and disabled lifeguard | 12:14 |
Pali | I'm going to investigare what happened | 12:14 |
Pali | ping jonwil ^^^^^ | 12:14 |
jonwil | wierd that its doing that, its working just fine for me | 12:15 |
jonwil | I am running the same .deb file as was uploaded to -devel | 12:15 |
Pali | I see that hildon-something crashing after I updated packages | 12:16 |
Pali | so maybe there is also 3rd app | 12:16 |
jonwil | maybe | 12:16 |
jonwil | all I know is that its working for me | 12:16 |
Pali | going to investigate | 12:16 |
kerio | Pali: ah fuck, bootloop | 12:17 |
*** Milhouse has quit IRC | 12:17 | |
kerio | and yes, it's connui-home-cellular | 12:17 |
Pali | kerio, enable R&D and disable lifeguard | 12:17 |
kerio | Pali: nah, where's the .so? | 12:17 |
jonwil | so there is a legit problem with connui-home-cellular as it stands in cssu-devel? | 12:17 |
jonwil | its definatly working for me with no crashes | 12:19 |
jonwil | or boot problems | 12:20 |
jonwil | which means the people having the problems (if there are any) will have to debug it themselves and come up with the solution... | 12:20 |
Pali | /usr/bin/hildon-home crashing | 12:21 |
Pali | and crashing after "Jan 5 11:15:03 Pali-Nokia-N900 cellular: csd[1275]: ISI_SMS .901580> incoming_cell_broadcast(): Incoming cell broadcast" | 12:21 |
Pali | jonwil ^^^ | 12:22 |
jonwil | Do you have the logging for the widget enabled? | 12:22 |
Pali | no | 12:22 |
Pali | going to enable it | 12:22 |
jonwil | please do | 12:22 |
Pali | but I do not remember gconf keys :D | 12:22 |
jonwil | enable it and get some logs of just how far it gets when you boot | 12:22 |
jonwil | look in the source code for the gconf keys :P | 12:23 |
Pali | I know | 12:23 |
kerio | huh, it reboots for me even without the connui-home-cellular .so | 12:23 |
*** Milhouse has joined #maemo-ssu | 12:23 | |
Pali | so then we need gdb + debug symbols | 12:23 |
Pali | hildon-home-dbg | 12:24 |
jonwil | or connui-home-cellular-dbg :) | 12:24 |
kerio | Pali: hildon-home crashed even without the .so | 12:24 |
jonwil | since its not broken for me, I cant help you fix it | 12:24 |
jonwil | going to go back to looking for interesting things to reverse engineer. I gave up on clockd because I have no way to test a number of its features to make sure they work (I don't want to reverse engineer anything I cant properly test locally) | 12:25 |
kerio | Pali: i have no connui-cellular-operator-home-item.so and it still crashes | 12:28 |
Pali | kerio, so need gdb and backtrace | 12:28 |
kerio | Pali: how do i type esc in recovery console? | 12:28 |
Pali | hm, I do not know | 12:29 |
Pali | kerio, enable R&D and disable lifeguard | 12:30 |
kerio | i don't wanna do that! :c | 12:30 |
Pali | then reboot cycling will be disabled | 12:30 |
kerio | what's signal 11? | 12:32 |
kerio | oh segv | 12:32 |
kerio | anyway, the problem isn't connui-home-cellular i think | 12:34 |
kerio | Pali: does hildon-home know about profiles? | 12:34 |
Pali | kerio, there is plugin in status menu | 12:34 |
kerio | the status menu is hildon-status-area or whatever | 12:35 |
kerio | hildon-home is hildon-home | 12:35 |
Pali | I started hildon-home & hildon-desktop in gdb | 12:35 |
Pali | nothing crashed | 12:35 |
Pali | and working fine | 12:35 |
kerio | :s | 12:35 |
Pali | going to start also hildon-status-menu | 12:36 |
kerio | dangling pointers ahoy | 12:36 |
kerio | fine, i'll do it your way pali >:c | 12:36 |
kerio | Pali: what flags do i need? no-lifeguard-reset? | 12:37 |
Pali | yes | 12:37 |
jonwil | ok, so the problem isn't connui-home-cellular? | 12:41 |
kerio | jonwil: it deleted it and it still crashes | 12:42 |
kerio | s | 12:42 |
kerio | seems like it's hildon-home somewere else | 12:42 |
kerio | *somewhere | 12:42 |
kerio | but of course it works fine when ran inside gdb :s | 12:42 |
kerio | no, wait, i got a sigsegv | 12:44 |
kerio | with no debug symbols | 12:44 |
kerio | how do i specify the debug symbols? | 12:45 |
kerio | Pali: ah crap, hildon-home is .launched | 12:47 |
kerio | how do we debug that? | 12:47 |
Pali | I do not know | 12:47 |
Pali | under gdb it not crashing | 12:47 |
Pali | and without gdb it crash | 12:47 |
kerio | how are you launching it? | 12:47 |
Pali | $ gdb --args maemo-summoner hildon-home.launch | 12:48 |
Pali | $ maemo-summoner hildon-home.launch | 12:48 |
kerio | oh, what's maemo-summoner? | 12:48 |
kerio | like maemo-invoker but in its own instance? | 12:48 |
Pali | yes | 12:48 |
kerio | well, it still crashes for me | 12:48 |
Pali | in gdb too? | 12:48 |
kerio | but maemo-summoner doesn't | 12:49 |
merlin1991 | bt plz | 12:49 |
Pali | backtrace backtrace | 12:49 |
merlin1991 | :D | 12:49 |
Pali | kerio, do you know how to get backtrace from crashed app? | 12:49 |
kerio | gdb thinks everything is fine | 12:49 |
Pali | there is some env in bash or something similar... | 12:49 |
Pali | but I do not remebmer | 12:49 |
*** NIN101 has joined #maemo-ssu | 12:49 | |
kerio | Pali: maemo-summoner doesn't crash but i'm not getting the icons | 12:51 |
kerio | when was hildon-home thumbified? | 12:53 |
kerio | Pali: the only thing i can think of is the new profiled | 12:58 |
kerio | but everything else works fine :S | 12:58 |
kerio | Pali: downgraded profiled, libprofile0 and profile-data to stock | 13:02 |
Pali | kerio, doing it too | 13:02 |
Pali | and not hepled | 13:02 |
kerio | :c | 13:02 |
kerio | i can't remember what else i changed in the system before this | 13:03 |
Pali | I upgraded also zlib1g and libpng12-0 | 13:03 |
kerio | oooh | 13:03 |
kerio | i bet it's zlib1g | 13:03 |
Pali | maybe some of that two libs | 13:03 |
kerio | i was running libpng12-0 for a while | 13:03 |
kerio | but zlib1g is new | 13:03 |
kerio | and fairly important | 13:03 |
Pali | but why nothing else crashed?? | 13:04 |
Pali | zlib is used everyhere | 13:04 |
kerio | Pali: nothing else crashed so far | 13:04 |
kerio | also, hildon-home is a fickle bitch | 13:04 |
kerio | loads all sorts of plugins | 13:05 |
Pali | all profile libraries are ok | 13:07 |
kerio | Pali: i has a hildon-home :D | 13:07 |
kerio | downgraded zlib1g | 13:08 |
Pali | operator is ok too | 13:08 |
Pali | libpng12-0 is ok too | 13:08 |
kerio | confirm | 13:08 |
Pali | so zlib1g crashing... | 13:08 |
kerio | freemangordon: you bootlooped up | 13:08 |
luf | kerio: what version of libxml2 do you have? | 13:09 |
kerio | luf: weren't you also involved? | 13:09 |
kerio | :D | 13:09 |
kerio | luf: 2.6.32.dfsg-5maemo4+0m5+0cssu0+thumb0 | 13:10 |
luf | That's the reason. | 13:10 |
kerio | freemangordon: you bootlooped up the correct dependencies | 13:10 |
luf | Fuck it shouldn't be possible to install zlib1g without correct libxml2 | 13:11 |
kerio | luf: see my previous line | 13:11 |
kerio | i suppose i should test if everything boots with both of those | 13:11 |
luf | kerio: It's my fault ;) | 13:11 |
kerio | while i'm still on rd mode | 13:11 |
luf | can you show me dpkg -s zlib1g ? | 13:12 |
kerio | luf: from which version? | 13:12 |
luf | The last from cssu-devel | 13:12 |
kerio | luf: no mention of libxml2 | 13:13 |
kerio | perhaps there should be a Breaks: libxml2 (<< 2.6.32.dfsg-5maemo4+0m5+0cssu1) | 13:13 |
Pali | $ sudo apt-get install zlib1g=1:1.2.3-9.osso9+0m5 | 13:13 |
Pali | this ^^^ fixing reboot loop | 13:13 |
luf | It's there in sources ... | 13:13 |
luf | No apt-get install libxml2 will fix the bootlook | 13:14 |
luf | *bootloop | 13:14 |
luf | (from cssu-devel) | 13:14 |
jonwil | ok, so connui-home-cellular is NOT broken? | 13:14 |
kerio | jonwil: i told you already | 13:15 |
kerio | you're completely blameless <3 | 13:15 |
jonwil | ok, great | 13:15 |
Pali | kerio, you can quit vim also with SHIFT+Z+Z | 13:15 |
kerio | not when i'm inputting text, surely | 13:16 |
kerio | otherwise it would be hard to type two capital Zs in a row | 13:16 |
Pali | so you do not need ESC :-) | 13:16 |
luf | Hmmm Breaks seems to be ignored during build ... | 13:16 |
Pali | and ctrl+C to go normal mode | 13:16 |
kerio | gregoa: halp | 13:16 |
kerio | Pali: oh, k | 13:16 |
luf | We have to put it into Conflicts. | 13:16 |
jonwil | ok, so can someone tell me which systemui packages have been cloned so far? | 13:16 |
kerio | Pali: still, isn't there a ctrl+something combination that does esc? | 13:16 |
kerio | luf: no we don't, yet | 13:16 |
kerio | Pali: like ctrl+i that's the same as tab | 13:17 |
luf | kerio: remap the keyboard ;) | 13:17 |
Pali | jonwil: alarm, tklock, dev, powerkeymenu | 13:17 |
luf | Shift+Enter is the empty combination ;) | 13:17 |
jonwil | are there any others that we need cloned?> | 13:17 |
Pali | jonwil, systemui daemon itself | 13:17 |
jonwil | ok, that I cant clone because of some of the things it does :P | 13:18 |
kerio | hrmpf, ctrl+[ | 13:18 |
kerio | Pali: is there a way to set rd flags from the device itself? | 13:18 |
Pali | kerio, yes writing own C application :-) | 13:19 |
kerio | can't 0xFFFF do that? | 13:19 |
Pali | I do not know, but maybe yes | 13:19 |
Pali | I do not know if I implemented it | 13:19 |
kerio | ...you wrote it! | 13:19 |
Pali | you can try | 13:19 |
Pali | I used some OSS cal parser in 0xFFFF | 13:19 |
Pali | so I think write support via proprietary libcal is not in 0xFFFF | 13:20 |
Pali | kerio, but I have prepared open source replacement for initrd-tools DEB package | 13:20 |
Pali | and in that package is cal-tool binary | 13:20 |
Pali | which write to stdout cal flags | 13:20 |
jonwil | There is a functionally identical clone of libcal out there | 13:20 |
Pali | I can add new options for storing cal flags | 13:20 |
kerio | Pali: btw, uboot sometimes fails to turn the backlight on | 13:20 |
Pali | kerio, uboot has no bakclight support | 13:21 |
Pali | kerio, backlight is controlled via SPI interface | 13:21 |
Pali | and I did not have time for creating driver for it in buoot | 13:21 |
luf | :q | 13:21 |
luf | :D wrong window ... | 13:22 |
kerio | luf: anyway, just wait for gregoa to come and tell us that our tools are older than dirt | 13:22 |
luf | :D | 13:22 |
kerio | and then give us a workaround | 13:22 |
luf | kerio: we have to fix it ASAP. | 13:22 |
jonwil | if the wiki was working I would update http://wiki.maemo.org/Fremantle_closed_packages with annotations on which ones are worth cloning and which ones are not (for example cloning nokia-maps-core is pointless when there are better maps apps for the N900) | 13:22 |
kerio | google-search-widged must be cloned! | 13:23 |
kerio | *widget | 13:23 |
kerio | otherwise how will i make a shortcut to the google homepage in hildon-home? | 13:23 |
jonwil | and for example I suspect cloning evolution-data-server-addressbook-backend-sim would be of limited usefulness | 13:24 |
kerio | jonwil: for which purpose? | 13:25 |
jonwil | thats what I mean, there is no point in cloning that package | 13:26 |
jonwil | because there is no use for having a clone of it | 13:26 |
kerio | jonwil: *limited* use for having a clone of it | 13:26 |
Pali | jonwil, yes, clone of evolution-data-server-addressbook-backend-sim is needed | 13:26 |
kerio | unless it's statically linked | 13:27 |
Pali | in maemo5 there is no support for copy to sim | 13:27 |
Pali | you can copy only from sim | 13:27 |
kerio | also, we might want to update evolution-data-server to a newer, incompatible version | 13:27 |
luf | kerio: welcome in the hell. | 13:27 |
Pali | kerio, our evolution server is fork of old version, but speed is still better then in upstream | 13:27 |
luf | Pali: no it's not | 13:28 |
Pali | not now? | 13:28 |
Pali | I remember that fremantle fork was faster | 13:28 |
jonwil | btw, I suspect that evolution-data-server-addressbook-backend-sim is not what we need for copy-to-sim | 13:28 |
luf | Pali: there were some measurement and fremantle was one of the slowest ... | 13:28 |
jonwil | I suspect we need to figure out how to use libsimpb0 | 13:29 |
luf | I forget where ... | 13:29 |
jonwil | i.e. cellular services daemon sim phonebook plugin | 13:29 |
jonwil | which may not even support copy-to-sim at that level | 13:29 |
jonwil | in fact, do we even know if cellmo supports copy-to-sim? | 13:29 |
Pali | jonwil, in osso addressbook there is only button copy from sim | 13:29 |
luf | freemangordon: Breaks is ignored by our build env :( | 13:29 |
kerio | freemangordon: you bootlooped up, son! | 13:29 |
luf | > grep Breaks debian/control | 13:30 |
luf | Breaks: libxml2 (< 2.6.32.dfsg-5maemo4+0m5+0cssu1) | 13:30 |
luf | > dpkg --info ../zlib1g_1.2.7.dfsg-13maemo1+0cssu1_armel.deb | 13:30 |
luf | new debian package, version 2.0. | 13:30 |
luf | size 80252 bytes: control archive= 1031 bytes. | 13:30 |
*** luf has quit IRC | 13:30 | |
jonwil | until we know if the cellmo and csd support copy-to-sim, doing anything with evolution doesn't help | 13:30 |
*** luf has joined #maemo-ssu | 13:30 | |
kerio | jonwil: csd even supports multiple connections at the same time! | 13:30 |
Pali | jonwil: libtime, clockd and systemui is priority | 13:30 |
kerio | luf: try with << | 13:31 |
kerio | Pali said there's a difference | 13:31 |
luf | kerio: it's my mistake (the bootloop) | 13:31 |
kerio | only your mistake? | 13:31 |
luf | Yes. | 13:31 |
*** Milhouse has quit IRC | 13:31 | |
jonwil | well I cant do libtime, clockd or systemui | 13:31 |
jonwil | so I am looking for something else to do | 13:31 |
kerio | freemangordon: apparently, you didn't bootloop up | 13:32 |
kerio | sorry about that | 13:32 |
kerio | luf: have you tried with << instead of <? | 13:32 |
*** Milhouse has joined #maemo-ssu | 13:32 | |
Pali | jonwil, libplayback | 13:33 |
Pali | or systeminfo daemon | 13:33 |
jonwil | whats special about libplayback? | 13:33 |
Pali | jonwil, libplayback is needed if you want to play some music (e.g in OMP) in silent profile | 13:33 |
jonwil | ok | 13:33 |
Pali | or if you want to pause music playback when there is incomming call | 13:34 |
jonwil | ok, so we know why we want to call this library. But whats the benefit of having an open clone? | 13:34 |
luf | merlin1991: can you please remove the zlib1g from cssu-devel? | 13:34 |
jonwil | There is already a -dev package for it | 13:34 |
Pali | jonwil, yes but for open source media player you need 3rd nokia closed libary | 13:35 |
Pali | and libplayback was reason why I needed to decompiled prolog code :D | 13:35 |
jonwil | ok, and for sysinfo daemon, whats the benefit of having that open? | 13:35 |
Pali | using phone info on other platforms (also on maemo) | 13:36 |
Pali | if you want to retreive IMEI (and other stuff) you can do it only wit sysinfo | 13:36 |
Pali | another closed daemon and library is omap3camd | 13:37 |
jonwil | all sysinfo stuff comes from CAL | 13:37 |
Pali | some SIM identifier not | 13:37 |
Pali | (IMSI) | 13:37 |
Pali | omap3camd is needed for camera | 13:37 |
jonwil | best thing for camera is to use fcam (which IIRC can do camera on N900 without omap3camd) | 13:37 |
Pali | really? | 13:38 |
jonwil | I think so | 13:38 |
Pali | omap3camd doing something with kernel | 13:38 |
Pali | if you kill it image quality is bad... | 13:39 |
Pali | but you still got image from camera | 13:39 |
Pali | another closed daemon is alsaped | 13:39 |
jonwil | I dont see anything in sysinfod (or rather the list provided by sysinfo-tool -l) that indicates IMSI | 13:44 |
merlin1991 | luf: what's wrong with it? | 13:44 |
jonwil | IMEI yes, IMSI no | 13:44 |
kerio | merlin1991: can cause a bootloop due to wrong metadata | 13:46 |
jonwil | IMSI comes from Cellular Services Daemon | 13:46 |
jonwil | not sysinfo daemon | 13:46 |
* merlin1991 starts the banhammer for zlib | 13:46 | |
*** Milhouse has quit IRC | 13:46 | |
kerio | namely, it's missing a Breaks: libxml12-0 (<< 2.6.32.dfsg-5maemo4+0m5+0cssu1) | 13:46 |
kerio | if you can add that instead, it would be also fine | 13:47 |
jonwil | I have checked and all things that sysinfod supports all come from CAL | 13:47 |
kerio | er, wtf | 13:47 |
kerio | libxml2 | 13:47 |
kerio | not sure where the 12-0 came from | 13:47 |
merlin1991 | luf: removed | 13:48 |
kerio | [general notice] if you installed the new zlib1g from cssu-devel, make sure you've also installed the latest libxml2 from cssu-devel, or you'll get a bootloop | 13:48 |
*** Milhouse has joined #maemo-ssu | 13:48 | |
jonwil | ok, so going through the list, alsaped was suggested as needing reverse engineering but its too complex for me. | 13:52 |
jonwil | bluetooth-sysinfo was done by me already :) | 13:52 |
jonwil | clockd I cant do because I cant test all its functionality | 13:53 |
jonwil | connui-* has thus far eluded my efforts to reverse engineer any of it | 13:54 |
jonwil | with regards to evolution-data-server-addressbook-backend-sim | 14:00 |
jonwil | it seems like that library is calling e_book_backend_set_is_writable | 14:01 |
jonwil | e_book_backend_set_is_writable(true) | 14:01 |
jonwil | so its already able to write to the SIM phonebook | 14:01 |
jonwil | so no changes are needed to evolution-data-server-addressbook-backend-sim to handle phonebook write, just need to change addressbook code itself to handle write to sim I think | 14:02 |
jonwil | moving down the list we have facebook-* which I wont be touching since I hate Facebook and refuse to have an account there :P | 14:02 |
*** Milhouse has quit IRC | 14:05 | |
jonwil | as for google-search-widget it would be easier (for someone with GTK skills) to write a new widget that does the same thing than it would be to reverse engineer it | 14:07 |
*** Milhouse has joined #maemo-ssu | 14:08 | |
jonwil | All the widget does is launches http://www.google.com/m?client=ms-nokia-maemo in the browser | 14:09 |
jonwil | all of the useful bits of hildon-im-* have been done already IIRC | 14:10 |
jonwil | libcityinfo is useless to clone without the underlying data that it uses (which we cant just copy from Nokia) | 14:11 |
luf | merlin1991: because package Breaks doesn't work :( | 14:16 |
*** Milhouse has quit IRC | 14:20 | |
*** Milhouse has joined #maemo-ssu | 14:25 | |
*** Milhouse has quit IRC | 14:32 | |
*** arcean has joined #maemo-ssu | 14:34 | |
*** Milhouse has joined #maemo-ssu | 14:36 | |
*** Milhouse has quit IRC | 14:42 | |
*** Milhouse has joined #maemo-ssu | 14:59 | |
luf | kerio: I uploaded fixed zlib package into cssu-devel. | 15:00 |
*** kolp has joined #maemo-ssu | 15:04 | |
luf | kerio: can you test refreshed zlib? I tested it and it can't install without proper libxml2 now. | 15:05 |
kerio | if you used Conflicts i'll be mad >:c | 15:05 |
luf | Feel free to be mad ;) | 15:05 |
luf | Better than bootloops. | 15:06 |
kerio | you could've waited for gregoa to wake up and report | 15:06 |
kerio | you *should've | 15:06 |
luf | And you can take a look into history that gregoa suggested Breaks or Conflicts (Breaks doesn't work) | 15:06 |
*** Milhouse has quit IRC | 15:06 | |
luf | kerio: you're bad boy :D | 15:06 |
luf | I'm going to work manually ... | 15:07 |
kerio | hm, i think he said that Breaks *should've* worked with our version of the tools | 15:08 |
*** Milhouse has joined #maemo-ssu | 15:09 | |
luf | kerio: ok I see that Conflicts ins't nice :D | 15:09 |
kerio | luf: why? what happens? | 15:11 |
luf | however it works : | 15:11 |
kerio | luf: hm, it's missing the Conflicts with zlib1 and zlib1g-udeb | 15:12 |
kerio | the stock zlib1g had that | 15:12 |
kerio | not sure if it'll change anything | 15:12 |
luf | http://pastebin.com/PsgvvDUU | 15:12 |
luf | leaving now | 15:12 |
kerio | heh, maybe Breaks would give out a better message? | 15:13 |
luf | kerio: Breaks doesn't work :( | 15:13 |
*** Martix has joined #maemo-ssu | 15:14 | |
kerio | Breaks doesn't pass properly between source package and .deb, but maybe it would still be parsed correctly | 15:14 |
*** Milhouse has quit IRC | 15:15 | |
Pali | please remove *udeb* packages | 15:20 |
Pali | that is used only by debian installer | 15:20 |
*** Milhouse has joined #maemo-ssu | 15:26 | |
*** Milhouse has quit IRC | 15:34 | |
*** Milhouse has joined #maemo-ssu | 15:34 | |
*** Milhouse has quit IRC | 15:40 | |
*** Milhouse has joined #maemo-ssu | 15:51 | |
*** xes has joined #maemo-ssu | 15:55 | |
*** luf has quit IRC | 15:59 | |
*** Milhouse has quit IRC | 16:00 | |
*** luf has joined #maemo-ssu | 16:01 | |
*** Milhouse has joined #maemo-ssu | 16:04 | |
*** Milhouse has quit IRC | 16:11 | |
gregoa | luf: "Breaks: libxml2 (<< 2.6.32.dfsg-5maemo4+0m5+0cssu1)" works in my scratchbox, i.e. it ends up in the binary package. "Breaks: libxml2 (< 2.6.32.dfsg-5maemo4+0m5+0cssu1)" leads to the warning "dpkg-gencontrol: warning: relation < is deprecated: use << or <=" and then I have "Breaks: libxml2 (<= 2.6.32.dfsg-5maemo4+0m5+0cssu1)" in the binary package. [tested with some random package I had lying around not with zlib] | 16:11 |
*** Milhouse has joined #maemo-ssu | 16:14 | |
kerio | gregoa: <3 | 16:14 |
kerio | gregoa: are you sure you have the original tools in your scratchbox, though? | 16:15 |
gregoa | kerio: that's indeed a good question -- what "original" means :) | 16:16 |
kerio | merlin1991: btw, can you add some CNAMEs to your dns records? and possibly some rewrite rules on your web server - i'd love to use urls like cssu-testing.merlin1991.at | 16:16 |
kerio | for the repos | 16:17 |
jonwil | still stuck for what to reverse engineer next... :P | 16:17 |
gregoa | kerio: I mean I'm quite sure I have a clean scratchbox but it's possible I've created it from a newer however-thats-called than luf | 16:17 |
kerio | so i can understand wtf is going on with apt-cache policy | 16:17 |
kerio | gregoa: do you reckon there's a way to make apt show the full repo URLs rather than just the hostname? | 16:18 |
gregoa | right, apt-cache policy output is mostly useless | 16:18 |
merlin1991 | kerio: hm what does a a browser send to the server when it follows a cname? | 16:18 |
gregoa | kerio: not sure, at least I don't remember right now, and I know that I had this question myself before. there might be something else ... | 16:18 |
kerio | merlin1991: the hostname - in this case, it would be cssu-testing.merlin1991.at | 16:19 |
kerio | not maemo.merlin1991.at :) | 16:19 |
merlin1991 | hm well for cssu-testing that isn't going to work | 16:19 |
merlin1991 | ham is a bitch ;) | 16:19 |
kerio | not without some configuration, yes | 16:19 |
kerio | on your part, i mean | 16:19 |
merlin1991 | nah I mean on the n900 hand | 16:19 |
merlin1991 | part | 16:19 |
*** Milhouse has quit IRC | 16:19 | |
kerio | well, we could eventually change the URLs in community-ssu-enabler | 16:20 |
gregoa | btw, what this "Breaks" question shows is IMO clearly that we want to have all packages for all repos (re)built on some autobuilder with a defined and clean environment. directly installing packages that someone has built on their machine has led to problems and will lead to problems | 16:20 |
merlin1991 | on my part it's a simple add cname ln -s in the /srv/http/ dir and rndnc reload merlin1991.at :D | 16:20 |
*** Milhouse has joined #maemo-ssu | 16:20 | |
kerio | merlin1991: but then the url will be community-devel.merlin1991.at/cssu/community-devel | 16:21 |
kerio | and that's ugly :( | 16:21 |
merlin1991 | err no | 16:21 |
kerio | oh you're going to move the directory too? neat | 16:21 |
merlin1991 | kerio: my setup is /srv/http/$hostname/STUFF | 16:22 |
kerio | if we want to follow the debian structure, it should be community-devel.merlin1991.at/maemo/ | 16:22 |
kerio | although it doesn't make much sense | 16:22 |
kerio | so maybe just on the / | 16:22 |
merlin1991 | so if I ln -s /srv/http/$oldpath /srv/http/$newhost then it's all nice and shiny | 16:22 |
Pali | merlin1991, luf, freemangordon, call 'echo 188.117.59.202 wiki.maemo.org >> /etc/hosts' to update maemo wiki | 16:23 |
kerio | ln -s /srv/http/maemo.merlin1991.at/cssu/community-devel /srv/http/community-devel.merlin1991.at | 16:23 |
Pali | and we can start use it | 16:23 |
merlin1991 | kerio: exactly :D | 16:23 |
kerio | also add one for extras-devel-light pls | 16:23 |
merlin1991 | pick a name ;) | 16:23 |
kerio | ln -s /srv/maemo.merlin1991.at/light/extras-devel/ /srv/extras-devel.merlin1991.at | 16:25 |
merlin1991 | hm I should rework my dns setup anyway, atm it's a bunch of A and AAAA entries, I should put cnames with a higher ttl and A for merlin1991.at | 16:25 |
jonwil | Pali: thanks for the link to the wiki | 16:27 |
jonwil | If I have that entry in /etc/hosts and am pointing to the new wiki | 16:27 |
jonwil | would it then be safe to edit it? | 16:28 |
*** xes_ has joined #maemo-ssu | 16:29 | |
jonwil | but yeah I think I will see what libplayback looks like :) | 16:30 |
kerio | merlin1991: and cssu-thumb too! | 16:30 |
*** Milhouse has quit IRC | 16:31 | |
*** xes has quit IRC | 16:32 | |
jonwil | damn, libplayback is using an opaque structure | 16:32 |
jonwil | but its definatly something reverse-engineerable I hope | 16:33 |
merlin1991 | http://extras-devel.merlin1991.at/ | 16:35 |
kerio | :D | 16:35 |
merlin1991 | the others I don't really want todo because you'd have to change the cssu enabler AND we'd run into the ipv6 problems again | 16:36 |
kerio | k | 16:36 |
*** Milhouse has joined #maemo-ssu | 16:37 | |
kerio | it doesn't really help with the apt-cache policy problem then, does it? :c | 16:39 |
kerio | meh, i suppose that now the packages with "thumb" in it are from -thumb and the ones without are from -devel | 16:39 |
kerio | so it did help, in a way | 16:39 |
*** Milhouse has quit IRC | 16:51 | |
merlin1991 | it got rid of 33% of the possible entries (in a way) | 16:53 |
*** Milhouse has joined #maemo-ssu | 16:53 | |
Pali | kerio, do you have bootmenu installed? | 17:22 |
Pali | I mean original or modified fanoush bootmenu script /bootmenu.sh | 17:22 |
kerio | i do | 17:23 |
kerio | with backupmenu and recovery console | 17:23 |
*** Milhouse has quit IRC | 17:23 | |
*** Milhouse has joined #maemo-ssu | 17:23 | |
Pali | I have open source replacement for text2screen, cal-tool and key_pressed binaries | 17:24 |
Pali | text2screen is used in bootmenu for rendering text | 17:24 |
Pali | do you want to try it? | 17:24 |
Pali | (text2screen & cal-tool comes from opendsme project, key_presses was written by me) | 17:24 |
Pali | I created debian package for it | 17:25 |
kerio | hm, not right now | 17:25 |
kerio | i'm pooping | 17:25 |
Skry | need scissors? | 17:41 |
jonwil | ok, I have been looking at libplayback and have come to the conclusion that its easiest to either use it as-is (the fact that its closed-source shouldn't make a difference to anyone except the most fanatical OSS zealots and I doubt anyone like that is using the N900) or to use the dbus interfaces (e.g. org.maemo.Playback) directly | 17:42 |
jonwil | Its clonable | 17:43 |
jonwil | but annoying | 17:43 |
jonwil | i.e. its too annoying to be worth the effort since changing it isn't something we have a reason to do | 17:43 |
jonwil | and since its really only a wrapper around some dbus stuff where the other end of the dbus link is open | 17:44 |
Skry | I prove your doubt about fanatic zealots not using N900 wrong | 17:45 |
jonwil | naybe | 17:45 |
jonwil | maybe | 17:45 |
jonwil | but there are far more important closed things on the N900 for people who care to complain about than libplayback :) | 17:46 |
jonwil | What we need to look at is cases where there is genuine value in cloning them | 17:47 |
jonwil | e.g. cloning clockd/libtime is definatly worth it (and I would do it if I could) | 17:47 |
Skry | sure, not complaining am I? I even welcome messy code with great joy and praise :) | 17:49 |
jonwil | :) | 17:54 |
*** Milhouse has quit IRC | 18:12 | |
*** Milhouse has joined #maemo-ssu | 18:14 | |
*** Milhouse has quit IRC | 18:22 | |
*** Milhouse has joined #maemo-ssu | 18:23 | |
*** luf has quit IRC | 18:39 | |
*** xes_ has quit IRC | 18:41 | |
*** dhbiker has quit IRC | 18:48 | |
*** jonwil has quit IRC | 18:51 | |
*** luf has joined #maemo-ssu | 18:55 | |
*** Milhouse has quit IRC | 19:02 | |
*** dhbiker has joined #maemo-ssu | 19:10 | |
*** luf has quit IRC | 19:19 | |
*** NIN101 has quit IRC | 19:23 | |
*** NIN102 has joined #maemo-ssu | 19:24 | |
*** luf has joined #maemo-ssu | 19:24 | |
*** NIN102 is now known as NIN101 | 19:34 | |
*** kolp has quit IRC | 19:57 | |
*** kolp has joined #maemo-ssu | 19:58 | |
luf | gregoa: Can you give me the package? I'll try it also. Maybe I did something wrong | 20:12 |
luf | gregoa: https://gitorious.org/community-ssu/zlib/ - here you can check how I did it (HEAD~2) | 20:12 |
luf | Pali: (zlib) udeb package is also in original maemo 1.2.3. So I did as few changes as possible. | 20:14 |
Pali | ok | 20:15 |
*** luf has quit IRC | 20:18 | |
*** luf has joined #maemo-ssu | 20:28 | |
luf | Pali: if you're sure about it I can try to comment it out (or remove) | 20:29 |
kerio | it's probably an upstream residual | 20:29 |
kerio | fremantle is loosely based on... etch? | 20:30 |
luf | kerio: I know but during upgrade I use the way of minimal changes ... | 20:33 |
kerio | luf: anyway, gregoa made the Breaks information work fine | 20:34 |
luf | kerio: I see he said it's ok for him. | 20:42 |
luf | But it doesn't work for me or freemangordon. | 20:42 |
kerio | gregoa: reveal us your magic | 20:43 |
luf | gregoa: https://gitorious.org/community-ssu/zlib/blobs/15518257be7ee967b07ea8fc7cc8b632fe593460/debian/control - this didn't work for me and freemangordon | 20:44 |
freemangordon | luf: hmm? | 21:10 |
freemangordon | aah, come on, not again :( http://support.microsoft.com/kb/2798897 | 21:11 |
*** M13 has joined #maemo-ssu | 21:13 | |
freemangordon | at least we have the sources tvis time https://gitorious.org/community-ssu/maemo-security-certman | 21:14 |
Sicelo | sucks | 21:16 |
freemangordon | merlin1991: http://googleonlinesecurity.blogspot.com/2013/01/enhancing-digital-certificate-security.html | 21:16 |
freemangordon | (all others too) ^^^ | 21:16 |
freemangordon | merlin1991: do you mind to post on TMO about it? | 21:20 |
Pali | I read about TURKTRUST too | 21:22 |
Pali | can we blacklist that CA? | 21:22 |
freemangordon | Pali: can we?!? WE MUST, if nor the whole CA, then at least the certificates in question. | 21:24 |
freemangordon | s/nor/not/ | 21:24 |
infobot | freemangordon meant: Pali: can we?!? WE MUST, if not the whole CA, then at least the certificates in question. | 21:24 |
freemangordon | though if goggle/M$/mozilla blacklist the root cert, we have to follow | 21:25 |
Pali | ok, do you know how to do in that maemo package? | 21:26 |
freemangordon | I don't, was just looking at https://gitorious.org/community-ssu/maemo-security-certman/commit/fc21ae95be7f44189be2360d445f4fe3f10ab712 | 21:27 |
freemangordon | this is for diginotar | 21:27 |
freemangordon | shouldn't be that hard | 21:27 |
Sicelo | debian is considering to remove the whole root | 21:28 |
freemangordon | well, maybe we should do the same then | 21:28 |
Pali | ok, we should release stable cssu ASAP | 21:28 |
Sicelo | http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=697366 | 21:28 |
Pali | with blacklisting that ca | 21:29 |
*** M13 has quit IRC | 21:29 | |
*** Milhouse has joined #maemo-ssu | 21:41 | |
freemangordon | Pali: stable? what about -testing :P | 21:44 |
freemangordon | going to make a thread on TMO, we can at least inform the "users" :) | 21:46 |
freemangordon | Pali: I think it is merlin1991/chemist to decide whether we will remove the whole CA or just compromised root certificates | 21:57 |
freemangordon | though TBH I tend to agree with the reporter on bogs.debian.org - this guys were aware of the issue back in 2011 and did nothing :( | 21:58 |
*** kolp has quit IRC | 22:01 | |
*** kolp has joined #maemo-ssu | 22:01 | |
Raimu | Humm, another cssu-devel update that messed up ext4 optfs today. | 22:03 |
Raimu | Sorry -- didn't mess as much as rewrote ext3 where ext4 was written and rebootlooped my n900. | 22:06 |
kerio | Raimu: eventually CSSU should fix the fstab situation | 22:06 |
kerio | hold on, can't ext4 be automatically identified? | 22:07 |
Raimu | Yup. It's easy to manually fix after updates for now. | 22:07 |
kerio | Raimu: it's just one place, right? | 22:07 |
kerio | rcS-late, where /home is mounted | 22:08 |
kerio | didn't apt-get ask for confirmation, though? | 22:09 |
freemangordon | Raimu: could you send a patch to Pali, I think it should be no problem to include it in CSSU | 22:09 |
Raimu | Naw, it overwrote genfstab.awk I think? | 22:09 |
freemangordon | kerio: afaik ^^^ is not a config file | 22:10 |
kerio | Raimu: ah yes, it does mount /home | 22:10 |
kerio | freemangordon: yeah, it's in /usr/lib | 22:10 |
kerio | Raimu: i just have if falsed the block that autogenerates the fstab | 22:10 |
freemangordon | why apt should ask then? | 22:11 |
kerio | freemangordon: i just figured out the problem | 22:11 |
kerio | rcS-late doesn't do the explicit mount | 22:11 |
freemangordon | aah, ok | 22:11 |
Raimu | kerio, suggest a way to make it a non-issue for me too if you can? | 22:12 |
Raimu | If-falsing which piece of scripting? | 22:12 |
kerio | Raimu: "if false; then" before # Generate fstab and mount /home | 22:12 |
kerio | "fi" after "rm -f $tmp_fstab" | 22:13 |
kerio | in /etc/event.d/rcS-late | 22:13 |
Raimu | Thanks. | 22:14 |
Raimu | freemangordon: I would if I was dapper enough with scripting | 22:14 |
gregoa | luf, kerio: sorry, I can't give you the package because it was just some random package in ~/src/ in my scratchbox, and I've removed the changes again. but I didn't do anything else than you, i.e. add the Breaks line to the binary package and run dpkg-buildpackage -uc -us | 22:14 |
kerio | gregoa: i didn't mean the package, i mean YOUR MAGIC | 22:15 |
kerio | or possibly the version of dpkg-buildpackage | 22:15 |
gregoa | kerio: re magic: :) | 22:16 |
gregoa | kerio: the question with the version is probably relevant indeed | 22:16 |
kerio | Pali: do you reckon we could add an official way to disable the fstab generation in rcS-late? | 22:16 |
kerio | like if the file /etc/default/no-auto-fstab exists | 22:17 |
gregoa | kerio, luf: ii dpkg-dev 1.14.25maemo3+0m5 | 22:17 |
luf | gregoa: same here | 22:24 |
luf | ii dpkg-dev 1.14.25maemo3+0m5 Debian package development tools | 22:25 |
Pali | FOR ALL: if you changing any file that is in /usr, first divert origin file and then edit it. system files which come from package system can be replaced without warnings!!! | 22:29 |
kerio | indeed | 22:29 |
kerio | well, you'll get a warning | 22:29 |
Pali | for diverting use: $ sudo dpkg-divert --local --rename --add <file> | 22:29 |
kerio | usually, apt-get telling you that it's about to update a package :) | 22:29 |
gregoa | luf: well, then I'm running out of ideas | 22:30 |
Pali | and about fstab problem: I think we can use autodetection for each partition (and fallback to default values if autodetection not work) | 22:30 |
kerio | Pali: when generating the fstab each time? | 22:31 |
Pali | ke-recv already doing it via HAL for SD card and for devices in usb host mode | 22:31 |
kerio | but i don't want to generate the fstab each time :( | 22:31 |
*** nox- has joined #maemo-ssu | 22:31 | |
Pali | only when not exists is good idea too | 22:31 |
kerio | why is it even autogenerated, anyway? | 22:32 |
Pali | no idea | 22:32 |
Pali | ask somebody from nokia | 22:32 |
luf | gregoa: can you try to make a zlib package from gitorious? | 22:32 |
Pali | or somebody who remember pr1.0 | 22:32 |
kerio | if we wanted to update it, we could just update base-files | 22:32 |
gregoa | luf: I can try but how does this help? | 22:33 |
luf | To see if it's in my env or mistake in zlib. | 22:33 |
luf | gregoa: HEAD~2 commit as I commited the workaround with Conflicts. | 22:34 |
gregoa | ok ... | 22:34 |
gregoa | luf: first attempt: fails to build with "dpkg-gensymbols: error: cannot open debian/zlib-core.symbols: No such file or directory" *sigh* | 22:38 |
luf | strange I rebuild it without problem. | 22:41 |
kerio | Pali: crazy: disable the autogeneration of /etc/fstab | 22:41 |
kerio | it's not like it can change from a boot to another, without changing the .awk file | 22:41 |
ShadowJK | It used to be able to change before devs started crying like little children that they don't know how to install stuff anywhere else than /usr, and Nokia had to move stuff to /home /opt to make room, after which the "use nand and usd if emmc is broken" logic stopped working and went away | 22:44 |
gregoa | luf: I now removed the 2 includes from the symbols file in order to build it. the resulting binary has this info: https://paste.debian.net/222025/ | 22:45 |
luf | gregoa: so the problem is in the build env ... | 22:45 |
luf | what dpkg version do you have? | 22:46 |
gregoa | luf: the same I pasted half an hour ago :) 1.14.25maemo3+0m5 | 22:46 |
luf | gregoa: You pasted dpkg-dev ... | 22:46 |
luf | But it's the same. | 22:46 |
gregoa | luf: yes because that's the relevant one :) (and the same) | 22:47 |
luf | gregoa: dpkg-buildpackage -uc -us - give me: dh_testroot: You must run this as root (or use fakeroot). | 22:48 |
luf | What sb env are you using? I'm using the virtual guest from Nokia. | 22:49 |
gregoa | puh, difficult question. I set this up years ago. with the debian packages for scratchbox and then some fremantle-target following some instructions | 22:50 |
gregoa | luf: does this help: http://paste.debian.net/222026/ ? | 22:51 |
luf | Yes. | 22:52 |
luf | Devkits: perl debian-etch qemu doctools svn git | 22:52 |
luf | This is the difference. | 22:52 |
gregoa | so s/etch/squeeze/ hm. interesting. | 22:53 |
gregoa | but still you see the same dpkg* version. -- and then something else is used behind your back? -- I guess we need someone who understand scratchbox better :/ | 22:54 |
freemangordon | gregoa: seems your SB is outdated, what is the version of scratchbox-core? | 22:54 |
gregoa | freemangordon: 1.0.27 | 22:54 |
freemangordon | the fuck?!? here it is 1.0.17 | 22:55 |
gregoa | freemangordon: and I had it in my sources.list until the host went down in december (?) ... | 22:55 |
freemangordon | gregoa: latest for SB1 is 1.0.17 AFAIK | 22:55 |
freemangordon | could it be you have SB2 installed? | 22:55 |
gregoa | not that I know of ... | 22:56 |
freemangordon | weird | 22:56 |
gregoa | at least I don't find any packages on my host with scratchbox2* or sb2* | 22:56 |
freemangordon | gregoa: though luf is right, now way to build in SB (here) without -frakeroot parameter to dpkg-buildpackage | 22:57 |
freemangordon | s/now/no/ | 22:57 |
infobot | freemangordon meant: gregoa: though luf is right, no way to build in SB (here) without -frakeroot parameter to dpkg-buildpackage | 22:57 |
gregoa | freemangordon: oh, interesting. | 22:58 |
gregoa | ah, the sb.org repo works again: https://paste.debian.net/222027/ | 22:59 |
freemangordon | gregoa: see, the first is 1.0.18 | 23:00 |
freemangordon | so you have SB2 | 23:00 |
gregoa | freemangordon: 1.0.x is sb2? wow | 23:01 |
freemangordon | SB1 is aphophis or something, can't remember the exact name | 23:01 |
*** Pali has quit IRC | 23:01 | |
gregoa | Apophis according to http://www.scratchbox.org/download/ | 23:01 |
freemangordon | gregoa: ofc I could be talking bullshit :D:D:D | 23:01 |
gregoa | ah, and this is Hathor: http://www.scratchbox.org/download/scratchbox-hathor/ | 23:02 |
freemangordon | right, so fremantle SB is apophis | 23:02 |
freemangordon | and you have hathor | 23:02 |
freemangordon | (at least it is apophis in vmware SDK images from nokia) | 23:03 |
gregoa | don't tell this to the fremantle_armel target in my SB :) | 23:03 |
freemangordon | ok, i'll keep it secret :P | 23:03 |
gregoa | well, yes, those images are not exactly fresh. and I followed some maemo instructions to set this up ~2,5 years ago ... | 23:03 |
gregoa | (and probably replaced etch with squeeze devkit while following along) | 23:04 |
freemangordon | gregoa: well, they are fresh(the images), SDK's been updated after PR1.3 | 23:04 |
freemangordon | latest SDK host OS is lucid | 23:04 |
kerio | >ubuntu | 23:05 |
freemangordon | yes | 23:05 |
gregoa | I guess that was http://wiki.maemo.org/Documentation/Maemo_5_Final_SDK_Installation#Installing_Maemo_5_SDK_on_32-.2F64-bit_Debian.2FUbuntu-based_distributions which points to http://www.scratchbox.org/download/scratchbox-hathor/ right at the beginning | 23:05 |
freemangordon | I see | 23:05 |
freemangordon | :( | 23:05 |
gregoa | lucid ... 10.04 -- just 2.5 years old :) | 23:06 |
kerio | well, it works-ish, isn't it | 23:06 |
kerio | *doesn't | 23:06 |
gregoa | I don't have problems with it, it's just that I can't reproduce luf's and freemangordon#s problems :) | 23:06 |
freemangordon | gregoa: well, I am lost then. And I am not exactly SB guru :) | 23:06 |
freemangordon | I don;t have problems ;) | 23:06 |
kerio | i doubt anyone understands SB | 23:07 |
freemangordon | luf: what problems had I? | 23:07 |
kerio | not even the SB devs | 23:07 |
kerio | freemangordon: Breaks isn't correctly parsed | 23:07 |
freemangordon | kerio: what? | 23:07 |
gregoa | freemangordon: I though you also couldn't get the Breaks added? | 23:07 |
freemangordon | well, i build the package as it was on gitorious. TBH didn't check for warnings | 23:08 |
freemangordon | which package was that? zlib? | 23:08 |
kerio | freemangordon: zlib1g | 23:08 |
freemangordon | i'll build it right now | 23:08 |
kerio | well, not right now | 23:08 |
kerio | it has Conflicts in the control file | 23:08 |
freemangordon | why? | 23:08 |
kerio | and not Breaks | 23:08 |
kerio | because Breaks didn't work for luf | 23:08 |
kerio | and didn't work in the past | 23:08 |
freemangordon | kerio: I have the version here which is in -devel | 23:09 |
luf | kerio: Breaks didn't work for freemangordon ;) and me too | 23:09 |
freemangordon | aah, i see | 23:09 |
luf | freemangordon: I reupload it due to missing Breaks. | 23:09 |
freemangordon | could someone check zlib in cssu-devel | 23:09 |
kerio | it caused a bootloop here and at pali's | 23:09 |
freemangordon | this is the version i built here | 23:09 |
kerio | it's been reuploaded, with a Conflicts | 23:09 |
luf | freemangordon: kerio ended in bootloop see the histor | 23:09 |
freemangordon | aah, ok | 23:09 |
kerio | but it's not the right thing - if you try to install it, it prompts to delete ~everything | 23:09 |
luf | Because zlib1g missing Breaks :( | 23:10 |
freemangordon | could it be because libxml2 was not pulled? | 23:10 |
kerio | a Breaks should be a bit more accomodating | 23:10 |
kerio | freemangordon: yep, because there was nothing in zlib1g referring to libxml2 | 23:10 |
gregoa | but still, a dist-upgrade should have installed at in parallel, no? | 23:10 |
kerio | gregoa: dist-upgrade is horrendously broken | 23:11 |
gregoa | what still confuses me is that we have the same dpkg* version _in_ sb | 23:11 |
freemangordon | dpkg-genchanges: warning: unknown information field `Breaks' in input data in package's section of control info file | 23:11 |
kerio | an upgrade should've installed it in parallel - but i have cssu-thumb pinned at 600 priority | 23:11 |
freemangordon | ok, sorry, my bad | 23:11 |
kerio | why is it unknown? :s | 23:11 |
freemangordon | i've missed that :( | 23:11 |
luf | freemangordon: my mistake ... | 23:11 |
kerio | unless dpkg-genchanges is somehow the one from outside scratchbox | 23:11 |
kerio | in which case, what the fuck SB | 23:11 |
gregoa | kerio: that's my only conclusion too (that dpkg-* from sb itself (?) is used and not the one inside). very weird. | 23:13 |
gregoa | ah, that's interesting: https://paste.debian.net/222031/ ← so there's a dpkg-genchanges in the scratchbox-devkits | 23:14 |
gregoa | hm, have a said before that I want an autobuilder for each package that hits my n900? instead of different scratchboxes on individual machines *sigh* | 23:15 |
kerio | a separate autobuilder? :o | 23:15 |
luf | Ok so you don't use nokia's dpkg-dev version :) | 23:15 |
gregoa | depends if the scratchbox-devkit-debian-squeeze package from scratchbox.org counts as nokia | 23:16 |
luf | dpkg-genchanges --version ? | 23:17 |
luf | Debian dpkg-genchanges version 1.13.25. | 23:17 |
gregoa | which of the three? host / sb / "inside" sb? | 23:17 |
kerio | what's "inside" sb? :o | 23:18 |
luf | Where you run the dpkg-buildpackage ;) | 23:18 |
kerio | isn't that scratchbox? | 23:18 |
gregoa | luf, kerio: here we go with all 3: https://paste.debian.net/222038/ | 23:20 |
kerio | the missing ssl cert for debian.net is annoying | 23:20 |
kerio | they could really get a real certificate or something | 23:20 |
gregoa | kerio: you could also use an OS that ships the cert :) | 23:21 |
gregoa | luf: so interestingly we still have different version _within_ the sb | 23:22 |
kerio | is it shipped on just os x? | 23:22 |
kerio | er | 23:22 |
kerio | wtf am i saying | 23:22 |
kerio | is it only shipped in debian? | 23:22 |
gregoa | probably | 23:22 |
luf | freemangordon: ping | 23:39 |
luf | Has someone stock bluez and can try to run: dbus-send --system --type=method_call --print-reply --dest=com.nokia.btcond /com/nokia/btcond/request com.nokia.btcond.request.rfcomm_bind string:00:00:00:00:00:00 string:FTP | 23:49 |
*** BCMM has joined #maemo-ssu | 23:56 | |
*** Estel_ has joined #maemo-ssu | 23:58 | |
*** Estel_ has quit IRC | 23:58 | |
*** Estel_ has joined #maemo-ssu | 23:58 | |
luf | Don't try it :D It's neat. Even in sb is Diablo only bt-dbus.h :D | 23:58 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!