**** BEGIN LOGGING AT Tue Aug 30 02:59:57 2011 Aug 30 07:38:03 * DocScrutinizer SIGHS on chandra Aug 30 07:38:55 * juri_ sighs in general. Aug 30 07:44:45 could anyone help me with installing SHR on a palm pre? Aug 30 08:04:54 within two minutes? yeah, very likely... [sigh] Aug 30 08:26:42 Sighday Aug 30 08:45:01 moinmoin Aug 30 09:01:35 ~seen morphis Aug 30 09:01:45 morphis <~morphis@dslb-092-076-169-214.pools.arcor-ip.net> was last seen on IRC in channel #openmoko-cdevel, 2d 15h 36m 43s ago, saying: 'ah ok'. Aug 30 18:16:08 I've done some bugreports... Aug 30 18:16:25 http://www.shr-project.org/trac/query?status=!closed&version=SHR-core Aug 30 18:17:00 two bugs were closed magically(I tested and I closed them),a lot more were opened.... Aug 30 18:17:25 dos1: hi, do you know phonefsod? Aug 30 18:17:39 gnutoo: in what sense? Aug 30 18:18:01 are you capable of bisecting a bug in it? Aug 30 18:18:13 http://www.shr-project.org/trac/ticket/1478 Aug 30 18:19:04 I tried to bisect it but failed because of runtime issues Aug 30 18:19:13 like require some lib that are gone etc... Aug 30 18:19:38 gnutoo: well, i can try, but i first look at it's sources i've had few days ago :P Aug 30 18:19:43 we alwasys notice stuff too late.... Aug 30 18:19:53 ok Aug 30 18:20:24 for instance the registration bug on the n900 was also noticed very late and we had to go back a lot in time.... Aug 30 18:20:34 gnutoo: i wanted to look at this issue anyway Aug 30 18:20:40 ok thanks a lot Aug 30 18:20:44 btw Aug 30 18:21:00 i have to register manually to network on N900 even after restarting phonefsod and entering pin Aug 30 18:21:05 my comments about the "noticying too late" means "how could we implement some process to fix that" Aug 30 18:21:19 ok Aug 30 18:21:25 it doesn't register until I call org.freesmartphone.GSM.Network.Register() Aug 30 18:21:36 ok Aug 30 18:21:45 it registers automatically if you have no PIN... Aug 30 18:21:45 gnutoo: proper test suite? Aug 30 18:21:57 I was rather thinking of user tests Aug 30 18:22:02 we need some users to test... Aug 30 18:22:22 the test suite would be something runtime then Aug 30 18:22:30 gnutoo: well, automatic test suite would be a great thing Aug 30 18:22:37 gnutoo: but i don't think we're able to correctly maintain it :x Aug 30 18:22:37 yes but what to test? Aug 30 18:22:50 phone registering at compilation? Aug 30 18:22:55 gnutoo: proper reactions from framework Aug 30 18:22:56 I don't think that's possible Aug 30 18:23:00 ah ok Aug 30 18:23:02 good point Aug 30 18:23:09 we have a test gsm backend Aug 30 18:23:24 but we would loose too much time in it Aug 30 18:23:29 gnutoo: of cource as runtime on device Aug 30 18:23:30 maybe some user could test then.... Aug 30 18:23:36 ah ok Aug 30 18:23:52 maybe we should make a call for testers on the blog.... Aug 30 18:24:00 and for bugreports after the tests... Aug 30 18:24:06 gnutoo: it can be even set of simple python scripts, it doesn't matter Aug 30 18:24:11 pespin: what do you think....? Aug 30 18:24:12 ok Aug 30 18:24:20 gnutoo: but it must be kept up-to-date Aug 30 18:24:29 ok Aug 30 18:25:28 replicant has no test suite either.... Aug 30 18:25:35 * pespin reading Aug 30 18:25:38 however I wonder how cyanogenmod handle that.... Aug 30 18:26:38 ofono has some test suite AFAIK Aug 30 18:26:44 ok Aug 30 18:26:48 gnutoo, nokia n900 testers? Aug 30 18:26:57 i've seen scripts when i was looking at their API Aug 30 18:27:02 no, SHR testers in general that do good bugreports... Aug 30 18:27:17 (it was also godd thing to look at when i wanted to learn how to use that api :D) Aug 30 18:27:37 s/godd/good/ Aug 30 18:27:38 dos1 meant: (it was also good thing to look at when i wanted to learn how to use that api :D) Aug 30 18:27:56 gnutoo, I think we should fix most visible and annoying bugs before letting testers fill lots of bug reports ;) Aug 30 18:28:05 ok Aug 30 18:28:10 such as shr-core stuff Aug 30 18:28:15 gnutoo, yeah Aug 30 18:28:17 even FSO has some test scripts Aug 30 18:28:26 in frameworkd.git repo Aug 30 18:28:27 first shr-core should work more a less correctly for users Aug 30 18:28:37 (but i guess out of date) Aug 30 18:28:37 ok Aug 30 18:28:39 then let users blame us for little details ;) Aug 30 18:29:43 well Aug 30 18:29:54 it doesn't matter if it's shr-core or shr-testing or anything else Aug 30 18:30:00 ok Aug 30 18:30:10 shr-testing worked ok Aug 30 18:30:13 in the past we were bitten by lots of bugs Aug 30 18:30:18 ok Aug 30 18:30:26 which were mostly regresions Aug 30 18:30:35 indeed Aug 30 18:30:48 and we spent a huge time dealing with them Aug 30 18:30:51 yup Aug 30 18:30:52 and we noticied them too late Aug 30 18:31:02 so the diff was huge.... Aug 30 18:31:18 the goal would be to notice them earlier Aug 30 18:31:22 when the diff is small.... Aug 30 18:31:33 so automated test suite would allow us to notice part of them almost immediately Aug 30 18:31:55 good idea: Aug 30 18:31:58 opkg upgrade Aug 30 18:32:09 then run test suite... Aug 30 18:32:15 or reboot and run test suite Aug 30 18:32:26 to be sure everything works fine... Aug 30 18:32:33 and then go out with your phone... Aug 30 18:32:38 that would be awesome Aug 30 18:32:47 we should find some good framework to handle tests Aug 30 18:33:08 we have real users like [RUI] that are beaten hard by bugs... Aug 30 18:33:13 i think there was something mentioned on FSOSHRUDCON Aug 30 18:33:18 ok Aug 30 18:33:29 maybe mickeyl knows something about test suites? Aug 30 18:33:43 i guess he knows something, frameworkd had some :) Aug 30 18:33:51 ok Aug 30 18:35:07 mostly handcrafted, if you take a look. you can test individual libraries fine with devices such as glib's gtest environment, you can also do simple dbus API compliance tests, however the real channel is integration / cross layer tests. If you want to automate, say a phone call, you will need quite some framework and external resources. this complicates things alot. Aug 30 18:35:29 s/channel/challenge/ Aug 30 18:35:29 mickeyl meant: mostly handcrafted, if you take a look. you can test individual libraries fine with devices such as glib's gtest environment, you can also do simple dbus API compliance tests, however the real challenge is integration / cross layer tests. If you want to a... Aug 30 18:36:08 http://build.shr-project.org/fsoshrudcon/FSOSHRUDCON09-dTest-DanielWillman.pdf Aug 30 18:36:18 hmm we want runtime tests Aug 30 18:36:39 like opkg upgrade;reboot;test-suite; Aug 30 18:38:29 hmmm Aug 30 18:38:46 maybe we need the UI to define some dbus calls Aug 30 18:38:51 like make a call Aug 30 18:39:01 something that would work at the UI level Aug 30 18:39:48 mickeyl: well, we don't need everything to be really automated Aug 30 18:40:39 it just has to be simple Aug 30 18:41:00 so it won't be pain for us to run it frequently Aug 30 18:41:49 I am thinking of a conf file + \ Aug 30 18:42:01 from SHR's point of few, we just need FSO to return what we expect and our apps to behave as we except based on this data Aug 30 18:42:04 some dbus methods in the GUI +\ Aug 30 18:42:04 s/few/view/ Aug 30 18:42:06 dos1 meant: from SHR's point of view, we just need FSO to return what we expect and our apps to behave as we except based on this data Aug 30 18:42:07 a test program Aug 30 18:42:22 the conf file would tell for instance what free number to call Aug 30 18:42:26 gnutoo: dbus methods in the GUI - i think it's wrong Aug 30 18:42:40 I mean like what's there is already now Aug 30 18:42:45 with phoneui Aug 30 18:42:57 gnutoo: if something can't be automated and needs user reaction, then it should have something like wizard Aug 30 18:43:07 ok Aug 30 18:43:16 "Phone should now Did it behave correctly? [yes] [no]" Aug 30 18:43:25 yes but for instance: Aug 30 18:43:38 wizard: warning test will be done Aug 30 18:43:48 mdbus2 phoneui call 4242 Aug 30 18:44:00 and then the user listens Aug 30 18:44:03 and hang up Aug 30 18:44:12 yup Aug 30 18:44:16 and then the wizard continues Aug 30 18:44:25 test app can then show "check if volume slider behaves correctly" Aug 30 18:44:29 like did it behave correctrly Aug 30 18:44:36 yes Aug 30 18:44:38 exactly Aug 30 18:44:57 and based on this data, it should generate some kind of report Aug 30 18:45:04 yes, something like that did the OM hardware folks do for factory testing... they hardcoded taiwanese S O S number though *cough* Aug 30 18:46:09 mickeyl: ech :/ sucks (hardcored SOS number) Aug 30 18:46:18 ok, but if we want to make such app Aug 30 18:46:26 lol Aug 30 18:46:34 we have to be sure that it won't became pain to maintain in future Aug 30 18:46:52 so it has to be easly extendable, but not overcomplicated Aug 30 18:46:59 like test doesn't run, then we need a test program for the test programs.... Aug 30 18:47:33 gnutoo: i think we are "test app for test app" - humans :D Aug 30 18:55:48 mickeyl, hi I'm updating some parts of elm.vapi in eflvala. In Elementary.h, I see "typedef struct _Elm_Object_Item Elm_Object_Item;" Aug 30 18:56:04 and the struct is not defined in that header, so the implementation looks hidden Aug 30 18:56:28 what do I need to set in vapi? do I have to look for the struct elements in other header files? Aug 30 18:56:55 not necessarily. sometimes the only reason to define structs is to have opaque types to give as parameters Aug 30 18:56:57 maybe use extern keyword? Aug 30 18:57:16 you may well define empty classes or structs Aug 30 18:57:24 ah ok :) Aug 30 18:57:27 just for the sake of making the type known Aug 30 18:57:40 oki, thanks! :) Aug 30 18:58:45 btw., if you want access to the E svn, i'm sure we can arrange something. I'm not really working with EFL/Vala anymore, so if someone else wants to step up maintaining the bindings, i'm all ears Aug 30 19:01:55 playya_: ping. could you update vala #604827. It's a bit of a shame that it hangs that long :) Aug 30 19:02:27 it seems that it can put the backlight off even with the backlight button in forbid after pressing the power button Aug 30 19:04:10 suspend even works.... Aug 30 19:04:15 the suspend button I mean Aug 30 19:07:19 arrg even without -O2 uboot fails to run the kenrel.... Aug 30 19:14:28 and JaMa is not there.... Aug 30 19:14:30 arrrg Aug 30 19:14:52 so I guess I'll have to start looking at that bug from scratch Aug 30 19:15:45 mickeyl, I usually send patches to playya_ when I send some updates. but could be cool to have access in case he is busy and we need to fix vala bindings quickly to have all shr feeds compiling :) Aug 30 19:27:45 adding -O0 fails too Aug 30 19:28:00 like CFLAGS_libs += O0 Aug 30 19:33:36 I'll try to rebase on top of current uboot Aug 30 19:34:30 freesmartphone.org: 03morphis 07specs * rddbe3c44a5a5 10/ (4 files in 4 dirs): Enable org.freesmartphone.Preferences.Service again and bump version Aug 30 19:34:50 freesmartphone.org: 03morphis 07gdbus * rf03a9c96e224 10libfso-glib/configure.ac: Bump version to catch up with fso-specs Aug 30 19:35:21 freesmartphone.org: 03morphis 07cornucopia * r16d80878d0e8 10/configure.ac: Bump version to catch up with fso-specs Aug 30 19:47:56 captainigloo, wow it seems naviframe is what I was looking for :D Aug 30 19:47:56 JaMa piing.... Aug 30 19:49:20 captainigloo, I already applied it to my app (and tit the bindings). efenniht doesn't look nice with it as it's not themed thouhg, that's a pity :( Aug 30 19:52:49 does somebody knows uboot here? Aug 30 19:53:16 basically newer compiler broke the uboot we use for the n900 Aug 30 19:53:25 sigh Aug 30 19:53:38 so oe-core's uboot is broken.... Aug 30 19:54:17 after trying to add -O0 at the end(it did add -O0) it still failled Aug 30 19:54:27 basically it fails at : Aug 30 19:54:59 Starting kernel... Aug 30 19:55:11 I've no indications whatsoever Aug 30 19:55:24 I guess that using md is not a good idea either...without the serial console Aug 30 19:58:59 anyway md blocks the device Aug 30 20:00:22 freesmartphone.org: 03morphis 07cornucopia * r3800213c742a 10/fsoeventsd/ (25 files in 8 dirs): fsoevents: add basic framework for implementation work Aug 30 20:00:23 freesmartphone.org: 03morphis 07cornucopia * r0a2ac0986eb9 10/fsoeventsd/vapi/libyaml-1.0.vapi: fsoevents: add vapi for libyaml-1.0 - should submitted upstream later Aug 30 20:00:23 freesmartphone.org: 03morphis 07cornucopia * re8a129e0702c 10/fsoeventsd/vapi/libyaml-1.0.vapi: fsoeventsd: libyaml-1.0.vapi: cleanup formatting Aug 30 20:02:14 why can't this uboot md.... Aug 30 20:13:10 the uboot patches autor doesn't respond.... Aug 30 20:13:18 I'm waisting my time...agin Aug 30 20:13:22 *again Aug 30 20:13:24 sighh Aug 30 20:20:51 gnutoo: i also hate u-boot :-) Aug 30 20:21:29 ok Aug 30 20:21:36 I don't hate it, Aug 30 20:21:45 I hate the fact that the patches are not mainline Aug 30 20:21:58 and I hate the fact that there is nobody there to answer Aug 30 20:22:04 and I've no serial port Aug 30 20:22:09 and no jtag Aug 30 20:22:27 and I hate that it has to be chainloaded Aug 30 20:22:31 i have jtag board for GTA02 Aug 30 20:22:38 and that nolo and the nokia xloader are proprietary Aug 30 20:23:20 basically I hate the n900 booting... Aug 30 20:23:35 gnutoo: i hate that i cannot build a u-boot for n900 which boots the original maemo kernel Aug 30 20:23:35 and I don't hate uboot Aug 30 20:23:42 uboot is nice and fine Aug 30 20:23:53 gnutoo: ok Aug 30 20:23:55 you can.... Aug 30 20:24:00 but not within oe.... Aug 30 20:24:31 at least not within oe-core Aug 30 20:24:34 i tried with maemo sdk and sourcery 8.1 8.3 Aug 30 20:24:43 ok Aug 30 20:24:51 with all the patches? Aug 30 20:25:07 including the gcc45 patch? Aug 30 20:25:33 gnutoo: no only the necessary patches Aug 30 20:26:09 http://al.robotfuzz.com/~al/maemo/u-boot/ Aug 30 20:26:21 try with all the patches from oe-core... Aug 30 20:27:37 gnutoo: but iam not using gcc 4.5 Aug 30 20:29:40 gnutoo: ok ill try again with sourcery 8.1 Aug 30 20:32:29 gnutoo: what are you currently working on u-boot ? Aug 30 20:32:48 I don't know.... Aug 30 20:32:50 I tried Aug 30 20:32:52 I failed Aug 30 20:33:01 and no one is there Aug 30 20:33:18 everyone like to says that SHR is not dead but when you need someone he's absent Aug 30 20:35:20 and who is the developer of the patches you mean ? Aug 30 20:35:33 JaMa touched uboot before me Aug 30 20:35:46 and the developer who developed the patches is: Aug 30 20:36:17 ali1234 in #meego Aug 30 20:36:39 gnutoo: i talked 2 days ago several hours with him Aug 30 20:37:25 gnutoo: look at #htc-linux :-) Aug 30 21:22:56 gnutoo: ping Aug 30 21:35:15 GarthPS, ping :) Aug 30 21:39:46 nschle85, hi Aug 30 21:40:40 I think I should quit the community and never come back... Aug 30 21:40:41 gnutoo: compiled u-boot with shr-core patches Aug 30 21:40:57 ok Aug 30 21:41:24 GNUtoo|htcdream: but the maemo kernel does not boot Aug 30 21:41:25 personally i broke my laptop's hdd and one of my 2 htcdream's screen Aug 30 21:41:51 GNUtoo|htcdream: why that ? Aug 30 21:42:04 rage against the machine... Aug 30 21:42:23 konsole crashed Aug 30 21:42:30 it was too much for me Aug 30 21:42:39 i was already very nervous Aug 30 21:42:47 now my laptop wont boot Aug 30 21:43:07 and i've no phone in a decent working state Aug 30 21:44:07 GNUtoo|htcdream: ups... Aug 30 21:44:20 indeed Aug 30 21:44:47 and i start working the day after tomorrow... Aug 30 21:44:50 GNUtoo|htcdream: I think you should drink a cold beer... Aug 30 21:45:16 or two Aug 30 21:45:29 i dont like beer... Aug 30 21:46:15 GNUtoo|htcdream: is it possible to buy a new phone ? Aug 30 21:47:41 yes Aug 30 21:47:53 howver i lost the laptop's data Aug 30 21:48:03 GNUtoo|htcdream: but what do i make wrong ?? all compiled u-boots starts but the maemo kernel wont start ! Aug 30 21:48:21 it had 2 hdds in raid0 Aug 30 21:48:40 same here.... Aug 30 21:49:03 maybe abandon before doing rage against the machine... Aug 30 21:49:18 compiler issues Aug 30 21:49:23 what's gcc version? Aug 30 21:50:15 GNUtoo|htcdream: gcc version 4.2.3 (Sourcery G++ Lite 2008q1-126) Aug 30 21:50:27 no idea then Aug 30 21:50:38 what are the obtimizations used? Aug 30 21:50:42 O2? Aug 30 21:50:44 gnutoo: this is the version ALI1234 also uses Aug 30 21:51:04 or O0 Aug 30 21:51:07 ok Aug 30 21:51:27 ask him maybe... Aug 30 21:52:53 GNUtoo|htcdream: this is one example line: Aug 30 21:53:03 GNUtoo|htcdream: arm-none-linux-gnueabi-gcc -g -Os -fno-common -ffixed-r8 -msoft-float -D__KERNEL__ -DTEXT_BASE=0x80e80000 -I/home/nschle85/u-boot-n900/u-boot/include -fno-builtin -ffreestanding -nostdinc -isystem /opt/arm-2008q1/bin/../lib/gcc/arm-none-linux-gnueabi/4.2.3/include -pipe -DCONFIG_ARM -D__ARM__ -marm -mabi=aapcs-linux -mno-thumb-interwork -march=armv5 -Wall -Wstrict-prototypes -fno-stack-protector \ Aug 30 21:53:03 -o cmd_bdinfo.o cmd_bdinfo.c -c Aug 30 21:53:54 -Os is wrong... Aug 30 21:54:13 it may be what breaks your build Aug 30 21:54:23 try that Aug 30 21:55:07 add CFLAGS_LIBS in board/nokia/somefile Aug 30 21:55:16 oops Aug 30 21:55:27 s/LIBS/libs Aug 30 21:57:37 ot grep for where to add it in board Aug 30 21:59:39 gnutoo: there is a makefile Aug 30 21:59:44 Makefile Aug 30 22:01:15 ok Aug 30 22:01:31 but there is already a CFLAGS += O0 defined Aug 30 22:02:17 did you grep for CFLAG_libs in board? Aug 30 22:02:28 they usually define it as O2 Aug 30 22:03:12 at least in u-boot master Aug 30 22:06:18 GNUtoo|htcdream: CFLAGS_libs and CFLAG_libs not found Aug 30 22:06:43 ok Aug 30 22:06:44 iam using the branch version which is also used in shr-core Aug 30 22:07:18 maybe it's only in uboot master then... Aug 30 22:07:21 i'll sleep Aug 30 22:08:59 GNUtoo|htcdream: for other boards its defined Aug 30 22:09:16 gnutoo: what is the default if not defined ? Aug 30 22:14:12 pespin: yes ? Aug 30 22:17:06 GarthPS, just to let you know that if you find new interesting info about the palm pre3 prices, please tell me as I'm interested :) Aug 30 22:17:26 pespin: ok will try Aug 30 22:17:26 GarthPS, and I wanted to ask you what's the support state of that device Aug 30 22:18:00 pespin: don't know :) or remember. I have to ask to morphis Aug 30 22:19:37 ok thanks anyway :) Aug 30 22:32:58 mickeyl, not this week. daily work + working @ innenhof-festival.de Aug 30 22:34:01 i wanted to update the vala bindings during holiday, but my friends made me go to the lake :( Aug 30 22:34:27 i didn't even fix my n900 and laptop yet **** ENDING LOGGING AT Wed Aug 31 02:59:57 2011