**** BEGIN LOGGING AT Wed Dec 30 02:59:59 2015 Dec 30 03:30:47 do we have anyone here who can program in Python? Dec 30 03:47:30 hi all. does anyone here have the ability to make .deb packages for n900 easily? Dec 30 03:47:48 i remember having tried with scratchbox in the past, but i failed Dec 30 03:48:29 i need to upgrade my rsync from 3.0.4 to 3.1.0 Dec 30 03:48:57 but i can't build it on my phone and i don't have a machine where i can use scratchbox Dec 30 07:25:53 KotCzarny: so N900 shutdown during the night, due to battery. anyway, 'slient earpiece' again Dec 30 07:25:56 :/ Dec 30 07:26:05 i wish i knew what the real cause is Dec 30 07:27:43 stop/restart pulseaudio does the trick, which makes it sound like a routing problem ... but still .. what causes it? :( Dec 30 07:28:28 Sicelo: did you try to --reinstall PA? Dec 30 07:28:30 apparently you have some application that does it Dec 30 07:29:09 the only times I have problems with sound was when I was using some platronics handsfree Dec 30 07:29:38 it was managing to block PA for > 30s Dec 30 07:30:09 sicelo, run dbus monitor in tha background and try to catch the offender? Dec 30 07:30:19 *the Dec 30 07:30:21 but I suspect it was the HF itself, repoting to be connected but without actually accepting data Dec 30 07:31:06 Sicelo: do you have something like call recoreder installed Dec 30 07:31:28 will do the reinstall. no, no call recorder Dec 30 07:31:36 i hardly get calls even :) Dec 30 07:31:45 :) Dec 30 07:31:46 or make them Dec 30 07:32:07 could be some broken file in /var/lib/PA... Dec 30 07:32:51 Sicelo: if reinstall does not help, you may try to enable PA debug messages and send them to syslog Dec 30 07:33:06 i know i've modified things there (but that's loooong ago for bluetooth a2dp). this has never affected earpiece. Dec 30 07:33:17 sure, will do that. thanks for tha suggestion Dec 30 09:14:52 freemangordon: ping Dec 30 09:15:00 Pali: pong Dec 30 09:15:25 freemangordon: can you resend that omapfb patch? Dec 30 09:15:34 sure Dec 30 09:15:55 you think tomi didn't see my ping? Dec 30 09:16:02 and also... are you include setting system_rev into that your reserve_rx51 mem patch? Dec 30 09:16:24 no, we'd better send another patch once atags patch is accepted Dec 30 09:16:26 I got email from Tomi asking you for resending it Dec 30 09:16:35 you did not have it? Dec 30 09:17:33 which patch, for "ARM: OMAPFB: panel-sony-acx565akm: fix missing mutex unlocks" ? Dec 30 09:17:59 as there are 2 I pinged Tomi for Dec 30 09:18:06 for ^^^ I got a reply Dec 30 09:18:12 and will resend it Dec 30 09:18:27 "Re: [PATCH v2] ARM: OMAPFB: panel-sony-acx565akm: fix missing mutex unlocks" Dec 30 09:18:36 but for the early memory allocator I don;t have a reply Dec 30 09:18:36 "Oh, I'm sorry, I must have forgotten about that. Please, send a new patch. Tomi" Dec 30 09:18:44 yes, I have that Dec 30 09:18:49 will send a patch Dec 30 09:19:01 ok Dec 30 09:19:06 but first i need a device to test on Dec 30 09:19:18 as I've lost my maemo setup on my dev device :) Dec 30 09:19:32 I mean - setup for upstream kernel Dec 30 09:20:17 Pali: could you provide url to your scripts again? Dec 30 09:20:29 Maemo with upstream kernel? Dec 30 09:20:33 yes Dec 30 09:20:40 searching for it.... Dec 30 09:20:48 hmm, or I'd rather boot it from NAND Dec 30 09:22:14 cssu-devel, rx51-bme-replacement, fix /etc/X11/Xsession.d/05connui-conndlgs-cellular, fix firmware to dir /lib/firmware/ti-connectivity, hal fix Dec 30 09:22:21 hal? Dec 30 09:22:32 there is script for it Dec 30 09:22:46 what's wrong with HAL, this is new to me Dec 30 09:22:56 https://lkml.org/lkml/2015/7/24/469 Dec 30 09:23:00 platform_soc_audio_logicaldev_input Dec 30 09:23:16 otherwise rtcom-call-ui goes into crash loop Dec 30 09:23:36 and probably fix for /etc/event.d/sgx is not needed.... do not know Dec 30 09:24:05 ah, I was just disabling call-ui autorestart :) Dec 30 09:24:40 ok Dec 30 09:24:45 this prevent crash, it is better Dec 30 09:24:55 yeah Dec 30 09:25:07 u-boot from extras is fine, ain't? Dec 30 09:25:41 yes, u-boot from extras-devel is fine for sure Dec 30 09:25:51 ok Dec 30 09:25:57 do not know what is in extras Dec 30 09:26:20 do I need KP or kernel-cssu should suffice? Dec 30 09:29:40 Pali: how to tell uboot to boot my kernel by default? Dec 30 09:30:54 freemangordon: ln -s /etc/bootmenu.d/ /etc/default/bootmenu.item Dec 30 09:31:23 thanks Dec 30 09:32:22 hi Dec 30 09:32:32 jonwil: hi! Dec 30 09:33:25 hi! Dec 30 09:33:58 Pali: and where is bme replacement? Dec 30 09:34:23 The repo is at http://maemo.merlin1991.at/cssu/bme-replacement/ Dec 30 09:34:24 http://maemo.merlin1991.at/cssu/bme-replacement/ Dec 30 09:34:29 thanks :) Dec 30 09:34:30 ~bme-replacement Dec 30 09:34:30 it has been said that bme-replacement is http://atrey.karlin.mff.cuni.cz/~pali/rx51-bme-replacement http://atrey.karlin.mff.cuni.cz/~pali/projects/maemo/bme-replacement.html See also: http://wiki.maemo.org/Bme_replacement . Please, use wiki page to report bugs/problems and/or solutions to them!, or http://talk.maemo.org/showthread.php?t=93183 Dec 30 09:34:43 Anyone here know how to program in Python? Dec 30 09:34:54 deb http://maemo.merlin1991.at/cssu/bme-replacement/ fremantle free Dec 30 09:34:54 I guess Pali knows Dec 30 09:35:23 I'm not python coder Dec 30 09:35:29 ok Dec 30 09:35:43 :) Dec 30 09:35:50 I can program in quite a few different languages but Python ain't one of them :( Dec 30 09:36:05 if you want quite a complete app example you can check oscp-remote.py Dec 30 09:36:17 with key handling, file handling, dbus handling etc Dec 30 09:36:42 I dont need it for an app on Maemo, I need it for some stuff related to maemo-security-certman Dec 30 09:36:48 freemangordon: "do I need KP or kernel-cssu should suffice" aren't you trying to boot with mainline? Dec 30 09:37:17 and more specifically the included set of root certificates Dec 30 09:37:17 yes, but bme replacement needs some kernel stuff Dec 30 09:37:21 bencoh: ^^^ Dec 30 09:37:38 and I need a kernel to boot maemo with if upstream does not boot Dec 30 09:38:08 kernel-power must work Dec 30 09:38:25 I know it works, the question was about kernel-cssu Dec 30 09:39:08 yeah, it needs KP Dec 30 09:41:04 freemangordon: ah :) Dec 30 10:01:01 Pali: what needs to be fixed in /etc/X11/Xsession.d/05connui-conndlgs-cellular? Dec 30 10:01:56 freemangordon: I commented that file as GSM modem did not worked correctly... Dec 30 10:02:11 at least that nokia stuff was not able to use modem Dec 30 10:02:28 and that script blocked Maemo booting until modem is working... Dec 30 10:02:40 oh Dec 30 10:02:45 that script ask for PIN code Dec 30 10:03:01 hmm, I don;t have sim card here Dec 30 10:04:33 WTF? NAND disappeared?!? Dec 30 10:05:00 LOL Dec 30 10:05:03 hm? Dec 30 10:05:06 rip Dec 30 10:06:00 Pali: can boot no more even KP after I've tried 4.4 Dec 30 10:06:25 what happen?? Dec 30 10:07:05 dunno, will boot rescueos to see Dec 30 10:07:42 4.4 progressed to the boint to boot maemo and hanged Dec 30 10:07:52 I did a power-down, and that's it Dec 30 10:09:02 Remind me never to run bleeding edge kernels on my everyday-use N900 :) Dec 30 10:11:42 ubifs - "failed to recover master node" :( Dec 30 10:11:55 Pali: any idea besides reflash? Dec 30 10:12:33 I did not see this message Dec 30 10:12:54 maybe hard power off damaged ubifs format? Dec 30 10:13:27 maybe Dec 30 10:13:51 that message is when i try to mount maemo under rescueos Dec 30 10:14:07 jonwil: what do you need done on python? Dec 30 10:14:17 https://patchwork.ozlabs.org/patch/162271/ Dec 30 10:14:59 Pali: but that would mean 4.4 should fix it, ain't? Dec 30 10:15:05 Well first I need the python script in https://packages.debian.org/source/sid/ca-certificates make to work with the Scratchbox version of Python then I will need some changes to it Dec 30 10:15:06 looks like Dec 30 10:15:21 hmm Dec 30 10:15:21 freemangordon: try to boot rescue OS with 4.4 kernel Dec 30 10:15:29 R/W mount ubifs Dec 30 10:15:38 and then umount it Dec 30 10:15:44 and after that do shutdown Dec 30 10:15:49 hmm, yes Dec 30 10:16:33 i think i can help with that, it's probably removing python2.7 specifics, as sb only has python2.5 Dec 30 10:16:59 ok, start by getting it to work correctly Dec 30 10:17:13 then I will talk about what changes I need made Dec 30 10:17:14 sure Dec 30 10:17:35 let me have breakfast before and then I'll get on with it Dec 30 10:17:38 :) Dec 30 10:17:55 ok :) Dec 30 10:19:51 Pali: can't open /dev/ttyS0 and then reboot Dec 30 10:20:14 rescueOS? Dec 30 10:20:16 gooing to reflash Dec 30 10:20:21 yes, in rescueos Dec 30 10:24:28 freemangordon: ttyO* ? Dec 30 10:24:58 bencoh: already flashed rootfs, no idea :) Dec 30 10:26:20 ah... maybe next time Dec 30 10:39:55 jonwil: mozilla/certdata2pem.py this one I guess? Dec 30 10:40:04 yep Dec 30 10:40:36 alrright, I'm on it Dec 30 10:43:31 guys, maybe document the process in README or wiki and put I link in the README, so we won't guess next time again Dec 30 10:48:28 yeah I have plans to put things into the repos so we have documentation of how to update the certificates going forward Dec 30 10:49:56 if i manage to port this thing we can provide this .deb package directly Dec 30 10:50:58 on the other hand Dec 30 10:51:07 i see that we can install python2.7 on scratchbox Dec 30 10:51:14 is that acceptable, jonwil? Dec 30 10:51:27 ceene: no Dec 30 10:51:27 because that'd make it work without modification Dec 30 10:51:42 2.7 is in extras, not in SDK Dec 30 10:51:49 oh, ok Dec 30 10:51:54 iirc Dec 30 11:01:40 i'm not quite sure this is all going to produce the same files Dec 30 11:02:02 there seems to be lots of utf8 encodings transformtaions and I don't know if python2.5 supports these the same way hasta 2.7 and 3 Dec 30 11:02:30 how about updating python to 2.7 then? Dec 30 11:03:13 (or using a real language with native utf8 support? :D) Dec 30 11:03:27 (nevermind, j/k) Dec 30 11:06:38 upgrading to 2.7 would be ideal in my opinion Dec 30 11:07:11 is there anything old glibc/maemo specific that would block the update? Dec 30 11:10:21 I do wonder how hard it would be to just rewrite this script in C++... Dec 30 11:10:25 or C Dec 30 11:10:33 or bash Dec 30 11:10:44 sslayer@collins:~/Proyectos/N900/ca-certificates (maemo)$ diff -urN mozilla /tmp/m27 Dec 30 11:10:49 no output Dec 30 11:11:03 that means all generated files are the same on python2.7 and sb's python :) Dec 30 11:11:33 i was a pessimistic lol Dec 30 11:11:56 what do you need now to modifiy, jonwil? Dec 30 11:12:22 Ok so it generated a bunch of certificate files? Dec 30 11:12:28 yep Dec 30 11:13:22 Pali: second try resulted with the same error Dec 30 11:13:23 What are the filenames its generating? Dec 30 11:13:30 something is terribly broken Dec 30 11:13:39 VeriSign_Universal_Root_Certification_Authority.crt Dec 30 11:13:39 Verisign_Class_1_Public_Primary_Certification_Authority.crt Dec 30 11:13:39 Verisign_Class_1_Public_Primary_Certification_Authority_-_G2.crt Dec 30 11:13:42 ok Dec 30 11:13:44 and a lot more, of course Dec 30 11:13:48 broken ubifs? Dec 30 11:13:51 yes Dec 30 11:13:54 Did it output anything about blacklisted certificates? Dec 30 11:14:03 fmg, badblock maybe? Dec 30 11:14:13 no Dec 30 11:14:13 ubi is not block device Dec 30 11:14:35 yes, several UNTRUSTED BUT NOT BLACKLISTED CERTIFICATE FOUND: Dec 30 11:14:39 Pali: can you try to boot with qemu and nand? Dec 30 11:14:47 ok Dec 30 11:14:51 in qemu it is working fine Dec 30 11:15:00 hmm Dec 30 11:15:06 that my small userspace (which I sent to you) Dec 30 11:15:16 it is not nand Dec 30 11:15:24 it is onenand Dec 30 11:15:34 on top ubi Dec 30 11:15:34 it creates stuff in /tmp Dec 30 11:15:38 and on ubi is ubifs Dec 30 11:15:56 it store stuff in /tmp/ Dec 30 11:16:02 and then generate mkfs.ubifs Dec 30 11:16:10 and then ubinize Dec 30 11:16:20 ah, yes Dec 30 11:16:29 after that it call qflasher to create one 256MB NAND image for qemu Dec 30 11:16:32 Ok so ceene, firstly, I want you to remove blacklist.txt code. Then I want you to change the output so it creates folder common-ca and puts non-blackisted certs in there and creates folder blacklist and puts blacklisted certs in there Dec 30 11:16:39 Pali: yeah, my bad Dec 30 11:16:55 qemu N900 machine can boot *only* from onenand Dec 30 11:17:02 ok Dec 30 11:17:07 ok, that's easy Dec 30 11:18:25 Pali: no ide what to do Dec 30 11:18:30 *idea Dec 30 11:18:59 bisect wrong commit in ubifs Dec 30 11:19:09 or write to ML about this problem Dec 30 11:19:20 needs flashing every time, that would take ages Dec 30 11:19:54 not to say that I would need to apply our patches every time Dec 30 11:20:09 resend that patch Dec 30 11:20:15 I think we tested it in past Dec 30 11:20:19 ceene: Do you know how to shell out to a program from Python and how to read whatever that program outputs to stdout? Dec 30 11:20:36 Pali: but I want to test it with upstream before resending Dec 30 11:20:43 i think so, yes Dec 30 11:20:57 jonwil: some.program.py | tee /tmp/output ? Dec 30 11:21:07 i mean, i know how but on 2.5 it may differ, but i'll learn :) Dec 30 11:21:11 ok Dec 30 11:21:19 (someprogram.py | tee /tmp/output) & Dec 30 11:21:53 ok, so what I want done is for each cert file, once its been written to disk, I want you to run the command "openssl x509 -in -text -noout" (without the quotes and with the certificate filename in place of the <> bit),. Dec 30 11:22:55 The output it generates will contain 2 lines Dec 30 11:22:56 X509v3 Subject Key Identifier: Dec 30 11:22:58 9B:E2:07:57:67:1C:1E:C0:6A:06:DE:59:B4:9A:2D:DF:DC:19:86:2E Dec 30 11:23:07 I need you to parse the output and find that string of digits Dec 30 11:23:28 Then the spaces and :s should be removed and the string made lowercase Dec 30 11:24:04 so e.g. 9be20757671c1ec06a06de59b49a2ddfdc19862e Dec 30 11:24:15 tr -d ' :' Dec 30 11:24:48 Then you rename the generated certificate file to that base number with a .pem extention Dec 30 11:25:17 But if there is already a file with that name, you need to append -1, -2 etc to the end of the filename Dec 30 11:25:25 ok Dec 30 11:25:33 so you might get f3a27298eeb81b82801c4db69a3027990a2f72e2.pem and f3a27298eeb81b82801c4db69a3027990a2f72e2-1.pem Dec 30 11:25:50 freemangordon: try 4.2 kernel if is working Dec 30 11:26:26 I also need each blacklist certificate (in the order they get written out/renamed) written to a file Dec 30 11:26:41 and each common-ca certificate (in the order they get written out/renamed) written to a different file Dec 30 11:26:53 there are just 11 ubifs commits between 4.2 and 4.4 Dec 30 11:27:08 some of them should be OK Dec 30 11:27:09 That way we preserve the order that the certificates appeared in the original input certdata.txt file Dec 30 11:27:25 ceene: Do you think you can do all the things that I asked? Dec 30 11:27:39 yes, jonwil Dec 30 11:27:43 ok, great Dec 30 11:27:44 it's all doable Dec 30 11:27:49 good. Dec 30 11:27:58 i'll have it soon, i expect Dec 30 11:28:06 great Dec 30 11:29:14 btw which version of python does your scratchbox have in it? Is /usr/bin/python a symlink to something else? Dec 30 11:29:36 has someone tried a recent build of otter-browser? We have an old version in qt5 repo that is very interesting nicely http://repos.fuhlbrueck.net/qt5-maemo5-examples/ Dec 30 11:29:48 freemangordon: just 4 relevant commits Dec 30 11:30:02 *and nicely running Dec 30 11:30:02 it should be 2.5 Dec 30 11:30:06 but i don't have /usr/bin/python Dec 30 11:30:14 ls -al $(which python) Dec 30 11:30:14 lrwxrwxrwx 1 root root 33 Feb 8 2011 /scratchbox/devkits/python-legacy/bin/python -> /scratchbox/tools/bin/python-host Dec 30 11:30:17 [sbox-FREMANTLE_ARMEL: /tmp/N900/ca-certificates/mozilla] > ls -al /scratchbox/tools/bin/python-host Dec 30 11:30:20 -rwxr-xr-x 2 root root 978728 Jan 17 2012 /scratchbox/tools/bin/python-host Dec 30 11:30:28 ok Dec 30 11:30:43 python on my box is /usr/bin/python Dec 30 11:30:45 freemangordon: maybe you can start reverting them: ab92a20bce3b4c21927cc1ae514c482f50ad3487 86ba9ed928f33cf6afbac433ff03dd8c6a0ceac4 be186cc4de35fcfe766d2938c13ddf3935d6c159 8f6983abe31f37d953197c3ede14b1d1321affee cf6f54e3f133229f02a90c04fe0ff9dd9d3264b4 Dec 30 11:30:46 which is weird Dec 30 11:31:23 do you have any packages showing up as installed when you do dpkg -l python*? Dec 30 11:31:42 ii python2.5 2.5.4-1maemo6 An interactive high-level object-oriented la Dec 30 11:31:45 ii python2.5-minimal 2.5.4-1maemo6 A minimal subset of the Python language (ver Dec 30 11:31:56 but i've had to install them myself by hand Dec 30 11:32:03 ok Dec 30 11:32:04 they were not on sb Dec 30 11:33:45 somehow I have package python 2.5.2-3maemo1 and package python2.5 2.5.2-11.1maemo3+0m5 Dec 30 11:34:14 and on my sb, that python script (unmodified from the debian original) fails with an error Dec 30 11:35:09 points to line 149 with a syntax error Dec 30 11:35:16 any 2.5 won't run the original version Dec 30 11:35:19 oh ok Dec 30 11:35:25 so you did need to modify it Dec 30 11:35:42 yes, several things Dec 30 11:35:46 ok Dec 30 11:36:01 i was afraid there would be some utf8 things that wouldn't work but it's been fine Dec 30 11:37:53 there are no untrusted files that match obj['CKA_CLASS'] == 'CKO_CERTIFICATE' Dec 30 11:38:00 so blacklist directory ends up being empty Dec 30 11:38:09 unless i keep blacklist.txt code Dec 30 11:42:33 also, generated files are .crt and you've asked me to run "openssl x509 -in -text -noout", but that expects a pem file, not a crt one Dec 30 11:42:46 so it fails: Dec 30 11:42:47 openssl x509 -in WoSign.crt -text -noout Dec 30 11:42:47 unable to load certificate Dec 30 11:42:47 11888:error:0D078079:asn1 encoding routines:ASN1_ITEM_EX_D2I:field missing:tasn_dec.c:500:Field=cert_info, Type=X509 Dec 30 11:42:50 11888:error:0906700D:PEM routines:PEM_ASN1_read_bio:ASN1 lib:pem_oth.c:83: Dec 30 11:43:49 can you post one of the generated .crt files? Dec 30 11:44:10 -inform der Dec 30 11:45:53 http://paste.debian.net/357020/ Dec 30 11:46:07 i don't know what that shit is Dec 30 11:46:10 but that ain't a certificate Dec 30 11:46:20 lol Dec 30 11:46:27 that's a base64 thing Dec 30 11:46:28 it's got way too many zeroes Dec 30 11:46:46 something got terribly wrong with the ASN.1 encoding Dec 30 11:46:56 let me check that i'm in fact generating the same thing that the original debian script Dec 30 11:46:59 maybe i did something wrong Dec 30 11:47:03 and it's all useless :/ Dec 30 11:48:01 yep :( Dec 30 11:48:19 i did something wrong and i did the wrong test to verify my modifications Dec 30 11:48:59 that's the utf8 thingie i feared, probably Dec 30 11:52:56 Its a pitty Juhani Mäkelä isn't around to share the "official" way to update those root certs from the Mozilla set (the way Nokia did it) Dec 30 11:56:50 i think i've found where it went wrong Dec 30 12:09:34 i'm almost there Dec 30 12:12:27 Great Dec 30 12:16:43 as for the blacklists, once get this this working, we can run it on the same version of certdata.txt that Juhani Mäkelä last run it on and see what the output looks like and figure out form there code changes to be made to the python script to get it to do the right thing with the blacklists Dec 30 12:21:15 there are little stupid discrepancies between equivalent functions of python2.5 and python2.7 that make this behave in stupid ways Dec 30 12:21:34 not important things but really stupid ones Dec 30 12:22:11 like p2.5 base64.encodestring already wrap the output to 77 columns Dec 30 12:22:59 is there any reason why openssl would want the certs to be wrapped to a certain number of chars? Dec 30 12:23:26 so it wouldn't care if instead of wrapping to 64 like the original version from debian it's wrapped to 77 Dec 30 12:28:39 ok, it works Dec 30 12:28:44 openssl now reads the .crt files Dec 30 12:30:43 jonwil: try to contact juhani Dec 30 12:30:53 i'll be gone now for a while, later I'll code the openssl output parsing Dec 30 12:30:59 ok Dec 30 12:31:03 thanks ceene Dec 30 12:31:13 i'm going to upload this somewhere before so you can take a look at it if you want Dec 30 12:31:15 I cant find any contact details for juhani other than an @nokia address Dec 30 12:31:20 ok Dec 30 12:31:26 he has gihub profile Dec 30 12:33:13 https://github.com/agamez/maemo-ca-certificates/commits/maemo Dec 30 12:33:20 do we know its the same guy and not someone else with the name? Dec 30 12:34:21 https://github.com/community-ssu/maemo-security-certman-applet/commits/master Dec 30 12:34:53 jonwil: it has paired email address from official commit messages to github profile Dec 30 12:34:57 great Dec 30 12:36:56 Looks like https://fi.linkedin.com/in/juhani-m%C3%A4kel%C3%A4-a9702a4 is also our guy Dec 30 12:37:14 mentions nixu (same as the github profile) and also nokia Dec 30 12:38:36 Ok so we have the right guy. I totally suck at writing the right kind of emails for contacts like this, do we have someone here who is better are writing things that are professional and well-written? Dec 30 12:46:52 Pali: looks like gcc is behaving :( Dec 30 12:47:42 or not? Dec 30 12:53:32 to be honest, i think our efforts should really go on the opposite direction... stop trying to backport half of debian to maemo and start porting maemo's specifics to debian, as Pali and freemangordon are doing with the kernel right now Dec 30 12:53:57 so we can install an standard debian and then apt-get install hildon with its dependencies Dec 30 12:56:11 ceene: certman is one part of maemo which we need... Dec 30 12:56:55 it is of course better to fix maemo stuff to be compatible with upstream (debian...) Dec 30 12:56:59 i was thinking on the whole ssl thing, this python backports, etc Dec 30 12:57:15 but ssl needs to be in maemo format too Dec 30 12:57:15 it's like going backwards all the time instead of forward Dec 30 12:57:24 btw, why do you need new python? Dec 30 12:57:52 well, this script jonwil is going to use to prepare new certs was written for 3.0 and 2.7 Dec 30 12:57:55 but we only have 2.5 Dec 30 12:58:06 so i've had to backport that script to use old python2.5 Dec 30 12:58:25 when the thing we should really aim for is to have python2.7 and 3 for maemo and dev environment Dec 30 12:58:34 you should try to contact that guy who worked on certman and certs Dec 30 12:58:50 instead trying to do something abnormal Dec 30 12:59:02 python2.7 is in extras, right? Dec 30 12:59:07 yep Dec 30 12:59:15 and that script is needed just at compile time of package, right? Dec 30 12:59:25 then download python2.7 from extras and build package Dec 30 12:59:30 no, its used only when we want to update the root certificate set Dec 30 12:59:30 i don't know exactly how jonwil wants to use it Dec 30 12:59:53 Pali: it is not gcc, unfortunately Dec 30 12:59:54 but yeah if its already been backported to work on Python 2.5, lets keep it Dec 30 13:00:12 jonwil: but that script is not used at runtime, or yes? Dec 30 13:00:14 I would try to contact Juhani but as I said, I suck at writing professional level emails of the sort one writes when you are trying to contact people like this Dec 30 13:00:20 no, the script isn't used at runtime Dec 30 13:00:23 mater node magic contains 0xFFFFFFFF Dec 30 13:00:27 *master Dec 30 13:00:49 jonwil: for the file renaming thing based on openssl i think we'd be better off with a simple shell script Dec 30 13:00:51 ceene: can you help jonwil to write that email? Dec 30 13:01:00 grep+tr and a couple things could be enough Dec 30 13:01:06 Yeah that would make sense Dec 30 13:01:16 well, i can try, but i don't know what i have to tell him :) Dec 30 13:01:34 My shell script knowhow is just as bad as my Python know-how though :P Dec 30 13:01:47 hi xyz, can you help us with maemo sw abc how to do thing cde? Dec 30 13:02:09 sorry, but this is what I would write :-) Dec 30 13:02:14 lol Dec 30 13:02:29 i can write it a bit better, i just don't know what it is thing cde we're trying to do Dec 30 13:02:38 today i'm just a coder under the orders of jonwil Dec 30 13:02:47 don't even know what i'm doing lol Dec 30 13:03:10 jonwil, maybe you should specify what we exactly need :-) as I'm lost too Dec 30 13:05:40 something like "hi Juhani, I am a software developer working on a community update for Maemo Fremantle and I found your name attached to https://github.com/community-ssu/maemo-security-certman/ where you updated the root certificates using the Mozilla root certificate set. Can you share the steps used to update those certificates so we can import the latest root certificate set into our... Dec 30 13:05:41 ...community patch?" Dec 30 13:05:49 but written much more proffessionally Dec 30 13:06:07 Does that give you the gist of what we are trying to do and what we want? Dec 30 13:06:52 its ok Dec 30 13:07:02 i find that friendly and professional enough, jonwil Dec 30 13:07:29 add a couple of thanks a lot and best regards at thend and you're good to go Dec 30 13:07:39 ok, I will send the email then Dec 30 13:08:53 openssl x509 -in common-ca/WoSign.crt -text -noout|grep 'X509v3 Subject Key Identifier:' -A 1|tail -1|tr -d ': ' | tr '[:upper:]' '[:lower:]' Dec 30 13:08:56 e166cf0ed1f1b34bb7062014fe8712d5f6fefb3e Dec 30 13:09:33 what should I use for the subject line of this email (to sound professional and to get the right attention)? Dec 30 13:10:50 maybe... "Maemo certificates question" ? Dec 30 13:13:41 How about "Question regarding Maemo Certificate Manager" Dec 30 13:14:00 sounds good! Dec 30 13:14:11 or better yet "Question about Maemo Fremantle Certificate Manager" Dec 30 13:14:27 to be clear that I refer to Fremantle and not some other version of Maemo Dec 30 13:14:50 Pali: do we know what is the latest kernel with workig ubifs? Dec 30 13:15:05 freemangordon: I think 4.2 worked Dec 30 13:15:11 on n900? Dec 30 13:15:13 from linux-n900 tree Dec 30 13:15:16