IRC log for #maemo-ssu on 20120310

01:02.23*** join/#maemo-ssu jonwil (~jonwil@27-33-137-199.static.tpgi.com.au)
01:08.30jonwilmerlin1991: ping
01:18.49merlin1991jonwil: pong
01:19.16jonwilOk, so has anything been decided on what will happen to my operator-name-cbs-widget?
01:20.15merlin1991will be included most probably
01:21.27jonwilin CSSU-T you mean?
01:31.27merlin1991yes
01:36.38jonwilok, good, lets hope a new CSSU-T comes out soon :)
01:38.27LaoLang_coolHi, I'm playing angry birds, which file is for the game recording? I want to back it up..
03:17.42*** join/#maemo-ssu amiconn_ (amiconn@rockbox/developer/amiconn)
04:53.52*** join/#maemo-ssu Kaptenen (~Kaptenen@81.216.60.47)
05:19.55*** join/#maemo-ssu LaoLang_cool (~LaoLang_c@221.226.175.142)
08:36.46*** join/#maemo-ssu Pali (~quassel@unaffiliated/pali)
09:17.23*** join/#maemo-ssu BCMM (~ben@unaffiliated/bcmm)
09:40.56*** join/#maemo-ssu NIN101 (~NIN@206.253.166.69)
10:13.17*** join/#maemo-ssu NIN101 (~NIN@206.253.166.69)
10:17.50*** join/#maemo-ssu DocScrutinizer (~halley@openmoko/engineers/joerg)
11:07.21*** join/#maemo-ssu andre__ (~andre@Maemo/community/bugmaster/andre)
11:19.01*** join/#maemo-ssu Free-MG (~test@p5DD7D537.dip.t-dialin.net)
12:02.03*** join/#maemo-ssu dafox (~dafox@ip51cc571d.speed.planet.nl)
12:26.08*** join/#maemo-ssu rd (~rd@p57B484A5.dip0.t-ipconnect.de)
12:52.14*** join/#maemo-ssu LaoLang_cool (~LaoLang_c@221.226.175.142)
12:57.50*** join/#maemo-ssu Guest87431 (~rd@p57B484A5.dip0.t-ipconnect.de)
13:29.21*** join/#maemo-ssu BCMM (~ben@unaffiliated/bcmm)
13:41.12*** join/#maemo-ssu DocScrutinizer51 (~lagrange@openmoko/engineers/joerg)
13:46.34*** join/#maemo-ssu Guest87431 (~rd@p57B484A5.dip0.t-ipconnect.de)
14:23.16*** join/#maemo-ssu trbs (~trbs@2001:470:d2ad:1:4a5b:39ff:fe7d:1623)
15:26.52*** join/#maemo-ssu Kaptenen (~Kaptenen@81.216.60.47)
15:44.44*** join/#maemo-ssu andre__ (~andre@ip-86-49-77-135.net.upcbroadband.cz)
15:44.44*** join/#maemo-ssu andre__ (~andre@Maemo/community/bugmaster/andre)
15:48.58*** join/#maemo-ssu arcean (~Arcean@aaeo169.neoplus.adsl.tpnet.pl)
17:00.38*** join/#maemo-ssu trumee_afk (~parul@46-64-48-176.zone15.bethere.co.uk)
17:08.31*** join/#maemo-ssu freemangordon (freemangor@130-204-50-168.2074046448.ddns.cablebg.net)
18:04.18*** join/#maemo-ssu M4rtinK (~M4rtinK@mail.melf.eu)
19:13.42*** join/#maemo-ssu mirandir (~valentin@2a01:e35:2e63:5170:7a2b:cbff:fed1:407c)
19:14.21*** part/#maemo-ssu mirandir (~valentin@2a01:e35:2e63:5170:7a2b:cbff:fed1:407c)
19:32.04*** join/#maemo-ssu nox- (noident@freebsd/developer/nox)
19:59.07*** join/#maemo-ssu arcean_ (~Arcean@aaeo29.neoplus.adsl.tpnet.pl)
20:07.48freemangordonPali, ping
21:54.51Palifreemangordon, pong
21:57.00freemangordonPali, I am testing ke-recv
21:57.05freemangordonso far so good
21:57.18Paliok, I'm using it too and no problem I see
21:57.37freemangordonbut my ext3 partition appears in filemanager as Nokia N900 :)
21:57.58Paliyes - but this is bug in osso-filemenager of in hildon
21:57.58freemangordonand it seems it tries to mount the swap partiyion (mmcblk1p3)
21:58.13Paliit try to mount all partitions
21:58.26freemangordonso I have a message "memory card format unsupported"
21:58.42Paliuh, so then swap must be skipped
21:58.49freemangordonPali, noone should try to mount swap partition
21:58.50freemangordonyep
21:58.57Palican you write syslog output?
21:59.08freemangordonlet mr try
21:59.11freemangordon*me
22:00.32Paliin ke-recv code is line for ignoring swap
22:01.29freemangordonke_recv[1578]: mount_volumes:950: trying mount /dev/mmcblk1p3 to /media/mmc1p3
22:01.29freemangordonMar 11 00:00:26 Nokia-N900 root: /usr/sbin/osso-mmc-mount.sh: /dev/mmcblk1p3 type '82' is not FAT32 or FAT16
22:01.29freemangordonMar 11 00:00:26 Nokia-N900 root: /usr/sbin/osso-mmc-mount.sh: mounting /dev/mmcblk1p3 read-write fs vfat to /media/mmc1p3, rc: 255
22:02.00Palinot enoght
22:02.14freemangordonwhat else do you need
22:02.15Palican you upload full ke-recv syslog to pastebin?
22:02.59Paliignoring swap is done in function which add paritition to volume list
22:03.01freemangordonwhat was the command from n900?
22:03.09freemangordonfor pastebin
22:03.29Palino idea if there is command for that
22:03.35freemangordonok
22:03.46freemangordonthere is, but i will use the site anyway
22:03.47Siceloapt-get install pastebinit
22:03.56freemangordonSicelo, 10x
22:04.26freemangordonPali, http://pastebin.com/bqzEURQ1
22:04.51freemangordonThat is the result after using usb mass storage mode
22:05.20freemangordon"/dev/mmcblk0p3     950401  974976   24576     786432   82  Linux swap / Solaris"
22:05.30freemangordonoops, sorry
22:05.57freemangordon"/dev/mmcblk1p3       1806+   1935     130-   1042168   82  Linux swap / Solaris" is the partitio in question
22:08.01Palifreemangordon, unplug card, restart ke-recv, plug it again and sent ke-recv syslog output again
22:08.24Palithis log say, that mmc part is already in list
22:09.18freemangordonPali, I have my swap on sdcard, need to move it on internal mmc first, give me a couple of minutes
22:09.32Palirestart ke-recv
22:09.41freemangordonok
22:09.54freemangordonwaiting for swapoff to complete
22:09.59Paliok
22:12.21Palialso give me output of lshal -u /org/freedesktop/Hal/devices/volume_uuid_130186e3_9585_40a8_954a_dc2ee500b23f
22:12.28freemangordonPali, how am i supposed to restart ke-recv?
22:12.42Palisudo /etc/init.d/ke-recv restart
22:12.45freemangordonok
22:13.05Paliis that above uuid swap?
22:13.21freemangordonPali, this device is my ext3 partition, are you sure you need lshal output for it?
22:13.25freemangordonno
22:13.37PaliI need to know which is swap
22:20.00freemangordonPali, http://pastebin.com/FMeKTxBP
22:23.08Paliis this your swap? /org/freedesktop/Hal/devices/volume_uuid_19ED_24CB
22:23.54freemangordonyes
22:24.21Paligive me output of lshal -u /org/freedesktop/Hal/devices/volume_uuid_19ED_24CB
22:25.40freemangordonhttp://pastebin.com/0MR7L0Gx
22:26.33freemangordonsomething in kernel?
22:26.51Paliso this output from HAL thinks that that partition is FAT32
22:26.53Palinot swap
22:26.58freemangordonI see
22:27.14freemangordonbut it is swap, that is for sure, id 82 :)
22:27.27Palitry mkswap
22:27.46Pali82 is only MBR ID
22:28.06Palibut HAL also look at first paritions magic chars...
22:28.07freemangordonPali, that swap partition is the only one I use
22:28.30freemangordonwhat is the idea of doing mkswap on it?
22:29.07PaliI think that mkswap create proper header on that part
22:29.45Palie.g. you need to create swap partition before using it via swapon
22:30.31Paliat least on that partition is some FAT32 header which some UID
22:30.35freemangordonsure, but why the hell would HAL think that ACTIVE (i.e. swapon-ed partition) is VFAT is beyond my imagination
22:30.47freemangordonPali, it is my ONLY swap
22:31.05freemangordonso it should be good enough for swapon
22:31.16Palino idea, but HAL is shit
22:31.20freemangordon:D
22:31.36freemangordonwell, maybe this should be workarounded somehow
22:31.52freemangordonin ke-recv scripts, based on partition id
22:32.36Palitry swapon -f <device>
22:32.51freemangordon-f stands for?
22:33.01Pali-f should reinitialize swap device
22:33.05freemangordonok
22:33.15Palimanpage tell that it call mkswap if needed
22:33.33freemangordonswapon: invalid option -- f :p
22:33.47Paliseems that busybox does not support -f :-(
22:33.50freemangordonwell, I will do mkswap
22:35.56Paliif you used that partition for FAT32 before swap and you did not run mkswap this is normal problem
22:36.35Palibecause swapon (without -f) maybe did not changed partition header
22:36.56freemangordoni did mkswap, let me try again
22:37.43Palimaybe again stupid HAL doing some bad things... or busybox
22:38.23freemangordonno joy, ke-recv tries to mount it
22:38.46freemangordonbtw HAL reports volume.partition.type = '0x82'  (string)
22:39.04freemangordonhow is that related to VFAT?
22:39.19Palivolume.partition is data from MBR
22:39.44Paliand volume.* is data from parition (or kernel or somewhat...)
22:39.55freemangordonPali, I think a workaround in /usr/sbin/osso-mmc-mount.sh is needed
22:40.04freemangordonto ingore swap partitions
22:40.31Paliproblem is that we cannot say that MBR id is always correct
22:40.43Palisome windows programs for formating does not update it
22:41.06Palibut I'd like to know why that stupid hal always think that part is FAT
22:41.27Paliand where HAL got that FAT UID (19ED_24CB)
22:41.53PaliDocScrutinizer, do you know how HAL is working?
22:42.04DocScrutinizera bit
22:42.11Palisee ^^^
22:42.12freemangordonwell, it is possible the stupid windoze I am connecting n900 to to make bad things
22:42.56freemangordonlike screwing up unknown partitions
22:43.31freemangordonnaah, scratch that
22:43.31Palifreemangordon, maybe HAL does not update info after running mkswap/mkfs.* try to reboot N900
22:43.32DocScrutinizersorry, no sufficent knowledge about HAL to contribute much
22:43.57Paliafter reboot HAL must loose its memory
22:44.09freemangordonPali, i've removed the backcover after making mkswap, that should be pretty enough
22:44.32freemangordondid I?
22:44.42freemangordonlet me try again, just to be sure
22:44.56Palitry to mkswap with closed cover
22:45.13freemangordon:)
22:45.50freemangordonwhat I did is mkswap, removed backcover, returned backcover on its place
22:46.05freemangordonno way I can do mkswap while de cover is removed
22:46.10freemangordon*the
22:46.20Paliok, see if lshal was changed
22:47.12freemangordon<PROTECTED>
22:47.15freemangordon:(
22:47.22Palirestart hal
22:47.44Palior can you create backup of first MB of your partition?
22:47.59freemangordondd?
22:48.05Paliyes with dd
22:48.34freemangordonyou think that bootsectors contains FAT signiture?
22:48.39freemangordon*bootsector
22:49.06Palimaybe
22:49.08freemangordonif that is th case then it is mkswap to blame
22:49.44Palior you can try to dd if=/dev/zero for first MB of your part
22:49.45freemangordonbut that won't solve our problem. BTW why do you think windows dows not update partition table?
22:49.59freemangordonok, will try that
22:50.21Palimaybe also mkswap/swapon in busybox is broken
22:50.41Paliso maemo did not updated partition correctly
22:51.35Paliwe found a lot of bugs in busybox, so this will be no surprise if swap is also broken...
22:52.14PaliDocScrutinizer, do you think that busybox can be safely use for swapon?
22:53.06DocScrutinizerAIUI it should, though it doesn't support priorities afaik
22:54.20freemangordon"dd if=/dev/zero of=/dev/mmcblk1p3 bs=512 count=2048" done the job :)
22:54.37freemangordonlets do mkswap on that
22:55.25DocScrutinizerfreemangordon: wtf?
22:56.18freemangordonDocScrutinizer wtf as in?
22:56.18*** join/#maemo-ssu gregoa (~gregoa@chello212186032184.406.14.vie.surfer.at)
22:56.28freemangordonI mean what did you find strange
22:56.29DocScrutinizerwhat are you doing?
22:56.45freemangordonzapping the fist meg of my swap partition
22:56.47DocScrutinizersecure erase?
22:56.54DocScrutinizerwhat for?
22:57.04freemangordonbecause HAL thinks it is VFAT
22:57.07freemangordon:D
22:57.08DocScrutinizerLOL
22:57.37DocScrutinizerI'd assume a simple mkswap should suffice
22:57.54freemangordonBut it seems it is not HAL to blame, after all it is mkswap not updating the first sector
22:58.05DocScrutinizermhm
22:58.13freemangordonPali ^^^
22:58.36DocScrutinizerwell, shouldn't this partition be marked as swap in MBR?
22:58.48Palibut HAL ignore MBR
22:58.53freemangordonit is
22:58.55Palifor detecting parittion
22:59.25Palie.g HAL tell you MBR ID, but it show that fstype is other that MBR ID
22:59.43DocScrutinizerdoesn't sound right
22:59.54freemangordonactually sound right to me
23:00.19freemangordonbecause fs type is written somewhere in the first sector
23:00.29PaliDocScrutinizer, you can see freemangordon lshal output, where is correct swap mbr id, but incorrect (fat) fstype
23:01.05Paliyou can detect fs type by first magic sequence on partition
23:01.31DocScrutinizerfreemangordon: nope, for a swap partition that's no requirement to have any special 1. sector afaik
23:02.00DocScrutinizerbtw same applies for superfloppy, obviously
23:02.25freemangordonPali, actually DocScrutinizer, is right, by no means we can guess what will be in swap partition first sector
23:02.49freemangordonor first sector is not used?
23:04.13freemangordonwe are not alone :) http://forums.fedoraforum.org/showthread.php?t=119948
23:05.10PaliI have no idea now how is that detecting...
23:05.19DocScrutinizerif HAL tries to guess fs by first sector for all partitions, then that's a bug for sure
23:05.33PaliHAL is one big bug
23:05.58Palinow I understand why more desktop distribution removed HAL
23:07.12Palifreemangordon, is now your swap and ke-recv ok?
23:07.26freemangordonseems like
23:08.00Paliok
23:08.37freemangordonbut we should solve the issue with "Nokia N900" appearing for sdcard partitions
23:09.16Paliyes, I think that problem is in hildon-fm
23:09.31freemangordonBTW yoy can tell the system where is sdcard FAT partition mounted by using an variable
23:09.38freemangordonlet me find the name
23:09.51Palinot working correctly
23:10.05Palibecause some apps has hardcoded /dev/mmc1
23:10.07freemangordonPali, I mean gconf value
23:10.13freemangordonsure?
23:10.18freemangordondamn
23:10.18Palinot exactly
23:10.33Palithere is env, gconf key and hardcoded values in apps
23:10.57freemangordon"/system/osso/af/mmc-mount-point"
23:11.06Paliso for FAT partition (write access) is good idea to use only /media/mmc1
23:11.13freemangordonok
23:11.45freemangordonso you think it is hildon-fm? then it should be fixable
23:11.48Paliand I saw that each app using other method (gconf, global env, or the worse hardcoded name to /media/mmc1)
23:12.17PaliI think that hildon-fm decide what to show in osso file manager (and also in select file dialog)
23:12.29Paliand also decide how to show name + icon
23:12.53Paliyou can try to use kp50 + last ke-recv for usb storage (host mode)
23:12.57freemangordonok, let me push updated debian/changelog for ke-recv an I'll look at hildon-fm
23:13.10Paliit will automount it to /media/usb/<mount_point>
23:13.28Paliand in filemanager will be nice usb icon with correct storage name
23:13.44freemangordonI don't have usb stick around :D
23:13.54Paliok
23:13.54freemangordonbut could try with second n900
23:14.02Pali:D
23:14.07freemangordon:D
23:17.12freemangordonPali, debian/changelog for ke-recv updated
23:17.26Paliok
23:23.57Palifreemangordon, did you find something new about thumb? or it is definitly broken on n900?
23:24.14freemangordondidn't have time to continue
23:24.26Paliok
23:24.54freemangordonPali, do you know who assigns "mmc-undefined-name" for volume name?
23:25.47Palino idea, but string "mmc-undefined-name" is not in ke-recv
23:27.01freemangordonbug #8904
23:27.03povbotBug https://bugs.maemo.org/8904 Memory card name changes to "mmc-undefined-name" after a couple of days
23:27.38Paliit is still in fremantle?
23:27.41Paligoogle say this: http://www.mail-archive.com/maemo-commits@maemo.org/msg01900.html
23:28.15Paliso libgnomevfs: +#define MMC_LABEL_UNDEFINED_NAME          "mmc-undefined-name"
23:28.22freemangordonlet me check something
23:28.53freemangordonyeah, that is what i wanted to check. we have that in cssu?
23:32.57Palisource code is on gitorious: http://maemo.gitorious.org/maemo-af/osso-gnome-vfs2
23:33.19Palibinary package libosso-gnomevfs2-0
23:35.21Paliand here is that patch: http://maemo.gitorious.org/maemo-af/osso-gnome-vfs2/blobs/master/debian/patches/07-mmc-card-name.patch
23:35.30freemangordonPali, no mmc-undefined-name find there
23:35.55Paliline 33
23:36.57freemangordonaaah, stupid me, i was skipping .patch files :)
23:37.34Palilook to branch fremantle, not master, it has more commits
23:39.05freemangordonyeah, lots of .patch files.
23:39.17freemangordonanyway, that should be fixable
23:41.01freemangordonhmm. that tries to open /tmp/.mmc-volume-label and searches there, i think it is ke-recv that writes that file
23:41.55Paliyou are right
23:42.20freemangordonso we are back at ke-recv :)
23:43.04Palibut in that patch is somethink which determinate name of entry in file manager
23:43.27Paliso it should be fixed to work also for /media/mmc1p*
23:44.41freemangordonwell, having several entries on filemanager with th same "Memory Card" labels is not much frendly
23:44.52Palifreemangordon, updating is done by shell script osso-update-mmc-label.sh (part of ke-recv)
23:45.02freemangordonyeah, I know
23:45.17Palithis is updating that /tmp/ files
23:45.44freemangordonbut we should put at lieast partition sizes next to "Memory Card", at least that is what ubuntu does
23:45.47Palibut fat, ext, ... also support labels
23:46.22Paliok
23:46.59freemangordonok, use the label if there is one and concat the size in Gb, concat only the size if there is no label
23:47.02Paliseems that using that osso-* for getting label working only for FAT and only for partitions in /etc/mtools.conf
23:47.16freemangordonyeah
23:47.31PaliHAL does not tell us label?
23:47.53freemangordonNFC, let me check
23:48.15Paliit should be volume.label in HAL key
23:48.36freemangordonI am not sure I have label set on my ext3 partition
23:49.06freemangordonyeah ,  volume.partition.label = ''  (string)
23:49.44Palino volume.label
23:50.13PaliI tested and it give me correct for ext4, swap and vfat
23:50.18freemangordonvolume.label = ''  (string)
23:50.43Palido you have set any label?
23:50.46freemangordonno
23:50.48freemangordon:)
23:51.05Paliso then it is ok
23:51.20freemangordonyes
23:51.21Paliwe can drop that stupid osso-* label script and use hal
23:51.35Palibecause hal told me also ext4 label
23:52.11Pali(drop = not call that script from ke-recv)
23:52.28Paliwe do not know if other programs in maemo does not depend on it
23:53.06freemangordonno, we can't as it updates /etc/mtools.conf, and I suspect that is used from /usr/sbin/osso-usb-mass-storage-enable.sh
23:53.21freemangordonand maybe by some others
23:55.11Paliok
23:55.40PaliI have other patches for pc-suite-enable.sh (and disable too) to enable usb network
23:55.58Paliwhat do you think can it be added to cssu?
23:56.42freemangordonwhat about sesection dialog, we should rewrite it too
23:56.48freemangordon*selection
23:56.55Paliis closed
23:57.00freemangordoni kno
23:57.04freemangordondamn
23:57.05freemangordoni know
23:57.18Palibut we can use pc-suite mode also for usb networking
23:57.30Palibecause pc suite mode using g_nokia gadget
23:57.50Paliwhich automatically add usb0 interface
23:58.07freemangordonare you sure? I think there is an additional gadget when connection to pc-suite, let me check
23:58.27PaliI have already patched that script and it is working :-)
23:58.55Paliit also startup dhcp server in n900, so I can easy connect it to my laptop
23:59.33freemangordonisn't qtmobilehotspot supposed to do that?
23:59.40Palino idea
23:59.44freemangordon:D
23:59.47Palisee fixed scripts: https://code.launchpad.net/~pali/+junk/maemo_pcsuite-usb-network

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.