**** BEGIN LOGGING AT Thu May 15 02:59:56 2008 May 15 03:06:32 liwei: means what it says :D May 15 03:07:10 * nab waves May 15 03:07:21 Greetings all. May 15 03:07:23 :-) May 15 03:07:38 mornin' May 15 03:07:48 Does anyone know how to increase the buffer size for audio player with the media player...? May 15 03:08:19 I really want to walk to the store listening to tunes using iSCSI/GPRS. :-) May 15 03:10:53 Using the 10 Kb/sec link it skips pretty bad. May 15 03:11:19 I am assuming the defaults are set to read from SD level I/O performance. May 15 03:11:34 iSCSI/Wifi is faster than that btw. May 15 03:11:44 10k/s can't work May 15 03:11:52 gprs is 5k/s May 15 03:12:29 GPRS ~= 56k modem ~= 7 KiB/s theoretical max :D May 15 03:12:59 Actually May 15 03:13:07 root@om-gta02:~# hdparm -t /dev/mmcblk0 May 15 03:13:07 Timing buffered disk reads: 4 MB in 4.85 seconds = 843.77 kB/sec May 15 03:13:10 i think gprs is slower than 56k May 15 03:13:17 root@om-gta02:~# hdparm -t /dev/sda May 15 03:13:17 Timing buffered disk reads: 2 MB in 5.23 seconds = 391.81 kB/sec May 15 03:13:22 /dev/sda == iSCSI LUN May 15 03:13:30 nab: not over GPRS May 15 03:13:38 SpeedEvil: I know May 15 03:13:42 Sorry, iSCSI/WiFI LUN May 15 03:13:43 :-) May 15 03:13:57 Oh - I diddn't realise GTA02 May 15 03:14:02 I wonder if that's a slow SD. May 15 03:14:03 But anyways, having to wait for 30 seconds before I can listen to mp3sis OK May 15 03:14:20 For iSCSI/GPRS May 15 03:14:29 Or if GTA02 is slower than GTA01 May 15 03:14:36 The mounts are completely stable over iSCSI/GPRS btw May 15 03:14:37 shitloads of local caching would be nice if possible May 15 03:14:38 I get 3MB/second over SD May 15 03:14:56 Now its just a matter of getting the apps to take into account the slow link. ;P May 15 03:15:03 SpeedEvil: Hrmm May 15 03:15:16 I am running git from 05132008 May 15 03:16:00 Linux om-gta01 2.6.24 #2 PREEMPT Thu Apr 24 23:51:06 CDT 2008 armv4tl unknown May 15 03:16:02 Anyways, anyone know where I should start looking for pre-buffer settings for audio with openmoko-media-player2 May 15 03:16:34 I think it uses pulseaudio for the backend May 15 03:17:17 * zash would grep -i buffer /path/to/openmoko-media-player2/src for starters May 15 03:17:32 * nab nods May 15 03:17:46 I see ~/.openmoko-mediaplayer May 15 03:18:00 But it appears to be non-text and related to the current playlist May 15 03:18:44 Ok, I see /etc/pulse May 15 03:19:27 Hrmm, nothing wrt pre-buffer size May 15 03:21:22 Ok, time to ask in #pulseaudio May 15 03:21:24 :-) May 15 03:23:54 raster: greetings May 15 03:24:11 raster: Hey, do you know anything about pulseaudio wrt openmoko-mediaplayer..? May 15 03:24:20 nada May 15 03:24:20 I don't think it's pulseaudio May 15 03:24:26 I am trying to figure out how to increase the 'pre-buffer before playback' for iSCSI/GPRS May 15 03:24:43 you don't want to do that in pulseaudio May 15 03:24:56 you want to do it in whatever is feeding the raw data May 15 03:25:02 not after you decode it May 15 03:25:10 Hrrmmm May 15 03:25:49 Who is the maintainer of openmoko-mediaplaye..? May 15 03:26:13 I will drop them a note and cc openmoko-devel May 15 03:26:28 * SpeedEvil looks at abraxa. May 15 03:26:50 * SpeedEvil picks abraxa_up, and drops him, to see if there is any response. May 15 03:27:25 * nab double checks http://wiki.openmoko.org/wiki/Media_Player May 15 03:27:34 Yep, nothing wrt a pre-buffer size May 15 03:27:57 I actually had this same problem with EZX devices May 15 03:28:10 As I understand the code, on an indepth reveiw (lasting 20s) it throws the file to mplayer May 15 03:28:20 I suppose in the pre-openmoko mobile world, audio is only coming from flash May 15 03:28:20 :-) May 15 03:28:29 Ahh yes. May 15 03:28:35 mplayer has -cache I believe May 15 03:28:41 yeah May 15 03:29:12 ah no May 15 03:29:14 gstreamer May 15 03:29:26 * nab looks May 15 03:29:32 stoomer May 15 03:29:44 line 287 playback.c May 15 03:29:53 omp_playback_play() May 15 03:31:08 goo-stewer May 15 03:31:48 Oh no! Rasters been driven mad by nasty graphics acceleration chips! May 15 03:31:59 SpeedEvil: Thanks May 15 03:32:05 :-) May 15 03:32:35 * raster stews SpeedEvil's goo May 15 03:32:58 // Create audio sink for PulseAudio May 15 03:32:58 omp_gst_audiosink = gst_element_factory_make("pulsesink", NULL); May 15 03:33:00 Hmmm May 15 03:33:32 // Let's have gstreamer and PulseAudio meet embedded requirements May 15 03:33:32 g_object_set(G_OBJECT(omp_gst_audiosink), "buffer-time", omp_config_get_pulsesink_buffer_time(), NULL); May 15 03:33:33 g_object_set(G_OBJECT(omp_gst_audiosink), "latency-time", omp_config_get_pulsesink_latency_time(), NULL); May 15 03:33:36 There we go. ;-) May 15 03:33:45 You don't care about that. May 15 03:33:50 That's on the output side. May 15 03:33:57 Oh.. May 15 03:35:52 are u playing audio.video from a network stream? May 15 03:36:13 raster: Its off a iSCSI/GPRS mounted R/O block device May 15 03:36:25 oooh May 15 03:36:33 ouch May 15 03:36:38 why opkg so sllow,and often download failed May 15 03:36:43 last time i checked gprs gets u about 40kb in theory May 15 03:36:45 20kbit in reality May 15 03:36:45 The mediaplayer works great coming off SD or iSCSI/Wifi, (so it assumes a few 100 Kb/sec bandwith) May 15 03:36:52 raster: Less than there I think May 15 03:36:56 and nasty-hairy-arse latency May 15 03:36:56 s/there/that May 15 03:37:07 raster: The mount is stable over iSCSI/GPRS however May 15 03:37:19 raster: I was consistently gettig 5K/s+ over GPRS for a gig or so May 15 03:37:27 (I added some WAN tuning bits to the Openmoko/iSCSI builds btw) May 15 03:37:33 u must be playing-ultra-low-quality audio.... May 15 03:37:38 i want to install python , and i opkg install python , but he couldn't download it May 15 03:37:41 raster: Some audio books May 15 03:37:46 SpeedEvil: 40kbit then :) May 15 03:37:52 SpeedEvil: thats the theory May 15 03:37:54 raster: But I have no problem to let it buffer up for 60 seconds to get the goods May 15 03:38:03 in my experience iw as lucky to go over 20kbit May 15 03:38:06 For the higher quality stuff. :-) May 15 03:38:12 tbut then it may have degraded due to reception and cell traffic May 15 03:38:22 And hopefully it will agreesively buffer the next track in the playlist too. :P May 15 03:38:31 * nab lives in the bay area May 15 03:38:40 I am used to it May 15 03:38:41 :P May 15 03:39:25 * nab notes that iSCSI connections (communication paths) can be brought up and down on the fly (doing from WiFi to GPRS and vice-versa) May 15 03:40:05 So the transition happens instantly from the users's perspective. May 15 03:40:38 But just waiting a single communication path's packets to get resent over the new link should work (with some small pauses) out of the box. May 15 03:40:39 As can iFAT May 15 03:40:44 ACtually May 15 03:40:52 Using vfat really sucks for iSCSI/GPRS May 15 03:41:04 Espically if you want anything above a few 100 MB May 15 03:41:34 I remember when I once tried to make the GTA01 do iSCSI/USB-net connected to a 100 GB USB device formatted vfat May 15 03:41:50 We waited 20 minutes for a directory listing (eg: one really big one) May 15 03:41:52 if u cant sustain download speeds on average at least at the playback rate - it's pointless buffering inthe long run May 15 03:41:53 if u can May 15 03:42:00 then u just need to cover up the bad patches May 15 03:42:08 But ext3 works fine for that stuff May 15 03:42:11 raster: May 15 03:42:20 i assume thats your problem? May 15 03:42:31 I never said iSCSI/GPRS wrt audio playback was easy. :-) May 15 03:42:35 hehehe May 15 03:42:44 raster: Well, I can get a steady GPRS link here. May 15 03:42:52 And like I said, the I/O is stable, just slow May 15 03:42:52 imho the player should adjust buffering May 15 03:42:56 NFS even might be better. May 15 03:43:00 raster: In a perfect world, yes May 15 03:43:05 SpeedEvil: Not even close. :-) May 15 03:43:11 NFS sucks over WAN May 15 03:43:12 rsync to pull new tracks May 15 03:43:17 it should learn what buffering is needed based on examining the data flow May 15 03:43:44 so if it g4etts stuttry, up the buffer size so when u have spare bw it fills it up a bit May 15 03:44:07 wrt GPRS network access, using block level access is the only (somewhat) useable setup according to my tests May 15 03:44:17 raster: Makes sense May 15 03:45:02 raster: I just figure doing a 512k buffer upfront or something May 15 03:45:21 I am patient. :P May 15 03:45:42 http streaming works just fine for me. May 15 03:46:41 SpeedEvil: Using openmoko-mediaplayer2..? May 15 03:46:58 I was trying to figure out how to access shoutcast stuff from Browser -> Media Player May 15 03:47:15 Ok, I will drop abraxa a line about the pree-bufferring bit May 15 03:47:31 I dunno - I had tried mplayer url://whatever May 15 03:47:36 and it sort-of-worked May 15 03:48:00 SpeedEvil: Gothca May 15 03:48:05 * nab looks for abraxa's email May 15 03:48:38 Ok, I am going to get a cold beer (It was 92 here today) May 15 03:48:50 I guess one more walk w/o iSCSI/GPRS is acceptable May 15 03:48:51 :P May 15 03:53:23 rememebr that for next time based on mountpoint May 15 03:53:23 or on network address May 15 04:09:24 * nab is still looking for *** Signoff: kdean06 ("I am free, no matter what rules surround me. I am free because I know May 15 04:09:24 +that I alone am morally responsible for everything I do) May 15 04:09:24 *** mbuf (n=shakthim@61.16.248.242) has joined channel #openmoko May 15 04:09:29 Ugh May 15 04:09:34 That was supposed to be; May 15 04:09:50 * nab is still looking for abraxa's email addy May 15 04:10:07 use the source! May 15 04:10:09 Anyone know that offhand..? May 15 04:10:17 Doh May 15 04:10:18 :-) May 15 04:10:50 SpeedEvil: Thanks again May 15 04:18:49 morning all May 15 04:19:52 hi Sup3rkiddo May 15 04:20:02 summatusmentis, y0 May 15 04:20:03 have you gotten your GSoC test machine yet? May 15 04:20:13 nope...not yet... May 15 04:20:37 :-/ May 15 04:37:36 SpeedEvil: Hey May 15 04:37:51 What is the quickest way to get mplayer on my device..? May 15 04:38:44 I would like to try with -cache 65536 (which I think the the defacto upper limiter for mplayer wrt pre-buffer May 15 04:38:59 And I can do some iSCSI/Wifi video playback too May 15 04:38:59 :-) May 15 04:44:33 hi! May 15 04:45:07 * nab waves May 15 04:45:08 Greetings May 15 04:45:19 what are the features the devs are working on right now? i read somewhere that its "Not suitable for end-users at release", what does this mean? May 15 04:45:57 I think they mean its not ready for grandma to use it for her daily phone. May 15 04:46:03 (at least not my grandma :-) May 15 04:46:26 But I can say that the basic bits seem to be quite stable now. May 15 04:46:35 meaning the userspace phone stuff May 15 04:46:37 haha May 15 04:46:43 yeah i'll manage May 15 04:46:56 but i'm just wondering whats not done yet May 15 04:46:57 But I am not working on that personally, so I can't say for 100% May 15 04:47:05 okey May 15 04:47:27 Xecuter: Feel free to ask in here and the folks with GTA02 hardware can tell you May 15 04:47:36 the important thing at release is ringing at sms... the rest can come as they develop... May 15 04:47:45 ok May 15 04:48:15 Also, you can look for specifics on the openmoko community list archive @ http://lists.openmoko.org/pipermail/community/ May 15 04:48:33 openmoko-devel is a good spot for questions too May 15 04:49:07 Xecuter: Actually, wrt sms.. Did you search the wiki..? May 15 04:49:20 no... should i? :P May 15 04:49:25 Of course May 15 04:49:30 :-) May 15 04:49:37 hehe May 15 04:50:19 I do see one "Unknown Sender" sms on my GTA02 sample May 15 04:50:50 Michael Shiloh does testing before sending them out to devs.. May 15 04:50:53 :-) May 15 04:51:10 May 15 04:51:33 * nab has had to work on sample hardware, that well, really did not work at all. May 15 04:51:33 hehe May 15 04:51:50 Thankfully the GTA02 has been doing everything I have been wanting to do with it. May 15 04:51:54 are you on the dev team? May 15 04:52:00 or part of production? May 15 04:52:02 3rd party developer May 15 04:52:07 ok May 15 04:52:10 3rd OSS devel. :-) May 15 04:52:11 so like you and i? May 15 04:52:24 Mostly kernel stuff for now May 15 04:52:38 what cool stuff can we see from nab? ^^ May 15 04:52:54 I have been doing a project for a number of years implementing the Internet Small Computer Interface for Linux v2.6 May 15 04:53:07 Both Initiator/Target (Client/Server) implementation May 15 04:53:24 The project is http://linux-iscsi.org May 15 04:54:04 Actually, the main page needs to be updated for the Freerunner stuff May 15 04:54:10 http://linux-iscsi.org/index.php/OpenMoko_Freerunner May 15 04:54:13 for the v2.6.24 builds May 15 04:54:20 cool May 15 04:54:25 There are also older Neo1973 builds, that I will be updating when I get my Neo back May 15 04:54:37 for v2.6.22.6 May 15 04:54:46 (for the kernel module) May 15 04:54:57 Also, the plan is to get OCFS2 up and running on the device May 15 04:55:11 So that Nokia/Maemo hardware can share a cluster FS over iSCSI/WiFi May 15 04:55:21 (Along with your R/W mounted iSCSI LIO-VMs ;-) May 15 04:55:25 i have noooo idea what you are talking about :P May 15 04:55:44 http://linux-iscsi.org/index.php/OCFS2/iSCSI_for_OS-2008 May 15 04:55:46 thats way to technical for me ^^ May 15 04:56:13 So, a cluster filesystem, or shared disk fileystem allows multiple computing nodes to access the same filesystem mount at the same time May 15 04:56:30 eg: If one node creates a directory and/or new files, the other nodes see this May 15 04:56:33 aaah May 15 04:56:50 This type of tech is usually used in high-end database and mission criticial infrastructure May 15 04:56:58 eg: linux-iscsi.org/builds run on OCFS2 May 15 04:57:13 But scaling it down to mobile stuff is quite an interesting project May 15 04:57:50 Espically when you can OpenMoko+Maemo+Wifi+VM+iSCSI May 15 04:57:53 :-) May 15 04:57:58 ^^ May 15 04:58:42 So when you get new mp3s, then can go on the cluster storage and can be accesed by all of the VM and Embedded nodes May 15 04:58:49 Or what-have you. May 15 04:59:06 I am currently working on accessing smaller FS mounts (sub 10 GB) over iSCSI/GPRS May 15 04:59:19 Which is up and running on the kernel side. May 15 04:59:47 * zash wants afs | coda | nfs-ish with caching for offline May 15 05:00:05 Now the question is how do you get an userspace media player to understand you are running on a network-backed storage device that can only do single digit K/b sec for transfer rates May 15 05:00:06 :-) May 15 05:00:17 zash: :-) May 15 05:00:26 Offline ops are very nice, no question May 15 05:00:41 So what we need is a non-networked filesystem that runs over iSCSI that does offline ops. May 15 05:01:03 Mabye we can bribe some of the btrfs folks to include this feature for their alpha-devel May 15 05:01:04 :P May 15 05:01:25 ^^ May 15 05:01:45 btrfs == next-generation Linux v2.6 filesystem out of Oracle May 15 05:02:00 how to install finger-keyboard May 15 05:02:50 i haven't any idea for 4 hours May 15 05:03:08 liwei: Ask on openmoko-devel May 15 05:03:14 Or the community list :-) May 15 05:04:48 thank you , i change channel May 15 05:05:02 liwei: I mean the mailing list May 15 05:05:03 :-) May 15 05:05:25 http://lists.openmoko.org/mailman/listinfo/ May 15 05:35:05 it's 15 already May 15 05:36:20 CVirus: I am sure the OM/FIC folks are busy. :P May 15 05:37:58 * nab notes that making a call while listening to audio over iSCSI/Wifi with the OM Media Player works! May 15 05:38:55 Same with incoming calls May 15 05:39:06 Well, the incoming ones don't kill the media player audio May 15 05:39:33 * nab wonders how iSCSI/GPRS would work for that May 15 05:40:21 * zash don't think that will work May 15 05:41:02 zash: well, with the R/O mounts killing the TCP connection will work May 15 05:41:06 And will re-connect when the GPRS link comes back online May 15 05:41:41 (Once gsmd and gprs can be used in parallel of course :-) May 15 05:41:57 if you are using gprs, how are you going to answer a call? May 15 05:42:09 * nab nods May 15 05:42:17 I suppose you will have to ask Harald that one May 15 05:42:18 :P May 15 05:44:31 afaik gprs is like old times modem May 15 05:45:52 * nab listens to voicemail with iSCSI/Wifi mp3s paused with OM MediaPlayer May 15 05:45:55 Cooooool May 15 05:46:03 zash: May 15 05:46:09 sub 10 Kb/sec May 15 05:46:15 data transfer rates May 15 05:50:18 and 3g was proprietary? May 15 05:51:27 zash: TIA I would imagine May 15 05:51:56 TIA == Trade Industry Association standards May 15 05:52:20 But if the use is large enough, that type of thing *usually* moves toward open standards May 15 05:52:49 The usb-net 3g modems work under linux, sooo. :-) May 15 05:53:02 I don't know about the actual wire protocol May 15 05:53:18 er, radio wire protocol May 15 05:53:19 :P May 15 05:53:34 :D May 15 05:55:10 Wow, the GTA02 power management looks pretty solid May 15 05:55:26 With mp3 iSCSI/Wifi usage May 15 05:55:43 I suppose the power to drive the LCD is the biggest draw May 15 05:56:17 Doing playback of 128 kbit audio over 802.11 b/g is the current test-bed May 15 05:56:59 nab: you have a GTA02 in hand right now ? May 15 05:57:04 CVirus: yes May 15 05:57:10 lucky you May 15 05:57:26 I assume you're an OM dev ... right ? May 15 05:57:42 CVirus: From my first few days with it, the OM devs have been spending alot of time on the OOB usage May 15 05:58:01 CVirus: well, linux-iscsi.org dev May 15 05:58:23 But with LIO code, you run into methods which the OM application code has not been used just yet. May 15 05:58:25 :-) May 15 05:58:37 That has historically been the case on other platforms as well May 15 05:59:11 Espically when we are talking about GPRS link speeds May 15 05:59:54 nab: how did you get your self a FreeRunner ? May 15 06:00:25 CVirus: Well, I think you need to be working on OSS projects that are going to be useful to the OM platform May 15 06:00:59 OM/iSCSI is going to be useful for devel/hackers and (hopefully) end-users May 15 06:01:02 :-) May 15 06:02:04 * nab plans to bribe raster to make an OM/iSCSI+iSNS client UI May 15 06:02:14 * nab updates linux-iscsi.org .plan May 15 06:02:15 :-) May 15 06:02:36 Actually, there is some working python+gtk core-iscsi code May 15 06:02:41 On the Maemo platform May 15 06:02:58 better have a lot of cash May 15 06:03:04 because i dont have any spare time :) May 15 06:03:05 * nab goes to the ATM May 15 06:03:10 hehe May 15 06:03:16 Ok, mabye just some questions May 15 06:03:16 :-) May 15 06:04:43 Or mabye one of the hunger-devels here in #openmoko would like to do some OM+E+iSCSI code May 15 06:04:44 :-) May 15 06:04:56 HUNGRY May 15 06:04:58 :-) May 15 06:05:50 * nab can provide active-active H/A + active/active iSCSI redundacy to OM devices May 15 06:05:51 :-) May 15 06:07:05 Even on non Linux based target platforms May 15 06:07:28 * nab will not speak such heresy May 15 06:07:48 Oh, the power of x86/virtualization May 15 06:08:09 :-) May 15 06:09:16 Is there a method to change volume via OM-UI yet..? May 15 06:13:32 * nab continues to listen to on-battery iSCSI/Wifi May 15 06:13:52 Very solid power usage May 15 06:14:03 * nab looks at /sys for exact usage May 15 06:14:57 /sys/devices/platform/bq27000-battery.0 May 15 06:15:09 Hmm, I don't see the output for the battery level applet May 15 06:15:22 Anyone know where that data is coming from..? May 15 06:15:55 * nab finds /sys/devices/platform/bq27000-battery.0/power_supply/bat May 15 06:15:56 :-) May 15 06:16:41 Actually, I have the Power Manager stuff completely disabled via the UI May 15 06:16:43 WOW May 15 06:16:53 The screen is still going dim though May 15 06:17:19 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# cat time_to_empty_now May 15 06:17:19 17400 May 15 06:17:41 What units are those again..? May 15 06:20:45 yeah, the PM is supposed to be disabled, according to the UI accessable while pressing and holding the power-button May 15 06:20:59 I would consider that a feature and != bug May 15 06:21:02 :-) May 15 06:21:02 Volts? May 15 06:21:24 * nab has no clue May 15 06:22:00 * nab wonders what the divisor is for Volate vs. Mins May 15 06:22:06 Volage too May 15 06:22:06 :-) May 15 06:22:14 ugh, VOLTAGE May 15 06:22:42 charge it to max, check value, let it drain until and see what min value is May 15 06:22:49 I must admit its pretty amazing to watch the hardware-devels to their thing May 15 06:23:12 zash: Or, we can look at the code May 15 06:23:28 or that :D May 15 06:23:30 Actually, I wonder if anyone has put the OM tree into LXR just yet May 15 06:23:36 LXR == Linux Cross Reference May 15 06:23:55 Which makes groking C code much quicking than $EDITOR/grep May 15 06:24:01 s/quicking/quicker May 15 06:24:23 Espically one as large as OM userspace May 15 06:24:42 http://lxr.linux.no/ May 15 06:24:49 If any openmoko.org devels are around May 15 06:24:50 ACtually May 15 06:24:57 I should send that to openmoko-community May 15 06:28:53 * nab sends May 15 06:33:08 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# cat time_to_empty_now May 15 06:33:08 11280 May 15 06:33:08 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# date May 15 06:33:08 Sat Apr 26 18:02:04 UTC 2008 May 15 06:33:08 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# cat time_to_empty_now May 15 06:33:09 10560 May 15 06:33:11 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# date May 15 06:33:13 Sat Apr 26 18:02:35 UTC 2008 May 15 06:33:40 So 720 decrease in ~30 seconds May 15 06:33:50 or 24 May 15 06:34:15 shit May 15 06:34:22 that is shocking May 15 06:34:37 The battery applet still shows 90% usage May 15 06:34:48 And I have been listening to 128 kbit iSCSI/Wifi audio for 45 mins..? May 15 06:34:49 I am assuming its slow on the uptake May 15 06:34:57 =/ May 15 06:34:58 Actually, closer to 60 mins May 15 06:35:26 write a script that takes the value at regular intervals and generate a nice graph :D May 15 06:35:30 patoh: Well, add a second or two for human CLI interaction May 15 06:35:32 :-) May 15 06:36:11 Actually, the applet just went to ~80% May 15 06:36:29 That is seriously good. May 15 06:36:52 With the LCD dim, you can listen to Wifi network audio for 4x hours! May 15 06:36:55 so it can last for around 10 hours of audio playing May 15 06:36:59 ok, I was way off :P May 15 06:37:08 I wonder how that would work with iSCSI/GPRS May 15 06:37:10 wonder what its like with the wifi off May 15 06:37:15 May 15 06:37:28 wait, with wifi, got a gta02? May 15 06:37:31 Actually 4+ hours May 15 06:37:46 :-) May 15 06:37:49 * nab nods May 15 06:38:00 It has been working very well May 15 06:38:18 nice, suspend? May 15 06:38:18 I still don't know if its a GTA02v5 or GTA02v6 May 15 06:38:40 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# cat /proc/cpuinfo | tail -n 3 May 15 06:38:40 Hardware: GTA02 May 15 06:38:40 Revision: 0350 May 15 06:38:40 Serial: 0000000000000000 May 15 06:38:51 patoh: I believe so, yes. May 15 06:39:09 nab: how long standby time does it have? May 15 06:39:12 And after the battery is drained, just plugging it into USB makes it reboot and go May 15 06:39:24 faern: Not sure May 15 06:39:38 Hopefully that can be estimated in the UI soon. :-) May 15 06:39:51 yes :P May 15 06:40:12 the battery life is my only concern about the neo May 15 06:40:28 http://wiki.openmoko.org/wiki/Neo_FreeRunner_GTA02_Hardware#Features May 15 06:40:51 "Battery life (Approximation/Ideal Target) Standby time 150-200 Hrs (GSM) Talk time (Backlight off) Up to 3-4 hrs(GSM)" May 15 06:41:10 Sounds good May 15 06:41:10 :-) May 15 06:41:19 That sounds similar to my N800 May 15 06:41:19 yes. that is ok May 15 06:41:24 not much talk time :( May 15 06:41:33 my 6300 can do about 8 hours May 15 06:41:39 than again it has a white screen of death too ;) May 15 06:42:15 :P May 15 06:42:30 * nab waits for N810 Wimax Ed. May 15 06:42:45 but it says "Ideal time" so I'm worried the powersaving is not too good implemented yet May 15 06:42:49 Hopefully we can get some OM-WiMax goodness by GTA05 May 15 06:42:58 Or via USB or BT May 15 06:43:07 But I figure that BT would be too slow for Mobile Wimax May 15 06:43:16 At least in the current incarnations. May 15 06:43:31 Well May 15 06:43:37 That would still faster than GPRS May 15 06:43:37 :-) May 15 06:46:23 so gta03 is been skipped? May 15 06:46:37 patoh: GTA03 is the GPS device May 15 06:46:42 The Dash May 15 06:47:52 GPS+WIFI device May 15 06:48:33 no, but there is a pic of the gta02 here: http://walkingice.twbbs.org/blog/?p=336 ;) May 15 06:48:58 i said gta03 May 15 06:54:27 heh, cool May 15 06:56:21 At ~75% usage with 128 kbit iSCSI/Wifi May 15 06:57:27 nab: have you tested your GTA02 with a voice call to another person? If so, any comments on the audio quality? May 15 06:57:52 mmontour: I pinged voicemail this evening May 15 06:58:07 (The dialer just called it for me) May 15 06:58:24 through the headset, and it sounds OK to non-audio engineer ears May 15 06:58:29 The real test is a live conversation with someone. On my GTA01 the other person hears an annoying echo of their voice May 15 06:58:46 * nab never tried that on his GTA01 May 15 06:58:51 I could never get iSCSI/GPRS working. :P May 15 06:58:56 (with the old device :-) May 15 06:59:16 I am guessing the GSM firemware updates will solve this May 15 06:59:46 And hopefully soon we will get the iSCSI/GPRS case working for all of the OM devices (including the DASH) May 15 06:59:47 :-) May 15 07:00:03 Last I heard there will be no GSM firmware updates in the field (although the mass production ones may have newer FW than the development samples) May 15 07:00:09 well, DASH for 802.11 b/g access May 15 07:00:23 mmontour: I sent mine to Michael Shiloh May 15 07:00:46 He is handling the OM devel relations in the US May 15 07:00:47 I'm not personally interested in GPRS - Canadian data rates are stupidly expensive (starting at $50/MB) May 15 07:01:00 (and all over from what I can tell :-) May 15 07:01:06 Ugh, that sucks May 15 07:01:24 Yeah, us US folks in/near big cities are pretty lucky May 15 07:01:33 hi May 15 07:01:53 I am interested in GSM for phone calls and messages, can use wifi when I need net :) May 15 07:01:53 I think the Mobile Wimax + Ebay Mobile-WiMax attenne is the way to go May 15 07:02:02 our GRPS rates are worse than mmontour's :( May 15 07:02:04 (for the 2010 stuff) May 15 07:02:22 nab: is the gta04 the suposed 3g device? May 15 07:02:29 GSM is nice now because of the obvious bit of the GSM carriers being seperated from the handset providers May 15 07:02:37 patoh: I believe that is the current talk on openmoko-kernel May 15 07:02:46 cool, I suspected so :) May 15 07:03:08 I know there is significant Mobile-WiMax investment in Taiwan May 15 07:03:31 * patoh wonders what syncing on linux will be like with openmoko May 15 07:03:35 devices May 15 07:03:41 I wonder if going native GTA04+Mobile-Mimax with BT 3-G would be viable May 15 07:04:04 Going with Intel silicon would be the obvious choice (which is what Nokia is using) May 15 07:04:12 On their Baxter Peak mobile platform May 15 07:04:46 We shall see how Xohn does this summer with their US launch May 15 07:04:50 Xohm May 15 07:05:40 Getting 1-2 Mbit (125 - 250 Kb/sec) for typical case is pretty convincing for Mobile-WiMax May 15 07:05:52 With 5-10 mile cell tower access May 15 07:06:21 Not to mention that w/ Xohm you can purchase Mobile-Wimax packets by the day/week/month May 15 07:06:37 (or so says the current new items) May 15 07:08:06 s/new/news May 15 07:08:07 :-) May 15 07:08:41 is the dash software (client and server side) going to be opensource? May 15 07:10:27 patoh: Have a look at http://lists.openmoko.org/pipermail/community/2008-May/017205.html May 15 07:11:59 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# date May 15 07:11:59 Sat Apr 26 18:41:23 UTC 2008 May 15 07:11:59 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# cat time_to_empty_now May 15 07:11:59 9960 May 15 07:12:22 Still at ~75% according to the battery appllet May 15 07:12:26 applet May 15 07:13:38 good morning May 15 07:13:43 * nab waves May 15 07:13:54 Goodmorning (GST) Hopscotch! May 15 07:14:14 :) May 15 07:33:01 * nab notes the battery applet says ~50% May 15 07:33:18 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# cat time_to_empty_now May 15 07:33:18 8640 May 15 07:33:18 root@om-gta02:/sys/devices/platform/bq27000-battery.0/power_supply/bat# date May 15 07:33:18 Sat Apr 26 19:02:42 UTC 2008 May 15 07:33:33 Actually, just jumped back to ~75% May 15 07:33:33 :-) May 15 07:33:53 (from eyeballing the UI :-) May 15 07:38:51 > tanuk: I am using an audio book @ 128 kbit May 15 07:38:51 > And am patient to wait 30-60 seconds for playback May 15 07:38:51 > :-) May 15 07:38:51 > (At least over GPRS) May 15 07:38:51 If the link speed is < 10 kB/s, you'll spend maybe about two minutes buffering for May 15 07:38:52 +every minute of playback. May 15 07:38:58 * nab hrrrmmms May 15 07:51:30 okay as i understood all the time i'm into this project, openmoko is about an OPEN platform [tm] - so why the hell is there NO official word on what GTA stands for??? (and no, it's not grand theft auto...) May 15 07:52:41 who cares May 15 07:52:47 everyone will call it a moko anyway May 15 07:52:51 haha May 15 07:53:08 it's just a mystery i /finally/ want to have solved May 15 07:53:18 only three words!! plz! May 15 07:53:46 i'm putting 20 on 'getting terrible acne' May 15 07:55:23 what's that program called Dinge? May 15 07:55:56 balrog-kun: have a look here: http://neo1973-germany.de/ May 15 07:56:09 balrog-kun: it's an alternative experimental gui May 15 07:56:19 based on python, efl and dbus May 15 07:57:31 josch|nsn: thaniks for the link May 15 07:58:46 balrog-kun: anyway you only distract from the GTA question to be answered - maybe this happened everytime one tried to solve this :P May 15 07:59:42 XY: hey what does gta stand for?, ZZ: hey XY have a look here!, XY: wow great! *walks away distracted* May 15 08:01:01 balrog-kun: if you have any questions you may join #neo1973-germany or ask emdete directly May 15 08:02:02 Generic Telephone Appliance May 15 08:02:11 rwhitby: for real? May 15 08:02:19 (the chinese design that it started from) May 15 08:02:59 rwhitby: wow - is this true? can i add this to the wiki? May 15 08:03:08 josch|nsn: it's to save your time, as i know we'll never figure out the GTA May 15 08:03:09 :) May 15 08:03:42 balrog-kun: watch what rwhitby has to reveal! May 15 08:03:47 josch|nsn: sure, you can add it to the wiki May 15 08:03:53 oh nice! May 15 08:04:17 (it'll be as accurate as most other stuff in wikis ...) May 15 08:04:32 hahaha May 15 08:04:53 you know that i will cite you? :P May 15 08:05:02 works for me May 15 08:05:26 unless someone gives you an official answer, my definition will be authoritative May 15 08:05:27 gneh! May 15 08:05:38 anyone build pulseaudio recently May 15 08:05:57 and if someone gives you an official answer, then you'll have your answer. so put it in the wiki and wait for someone to correct it ... May 15 08:06:36 rwhitby: i think you are authoritive enough that one may believe your word ^^ May 15 08:06:38 Actually, I think it was "GNU Telephony Appliance" May 15 08:06:46 lol rly`? May 15 08:07:01 heck, put em both in the wiki ... May 15 08:07:43 Google Takes All May 15 08:08:03 Gates Telephone Anti May 15 08:08:16 Giant Toad Arse May 15 08:08:34 Go Try Again May 15 08:08:54 "GSM Telephone (Allegedly)" May 15 08:09:18 [G]I [T]hate [A]mtn May 15 08:09:18 Grand Telephony Alliance May 15 08:09:27 s/Grand/Great May 15 08:09:28 there we are again... May 15 08:10:00 * mmontour is still waiting to hear from a GTA02 owner who has used it to talk to another real person, and who can comment on the audio quality May 15 08:10:11 | /builder/newBuild/build/tmp/staging/arm-angstrom-linux-gnueabi/usr/lib/libavahi-common.so: undefined reference to `__stack_chk_fail@LIBSSP_1.0' May 15 08:10:15 dagnabit May 15 08:11:25 sure hope GTA02 is faster than the slowwwwwwwwwwwwww GTA01 video i saw on scaredycats website May 15 08:11:44 * nab is up for doing some video playback over iSCSI/Wifi May 15 08:12:12 I thought libssp wasn't being built May 15 08:12:17 Where can I find a video player .ipk and what resolution do I need..? May 15 08:12:37 I have been using 320x240 .avi for Nokia/Maemo hardware with iSCSI/WifI May 15 08:13:45 Anyone..? May 15 08:14:32 The iSCSI/WiFi traffic is ~500 KB/sec May 15 08:14:44 (to a mounted filesystem) May 15 08:15:57 I assume I am looking for mplayer May 15 08:16:19 Considering openmoko-mediaplayer2 calls mplayer for video decoding. :-) May 15 08:16:35 * nab searches for a mplayer .ipk May 15 08:19:46 nab ipkg install mplayer should suffice May 15 08:20:30 nab: i wrote my experiences and some mencoder lines in this wiki article: http://wiki.openmoko.org/wiki/Video_Player May 15 08:21:59 josch|nsn: I assume h264 is out of the question? May 15 08:22:23 patoh: mpeg4 video is played just fine in qvga May 15 08:22:46 josch|nsn: but mpeg4 usually uses dvix/xvid to encode does it not? May 15 08:22:48 josch|nsn: Great, thank you!! May 15 08:23:15 patoh: you can use divx/xvid, yes May 15 08:23:38 josch|nsn: so MPEG4 part 2 works? May 15 08:23:43 400kbps mpeg4 works on a 200Mhz neo (without sound) on prerotated 240*320 ffmpeg codec videos, at 25fps, using stock mplayer. May 15 08:23:52 patoh: yepp in qvga May 15 08:24:03 cool, wonder how well part 10 will work iwth qvga :) May 15 08:24:20 patoh: the mpeg4 decoder is really amr optimized in mplayer - thats the reason it works so well May 15 08:24:24 * nab find some 320x240 .avi to copy onto iSCSI/WiFI May 15 08:24:26 :) May 15 08:25:38 * nab notes its pretty cool to be able to do opkg install from a device in your pocket May 15 08:25:54 patoh: you get the optimal results with the mencoder lines i noted in the wiki May 15 08:26:31 nab: /me uses debian on his neo and thinks being able to do apt-get install is even cooler May 15 08:26:50 josch|nsn: I was using some win32 encoder going from DVD -> .avi that gave me the best results on the Nokia 770 with iSCSI/WiFi video playback May 15 08:27:21 Something that someone in my old team recommended. May 15 08:27:29 I will have to look into mencoder though. May 15 08:27:31 :-) May 15 08:31:23 Cap_J_L_Picard: ping May 15 08:31:57 nab: srsly.... use mencoder or ffmpeg directly - far better than those M$ crap May 15 08:32:20 mencoder seems to be one of the best video encoders out there May 15 08:32:29 seconded May 15 08:32:32 does not have the best set of filters but :( May 15 08:32:44 avisync for windows has some great filters floating around on the internet May 15 08:32:54 patoh: what filters do you need? May 15 08:33:09 josch|nsn: Nod May 15 08:33:17 ?? May 15 08:33:20 It was what one of my guys in my old team recommended May 15 08:33:24 hey guys, does anyone have the alsactl state files for the gta02 hardware? May 15 08:33:29 (A win32 guy, but a good guy :-) May 15 08:33:39 * nab just runs on non Linux stuff in a VM May 15 08:33:45 s/on/all May 15 08:33:56 But anyways. May 15 08:34:00 * nab copies videos May 15 08:34:08 nab: whats nod? May 15 08:34:14 ah May 15 08:34:19 nod May 15 08:34:21 ^^ May 15 08:34:40 * josch|nsn thought it was an abbr. of some kind XD May 15 08:35:04 :-) May 15 08:35:28 then /me thought about the gdi/nod thingy o0 May 15 08:35:29 josch|nsn: http://forum.doom9.org/showthread.php?t=108681 May 15 08:36:19 -rwxr--r-- 1 root root 181677540 2008-05-15 01:36 draw_the_line.avi May 15 08:36:19 -rwxr--r-- 1 root root 211409090 2008-05-15 01:36 more.avi May 15 08:36:19 -rwxr--r-- 1 root root 406894458 2008-05-15 01:34 wakinglife.avi May 15 08:36:19 [root@ps3-cell mnt]# file *.avi May 15 08:36:19 draw_the_line.avi: RIFF (little-endian) data, AVI, 352 x 208, ~15 fps, video: DivX 4, audio: MPEG-1 Layer 3 (stereo, 48000 Hz) May 15 08:36:21 more.avi: RIFF (little-endian) data, AVI, 352 x 208, ~15 fps, video: DivX 4, audio: MPEG-1 Layer 3 (stereo, 48000 Hz) May 15 08:36:24 wakinglife.avi: RIFF (little-endian) data, AVI, 352 x 208, ~15 fps, video: DivX 4, audio: MPEG-1 Layer 3 (stereo, 48000 Hz) May 15 08:36:28 josch|nsn: Hrmmm May 15 08:36:42 Yeah, I encoded these following the requirement's for maemo's media player May 15 08:36:58 patoh: -> The binary... <= HAHAHAHA May 15 08:37:08 patoh: the source is missing it seems :P May 15 08:37:37 josch|nsn: go a few posts down May 15 08:37:42 there is a improved version of it with source May 15 08:37:51 someone pm'd him for it and released it under his permission May 15 08:37:53 Ugh May 15 08:38:01 mplayer is doing remote X11 for me May 15 08:38:02 :-) May 15 08:38:10 * nab relogs-in without ssh -X May 15 08:38:38 er, unless there is another method to output to glamo when running X apps remotely..? May 15 08:39:12 There we go. :-) May 15 08:39:34 Ok, the video decoding is definately off with my encode May 15 08:39:36 And no audio May 15 08:39:38 patoh: the source file is not available anymre - so fail May 15 08:39:50 josch|nsn: Can you post a vid for me to test with..? May 15 08:39:54 with mplayer..? May 15 08:40:08 nab: sure i have one uploaded already May 15 08:40:14 josch|nsn: funny, it was there the other day May 15 08:40:37 So, I am actually getting video playback with one of mine over iSCSI/WiFi, it just looks like the player is having trouble reading the encode May 15 08:40:47 josch|nsn: Excellent, URL..? May 15 08:41:15 josch|nsn: http://kosmos.kawaii-shoujo.net/gradfun2db/gradfun2db-v1.0.zip :) May 15 08:42:01 I will save it over r/w EXT3 on iSCSI/WiFI May 15 08:42:02 :-) May 15 08:42:05 its gpl'd May 15 08:42:25 patoh: oh and it's gpl'ed - i'm impressed May 15 08:42:42 josch|nsn: What was that link..? May 15 08:42:47 josch|nsn: yeah, would take a bit of porting to mplayer, its done in cpp, but the asm part looks _ok_ May 15 08:42:52 nab: just searched for it but its not there anymore :-( May 15 08:43:08 You got a HTTP server handy..? May 15 08:43:16 You can just email it to me too May 15 08:43:26 nab linux-iscsi.org May 15 08:43:37 (thanks) May 15 08:43:47 nab: i'm at work atm (windows pc) and as far as i see it's not on one of my servers anymore May 15 08:44:11 josch|nsn: What would you recommend for me to encode my own..? May 15 08:44:24 I am using Debian/Lenny i386 May 15 08:44:32 you can encode /everything/ because mplayer eats everything May 15 08:44:40 nab: maybe just load a video from youtube May 15 08:44:47 .flv..? May 15 08:44:49 Ok May 15 08:44:59 * nab downloads May 15 08:45:00 :-) May 15 08:45:01 at least ffmpeg works well for flv May 15 08:45:44 * nab runs youtube-dl May 15 08:45:54 yeah youtube-dl ftw May 15 08:48:49 flv is simply a container format with h263 video and mp3 audio in it so it should be no problem for mplayer but as i said i onl use ffmpeg and vlc for flv May 15 08:49:04 Hmm, video output is still off (but playing) and no sound May 15 08:49:14 What should I be using for -vo ..? May 15 08:49:38 root@om-gta02:/media/sda# mplayer Coconut_Records-West_Coast.flv May 15 08:49:38 MPlayer 1.0rc2-4.1.2 (C) 2000-2007 MPlayer Team May 15 08:49:38 CPU: ARM May 15 08:49:38 Playing Coconut_Records-West_Coast.flv. May 15 08:49:38 libavformat file format detected. May 15 08:49:38 [lavf] Video stream found, -vid 0 May 15 08:49:40 [lavf] Audio stream found, -aid 1 May 15 08:49:42 VIDEO: [FLV1] 320x240 0bpp 29.917 fps 0.0 kbps ( 0.0 kbyte/s) May 15 08:49:44 vo: couldn't open the X11 display ()! May 15 08:49:46 vo: couldn't open the X11 display ()! May 15 08:49:48 VO XOverlay need a subdriver May 15 08:49:50 vo: couldn't open the X11 display ()! May 15 08:49:52 [VO_SDL] Using driver: fbcon. May 15 08:49:54 vo: couldn't open the X11 display ()! May 15 08:50:12 It looks like its picking up -vo sdl May 15 08:50:20 nab: use nopaste plz May 15 08:50:40 (sorry :-) May 15 08:51:18 * nab trys switching to landscape mode May 15 08:51:25 try -vo help to get a list of available modules May 15 08:51:53 mplayer video output looks the same. May 15 08:53:12 [AO OSS] audio_setup: Can't open audio device /dev/dsp: Device or resource busy May 15 08:53:13 Hrrmm May 15 08:53:19 nab: try fbdev or similar May 15 08:53:21 I thought I closed openmoko-mediaplayer2 May 15 08:53:44 you can find out what uses sdp by doing lsof | grep dsp May 15 08:53:59 josch|nsn: Ok, its going. :-) May 15 08:54:06 nice May 15 08:54:12 In a small section of the LCD. :-) May 15 08:54:18 yepp May 15 08:54:25 you have to enable scaling May 15 08:54:32 and rotation May 15 08:54:33 OR May 15 08:54:34 * nab now looks at which mediaplayer process is holding /dev/dsp May 15 08:54:39 you switch to qvga first May 15 08:54:57 it's all explained here: http://wiki.openmoko.org/wiki/Video_Player May 15 08:55:20 josch|nsn: Thanks I will take another look May 15 08:55:24 :-) May 15 08:55:54 nab: if i explained something now that is NOT in the wiki feel free to add it May 15 08:58:05 josch|nsn: May 15 08:58:09 :-) May 15 08:58:15 hehe May 15 08:58:24 nab: May 15 09:03:38 josch|nsn: The other encodes are working with -vo fbdev just fine btw May 15 09:05:51 nab: you can set fbdev as the standard output module in mplayer config in your home directory May 15 09:15:33 OK, after stopping pulseaudio audio is working May 15 09:17:28 hehe May 15 09:18:50 htts://linux-iscsi.org/index.php/Image:ISCSI-WiFI-Mplayer-Freerunner.png May 15 09:18:59 er May 15 09:19:15 http://linux-iscsi.org/index.php/Image:ISCSI-WiFI-Mplayer-Freerunner.png May 15 09:19:16 :-) May 15 09:19:40 * nab reads how to go fullscreen May 15 09:22:44 FATAL: Could not initialize video filters (-vf) or video output (-vo). May 15 09:22:46 Hmm May 15 09:23:01 Using the 'autosync 30 -vf scale -zoom -xy 640 -vf rotate=1' May 15 09:23:10 as additional mplayer parameters May 15 09:23:25 Do I have to switch the device to landscape w/ rotate=1..? May 15 09:25:10 Ok, mplayer -sws 0 -nodouble -vo fbdev -fs -framedrop file.flv -vf rotate=1 is going.. May 15 09:25:17 Still the small video res May 15 09:25:53 * nab trys with 320x240 avi May 15 09:27:32 Hmmm, still the small decode (but properly in full screen :-) May 15 09:27:45 I suppose that is some additional parameter for -vo fbdev May 15 09:27:58 Time to get sdl installed too May 15 09:29:05 mplayer playback with audio for the iSCSI/WiFI .avi decode sounds really good May 15 09:29:11 Espically through headphones May 15 09:29:12 :-) May 15 09:33:58 * nab lets the .avi decode play before trying mplayer audio playback over iSCSI/GPRS May 15 09:44:50 http://linux-iscsi.org/index.php/Image:ISCSI-WiFI-FBdev-Landscape.png May 15 09:46:45 * nab starts charging the Freerunner sample on a PS3 May 15 09:50:30 what is the point with iSCSI/GPRS? May 15 09:50:47 kario: The LIO-devs want to be able to go snowboarding and listen to audio books May 15 09:50:50 On the mountain May 15 09:51:22 And not have to wait until 2010 for Mobile-Wmiax May 15 09:51:23 :-) May 15 09:55:50 Ok, 30 minutes into .avi player is stable over iSCSI/WiFI May 15 09:56:38 point taken May 15 09:57:02 or skiing if that is your thing. May 15 09:57:03 :-) May 15 10:23:03 hm just black screen and some icons bt and wlan o the top... light is on? sems i installed some WM too much... but r moving e-wm didt work. May 15 10:25:00 * nab waves May 15 10:25:16 abraxa__: You about..? May 15 10:25:21 nab: I got disconnected earlier, unfortunately May 15 10:25:27 Did you get my comments? May 15 10:25:27 * nab greetings May 15 10:25:33 I did not May 15 10:25:39 Oh May 15 10:25:46 nab: Got your message, I'll reply in a little May 15 10:25:46 If you want to read ahead, check out http://gstreamer.freedesktop.org/data/doc/gstreamer/head/gst-plugins-base-plugins/html/gst-plugins-base-plugins-playbin.html under "Buffering" May 15 10:25:46 Currently I play asap and disregard the buffer messages, so that would have to change May 15 10:25:51 Did you reply to the list..? May 15 10:26:05 * nab reads May 15 10:26:08 Excellent. :-) May 15 10:27:25 Actually that's pretty much all of the response, I suppose May 15 10:27:56 Do you want to have a try implementing the change or would you prefer if I do it? May 15 10:28:05 !!! May 15 10:28:12 I was just moving some more iSCSI/WiFI packets with mplayer May 15 10:28:14 what, you mean.... supporting code! May 15 10:28:18 abraxa__: That would be great! May 15 10:28:47 * nab can re-enable his iSCSI/GPRS to test any new changes May 15 10:28:48 :-) May 15 10:28:55 nab: Alrighty, I'll have a look in a bit then May 15 10:29:07 abraxa__: Great, thank you!! May 15 10:30:21 ScaredyCat: It's not about "supporting code", it's asking whether he wants to come up with a patch himself or not - of course I'll help if requested May 15 10:30:47 [2008-05-15 11:27:50] Do you want to have a try implementing the change or would you prefer if I do it? May 15 10:31:50 abraxa__: Btw, I was missing the pixmaps for the media-player buttons after grabbing via opkg install May 15 10:31:59 What package am I missing...? May 15 10:32:12 ScaredyCat: Maybe you interpretate it differently than I am? *scratches his head* May 15 10:32:23 (It looks like it is missing a dep :-) May 15 10:32:48 (I did this from a fresh ASU install on Freerunner hardware May 15 10:33:07 that suggests 'he' could write the code or you would do it form 'him' May 15 10:33:07 nab: It should be in the openmoko theme package, however the theme has been changed by thos rather recently, so maybe he omitted the media player pixmaps as they don't match that style May 15 10:33:17 s/form/for/ May 15 10:33:18 ScaredyCat meant: that suggests 'he' could write the code or you would do it for 'him' May 15 10:33:22 root@om-gta02:~# cat /etc/issue | tail -n 2 May 15 10:33:22 Openmoko P1-Snapshot-20080418 \n \l May 15 10:33:37 And opkg install openmoko-mediaplayer2 is what I used IIRC May 15 10:34:13 abraxa__: May 15 10:34:20 ScaredyCat: Hmm, honestly I fail to see what you're trying to say :) May 15 10:34:28 * nab remembers the button layout May 15 10:34:29 :-) May 15 10:34:35 hehe May 15 10:34:44 That reminds me May 15 10:34:54 I have an .nfo and .jpg in my audio book directory May 15 10:35:11 abraxa__: nm - it's not life threatening May 15 10:35:22 And after queueing up the directory to the media player, I will get gstreamer errors when the playback hits that non audio file May 15 10:35:41 Could you add some logic to keep going by default for tracklist playback..? May 15 10:35:45 ScaredyCat: I just wanna make sure you don't get the idea that I wouldn't support my own software ;) May 15 10:35:49 (When you get a bit :-) May 15 10:36:20 * nab will probably just shutup and start spending patches May 15 10:36:24 :P May 15 10:36:47 Or mabye, on the discan, o-mit the non media files May 15 10:36:56 s/discan/dirscan May 15 10:37:17 It would be cool to start up quiver to display the image files for the covers May 15 10:37:38 nab: Well, that's a feature I always wanted to add (i.e. file-type detection based on available gst plugins) but with the downfall of the GTK application suite that's rather unlikely to happen, I'm afraid May 15 10:37:39 When a image file was detected in the tracklist May 15 10:37:40 :-) May 15 10:37:50 abraxa__: Ahh, gotcha May 15 10:38:27 abraxa__: What have you been considering for your new work..? May 15 10:38:49 My goal is to re-write the media player using python after my PIM daemon fulfills the immediate needs May 15 10:39:01 Very interesting May 15 10:39:29 There where some LIO-devs working on python-gtk stuff on Maemo for iSCSI-Client-UI May 15 10:39:42 One of the motivations is that people can then edit the application on the device with a text editor, not needing the SDK and an external PC May 15 10:39:47 They have been waiting to see how OM stuff is working.. May 15 10:39:55 (very good :-) May 15 10:40:05 abraxa__: Makes sense May 15 10:40:13 I thought so, too :) May 15 10:40:29 And what does LIO stand for? May 15 10:40:30 hi. how can I enforce that the virtual keyboard will be shown? (eg. starting a program or such) May 15 10:40:41 LIO == Linux-iSCSI.org May 15 10:40:46 Ahh, thanks May 15 10:41:06 LIO-Target is the upstream project that the OpenMOko/iSCSI stuff (as well as mission critical stuff) connects to May 15 10:41:47 Please let me know if you have an issues getting an OpenMoko/iSCSI environment setup with LIO-VM May 15 10:42:09 LIO-VM == Running an Linux/iSCSI Target in an x86 virtual machine on Linux/Mac/Win May 15 10:42:28 You can basically just plug in a USB drive and point your OpenMoko/iSCSI client at the target IP May 15 10:43:05 Or any Type == Direct-Access (DISK) really May 15 10:43:25 With ext3 mounts, the Freerunner is looking very stable May 15 10:43:30 both r/o and r/w May 15 10:44:08 Your first iSCSI LUN will automatically be mounted by the OM device on /media/sda May 15 10:44:13 Interesting.... May 15 10:44:28 And accessable via the playlist tab with media-player, etc. May 15 10:44:34 Does it offer automounting/autoplugging stuff? May 15 10:44:41 abraxa__: Which part..? May 15 10:44:48 iSCSI May 15 10:45:14 For LIO-VM, you have to grab focus on the VMware machine, and re-run LIO-demo.sh May 15 10:45:18 But I am working on that May 15 10:45:19 Well, never mind May 15 10:45:28 I wasn't thinking enough :) May 15 10:45:32 On the client side, core-iscsi is still broken wrt to that May 15 10:45:39 Ahh okay May 15 10:45:42 With Open/iSCSI however, you can rescan an existing iSCSI TCP connection May 15 10:45:52 And detect new iSCSI LUNs and filesystems etc. May 15 10:46:05 Something I imagine will end up as a AUX function May 15 10:46:25 Along with the iSCSI/iSNS config for your iSCSI enabled mobile setup. May 15 10:46:30 I was more thinking along the lines of "I come home and Openmoko sees my WiFi-AP at home, so it can automatically connect to the media share on the NAS" May 15 10:46:41 abraxa__: Ahh yes.. May 15 10:46:42 Well May 15 10:47:01 Traditional iSCSI does define the concept on multiple communication paths per iSCSI storage nexus May 15 10:47:14 nexus == mapping between Initiator and Target storage object mappings May 15 10:47:37 You can start/stop these on the fly across multiple interfaces with core-isci May 15 10:47:40 core-iscsi May 15 10:47:43 and lio-target May 15 10:48:05 The other option would be using iSCSI/SCTP and multi-homing May 15 10:48:18 I hope the overhead isn't as high as it sounds ;) May 15 10:48:42 To prevent any service interruption to the media-player end-user May 15 10:48:51 But with GPRS that will be tricky May 15 10:48:53 :-) May 15 10:49:01 yeah gprs is slooow May 15 10:49:05 in oz anyway May 15 10:49:06 But not impossible May 15 10:49:20 Anyway, bbl May 15 10:50:00 abraxa__: Actually, the login/logout overhead is very small May 15 10:50:10 As it the other advanced features of iSCSI. May 15 10:50:13 s/it/is May 15 10:50:49 (Stuff not supported just quite yet on the client side) May 15 10:50:57 But we are getting close. :-) May 15 11:00:19 root@om-gta02:/media/sda# opkg install libsdl May 15 11:00:19 Nothing to be done May 15 11:00:19 An error ocurred, return value: 4. May 15 11:00:19 Collected errors: May 15 11:00:19 Cannot find package libsdl. May 15 11:00:20 Hmmm May 15 11:02:19 do a : opkg update May 15 11:02:27 then try the install May 15 11:02:48 Doh May 15 11:03:55 nab: why do you need libsdl? May 15 11:04:07 you wanna play wesnoth? ;) May 15 11:04:29 josch|nsn: Will SDL give me fullscreen..? May 15 11:04:40 hm where do i get opkg .. i have ipkg here, at my a780 May 15 11:04:43 Or mabye my fbdev parameters are just wrong May 15 11:05:00 mplayer -sws 0 -nodouble -vo fbdev -fs -framedrop wakinglife.avi -xy 640 -vf rotate=1 May 15 11:05:45 An error ocurred, return value: 4. May 15 11:05:45 Collected errors: May 15 11:05:45 Failed to download http://buildhost.openmoko.org/daily-feed/all/Packages.sig, error 404 May 15 11:05:45 Failed to download http://buildhost.openmoko.org/daily-feed/armv4t/Packages.sig, error 404 May 15 11:05:45 Failed to download http://buildhost.openmoko.org/daily-feed/neo1973/Packages.sig, error 404 May 15 11:05:45 Failed to download http://buildhost.openmoko.org/daily-feed/om-gta02/Packages.sig, error 404 May 15 11:05:47 Hrrmmm May 15 11:05:58 nab: i told you to read http://wiki.openmoko.org/wiki/Video_Player May 15 11:06:29 nab: you are missing -vf scale -zoom May 15 11:06:32 * nab looks again for QVGA May 15 11:06:49 640 is NOT qvga May 15 11:06:50 For switching between QVGA and VGA mode do the following with a compatible kernel: May 15 11:06:57 What does the last bit entail..? May 15 11:07:07 nab: sdl fullscreen works May 15 11:07:13 do you want to use qvga or resize your video? May 15 11:07:29 nab: if you have installed a devel image run kobodeluxe (works in either orientation) May 15 11:07:29 rschuster: as mentioned in the wiki article... May 15 11:08:13 Ok, I am missing /sys/devices/platform/s3c24xx-spi-gpio.1/spi0.0/state May 15 11:08:37 josch|nsn: Are the QVGA stuff changes to fbdev, or a seperate mod..? May 15 11:08:46 pff... May 15 11:08:46 rschuster: Thanks May 15 11:08:55 nab: its all in the wiki article May 15 11:09:01 if it's not, tell me May 15 11:09:02 josch|nsn: Ok, thanks May 15 11:09:03 :-) May 15 11:09:12 nab: Try using xrandr to change resolution, the sysfs stuff in the video player article is outdated May 15 11:09:14 nab: i said so hours ago... May 15 11:09:29 abraxa__: it is? why didnt you update it? May 15 11:09:46 I didn't have time :( May 15 11:09:57 abraxa__: so xrandr finally works for resolution switching? May 15 11:10:01 josch|nsn: Sorry, my video hardware knowledge is some-what limited May 15 11:10:02 :-) May 15 11:10:17 josch|nsn: I'll try as soon as my Neo has booted up May 15 11:10:23 hehe May 15 11:10:39 yay for xrandr modechanges May 15 11:11:14 Where can I find the libsdl*.ipk stuff directly..? May 15 11:11:30 * nab looks at opkg repo info May 15 11:12:13 nab: buildhost.automated.it is a nice repo (ScaredyCat maintains it) May 15 11:12:30 Hmm May 15 11:12:38 I can't seem to find the libsdl packages on http://buildhost.openmoko.org/daily-feed/ May 15 11:12:43 abraxa__: OK, thanks May 15 11:13:42 josch|nsn: xrandr doesn't work for me, I'll see what I can find May 15 11:14:23 Maybe it does work for nab, however, as he has a GTA02 May 15 11:14:29 Ok, I am grabbing libsdl-image and libsdl-mixer May 15 11:14:52 And libsdl-1.2 May 15 11:15:23 nab: Mind doing a "DISPLAY=:0 xrandr -s 320x240" to see whether that works for you? May 15 11:16:09 Hmm, I think the UI just crashed w/ pcf50633_battvolt timeout :-( May 15 11:16:19 That is the actualy console message May 15 11:16:19 abraxa__: nab has a gta02???? May 15 11:16:20 :-) May 15 11:16:34 The mplayer video is still doing however May 15 11:16:35 nab: After doing the xrandr or when? May 15 11:16:37 s/doing/going May 15 11:16:41 abraxa__: Nope May 15 11:16:45 oh May 15 11:16:46 I guess I will have to reboot May 15 11:16:47 abraxa__: you mean xrandr -s 240x320 May 15 11:16:51 one sec May 15 11:16:52 :-) May 15 11:17:10 nab: And the pcf50633 is just a kernel debug message May 15 11:17:19 abraxa__: The screen dim timeout just kicked off May 15 11:17:40 josch|nsn: Yeah, oops May 15 11:17:55 nab: why doyou have a freerunner?? May 15 11:17:57 And I can't bring it back from dim.. (with the video player stuff running :-) May 15 11:18:14 Time to take a note and reboot May 15 11:18:15 :-) May 15 11:20:40 josch|nsn: I am providing LIO packages for the device May 15 11:21:06 So you can access 900 GB filesystem mounts over the network May 15 11:21:15 And other fun stuff like that May 15 11:21:37 Hmmm May 15 11:21:45 * nab notes the screen is still dim May 15 11:21:54 Mabye time for a "cold-reboot" May 15 11:22:24 There we go May 15 11:22:29 nab: LIO? i used to do that kind of stuff via ssh May 15 11:22:43 josch|nsn: What stuff? May 15 11:22:56 LIO == Linux-iSCSI.org May 15 11:22:56 why do i have black, lit screen, and just a non-working panel after installing like e-wm and matchbox-panel? May 15 11:22:58 mounting distant file systems May 15 11:23:04 Ah yes, sshfs May 15 11:23:06 Wel May 15 11:23:16 with fusefs May 15 11:23:17 No one really runs that in production (from what I know :-) May 15 11:23:24 But still pretty cool May 15 11:23:27 sshfs and httpfs May 15 11:23:36 LIO url plz May 15 11:23:43 Linux-iSCSI.org May 15 11:25:20 nab: encryption? May 15 11:26:41 josch|nsn: For the iSCSI traffic..? May 15 11:27:13 The IPS TWG defined ipsec AH/ESP for securing bulk traffic May 15 11:27:59 And IKE for the key exchange protocol May 15 11:28:22 Ok, the device is back up May 15 11:28:32 * nab trys xrandr -s 240x320 May 15 11:28:46 root@om-gta02:/media/sda# xrandr -s 240x320 May 15 11:28:46 Can't open display May 15 11:28:55 Do I have to not be in X..? May 15 11:29:03 DISPLAY=:0 xrandr ... May 15 11:29:04 :) May 15 11:29:06 er duh May 15 11:29:11 (Its getting late over here :-) May 15 11:29:14 No worries May 15 11:29:31 * nab retires May 15 11:29:37 er, retries even May 15 11:30:50 nab: does your solution use native encryption May 15 11:32:12 VO: [fbdev] 208x352 => 190x320 BGR 16-bit [fs] May 15 11:32:12 screensize is smaller than video size May 15 11:32:12 FATAL: Cannot initialize video driver. May 15 11:32:19 Hmmm May 15 11:32:36 josch|nsn: IPSEC is done at the network layer, so yes. May 15 11:32:58 But the to really make it work you need the MobilIKE stuff, yes. May 15 11:33:19 * josch|nsn finds the sshfs version more convenient May 15 11:33:32 * josch|nsn does everything through ssh - even tunneling May 15 11:33:51 (real tunneling with tun0 device ) May 15 11:33:55 Ok, fullscreen .flv playback is doing May 15 11:33:56 :-) May 15 11:34:01 nab: congrats! May 15 11:34:02 GOING May 15 11:34:11 me kills pulseaudio again May 15 11:34:19 sshfs fails miserably over time May 15 11:34:47 with large amounts of data it just stops working May 15 11:35:24 ScaredyCat: it worked fine for me so far and i copied whole dvd images May 15 11:35:41 Playback looks pretty decent over iSCSI/WiFI May 15 11:35:48 well, that's not what I'd call 'large amounts of data' May 15 11:35:59 nab: you rly can stream your file? May 15 11:36:00 4-10gb of dvd ... May 15 11:36:03 great! May 15 11:36:04 josch|nsn: I know a bit about that type of stuff.. May 15 11:36:17 I'm talking about 100's of GB May 15 11:36:27 ScaredyCat: in one file? May 15 11:36:31 * nab presented iSCSI/HD from PS3-Linux @ LWCE last year May 15 11:36:42 when it works, it's grat, when it stops working it's not... May 15 11:36:52 Doing playback from BD/HD-DVD shit, etc. May 15 11:36:54 no, not in one file, multiple May 15 11:37:44 * nab wonders how to get his .avi to playback using xrandr May 15 11:38:25 google:/home/nab/haakon-ppc/nab# file more.avi May 15 11:38:26 more.avi: RIFF (little-endian) data, AVI, 352 x 208, ~15 fps, video: DivX 4, audio: MPEG-1 Layer 3 (stereo, 48000 Hz) May 15 11:39:00 And I was using xrandr -s 240x320 May 15 11:39:14 * nab notes that 'file' does not extract the resolution for .flv files May 15 11:39:31 ScaredyCat: well this workes also - bavkuped 800GB of data through sshfs once May 15 11:39:40 ScaredyCat: is the issue bugreported? May 15 11:40:06 iirk it's a know issue, as yet unfixed May 15 11:40:09 iirc May 15 11:40:17 known May 15 11:41:29 * josch|nsn never used iscsi - what are the alternatives? does it integrate as a mountable device? May 15 11:43:37 josch|nsn: SAN storage transports encapsulate the I/O request below the host OSes block layer of the storage stack May 15 11:43:42 ataoe May 15 11:44:00 Which makes the SAN struct block_device (eg: /dev/sda) appear at a locally connected evice. May 15 11:44:50 fdisk, mkfs and mount work just like any other storage device May 15 11:47:08 ah nice lowlevel then May 15 11:47:38 and i assume i can outomount it in through fstab? May 15 11:48:03 josch|nsn: There are a couple of different methods to do that yes.. May 15 11:48:24 The OM stuff will automount /dev/sda in /media/sda May 15 11:48:31 and so on. May 15 11:48:48 Actually, I don't know what it does for /dev/sda1 May 15 11:48:49 /dev/sda being the remote drive May 15 11:48:53 Correct May 15 11:49:04 For example, I am using FILEIO from a PS3-Linux LIO-Target May 15 11:49:13 wow rly nice - dont know why i didnt recognize the stuf yet May 15 11:49:25 FILEIO == A file on a mounted filesystem using buffered level kernel I/O May 15 11:49:58 The LIO-Target Storage Engine can also send raw blocks to a struct block_device (BLOCK layer) or struct scsi_device (SCSI layer) May 15 11:50:16 The former includes logical volumes, MD RAID arrays, etc. May 15 11:50:21 You get the idea. :-) May 15 11:50:42 yeah May 15 11:50:50 have to try that @ home May 15 11:51:05 is it in some repos like debian? May 15 11:51:24 josch|nsn: May 15 11:51:39 nab: great will report back on success May 15 11:51:42 http://linux-iscsi.org/[debian,ubuntu] May 15 11:52:06 josch|nsn: Great, I am going to have to head-out soon, but post any questions to the LIO-Target devel list May 15 11:52:16 There is a link to the google group on the main page May 15 11:52:25 lol google May 15 11:52:34 :-) May 15 11:58:20 nab: damn only i386 and amd64 packages - have to build myself then May 15 11:59:21 josch|nsn: Use m-a -vt a-i target to build it for your running kernel May 15 11:59:28 it == iscsi_target_mod.ko May 15 11:59:34 ER Sorry May 15 11:59:58 The OM client stuff is in http://linux-iscsi.org/builds/openmoko/freerunner/core-iscsi/ May 15 12:00:11 The target stuff is in the debian/ubuntu repos May 15 12:00:32 http://linux-iscsi.org/svn/trunk/README May 15 12:00:41 On howto build the target from source May 15 12:01:06 nice docu thx May 15 12:08:26 josch|nsn: The stacks have been run on pretty much every arch under the sun May 15 12:08:55 But if you got some new builds, let the list know. :-) May 15 12:09:03 nab: what about my old sparc? May 15 12:09:16 * nab has an Ultra 10 sitting close by May 15 12:09:34 nab: would be nice to have also binaries for other arcs May 15 12:09:43 josch|nsn: May 15 12:09:47 It will build on all of them May 15 12:09:48 i assume you autobuild them so adding more wouldnt be more work May 15 12:09:55 I use a G4 Cube running Debian/Lenny for my workstation May 15 12:10:04 nice May 15 12:10:20 josch|nsn: I do the builds in some VMs running iSCSI/OCFS2 May 15 12:10:30 i love the fact that this ultra portability is only possible with oss May 15 12:10:34 eg: The repo is self hosted on top of iSCSI May 15 12:10:46 josch|nsn: :-) May 15 12:11:01 * nab has an a few Alpha nearby too May 15 12:11:14 nab: but you only support linux as a kernel? May 15 12:11:27 josch|nsn: Well, the LIO-Target design is OS independent May 15 12:11:40 The plugins talk with the local host storage subsystems. May 15 12:11:50 But you can easily run it inside of a VM on non Linux systems May 15 12:11:52 And export stuff May 15 12:11:57 kk May 15 12:12:10 The design has been ported in the past to non Linux OSes however. May 15 12:12:24 But for LIO-Target its all Host local + VM linux May 15 12:12:25 :-) May 15 12:12:51 nab: who are the main devs? May 15 12:13:01 josch|nsn: Well, mostly myself May 15 12:13:10 wow, then kudos to you May 15 12:13:15 this all sounds great May 15 12:13:21 But the project initially was a funded start-up etc. May 15 12:13:22 http://www.geograph.org.uk/photo/450784 May 15 12:13:24 oops May 15 12:13:35 Then last near I bought the code and turned it into an OSS thing May 15 12:14:06 you "bought" it?? May 15 12:14:10 May 15 12:14:23 Well, I was the co-founder and chief architect for it May 15 12:14:24 :-) May 15 12:14:30 ah kk... May 15 12:14:47 And people got tired of listening to me about binary kernel modules May 15 12:14:48 Soooo May 15 12:14:49 :-) May 15 12:14:54 hahaha May 15 12:14:56 great May 15 12:15:05 thanks for that May 15 12:15:11 :- May 15 12:15:12 :-) May 15 12:15:24 The LIO-Target stuff is destined for upstream kernel.org before too long May 15 12:15:27 are there alternatives for your work? May 15 12:15:38 Sure May 15 12:15:51 But the LIO-Target project has been around the longest on the server side May 15 12:15:56 And thats all I really know, soo. May 15 12:16:11 On the client, there is support in kernel.org for the client side with open/iSCSI May 15 12:16:22 And then the other x86 OSes have their iSCSI clients May 15 12:16:26 Some RISC UNIX platforms too May 15 12:16:36 And of course iSCSI HBAs and whatnot. May 15 12:16:55 The LIO stuff supports a bunch of extra stuff not usually implemented by vendors. May 15 12:17:16 Like active-active connection recovery so you can multi-plex across multiple backbone providers coast-to-coast May 15 12:17:22 so what is currently in the kernel and what is not? May 15 12:17:44 And as you loose links, the nexus between the storage objects never stops, etc. May 15 12:18:04 josch|nsn: The Open/iSCSI Initiator client is upstream May 15 12:18:15 And the Target engine folks are still duking it out May 15 12:18:25 eye is in your kernelz stealinz yor bytses! May 15 12:18:25 * nab presented at LSF 08' for the project this year May 15 12:18:48 http://iou.parisc-linux.org/lsf2008/IO-iSCSI_and_Target_Mode-Nicholas_Bellinger.pdf May 15 12:18:52 (If anyone gets bored :-) May 15 12:19:09 OpenMoko did have a mention in the discussion fo course May 15 12:19:25 Under "Sexy Embedded Linux Hardware" I believe May 15 12:19:27 LSF = Linux Storage & Filesystem Workshop May 15 12:19:35 josch|nsn: May 15 12:19:40 where? May 15 12:19:47 San Jose May 15 12:19:50 , CA May 15 12:19:53 oups May 15 12:20:01 * josch|nsn = europe May 15 12:20:25 ewwwop May 15 12:20:31 raster: provide the lolcat image and you are done for today May 15 12:20:37 hahaha May 15 12:20:51 josch|nsn: It's a USENIX event May 15 12:21:04 It was alot of fun this year. :P May 15 12:21:10 Lots of talk about SSD drives May 15 12:21:21 Lots of storage industry folks, and the usual suspects May 15 12:21:24 :-) May 15 12:23:13 a meeting of all those guys that think they are more intelligent with their strage drivers than the drives internal logic :P May 15 12:23:44 josch|nsn: Heh May 15 12:23:47 Something like that May 15 12:24:11 ^^ May 15 12:24:13 And very different personality and not everyone speaks the same languages May 15 12:24:18 personalities May 15 12:24:27 And lots of free-drinks May 15 12:24:29 :P May 15 12:24:37 oh free is good May 15 12:24:47 Espically once LSF ends and the academics come to town May 15 12:24:48 even if it's free beer instead of speach May 15 12:24:48 :P May 15 12:25:05 speech even May 15 12:25:07 :-) May 15 12:25:46 * nab is looking forward to driving some iSCSI mobile devices next year May 15 12:26:20 n800/810, freerunner... May 15 12:26:25 josch|nsn: I think doing the shared storage FS stuff is the most interesting so far for the LIO-Work May 15 12:26:28 May 15 12:26:35 pandora May 15 12:26:59 I believe that > v2.6.22 has the Distributed Lock manager (DLM) v8 wire protocol bits for doing traffic across the devices May 15 12:27:15 * nab will be taking a look at getting builds for this soon May 15 12:27:38 Since the ocfs2+dlm kernel modules can be easily enabled in v2.6.24. :-) May 15 12:28:32 http://linux-iscsi.org/index.php/OCFS2/iSCSI_for_OS-2008 for the project page May 15 12:29:07 So to get the Freerunning and OS-2008 devices up, they both need > 2.6.22 kernel May 15 12:29:25 I know there are folks on linux-omap working on the latter device May 15 12:29:35 s/Freerunning/Freerunner May 15 12:29:37 :-) May 15 12:29:59 Having a few x86 LIO-VMs and those devices on a shared storage mount would be pretty slick May 15 12:30:22 (There are still obvious changes needed for R/O OCFS2 access of course) May 15 12:30:43 The internal logic will still fence nodes for R/O filesystem mounts when cluster nodes go away May 15 12:30:56 internal DLM logic May 15 12:31:18 But that all currently depends on timeout settings and wireless connectivity May 15 12:31:44 But making OCFS2/DLM more WiFI friendly is a long-term project May 15 12:32:02 And will require some kernel patches May 15 12:32:04 ;-) May 15 12:32:38 But I plan to use R/O and R/W ext3 for now, and see how the OCFS2/WiFi bits come along May 15 12:32:40 i would rather like to have a distributed and redundant filesystem like ceph, glusterfs or gfarm. I don't know, somehow i don't like shared storage on a block layer. May 15 12:33:00 DPThuught: Well, its a matter of making that logic lightweight I think. May 15 12:33:25 And portable to a number of different environments (at least in the storage case) May 15 12:33:51 DPThuught: Btw, OCFS has done through to different major versions May 15 12:33:55 intermezzo sounds like a really good idea for a nomadic user btw. Sadly it has really poor online performance and offline work depends on local disk space which is usually sparse May 15 12:34:03 The first used block READs to access metadata stored on the disk May 15 12:34:21 Which the OCFS devs found did not scale well to larger numbers of cluster nodes May 15 12:34:27 * ScaredyCat awns May 15 12:34:28 nab: yea, thats a really good point. all cluster filesystems i checked out so far are not very trustworthy May 15 12:34:29 then May 15 12:34:32 * ScaredyCat yawns May 15 12:34:46 So along came OCFS2 and put the metadata onto the network fabric May 15 12:35:05 (Which is what most the major cluster fses do) May 15 12:35:09 But anyways May 15 12:35:10 :-) May 15 12:35:23 i think i have to read more about OCFS2... :) May 15 12:35:25 DPThuught: OCFS2 is great because the config is so simple May 15 12:35:57 And has been included upstream for a while, and can run in BE/LE 32/64 mode, etc May 15 12:36:14 And inside of VMs, etc May 15 12:36:20 But you get the idea May 15 12:36:31 I just like the idea of multiple machins all having free disk space coming toghether and forming a filesystem/namespace cloud from the collected free space. Of course with redundancy. May 15 12:36:34 There are some wikipedia links for cluster/shared storage on the OCFS2/iSCSI OS 2008 page May 15 12:36:37 Check them out May 15 12:36:49 DPThuught: May 15 12:36:57 But in reality, you are doing to have multiple pieces of storage May 15 12:37:02 Some small FS mounts for R/W access May 15 12:37:24 And on the backend you can have VMs moving files around to your shared storage filesystems May 15 12:37:47 So much of it comes down to network bandwith and connectivity May 15 12:37:54 Esp wrt R/W access May 15 12:38:11 For now we assume the OCFS2 mount on the mobile devices will be read-only May 15 12:38:24 But other VMs with R/W can put files onto the mount May 15 12:38:34 And the mobile device will see the new dirs/files, etc May 15 12:38:51 And that userspace will also see the updates May 15 12:39:08 (Which is not the case with the Maemo devices and OCFS2) May 15 12:39:20 At least not with the default file manager. May 15 12:39:22 :-) May 15 12:39:24 But you get the idea. May 15 12:56:36 ScaredyCat: Pong? May 15 12:56:49 oh.. blimey! May 15 12:56:56 that was hours ago :) May 15 12:56:59 * rwhitby adds an OM_BRANCH variable to MokoMakefile to enable building the org.openmoko.april-update branch ... May 15 12:57:39 Cap_J_L_Picard: I jst came across the libssp error again. May 15 12:58:00 I don;t think anyone's really don a clean build for a long time May 15 12:58:20 hmm... May 15 12:58:23 yeah... May 15 12:58:48 * XorA did one 2 days ago May 15 12:58:48 I've added what I did to fix it to that bug May 15 12:58:57 * ScaredyCat slaps XorA May 15 12:58:59 * Cap_J_L_Picard tries to remember how he got pas it last time... May 15 12:59:02 really? May 15 12:59:03 * rwhitby wonders how to reproduce http://quickstart.openmoko.com/recommended/ builds ... May 15 12:59:14 s/pas/past/ May 15 12:59:16 so how did you get past the libssp errors then XorA? May 15 12:59:25 there were no libssp errors May 15 12:59:48 well, you didn't build form sane src revs and the current .bb files then May 15 13:00:00 or he had a different host setup May 15 13:00:00 because they have errors in that cause the libssp issue May 15 13:00:17 ScaredyCat: OM people use moko-autorev May 15 13:00:25 * XorA didnt use autorev May 15 13:00:51 the libssp errors cause avahi to fail, which causes gnome-vfs, pulseaudio etc to fail. May 15 13:02:29 'libssp errors' == avahi failing to link because libssp not built (disabled in gcc EXTRA_OECONF) May 15 13:03:52 avahi doesnt require ssp anyway May 15 13:04:37 !ombug 1311 May 15 13:04:38 * * Bug 1311, Status: NEW, Created: Unknown May 15 13:04:39 * * scaredycat(AT)automated.it: libssp.so.0 is required for oh-puzzle but is not built... May 15 13:04:40 * * http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=1311 May 15 13:10:44 rwhitby: I thought the whole point of oe was that it was distro agnostic, hence all the native stuff May 15 13:12:39 ScaredyCat: indeed May 15 14:40:53 re May 15 15:23:24 * ScaredyCat sighs May 15 15:24:04 * SpeedEvil passes ScaredyCat an angry crocodile, to take his mind off his troubles. May 15 15:25:40 * XorA laughs when he sees the new bootsplash May 15 15:25:42 * ScaredyCat fears even an angry crocodile could not achieve that. May 15 15:29:32 * CM is half way to a booting pc :) May 15 15:30:26 finally got all the parts then CM? May 15 15:31:40 Yepp, botoed with a livecd and about to install now May 15 15:32:05 Using my old disk, so basically just have to change some configs and build a new kernel :) May 15 15:32:14 cool May 15 15:32:21 XorA: pic May 15 15:33:01 PIC18F628A May 15 15:33:56 * ScaredyCat thows a pic axe at SpeedEvil May 15 15:34:27 so this freerunner phone, will it be USABLE? May 15 15:34:30 ScaredyCat: exquisite May 15 15:34:33 That is, can I actually place phone calls on it. May 15 15:34:36 If so, I'm all over it. May 15 15:35:09 wasabi: basically. On release it's at best a basic phone and SMS thingy May 15 15:35:12 * ScaredyCat rubs wasabi all over XorA May 15 15:35:27 wasabi: there will be a number of problems that mean it's not a general user-ready phone. May 15 15:35:44 wasabi: power saving May 15 15:35:46 I see. May 15 15:35:51 So why charge $400 for it? May 15 15:35:54 wasabi: general state of software readyness, and lack of core apps such as light-seber noise. May 15 15:36:15 becuase $1 is worth nothing atm May 15 15:36:19 wasabi: because that's broadly what it costs to make, and make a small profit. May 15 15:36:20 hehe May 15 15:36:29 $400 is worht just about something May 15 15:36:32 Oh well. See... I'd love to get one, and hack on it. May 15 15:36:41 But I'm not going to spend $400 if it can't do the basic stuff. May 15 15:36:42 we all would May 15 15:37:07 I say 'we' .... May 15 15:37:17 wasabi: At the moment, it's a PII/200 class machine, with a 16 bit ISA graphics card, and stuff, all packed into a little case. May 15 15:37:48 wasabi: As a small linux device that is open-source, and is dev friendly, there is little competition. May 15 15:37:51 What's wrong with the power saving? May 15 15:38:01 wasabi: it's not really implemented fully May 15 15:38:55 so... when will that be solved? May 15 15:39:50 wasabi: When will birds suddenly appear? May 15 15:40:08 when you are near? May 15 15:43:03 why is everyone speaking in riddles? :D May 15 15:43:18 :D May 15 15:43:22 isnt that Carpenters lyrics? May 15 15:43:36 * mwester doesn't listen to the Carpenters... May 15 15:43:37 nearly May 15 15:43:48 wasabi: basically - when someone codes it May 15 15:43:54 oh well. i really want to buy and hack on one... but if it's going to sit at home all day and i'm not going to be able to actually place calls and stuff on it, it's just not going to happen. sort of the same position my maemo device is in. May 15 15:44:01 awesome device... but i have no reason to carry it around. May 15 15:44:14 and so it sits at home ignored May 15 15:44:22 Eh, i thought you can call and send sms with openmoko, when you open it from case May 15 15:44:23 It can place calls and stuff May 15 15:44:37 well, how long does it run without power saving? May 15 15:44:37 Power saving is an iterative process; there is *always* something that can be improved, a new algorithm to try, etc. Power saving basics will be working shortly (IMO); it will be refined for much time to come. May 15 15:44:43 However, it does not have many important apps that would make it actually usable. May 15 15:44:55 I mean fot the general user May 15 15:44:55 how so? dialer and sms is all I use my real phone for. May 15 15:44:59 SpeedEvil: what apps? May 15 15:45:05 wapss May 15 15:45:08 wapps May 15 15:45:10 A nice polished browser. May 15 15:45:25 SpeedEvil: I think that's all being done in secret May 15 15:45:30 A polished audio/video player May 15 15:45:32 whats going to happen when android finally has hardware? May 15 15:45:32 SpeedEvil: epithany-webkit :-) May 15 15:45:41 have you not noticed bugger all going on? May 15 15:45:44 we got a polished audio player May 15 15:46:00 either it's not goig on, or they're deving in the background May 15 15:46:04 XorA: sure - I mean a video player that knows the limits of the GTA02 hardware May 15 15:46:24 I suspect the latter from some of the comments on the commuity list May 15 15:46:46 org.openmoko.april-update is current stuff May 15 15:47:13 and gtk is going... or at least not being used May 15 15:47:18 I also suspect the latter; the lack of discussion on some of the important issues is leading me to think that they are already resolved in some other codebase. May 15 15:48:08 mwester: no, I think its that there is only one guy who can help you and he is massively busy May 15 15:48:10 or they just dont't give a fuck :) May 15 15:48:48 :( May 15 15:48:50 whatcha mean about gtk? May 15 15:49:44 over there --> gtk May 15 15:50:02 hmm maybe I could hack up a mcclim-etk backend May 15 15:50:13 they're not using gtk anymore? May 15 15:50:14 there is already a mostly working gtk/cairo backend though May 15 15:50:17 there is no secret codebase inside OM May 15 15:50:21 EFL May 15 15:50:35 unless its so secret I dont know about it May 15 15:50:42 are there any actual people who develop in here? May 15 15:50:53 wasabi: yes May 15 15:51:40 wasabi: http://unknownlamer.org/img/misc/lj/clim-accepting-gadgets-working.png I'm hacking on that so I can write an openmoko-view and get purely gadget/touch based UIs working in CLIM May 15 15:51:59 * balrog-kun develops a neck ache May 15 15:52:31 * ScaredyCat oils down a ferret for balrog-kun May 15 15:53:18 NOTE: package openmoko-image-1.0: completed May 15 15:56:51 * ScaredyCat dies a little inside May 15 15:56:54 * XorA sees no libssp problem, and has no libssp in tmp/ anywhere May 15 15:57:12 hm.. libssp was what exactly? May 15 15:57:14 oh well. i guess i'll wait another year and see what happens then. android might have hardware at that point May 15 15:57:28 wasabi: check back in a month May 15 15:57:30 android will have h/w this year May 15 15:57:36 wasabi: something may have happened May 15 15:58:00 if you want the logs XorA, you're welcome May 15 15:59:33 ScaredyCat: find tmp/ -name "*libssp*" May 15 16:01:14 'ERROR: '/mnt/data/om/openembedded/packages/glibc/glibc_2.6.1.bb' failed May 15 16:01:19 http://pastebin.ca/1019129 May 15 16:01:19 thoughts? May 15 16:01:25 It's been a while since I built May 15 16:01:32 not from just that SpeedEvil... May 15 16:01:44 give us 5 lines back May 15 16:01:48 from there May 15 16:01:52 * mwester moves his tmp aside, and syncs up to see what happens to his build (it worked fine a few days ago). May 15 16:02:02 ERROR: function do_prep_locale_tree failed May 15 16:02:06 ScaredyCat: mtn head May 15 16:02:07 seems to be the original May 15 16:02:57 you're not using om XorA? May 15 16:03:11 ScaredyCat: I am using om May 15 16:03:17 After binary locale generation May 15 16:03:40 SpeedEvil: paste it in pastebin May 15 16:03:46 For anyone wondering whether gtk will be around, or what else will be around, look at http://wiki.openmoko.org/wiki/OpenmokoFramework May 15 16:04:18 yes, head May 15 16:04:18 esp. the diagram mid-page. May 15 16:04:43 ( XorA ) May 15 16:04:56 http://pastebin.com/m10a46988 May 15 16:04:58 om head... May 15 16:06:11 do you have ENABLE_BINARY_LOCALE_GENERATION = "0" May 15 16:06:16 in your loal.conf ? May 15 16:06:17 auto-rev or srcrev? Which is XorA using? (I'm using auto, and aside from the occasional expected breakage, it builds ok) May 15 16:06:19 local May 15 16:06:38 ScaredyCat: no May 15 16:06:47 * SpeedEvil searches for local.conv May 15 16:07:11 mwester: I blew away my tmp for a fresh build.. (there is an issue with autorev afaik) May 15 16:07:25 ./build/conf/local.conf ? May 15 16:07:29 SpeedEvil: ${OMDIR}/build/conf/local.conf May 15 16:07:33 Ok. Well I'll find out shortly. May 15 16:07:55 ScaredyCat: added, new build kicked May 15 16:08:41 mwester I switched to sane srcrev May 15 16:09:02 hmm... i was pinged overnight, and I have no idea by who, or in what channel May 15 16:09:02 can we assume glib will always exist in the standard images? for example if u write some dbus programs and use the glib-dbus bindings May 15 16:09:11 mwester: Im using no autorev and head of OM mtn May 15 16:09:13 to get newer versions May 15 16:09:52 isn't the default sane then XorA May 15 16:09:52 ? May 15 16:10:23 what default? the default is no autorev May 15 16:10:46 yes, so the versions will be got from sane srcrev May 15 16:10:47 zedstar_: D-Bus will be used extensively, as will glib May 15 16:11:16 ${OMDIR}/openembedded/conf/distro/include/sane-srcrevs.inc May 15 16:12:07 summatusmentis: Wasn't in this chan, as far as my logs show May 15 16:12:20 surely that's the defaults, since any other just overrides those... ?= means set if not set - right? May 15 16:12:24 abraxa_: mm... I scrolled back as far as I could, thanks May 15 16:12:31 !logs May 15 16:12:33 Channel logs for #openmoko are archived at: May 15 16:12:34 http://hentges.net/tmp/logs/irc/%23openmoko May 15 16:12:35 Live-logs are available at May 15 16:12:37 http://hentges.net/tmp/logs/irc/livelogs/%23openmoko.livelog May 15 16:12:39 See ?? help-logs for usage instructions May 15 16:12:54 abraxa_ ok good thanks May 15 16:12:59 summatusmentis: Actually, I was mislead - Sup3rkiddo pinged you - supposedly to say hello May 15 16:13:15 * abraxa_ gets confused and shuts up May 15 16:13:18 I was still there for that, though, I think May 15 16:13:40 * ScaredyCat prods XorA May 15 16:14:00 ScaredyCat: your mumbling incoherentlu May 15 16:14:07 summatusmentis: Yeah, I speak too fast. May 15 16:14:20 XorA: sorry? May 15 16:14:21 lol, no worries May 15 16:15:00 DISTRO="openmoko" May 15 16:15:00 INHERIT += "rm_work" May 15 16:15:00 GLIBC_GENERATE_LOCALES = "en_GB.UTF-8 en_US.UTF-8" May 15 16:15:07 bitbake openmoko-image May 15 16:15:16 thats my setup May 15 16:15:34 right, but the versions it will use to build come from ${OMDIR}/openembedded/conf/distro/include/sane-srcrevs.inc May 15 16:16:20 ScaredyCat: obviously May 15 16:16:45 that's what I said May 15 16:17:24 no autorev == sane-srcrevs May 15 16:18:41 anyway I give up for today May 15 16:19:09 * ScaredyCat slaps XorA|gone May 15 16:20:24 * CM hands ScaredyCat a slimy trout ;) May 15 16:20:37 ~trout XorA|gone May 15 16:20:38 * apt slaps XorA|gone around a bit with a large trout! May 15 16:20:46 Hehe May 15 16:20:48 * SpeedEvil passes ScaredyCat a book entitled 'Cooking Trout'. May 15 16:20:59 ~cook trout May 15 16:21:00 * apt throws trout in a big pan with veggies inside and cooks trout on 350 for an hour May 15 16:22:09 only another 1283 tasks to go May 15 16:31:37 an hour??! Not fish, for heavens sake! May 15 16:32:32 ~cook FreeRunner May 15 16:32:33 * apt throws FreeRunner in a big pan with veggies inside and cooks FreeRunner on 350 for an hour May 15 16:32:40 haha May 15 16:46:48 is http://gypsy.freedesktop.org/wiki/ broken May 15 16:51:29 idarwin: that picture looks wrong May 15 16:51:41 ScaredyCat: looks like it **** ENDING LOGGING AT Thu May 15 17:18:19 2008 **** BEGIN LOGGING AT Thu May 15 17:20:14 2008 May 15 17:35:38 Hi All! Where I can to make an order for GTA02? May 15 17:35:49 * SpeedEvil passes green a TARDIS. May 15 17:37:19 green: GTA02 isn't on sale yet, I'm afraid May 15 17:37:37 SpeedEvil: that would be 'when' surely May 15 17:37:39 SpeedEvil, heh, thanks May 15 17:39:12 I see in topic: "should start mass production before May 16", or this is a joke? May 15 17:39:26 green: "should" May 15 17:39:28 green: according to previous information no. May 15 17:39:34 no, based on real info May 15 17:39:38 However, we have had no information aas to the progress of this. May 15 17:39:52 did your build wokr SpeedEvil? May 15 17:40:00 And it's probably going to be at least amonth before successful production leads to shipment. May 15 17:40:06 NOTE: Running task 877 of 5854 (ID: 1700, /mnt/data/om/openembedded/packages/gtk+/gtk+-fastscaling_2.10.14.bb, do_populate_staging) May 15 17:40:16 Further than it was. May 15 17:40:22 oh, so only another 7 hour then May 15 17:42:10 sadly(( May 15 17:48:42 'packagekit' build is failing for me: Requested 'libopkg = 0.1.3' but version of libopkg is 0.1.4 May 15 17:49:41 have you updated? May 15 17:51:47 Yes, I just did an update (with autorev) May 15 17:52:51 (the packagekit dependency was from openmoko-qtopia-x11-image; I was able to build the regular openmoko-devel-image this morning) May 15 17:55:14 override the libopkg preferre version in your conf May 15 17:56:09 PREFERRED_VERSION_libopkg = "0.1.4" May 15 17:56:39 unless it sets it in the image .bb it should work May 15 18:17:30 ScaredyCat: Thanks, but it didn't help. http://www.pastebin.ca/1019256 has the full log. May 15 18:23:36 you could try modding the .bb ( /home/mmontour/moko/openembedded/packages/packagekit/packagekit_git.bb ) May 15 18:24:26 seth the opkg entry to "opkg (>=0.1.3)" without quotes May 15 18:24:43 seth... May 15 18:24:44 set May 15 18:24:50 I'll have a looksie May 15 18:37:58 hey guys, can I by a freerunner yet? May 15 18:38:05 s/by/buy/ May 15 18:38:06 XorA|gone meant: hey guys, can I buy a freerunner yet? May 15 18:38:14 * ScaredyCat slaps XorA|gone May 15 18:38:24 you get one free don;t you May 15 18:41:11 scaredycat what looks wrong about the picture? May 15 18:41:33 oh, hehe... nothing now... May 15 18:41:50 I didn;t set the high level thingys as services until after May 15 18:42:18 I saw them as apps, hence why are they below gtk et al May 15 18:43:10 but apparently I talk unitelligable giberish anyway May 15 18:43:17 ScaredyCat has been reflashed with new understanding from Mickey's picture May 15 18:43:28 yes... May 15 18:43:46 * ScaredyCat still isnt sure they should be python services though May 15 18:43:52 well May 15 18:43:58 ;o May 15 18:43:59 see "the role of python" May 15 18:44:03 yoink! May 15 18:44:06 i'm not anal 'bout that May 15 18:44:24 if someone comes up with more efficient implementation serving the same API - be my guest May 15 18:44:57 yeah, I see that you're using it to prototype it... I'm just not sure that it'd ever end up as a C app - you know, "It works as is, why change it" May 15 18:45:19 just compile the python, job done May 15 18:45:40 that fish clearly wasn't big enough ;) May 15 18:45:47 ~prod ScaredyCat May 15 18:45:48 * apt zaps ScaredyCat with a high voltage cattle prod May 15 18:46:52 * XorA|gone needs food, hacking ASU makes me hungry May 15 18:47:39 ScaredyCat: your concern is very valid, but then again... if it works as is why change it? :D seriously, I'm expecting the C ones will hardly give any benefit except memory footprint, but will cost 5 times the implementation time. May 15 18:48:34 I agree, the memory footprint and supporting libs needed are my 2 'issues' May 15 18:49:13 but it's all dbus anyway, right? So my apps can be C, bash script whatever .... May 15 18:49:24 just making the dbus calls... May 15 18:49:46 I just don;t want the supporting services eating all the ram / flash May 15 18:50:09 ~cook roast beef for XorA|gone May 15 18:50:10 * apt puts on his apron and fires up the stove. "Your roast beef will be ready in a little while, XorA|gone." May 15 18:50:11 I think the time lost due to interpreting python will be dwarfed by the time taked to blit images around a gta02 screen :-) May 15 18:50:43 :( May 15 18:51:43 where's that new boot splash anyway XorA|gone May 15 18:51:44 ScaredyCat: *nod*, that's the idea. write your dialer using the 'dialog' program :) May 15 18:51:51 lol May 15 18:52:27 actually looking at building gtk-server for it - much more flexible :) May 15 18:52:37 right May 15 18:53:24 ScaredyCat: it will be in the ASU branch when I finish merging it May 15 18:53:35 but yes, the dbus stuff is flexible enough for all sorts May 15 18:53:50 ScaredyCat: maybe tomorrow if I act like a genius May 15 18:54:01 ASU? April Surprise Undoing? May 15 18:54:26 April Software Update May 15 18:54:43 ... it's not April May 15 18:54:54 * ScaredyCat changes it May 15 18:55:01 its 11 months until April so a good chance we will deleiver early :-) May 15 18:55:02 s/April/Another/ May 15 18:55:02 ScaredyCat meant: ... it's not Another May 15 18:55:08 lol May 15 18:55:12 stupid bot May 15 18:55:22 XorA|gone: roflmao May 15 18:56:05 [2008-05-15 19:53:44] ScaredyCat: maybe tomorrow if I act like a genius May 15 18:56:19 acting in't your forte though, is it ;) May 15 18:56:53 sshh ssh he's back May 15 18:57:00 ScaredyCat: well linux companies hire me although Im a total windows fanboy :-) May 15 18:57:23 hehehe May 15 18:57:53 I heard microsoft pay you to slow down linux projects... May 15 18:58:29 they do May 15 18:58:42 yeah, I can see it works :P May 15 18:58:50 * mwester does that for free :p :D May 15 18:58:58 lol May 15 19:00:46 XorA|gone: You're a Windows fanboy? May 15 19:01:08 yeah, windows 3.1 too May 15 19:01:40 Wow, what a masochist May 15 19:02:35 a true masochist would prefer Vista ;-p May 15 19:04:22 openmoko(freerunner) will come with pulseaudio like on that:http://wiki.openmoko.org/wiki/NeoSoftwareStack so i will be able to make the audio from my computer go to the phone that will be connected to my hifiset...or is there any complications? May 15 19:04:25 oops hello May 15 19:06:11 GNUtoo: Yeah, the complication is about how to attach the phone to your head so you can use it as a headphone May 15 19:07:41 :) May 15 19:08:44 * SpeedEvil pases abraxa_ a selection of galvanised nails. May 15 19:09:52 SpeedEvil: Why me? I'd use duct tape ;) May 15 19:10:44 abraxa_, it's not a standard jack? May 15 19:12:16 GNUtoo: You want to connect the FR to your HiFi amp? Sure you can, but... why, if you have a computer May 15 19:13:24 abraxa_, if i use the computer i must pass a cable between it and the hifi amp which is not possible because i would have the risk of falling because of the cable... May 15 19:14:01 bluetooth A2DP works May 15 19:14:03 So instead you want to leave the phone on the amp and feed it audio via BT/WiFi? May 15 19:14:21 From the phone to the amp that is. May 15 19:14:32 abraxa_, yes May 15 19:14:42 There are any number of BT->3.5mm jack thingies available. May 15 19:14:43 SpeedEvil: Do you remember if the 2.5mm jack is mono or stereo? I forgot May 15 19:14:49 abraxa_, wifi with pulseaudio... May 15 19:14:50 stereo May 15 19:15:00 GNUtoo: Then that works May 15 19:15:06 There is one speaker. May 15 19:15:18 abraxa_, thanks a lot.... May 15 19:15:40 GNUtoo: Personally I'd just buy a dedicated A2DP receiver and a BT dongle for the computer but... yeah May 15 19:16:14 abraxa_, as i will buy the openmoko...i can also use it for that... May 15 19:16:22 Yup :) May 15 19:25:40 have there been any efforts to set a release date for software? May 15 20:14:49 stefan_schmidt: random question, are you related to Soren Schmidt at all? May 15 20:15:08 cb22: no May 15 20:17:40 stefan_schmidt: but I bet you are related to many people why have left billions of dollars in secure accounts in 3rd world countries May 15 20:18:32 openmoko: 03thomas * r4436 10/trunk/src/target/opkg/src/Makefile.am: opkg: fix remaining accidental commit May 15 20:19:22 XorA|gone: heh May 15 20:21:21 I have no such luck with secure accounts. However, I *have* been personally selected by a deposed politician in Nigeria to help him recover his wealth. May 15 20:21:43 hello you guyes May 15 20:22:44 anyone knows how you get freerunner hardware in qemu with mokomakefile? May 15 20:23:14 dnh: you can only get neo1973 hardware emulation in qemu. May 15 20:23:47 mwester: no - you need to build the matter replicator module. May 15 20:23:52 Then you get the hardware. May 15 20:24:25 speedevil how do you do that? May 15 20:24:52 Sorry - it was not a serious comment.. May 15 20:25:29 aha, now i get it May 15 20:26:42 the guide says: Neo FreeRunner (GTA02) May 15 20:26:42 The instructions for running the emulator are the same for FreeRunner, but the final command line to invoke qemu changes. Instead of "-M gta01" as the machine name "-M gta02fake" has to be used May 15 20:27:40 how do i do this with the makefile? May 15 20:41:08 Can anyone help with a failure in the git fetcher for linux-openmoko? http://www.pastebin.ca/1019412 May 15 21:09:22 * SpeedEvil wonders why tangogps seems to occasionally lose config. May 15 21:45:14 last quarter hour "before May 16." :) May 15 21:45:37 haha May 15 21:45:47 (At my timezone!) May 15 21:45:47 it's May 16 here already May 15 21:45:53 c-c May 15 21:45:55 :) May 15 21:46:13 it's 45 minutes past May 16 May 15 21:57:25 lets just hope everything went well May 15 21:57:44 somebody does know May 15 21:58:00 :) May 15 21:58:36 y let the linux slinging begin May 15 22:00:18 * mwester thinks OM should just collect a bunch of GTA01's and send them off to the too-eager can't-wait consumers... May 15 22:00:56 I want a broken screen GTA01 to try adding wifi to. May 15 22:01:12 By the time they arrive and someone realizes, there will have been time to do the GTA02 production *right* May 15 22:01:18 I think I have a solution. May 15 22:01:25 For internal wifi. May 15 22:01:39 I just want the GSM to work. It's a phone, after all. May 15 22:01:43 yeah May 15 22:02:26 There's room above the display by the debug connector; that might work for a usb wireless device. May 15 22:05:48 There really isn't May 15 22:06:20 :) Ok... May 15 22:06:23 The earpiece gets in the way May 15 22:06:31 so where else is there room? Battery? May 15 22:06:44 I was thinking of deleting and replacing the GPS antenna May 15 22:06:48 with a smaller one May 15 22:06:53 And putting it there May 15 22:06:54 Ah! May 15 22:07:13 how do you 'delete' hardware? May 15 22:07:19 It slides right out. May 15 22:07:21 Wiresnips. May 15 22:07:31 And hammers. May 15 22:07:52 lasers. May 15 22:07:55 It is a rather large space, most of that pcb is wasted (looks to exist only to secure the little ceramic antenna) May 15 22:08:19 It's not completely simple. May 15 22:08:37 It's quite probable that it'd end up with GPS only working when wifi was off. May 15 22:09:13 Which is not an unreasonable compromise, if one considers. May 15 22:11:02 not being able to use both together also ensures a battery life of > 30 minutes :P May 15 22:11:50 It's not _quite_ that bad May 15 22:12:07 But, yes, it has an impact. May 15 22:26:53 call fic an find out for us May 15 22:40:50 Today is the last day before May 16th. Hope the phone is in production. :) May 15 22:41:33 Actually, it's May 16th in China now. They must be shooting off the assembly line by the tens of thousands. May 15 22:41:39 Surely. May 15 22:51:18 we hope May 15 22:54:42 Where in china is this plant? May 15 22:54:53 It's quite far I hope from the recent quake? May 15 22:55:19 My impression is that most of the manufacturing of this type occurs near the coast. May 15 22:58:16 SpeedEvil: http://laforge.gnumonks.org/weblog/2007/06/15/#20070615-gta01-factory-trip - should be the same for GTA02, I'd say May 15 23:23:49 hmm May 15 23:24:00 * mickeyl wonders about broken XMLRPC May 15 23:24:05 CIA-44: is your daddy dead? May 16 00:10:40 wait...may 16th...That's tomorrow... May 16 00:10:52 Today May 16 00:10:58 For an hour. May 16 00:11:01 (here) May 16 00:11:05 (uk) May 16 00:11:20 oh no! May 16 00:13:24 * Sonicadvance1 dies May 16 00:14:33 * SpeedEvil lights a ultrabright red LED in Sonicadvances memory. May 16 00:15:55 * Sonicadvance implodes May 16 00:16:11 and thus the desktop die May 16 00:28:34 oh yay someone who knows ARM architecture offered to fix ffcall and libsigsegv May 16 00:28:46 mmm clisp with FFI and generational GC May 16 00:29:33 :D May 16 00:29:36 "ground cover" 5cm May 16 00:29:38 oops May 16 00:30:30 foggy out :P May 16 01:10:48 freesmartphone.org: 03mickey * ra39442d8db86 10/ophoned/ (gsm/channel.py gsm/mediator.py objects.py ophoned): python-ophoned: add test commands and fix decoding for phonebook entries != ASCII May 16 01:10:48 freesmartphone.org: 03mickey * r4309b910b9ed 10/ophoned/ (5 files in 2 dirs): May 16 01:10:48 freesmartphone.org: python-ophoned: more goodies May 16 01:10:48 freesmartphone.org: * use seconds as timeout metric May 16 01:10:50 freesmartphone.org: * start using regular expressions for parsing May 16 01:10:52 freesmartphone.org: * add more of org.freesmartphone.GSM.Network May 16 01:14:40 so... with mickeyl's most recent blog post, do we know anything about when a usable software set might be ready? May 16 01:15:28 5 hours? :D May 16 01:15:36 np: Annie - The soundtrack - Tomorrow. **** ENDING LOGGING AT Fri May 16 02:59:57 2008