**** BEGIN LOGGING AT Sat Aug 16 12:48:14 2008
Aug 16 12:57:14 <roh> n8
Aug 16 13:00:07 <CIA-19> freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rdd440a212e35 10/odeviced/src/plugins/accelerometer/ (accelerometer-helpers.c accelerometer.vala): Accelerometer: fix types and ignore EV_SYN correctly
Aug 16 13:00:07 <CIA-19> freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * rcd8ff3fc8f08 10/odeviced/ (13 files in 5 dirs): 1.) Refactor wifi plugin into a generic Powercontrol plugin TODO: add Bluetooth,GSM,GPS
Aug 16 13:00:08 <CIA-19> freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r7b87e22f9589 10/odeviced/src/plugins/powercontrol/ (powercontrol.plugin powercontrol.vala): Add GSM, GPS and Bluetooth Powercontrol
Aug 16 13:00:09 <CIA-19> freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r2e6564f3644d 10/opimd/ (13 files in 2 dirs): Merge branch 'master' of git@git.freesmartphone.org:openmoko-gsoc2008
Aug 16 13:00:57 <eartoast> moin
Aug 16 13:02:54 <lindi-> any ideas on how to debug xorg video modules?
Aug 16 13:03:10 <lindi-> ginger:~# file /usr/lib/xorg/modules/drivers/fbdev_drv.so
Aug 16 13:03:10 <lindi-> /usr/lib/xorg/modules/drivers/fbdev_drv.so: ELF 32-bit LSB shared object, ARM, version 1 (SYSV), dynamically linked, not stripped
Aug 16 13:03:22 <lindi-> yet i get 'Function "FBDevPointerMoved" not defined.' in gdb
Aug 16 13:04:14 <eartoast> in the api docs, there is nowhere specified in what form incoming call numbers are handed to the application... meaning national/international notation
Aug 16 13:31:17 <CIA-19> freesmartphone.org: 03sudharsh 07openmoko-gsoc2008 * r99602a8f4c13 10/odeviced/src/daemon/helpers.vala: Fix a bug where ListObjectsByInterface didn't return in case of plugins registering a single dbus_path
Aug 16 14:21:01 <easy_> argl, my pointer coords are wrong.
Aug 16 14:53:25 <lindi-> easy_: xorg?
Aug 16 14:53:55 <easy_> lindi-, yeah, but fixed now.
Aug 16 14:54:07 <lindi-> easy_: in landscape too?
Aug 16 14:54:38 <easy_> Dunno, because after entering my pin it looks like the X is frozen...
Aug 16 14:55:03 <easy_> hrm reboot (again)...
Aug 16 15:01:01 <trupheenix> yay!!!!
Aug 16 15:01:10 <trupheenix> developers channel!! i found it
Aug 16 15:01:17 * trupheenix is very happy
Aug 16 15:02:06 <easy_> hrm, looks like zhone is frozen. It don't react anymore after I entered my pin...
Aug 16 15:04:25 <trupheenix> any reason why om2008.08 is built on QTOPIA? i don't see where's the "open"-ness in Qtopia. you need to pay for the damn thing to open it
Aug 16 15:05:31 <edistar_> trupheenix: qtopia is gpl
Aug 16 15:05:43 <lindi-> gplv2-only
Aug 16 15:06:08 <lindi-> easy_: /etc/init.d/frameworkd stop
Aug 16 15:06:14 <lindi-> easy_: then run frameworkd from console
Aug 16 15:06:23 <lindi-> easy_: to see what it talks to zhone
Aug 16 15:06:48 <trupheenix> edistar_: i hate Qt!!! 
Aug 16 15:06:58 <easy_> lindi-, hrm. I just want to install debian using the "official" install-script instead of install.txt.gismo.
Aug 16 15:07:00 <edistar_> trupheenix: me too, but that doesn't mean it's not open ;)
Aug 16 15:07:14 <trupheenix> the only reason i like openmoko is because it used GTK+
Aug 16 15:07:15 <lindi-> easy_: yes, this is what i am doing too?
Aug 16 15:07:36 <edistar_> trupheenix: it will abandon qtopia again.. don't worry
Aug 16 15:07:38 <lindi-> trupheenix: so use gtk, debian has them all
Aug 16 15:08:02 <trupheenix> lindi-: debian has them all?
Aug 16 15:08:10 <lindi-> trupheenix: debian has both gtk and qt
Aug 16 15:08:19 <trupheenix> lindi-: are u asking me to pt debian atmel on the neo?
Aug 16 15:08:26 <lindi-> yep
Aug 16 15:08:36 <trupheenix> lindi- hmmmm... interesting
Aug 16 15:08:41 <lindi-> or well, not asking but saying that i am doing it
Aug 16 15:11:21 <edistar_> lindi-: there is really no reason at all not to use debian on the FR, now that frameworkd works (except space, but that's not a big deal with sdhc)
Aug 16 15:11:44 <lindi-> edistar_: glamo is not included yet
Aug 16 15:11:54 <lindi-> edistar_: and i have trouble getting touchscreen working in rotated fbdev
Aug 16 15:12:07 <lindi-> i don't understand which part of xorg is supposed to do rotation and which is not :)
Aug 16 15:12:16 <lindi-> both tslib and fbdev drivers seem to rotate things :)
Aug 16 15:13:02 <edistar_> lindi-: glamo isn't included with the normal OM releases either
Aug 16 15:13:18 <edistar_> if you rotate both, what happens?
Aug 16 15:14:03 <lindi-> edistar_: with http://rafb.net/p/27A7mx32.html scaling is wrong
Aug 16 15:14:33 <lindi-> edistar_: if i interchange width/height in tslib scaling is right but i have an offset of 160 in y direction
Aug 16 15:14:52 <edistar_> still a bug in tslib?
Aug 16 15:15:01 <edistar_> afaik the values were all hard-coded?
Aug 16 15:15:02 <lindi-> edistar_: i am not sure where the bug is
Aug 16 15:15:09 <lindi-> edistar_: i am not sure what "width" is supposed to mean :)
Aug 16 15:15:24 <edistar_> what about if you leave out those options?
Aug 16 15:15:32 <edistar_> because width and height have to be swapped..
Aug 16 15:15:39 <lindi-> edistar_: they do?
Aug 16 15:15:43 <edistar_> I think so..
Aug 16 15:15:51 <lindi-> edistar_: what does width mean?
Aug 16 15:16:09 <lindi-> is it property of the touchscreen hardware?
Aug 16 15:16:13 <edistar_> it tells tslib how wide the ts is
Aug 16 15:16:17 <edistar_> yep
Aug 16 15:16:27 <lindi-> edistar_: is the touchscreen resolution really 640?
Aug 16 15:17:08 <edistar_> dunno how it's translated.. must be though
Aug 16 15:17:12 <edistar_> for the input device
Aug 16 15:17:25 <lindi-> too many rotations and translations going on :)
Aug 16 15:17:40 <edistar_> usual for X ;0
Aug 16 15:18:40 <edistar_> in principle everything should be transparent.. e.g. tslib shouldn't care how the screen is rotated
Aug 16 15:18:57 <lindi-> edistar_: ok
Aug 16 15:19:08 <lindi-> edistar_: so what is this rotation option for?
Aug 16 15:19:40 <edistar_> not sure
Aug 16 15:20:26 <edistar_> how does OM rotate? don't they use kdrive?
Aug 16 15:21:19 <lindi-> yes
Aug 16 15:21:48 <edistar_> what do those configs look like?
Aug 16 15:22:00 <lindi-> it is configured by editing source code
Aug 16 15:22:13 <lindi-> to save space :P
Aug 16 15:22:53 <edistar_> really? never looked into kdrive.. thought it had config files but then compiled them in
Aug 16 15:23:04 <edistar_> saves about 10kb maybe?
Aug 16 15:23:13 <lindi-> i am not sure
Aug 16 15:23:36 <lindi-> now i have debug symbols in both tslib and fbdev
Aug 16 15:24:02 <trupheenix> lindi- what exactly is the FSO image do?
Aug 16 15:24:40 <lindi-> trupheenix: no idea, haven't managed to compile it
Aug 16 16:11:43 <lindi-> edistar: any experience on quilt for debian patch management?
Aug 16 16:11:53 <lindi-> edistar: xf86-input-tslib seems to use it
Aug 16 16:12:09 <edistar> no, sry
Aug 16 16:13:14 <lindi-> i think i have a solution but i can't apply it because of this :P
Aug 16 16:13:41 <edistar> oh
Aug 16 16:50:32 <lindi-> hurray, fixed :)
Aug 16 17:19:00 <CIA-19> freesmartphone.org: 03mickeyl * r298 10/ (6 files in 5 dirs): clean
Aug 16 18:09:09 <lindi-> nomeata: hi, what's the status of http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=494724 (or http://trac.freesmartphone.org/ticket/90)? could debian include a quick fix before it is in upstream? this seems to make it impossible for people to make calls when they first install debian at the moment
Aug 16 18:10:06 <lindi-> i'm thinking about going through the installation script and check that all known bugs that i have workarounded on my local system have been fixed
Aug 16 18:12:13 <nomeata> mickey|dinner: I’m wondering about the patches that I submitted on smartphone-standards, especially the number matching and newline fixes (libpam-dbus probably needs thought): are you the one that will appy or reject them, or should I kick shoragan instead?
Aug 16 18:12:13 <nomeata> (kicking him is easy ATM, I just have to straighten my knee)
Aug 16 18:14:49 <nomeata> lindi-: I’d like to see it applied in the upstream git repositry and then cherry-pick the change, to make sure we get the same fix as upstream
Aug 16 18:17:59 <lindi-> nomeata: very good point yes. it's just that a dozen people are hitting it at the moment
Aug 16 18:23:48 <nomeata> lindi-: true
Aug 16 18:24:52 <nomeata> lindi-: shoragan just told me that we should first find out why this bug only seems to occur on Debian – or was it reported on FSO as well (I’m not sure)
Aug 16 18:26:29 <lindi-> we could fix it first and then figure it out :P
Aug 16 18:27:21 <nomeata> good point as well.
Aug 16 18:27:36 <shoragan> lindi-, i've got new images at http://134.169.172.106/~jluebbe/images/neo1973/
Aug 16 18:27:58 <lindi-> shoragan: FSO images
Aug 16 18:28:10 <lindi-> ok maybe i should compare against those
Aug 16 18:39:41 <nomeata> lindi-: the thing is, the code for the SIM passing is not in anyway special, so if there is a bug, it ought to be fixed centrally at the same point
Aug 16 18:40:20 <lindi-> yes sure
Aug 16 18:42:15 <lindi-> hmm, did something break with evas? 'svn co http://barbieri-playground.googlecode.com/svn/efl-tests/vkbd/' worked yesterday but today i get 'TypeError: Argument 'canvas' has incorrect type (expected evas.c_evas.Canvas, got NoneType)'
Aug 16 18:51:06 <nomeata> shoragan: http://dbus.freedesktop.org/doc/dbus-python/doc/tutorial.html#return-values-and-the-byte-arrays-and-utf8-strings-options aber auch nicht viel neues
Aug 16 20:34:02 <nomeata> lindi-: still here?
Aug 16 20:36:04 <lindi-> nomeata: yes
Aug 16 20:36:16 <CIA-19> freesmartphone.org: 03stefan 07artwork * r7b1b64007cbb 10/illume/illume/ (25 files): illume-theme: Sync theme with upstream illume. Add pixmaps for busywin.
Aug 16 20:36:16 <CIA-19> freesmartphone.org: 03stefan 07artwork * r1e497e1b9f2e 10/illume-config/ (e.src module.battery.src module.illume.src): 
Aug 16 20:36:16 <CIA-19> freesmartphone.org: illume-config: Upstream version of the illume config files. Need to adjust to
Aug 16 20:36:17 <CIA-19> freesmartphone.org: FSO needs
Aug 16 20:38:23 <nomeata> lindi-: how was it: does making phone calls cause the same error as entering a PIN?
Aug 16 20:39:06 <CIA-19> freesmartphone.org: 03stefan 07artwork * r028de6a7036b 10/illume-config/e.src: illume-config: Disable clock module
Aug 16 20:39:28 <lindi-> nomeata: i didn't get past pin entry so i did not try making calls at that time
Aug 16 20:40:44 <lindi-> nomeata: it's kind of hard to test. if i fix the bug i can enter pin and the unicode string won't cause a problem with making calls either
Aug 16 20:41:31 <nomeata> lindi-: hmm, I could reproduce it here
Aug 16 20:41:48 <nomeata> lindi-: it’s really strange: the (assumed) cause was a change in the str/unicode handling in zhone
Aug 16 20:42:09 <nomeata> lindi-: and it’s surprising that it should be propagated through dbus
Aug 16 20:42:17 <lindi-> :)
Aug 16 20:44:47 <lindi-> nomeata: do you know how matchbox-keyboard can stay non-fullscreen with matchbox while all other apps are fullscreen?
Aug 16 20:45:06 <rwhitby> Ainulindale: http://shr.bearstech.com/fso-testing/ is now up
Aug 16 20:45:44 <nomeata> lindi-: no, but It’s probably some matchbox-specific X atom appended to the window.
Aug 16 20:45:53 <nomeata> lindi-: I’m sure you’ll easily find it in the code
Aug 16 20:45:54 <MarcOChapeau> rwhitby: nice :-)
Aug 16 20:46:37 <rwhitby> MarcOChapeau: fso-milestone2 and fso-unstable are still building.  ignore the fso-dev directory that's there, it's disappearing soon.
Aug 16 20:46:51 <lindi-> nomeata: yeah i am reading through it right now
Aug 16 20:47:18 <MarcOChapeau> rwhitby: no problem. do the ipk for the shr stuff get built too ?
Aug 16 20:48:44 <rwhitby> MarcOChapeau: I'm intending to build task-openmoko-feed for each
Aug 16 20:49:09 <rwhitby> but that's probably a day or two away - looks like no-one has built that in a long time.
Aug 16 20:49:09 <MarcOChapeau> ok
Aug 16 20:53:20 <nomeata> lindi-: ah, mickey|dinner is looking into the encoding bug
Aug 16 20:54:29 <lindi-> i am trying to figure out how to build a virtual keyboard :)
Aug 16 21:11:18 <nomeata> lindi-: ok, I uploaded a version of fso-frameworkd with the fix. care to do a quick test?
Aug 16 21:14:21 <lindi-> nomeata: "apt-get update -o Acquire::PDiffs=false" is running
Aug 16 21:25:17 <CIA-19> freesmartphone.org: 03stefan 07artwork * r551dd7dc53f0 10/illume/illume/freerunner.edc: illume-theme: Remove flaunch
Aug 16 21:30:24 <lindi-> now rebooting
Aug 16 21:35:01 <lindi-> nomeata: as expected http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=494724 does not occur anymore
Aug 16 21:35:12 <lindi-> nomeata: i only have to workaround http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=494781 now
Aug 16 21:35:25 <lindi-> (or i hit another assertion failure)
Aug 16 21:35:49 <nomeata> lindi-: ok, thx
Aug 16 21:36:49 <lindi-> there seems to be a lot to learn about matchbox and X internals
Aug 16 21:38:19 * nomeata has a quick look
Aug 16 21:40:26 <lindi-> the iphone style keyboard demo seems to be quite usable, i can type quite fast with it. the only problem is that it's just for demonstration. it is fullscreen and only shows the text in its own window and doesn't know how to send it to other apps
Aug 16 21:41:37 <nomeata> this looks interesting:
Aug 16 21:41:39 <nomeata>   wm_hints = XAllocWMHints();
Aug 16 21:41:39 <nomeata>   if (wm_hints)
Aug 16 21:41:39 <nomeata>     {
Aug 16 21:41:39 <nomeata>       DBG("setting no focus hint");
Aug 16 21:41:39 <nomeata>       wm_hints->input = False;
Aug 16 21:41:40 <nomeata>       wm_hints->flags = InputHint;
Aug 16 21:41:42 <nomeata>       XSetWMHints(ui->xdpy, ui->xwin, wm_hints );
Aug 16 21:41:44 <nomeata>       XFree(wm_hints);
Aug 16 21:41:46 <nomeata>     }
Aug 16 21:41:48 <nomeata> in ./src/matchbox-keyboard-ui.c
Aug 16 21:44:49 <lindi-> hmm
Aug 16 21:45:37 <lindi-> "http://barbieri-playground.googlecode.com/svn/efl-tests/vkbd/" uses enlightenment libraries and is written in python. i hope there's some way to set those flags without C
Aug 16 21:46:18 <nomeata> probably hard
Aug 16 21:47:32 <lindi-> http://www.youtube.com/watch?v=TrWpUBTDDr0 seems to demo the keyboard if you want to see (youtube-dl it, didn't work in gnash today)
Aug 16 22:46:26 <Sargun> Question: Why don't we use Debian as a base for FSO/SHR instead of OE?
Aug 16 22:57:15 <mwester> Because Debian/ARM is not well-tuned for small devices such as the neo; we would end up doing much repackaging and custom work -- that has all been done already in OE to a great extent.
Aug 16 22:59:46 <rwhitby> http://tinderbox.openembedded.net/builders/shr.bearstech.com/
Aug 16 23:01:15 <mwester> :)
Aug 16 23:02:21 <Sargun> mwester, the Neos pretty large...
Aug 16 23:02:43 <Sargun> I mean, I've ran Debian on 64 Mbyte/133Mhz before.
Aug 16 23:02:56 <Sargun> (X86)
Aug 16 23:03:04 <Sargun> mwester, you head of Scratchbox and/or Maemo?
Aug 16 23:04:38 <rwhitby> Since SHR just uses GTK+ and FSO framework, there is no reason why it shouldn't run identically on either a Debian or OpenEmbedded base system.
Aug 16 23:08:58 <Sargun> uhm.
Aug 16 23:10:14 <Sargun> What does OE have that debian doesn't?
Aug 16 23:13:22 <rwhitby> optimized recipes for small-resource devices
Aug 16 23:14:59 <rwhitby> the fact that you can't run a Debian rootfs of any sort from internal flash shows the quantum difference in resource usage.
Aug 16 23:16:43 <Sargun> Good point.
Aug 16 23:16:57 <Sargun> You need about 1Gig-ish for Debian.
Aug 16 23:17:02 <Sargun> but NAND is cheap.
Aug 16 23:17:45 <rwhitby> internal RAM is the most precious resource.  Debian is not optimised for small-RAM devices.
Aug 16 23:17:52 <Sargun> RAM is expensive.
Aug 16 23:18:04 <Sargun> Flash is cheap(ish)
Aug 16 23:18:27 <rwhitby> the RAM footprint of Debian rootfs is much larger than an OE one too
Aug 16 23:18:29 <Sargun> rwhitby, How much RAM does base debian really use? 
Aug 16 23:18:29 <rwhitby> that's the point.
Aug 16 23:19:18 <rwhitby> on an NSLU2, for instance, the OE-built rootfs uses about half as much RAM after boot as the Debian one.
Aug 16 23:19:55 <Sargun> How much RAM?
Aug 16 23:20:07 <rwhitby> on the nslu2, it's 32MB.
Aug 16 23:20:23 <rwhitby> and if you say FR has more ram, then you don't get the point :-)
Aug 16 23:20:55 <Sargun> Haha.
Aug 16 23:21:01 <rwhitby> (since it's about efficiency, not actual current limits)
Aug 16 23:21:01 <Sargun> No, I'm saying how much RAM does OE use?
Aug 16 23:22:02 <rwhitby> SlugOS runs in 16MB of RAM
Aug 16 23:23:03 <rwhitby> don't get me wrong - I think Debian is great.
Aug 16 23:23:04 <Sargun> hm.
Aug 16 23:23:06 <Sargun> Neat.
Aug 16 23:23:21 <Sargun> I wish I could run Gentoo on it, but Gentoo eats RAM like a bitch.
Aug 16 23:23:30 <rwhitby> But the mere fact that Emdebian exists shows that normal debian is not optimised for embedded devices
Aug 16 23:23:51 <rwhitby> (I use both OE-based and Debian-based distros on embedded devices I develop for ...)
Aug 16 23:24:34 <Sargun> hm
Aug 16 23:24:44 <Sargun> Do you work for FIC?
Aug 16 23:24:54 <mwester> JMO, but I think Debian on the neo is a great idea for those who are looking for a very small PC.  It's not what I would consider useful for a phone, though...
Aug 16 23:25:08 <Arcnet> hi
Aug 16 23:25:08 <rwhitby> Sargun: no, I run nslu2-linux.org
Aug 16 23:25:14 <Sargun> Ah, neat.
Aug 16 23:25:28 <rwhitby> FIC can't afford me ;-)
Aug 16 23:25:31 <Sargun> hehe.
Aug 16 23:25:55 <Sargun> I don't develop on embedded hardware...ever.
Aug 16 23:26:26 <Sargun> I do server side stuff mostly
Aug 16 23:26:49 <Arcnet> The partition part of the debian installer seems to cause SD card corruption, at least i think it messed up my partition table... Did anyone had the same problem?
Aug 16 23:27:19 <Sargun> Though, I did do some fun reverse engineering with the DE
Aug 16 23:27:57 <Sargun> I wonder if they are ever going to release the HXD8 independently from DE
Aug 16 23:28:06 <Arcnet> (or: Is it a known bug? are any debian people here?)
Aug 16 23:28:23 <Sargun> rwhitby, What's your day job?
Aug 16 23:29:00 <Sargun> Anyone know?
Aug 16 23:29:33 <rwhitby> Sargun: CTO
Aug 16 23:29:45 <Sargun> For.....
Aug 16 23:29:54 <rwhitby> Google knows :-)
Aug 16 23:32:25 <Sargun> hehe
Aug 16 23:32:28 <Sargun> No it doesn't...
Aug 16 23:32:32 <Sargun> http://wiki.openmoko.org/wiki/HXD8
Aug 16 23:32:36 <rwhitby> http://www.linkedin.com/in/rwhitby
Aug 16 23:33:02 <Sargun> Oh
Aug 16 23:33:11 <Sargun> I thought you meant Google knows about the HXD8
Aug 16 23:33:30 <Sargun> Damn Aussy
Aug 16 23:33:31 <Sargun> :-P
Aug 16 23:37:35 <Sargun>  be right back
Aug 16 23:43:17 <Weiss> is there a sensible reason why http://www.angstrom-distribution.org/unstable/sources/e17.libs.eet_anoncvs.enlightenment.org__200808110200.tar.gz (which my OE build wants) doesn't exist at the moment? have i done something wrong?
Aug 16 23:44:00 <Weiss> i fixed the problem by changing EFL_SRCDATE to 20080811 (trimming off the '0200' in sane-srcdates.inc, but that doesn't feel right
Aug 16 23:45:50 <Weiss> also, at what point is local.conf read? it doesn't seem to be paying attention to my INHERIT += "rm_work", though i did interrupt the build to add that when i realised i was going to run out of disk space..
Aug 16 23:52:08 <rwhitby> Weiss: it is normal that it does not exist
Aug 16 23:52:22 <rwhitby> it's a cache if the autobuilder has built it before you have
Aug 16 23:52:41 <rwhitby> if it doesn't exist your build will just check it out directly from cvs
Aug 16 23:53:12 <Weiss> rwhitby: i saw that, but it tried to check it out with the date '200808110200', and complained about an invalid date specifier
Aug 16 23:53:14 <rwhitby> rm_work actually only operates at the end of the build
Aug 16 23:53:50 <Weiss> hmm.. i thought the point was that it got rid of everything when it was finished with?
Aug 16 23:53:51 <rwhitby> Weiss: you should report that in the OE bugtracker then
Aug 16 23:54:05 <rwhitby> it does, but only at the end of the build
Aug 16 23:54:24 <rwhitby> dunno why they decided to do it that way
Aug 16 23:54:40 <Weiss> grr... i'm going to have to move things around to free up a load of disk space in that case.. :(
Aug 16 23:55:06 <rwhitby> yup
Aug 16 23:56:13 <rwhitby> those interested in fso-testing task-openmoko-feed build can track it here: http://tinderbox.openembedded.net/builds/20658/
Aug 17 00:10:19 <Weiss> hmm, i can free up the required space by taking a chunk out of my used-about-twice-ever Windows partition.... is 21Gb-ish a sensible amount?
Aug 17 00:22:27 <mwester-laptop> That should do it. 
Aug 17 00:23:15 <mwester-laptop> I grew weary of fighting to make room all the time, so now my build partition is 250GB.
**** ENDING LOGGING AT Sun Aug 17 02:59:57 2008