**** BEGIN LOGGING AT Tue Nov 09 02:59:57 2010
Nov 09 03:03:02 <ashi_todd> halfhalo: i have no trouble changing channels... alt+number or /window #
Nov 09 03:03:09 <halfhalo> hmmm... err: Could not retrieve catalog from remote server: undefined method `closed?' for nil:NilClass
Nov 09 03:03:19 <halfhalo> most. useless. error. ever.
Nov 09 03:03:25 <ka6sox> bip
Nov 09 03:03:37 <SineOt> No, the old Error 11 on classic Mac OS was. :B
Nov 09 03:04:07 <halfhalo> stupid puppet in ubuntus repo has an issue where ruby hides its errors
Nov 09 03:04:35 <SineOt> "dsMiscErr: miscellaneous hardware exception error"
Nov 09 03:05:35 <ashi_todd> might as well say "oops"
Nov 09 03:08:01 <halfhalo> WHY WON'T YOU TELL ME WHY YOU CAN'T CONNECT YOU [CENSORED FOR THE CHILDREN]
Nov 09 03:12:57 * halfhalo slaps ubuntu upside the head for not liking the lenny backports repo he is trying to add
Nov 09 03:15:11 * halfhalo throws puppet out the window due to: err: Could not retrieve catalog from remote server: hostname was not match with the server certificate
Nov 09 03:23:03 <scoutcamper> ~halfhalo
Nov 09 03:23:03 <infobot> methinks halfhalo is violent today
Nov 09 03:23:16 <halfhalo> today?
Nov 09 03:23:35 <scoutcamper> well, excesivly so today :D
Nov 09 03:24:00 * scoutcamper waits for the webossucks troll to return, because he is bored :D
Nov 09 03:24:14 * halfhalo gets violent when programs misbehave
Nov 09 03:26:31 <halfhalo> oh you son of a.... if you are doing what I think you are doing I am going to beat you with your lifeless rake tasks
Nov 09 03:26:47 <scoutcamper> lol
Nov 09 03:31:43 <Loudergood> the beautiful thing about webos cards is you can smack them back and forth
Nov 09 03:32:08 <scoutcamper> then throw them away if they are really bad :D
Nov 09 03:34:46 <ashi_todd> :)
Nov 09 03:36:05 * halfhalo shoots openssl
Nov 09 03:40:31 <halfhalo> it set the cert start time in the FUTURE
Nov 09 03:40:47 * halfhalo got puppet working though
Nov 09 03:41:36 <Loudergood> on your pre?
Nov 09 03:42:11 <halfhalo> THAT IS A GREAT IDEA!!!
Nov 09 03:42:44 <Loudergood> (what have I done???)
Nov 09 03:44:31 <RobiPre> or chef :)
Nov 09 03:45:07 <RobiPre> at talk w/ guy from opscode
Nov 09 04:02:09 <destinal> this is really clever:   DIY dry-erase-form-factor conductive stylus for iPad   http://www.therussiansusedapencil.com/post/1521469842/the-right-feel-for-an-ipad-stylus/   (and more hygenic than sausage)
Nov 09 04:02:39 <destinal> I'll have to remember that for when webos tablets come around
Nov 09 04:08:43 <PuffTheMagic> thats sweet
Nov 09 04:11:39 * halfhalo has pixi sticks
Nov 09 04:15:21 <Davide-NYC> Has anyone been successful at running webos2.0 on an original pre?
Nov 09 04:15:42 <rwhitby> no
Nov 09 04:15:53 <destinal> Davide-NYC: not quite succesful, the GUI crashes infinitely
Nov 09 04:16:03 <rwhitby> well, it runs but crashes before the GUI displays anything
Nov 09 04:16:53 <Davide-NYC> what is the general feeling? will there be success? or is the hardware too different...
Nov 09 04:17:59 <destinal> Davide-NYC: it may be extremely difficult without kernel source from 2.0, or it may be as simple as swapping out some libraries depending on architecture.
Nov 09 04:18:11 * destinal pulls out the victim pre again
Nov 09 04:20:40 <destinal> I had stopped before because I reflashed this particular pre with 1.4.5 to try installing the flash plugin from 2.0 on it, but I didn't make much headway there
Nov 09 04:25:13 <ashi_todd> victim pre, lol, good name for it.
Nov 09 04:26:58 <Davide-NYC> It's probabl just best to wait for something a little more official.
Nov 09 04:28:37 <destinal> Davide-NYC: for end users, you're probably right, even if we can get it working they probably shouldn't do such a foolish thing to their primary device.  but for hackers the answer to "why" is like for mountain climbers  .. because it's there  (and in this case it would be cool)    ;)
Nov 09 04:30:15 <destinal> unfortunately I let the victim pre discharge so I have to wait a few minutes even charging at 1 amp   (oops)
Nov 09 04:32:02 * destinal goes rummaging around comparing doctors
Nov 09 04:36:57 <destinal> sadly I think the lib I was thinking of swapping with something from 1.4.5, doesn't exist in 1.4.5.  alas.
Nov 09 04:37:19 <destinal> what breaks I believe is unique to 2.0 so no drop in replacements to be found
Nov 09 04:40:53 <destinal> or .. maybe not.  /usr/lib/egl may be worth swapping
Nov 09 04:45:31 * phil_bw wanders by
Nov 09 04:47:41 <destinal> phil_bw: hey
Nov 09 04:47:52 <phil_bw> how's it goin?
Nov 09 04:48:20 <destinal> phil_bw: rebuilding my frankendoctor
Nov 09 04:48:44 <phil_bw> ah very nice, I was reading about that a little, how's it coming?
Nov 09 04:49:21 <destinal> phil_bw: it's probably a lost cause in any kind of reasonable time scale but I had a thought about swapping an EGL library that may or may not relate to the problem
Nov 09 04:49:42 <phil_bw> I see
Nov 09 04:50:09 <phil_bw> does 2.0 have significantly different EGL stuff going on?
Nov 09 04:50:12 <destinal> everything's hard because all this is closed source except the kernel used in 2.0 which hasn't been released yet so for the moment may as well be
Nov 09 04:51:01 <phil_bw> ah
Nov 09 04:51:21 <phil_bw> and I thought writing an OpenGL based app was hard, sounds like you're dealing with more diffcult stuff
Nov 09 04:51:24 <destinal> and if it involves reverse engineering and binary patching of proprietary userland binaries and proprietary kernels it could be rather challenging
Nov 09 04:51:46 <destinal> I hope it doesn't come to that because if it does, may as well just wait for 2.0 :)
Nov 09 04:52:45 <phil_bw> yeah, I have a feeling 2.0 will come around shortly after anything really useful comes from homebrew land
Nov 09 04:52:50 <destinal> so right now I'll take a couple of shots in the dark and hope to get lucky :)
Nov 09 04:53:13 <phil_bw> fun stuff
Nov 09 04:53:19 <phil_bw> well I better go to sleep
Nov 09 04:53:29 <phil_bw> tired of playing with texture maps for one day
Nov 09 04:53:54 <destinal> phil_bw: yeah, I'm not qualified to do 3d opengl programming though I'd like to learn someday
Nov 09 04:54:20 <phil_bw> oh 3d?  I'm no where near that yet, I'm still working on 2d myself
Nov 09 04:54:40 <phil_bw> got some naked sprites, now I just need to map some nice pictures to them :P
Nov 09 04:54:52 <destinal> as to that, I've read some opengl 2d stuff and understood some of it but never written any
Nov 09 04:55:07 <phil_bw> it's all quite painful
Nov 09 04:55:08 <scoutcamper> gah
Nov 09 04:55:25 <scoutcamper> graphics in photoshop=yay, in anything else =?
Nov 09 04:55:42 <phil_bw> haha
Nov 09 04:55:55 <phil_bw> if you ever feel like a headache look into OpenGL ES 1.1
Nov 09 04:55:59 <scoutcamper> lol
Nov 09 04:56:07 <phil_bw> if you want your head to flat out explode try OpenGL ES 2.0
Nov 09 04:56:15 <phil_bw> anyhow, 'night everybody
Nov 09 04:56:19 <destinal> usually when i feel like a headache I open up ARM binaries in IDA pro :)
Nov 09 04:56:23 <scoutcamper> ill stick with my mastery of photoshop CS5 :D
Nov 09 04:56:35 <destinal> but that's mainly because it's mostly greek to me at this point
Nov 09 04:57:07 * phil_bw wanders off
Nov 09 04:57:14 <destinal> phil_bw: cheers
Nov 09 05:00:11 * scoutcamper goes back to working on a WI shirt design
Nov 09 05:01:01 <scoutcamper> ~WI
Nov 09 05:01:01 <infobot> rumour has it, wi is webos-internals
Nov 09 05:02:04 <SineOt> scoutcamper: are you using MS Paint? XD
Nov 09 05:02:27 <scoutcamper> lol, nope, CS5 photoshop, much much beter :D
Nov 09 05:02:35 <SineOt> Good XD
Nov 09 05:03:33 <destinal> ms paint rules
Nov 09 05:03:38 <scoutcamper> lol
Nov 09 05:03:43 <scoutcamper> for n00bs
Nov 09 05:04:12 <SineOt> MS Paint is awesome if you want to make something ~ironically~ look terrible.
Nov 09 05:04:40 <scoutcamper> SineOt, good point
Nov 09 05:05:19 <SineOt> I've used MS Paint to make some sweet fan art before.
Nov 09 05:07:12 * scoutcamper shudders at MS Paint, then goes back to Photoshop, all is good :D
Nov 09 05:07:40 <SineOt> get in touch with your inner ironic hipster self
Nov 09 05:07:56 <SineOt> MS Paint is the only true way to create a sparkle dog
Nov 09 05:08:12 <scoutcamper> no, i dont wanna spoil my computer by starting that program
Nov 09 05:08:57 <oil> spraypaint tool ftw
Nov 09 05:09:18 <ashi_todd> heh, i could run ms paint if i fired up virtualbox... doesn't seem worth it. :)
Nov 09 05:09:41 <SineOt> run it in WINE!
Nov 09 05:10:08 <ashi_todd> lol
Nov 09 05:10:19 <ka6sox> Merlot?
Nov 09 05:12:30 <scoutcamper> ~mspaint
Nov 09 05:12:31 <infobot> from memory, mspaint is the world worst graphics prgram, and will continue to be
Nov 09 05:13:00 <SineOt> MSPaint in Windows 7 has the Office Ribbon, lol
Nov 09 05:15:56 <scoutcamper> Shirts-do you want front and back or just back or just front?
Nov 09 05:16:35 <ashi_todd> ooh a shirt
Nov 09 05:19:33 <oil> relevant to mspaint: http://www.collegehumor.com/video:1771096
Nov 09 05:22:18 <halfhalo> THAT IS RELEVANT TO MY INTERESTS
Nov 09 05:22:44 <tmzt> and lacks either working replace or transparent
Nov 09 05:22:53 * halfhalo has gone insane
Nov 09 05:23:07 <tmzt> but really, paintbrush was ood, replace is awesome, not dpi independent though
Nov 09 05:23:09 * oil wonders what took halfhalo so long to figure that out
Nov 09 05:23:31 * halfhalo demands oil goes and deals with the perils of puppet
Nov 09 05:23:53 <scoutcamper> ~halfhalo
Nov 09 05:23:53 <infobot> extra, extra, read all about it, halfhalo is crazy
Nov 09 05:24:09 <halfhalo> Didn't I change that to a real description?
Nov 09 05:24:42 <oil> thankfully not
Nov 09 05:24:45 <halfhalo> I could have sworn I did....
Nov 09 05:25:34 <tmzt> hmm, what's the issue with 2.0 on pre? why does that graphics server crash?
Nov 09 05:26:23 <scoutcamper> is there an official webos-internals font?
Nov 09 05:26:58 <oil> scoutcamper: comic sans
Nov 09 05:27:12 <halfhalo> ...
Nov 09 05:27:18 <scoutcamper> oil, thanks!
Nov 09 05:27:19 <oil> haha
Nov 09 05:27:23 * oil was joking!
Nov 09 05:27:31 * scoutcamper thinks it looks good
Nov 09 05:27:38 <oil> NOOOOO!
Nov 09 05:28:05 <halfhalo> infobot, no, halfhalo is a web programmer specializing in ruby on rails and node.js server side applications, and also is dabbling in sysadmin work as well.  He is also insane in the membrane, and drives way to fast on the streets.
Nov 09 05:28:06 <infobot> okay, halfhalo
Nov 09 05:28:37 * halfhalo DID set it... and SOMEONE overwrote it
Nov 09 05:28:52 <oil> isnt setting it so something you like against hte rules?
Nov 09 05:29:08 <rwhitby> OK, anyone know how to turn an OpenOffice presentation with simple reveal/hide animations in it into a PDF which retains the animations on click?
Nov 09 05:29:18 <dtzWill> scoutcamper: prelude is the font I believe.
Nov 09 05:29:28 <halfhalo> pdf's have animations?
Nov 09 05:29:39 * dtzWill didn't know pdfs could be animated
Nov 09 05:29:47 <dtzWill> oh do you mean splits animations into static slides?
Nov 09 05:29:58 <dtzWill> anyway i don't know, if that's not what 'export to pdf' does in OO :(
Nov 09 05:30:04 <halfhalo> oil: its ok if its true
Nov 09 05:31:25 <destinal> hmm, even after swapping out /usr/lib/egl with a 1.4.5 copy, still
Nov 09 05:31:25 <destinal> http://webos.pastebin.com/R40BMiNG
Nov 09 05:31:53 <destinal> ** (LunaSysMgr:3516): CRITICAL **:  FATAL:(TextureSharedCompositingWindow::TextureSharedCompositingWindow(int,  int, unsigned int, HostWindowDataOpenGLTextureShared*):66) gc
Nov 09 05:31:53 <destinal> --- SIGSEGV (Segmentation fault) @ 0 (0) ---
Nov 09 05:32:50 <tmzt> rwhitby: javascript, or, alternaitvely, you'll go insane trying
Nov 09 05:33:07 <tmzt> oh
Nov 09 05:33:14 <tmzt> click as in advance
Nov 09 05:35:08 <halfhalo> try google presentation?  maybe it could do something with it?
Nov 09 05:36:45 <destinal> better yet:  http://webos.pastebin.com/9q4rwKGj
Nov 09 05:38:07 * halfhalo attempts to macguyver something out of the items on his desk for kicks
Nov 09 05:38:58 <halfhalo> I have... a butane lighter, compressed air, needle(s), glass of ice, butane soldering iron, and a laptop.... hmmmm.....
Nov 09 05:39:18 <scoutcamper> who is the orriginal tux+prE+screwdriver artist?
Nov 09 05:40:14 <dtzWill> destinal: nice. keep'em coming, bunch of interesting stuff there :)
Nov 09 05:40:37 <tmzt> so luna isQt based now? or am I reading that wrong
Nov 09 05:40:51 <rwhitby> yep
Nov 09 05:40:54 <dtzWill> tmzt: you're reading it right :D
Nov 09 05:40:58 <rwhitby> looks accelerated too
Nov 09 05:41:07 <destinal> http://webos.pastebin.com/s1fzEpEd   I guess /usr/plugins/gfxdrivers/libqgfxegl.so   /usr/lib/egl/libeglsubdrv.so    /usr/lib/libpvr2d.so /usr/lib/libEGL.so" /usr/lib/libIMGegl.so /usr/lib/libGLESv2.so" are the most likely to be problematic
Nov 09 05:41:12 <dtzWill> HW egl render mode requested
Nov 09 05:41:16 <dtzWill> (from the pastebin)
Nov 09 05:41:17 <dtzWill> :D
Nov 09 05:41:50 <destinal> let me try overwriting a few
Nov 09 05:42:16 <tmzt> so maybe there share this stuff with meego
Nov 09 05:42:21 <tmzt> in one of the enerations
Nov 09 05:42:25 <tmzt> generations
Nov 09 05:43:01 <tmzt> what interface is exposed by the kernel for gles?
Nov 09 05:48:00 <halfhalo> so... how bout them continuums fake two screens?
Nov 09 05:49:24 <destinal> replaced  /usr/lib/libGLESv2.so usr/lib/libEGL.so /usr/lib/libIMGegl.so /usr/lib/egl/libeglsubdrv.so /usr/lib/libpvr2d.so  from 1.4.5, same result  :(     I don't have usr/plugins/gfxdrivers/libqgfxegl.so in 1.4.5 so no help there
Nov 09 05:51:17 <tmzt> find the nULL find the prize?
Nov 09 05:51:39 <tmzt> you should be able to get that from qt, I don't think it's palm specific
Nov 09 05:52:12 <dtzWill> lots of missing files from the logs it seems? :/
Nov 09 05:52:47 <destinal> yeah, lots of staging related things, looks like it wasn't cleaned up for release
Nov 09 05:52:56 <destinal> open("/home/reviewdaemon/projects/nova/oe/BUILD-roadrunner/staging/arm-none-linux-gnueabi/lib/fast-mult/libcjson.so",  O_RDONLY) = -1 ENOENT (No such file or directory)
Nov 09 05:52:56 <destinal> etc
Nov 09 05:53:07 <dtzWill> well but also things that look less staging-y
Nov 09 05:54:04 <dtzWill> com.palm.ftp? hehe
Nov 09 05:54:27 <halfhalo> com.palm.launchthenukes
Nov 09 05:55:15 <destinal> etc/prefs/properties/buildName  usr/share/locale/locale.alias usr/lib/charset.alias usr/lib/gconv/gconv-modules.cache /etc/palm/luna-platform.conf  etc/xdg/Trolltech.conf usr/plugins/codecs/ /tmp/qtembedded-0/QtEmbedded-0 /usr/share/fonts/fontdir
Nov 09 05:55:41 <tmzt> they aren't missing, there's no fast-mult version
Nov 09 05:56:04 <dtzWill> com.palm.app.vpn :)
Nov 09 05:56:30 <destinal> yeah, ipsec and cisco anyconnect clients  IIRC
Nov 09 05:57:39 <destinal> so what is the compositingwindow issue I wonder.
Nov 09 05:57:52 <destinal> comes right after accessing shared memory
Nov 09 05:58:04 <tmzt> destinal: if you can attach gdb look for a NULL
Nov 09 05:58:15 <tmzt> in one of the data structures it uses
Nov 09 05:58:44 <dtzWill> access("/var/preferences/com.palm.bus/public_bus_default", R_OK) = -1 ENOENT (No such file or directory)
Nov 09 05:58:48 <destinal> hmm, where's that optware bootstrap..
Nov 09 05:58:58 <dtzWill> happens a /lot/ and doesn't sound good :(
Nov 09 05:59:21 <tmzt> they replaced dbus again?
Nov 09 05:59:45 <tmzt> dude, strace always has missing files that seems to be normal/acceptible in every desktop system I've seen
Nov 09 06:00:07 <dtzWill> tmzt: okay
Nov 09 06:00:12 <tmzt> maybe it should have stat'd first, but it doesn't make much difference
Nov 09 06:00:51 <destinal> http://gitorious.org/webos-internals/bootstrap/blobs/raw/master/optware-bootstrap.sh ?
Nov 09 06:01:21 <destinal> oh, heh, reverse tethering first is important
Nov 09 06:01:25 <halfhalo>  /me goes to buy an oil based popcorn maker
Nov 09 06:01:56 <dtzWill> from your strace destinal: https://pastee.org/8nc7y
Nov 09 06:02:28 <destinal> dtzWill: oof, yeah that doesn't sound good
Nov 09 06:02:31 <dtzWill> seems like first real error before things go south.
Nov 09 06:02:57 <dtzWill> that's on ~line 3200, doesn't segfault until ~5230
Nov 09 06:03:38 <destinal> notice "Available:" is empty for displays
Nov 09 06:03:41 <destinal> also not good
Nov 09 06:03:51 <dtzWill> that isn't good :(
Nov 09 06:04:08 <dtzWill> also badness on the ioctl to pvrsrvkm :/
Nov 09 06:04:32 <tmzt> svr?
Nov 09 06:04:40 <destinal> yeah, a lot of it must have to do with the ioctls and our kernel probably doesn't have that device
Nov 09 06:05:07 <dtzWill> open("/dev/pvrsrvkm", O_RDWR)           = 43
Nov 09 06:05:21 <tmzt> can you compile a simpler program against it?
Nov 09 06:06:10 <destinal> dtzWill:  is that a good result?  the driver does in fact seem to exist for me, I wonder if it's just changed between 1.4.5 and 2.0 kernels
Nov 09 06:06:21 <destinal> s/driver/device/
Nov 09 06:06:21 <infobot> destinal meant: dtzWill:  is that a good result?  the device does in fact seem to exist for me, I wonder if it's just changed between 1.4.5 and 2.0 kernels
Nov 09 06:06:54 <destinal> time to check out kernel source
Nov 09 06:07:25 <dtzWill> destinal: no i was responding to tmzt that it was srv not svr--not that i know what that is exactly (other than some device for graphics, but svr vs srv? idk the details)
Nov 09 06:07:46 <dtzWill> destinal: it opens it and the first immediate ioctl fails then we get a pretty print telling egl failed to create its surface.
Nov 09 06:08:22 <dtzWill> looks like it exists, but that yeah the driver changed or maybe the pre 2 has different hardware that our userspace was built for...idk
Nov 09 06:08:32 <destinal> so pretty clearly it's trying to talk to the powervr and the version in our kernel has no idea what it's asking for
Nov 09 06:08:54 <dtzWill> shame that's the one piece that we don't have the source for :(
Nov 09 06:09:00 <destinal> not a good sign of us being able to do much without kernel source
Nov 09 06:09:16 <dtzWill> oh /do/ we have the source for it? thought it was proprietary.
Nov 09 06:09:25 <tmzt> the kernel side
Nov 09 06:09:33 <tmzt> it's the use side that's closed I think
Nov 09 06:09:52 <dtzWill> tmzt: ah, okay. ty.
Nov 09 06:09:52 <destinal> well even so we don't have 2.0.0 yet to even reverse the binary module
Nov 09 06:10:10 <dtzWill> errr the doctor doesn't contain the newer module?
Nov 09 06:10:11 <destinal> palm hasn't posted source for any 2.0.0 stuff yet
Nov 09 06:10:14 <dtzWill> yeah :/
Nov 09 06:10:34 <destinal> well yeah we could in fact reverse the arm binary.. hmm
Nov 09 06:11:18 <destinal> but if that source code release would happen already life would be much easier :)
Nov 09 06:12:42 <destinal> anyone have any idea where to find this device in the kernel patch?
Nov 09 06:13:53 <tmzt> drivers char or drivers video gpu
Nov 09 06:14:14 <tmzt> why not just updated qt?
Nov 09 06:14:26 <destinal> tmzt: huh?
Nov 09 06:14:29 <dtzWill> http://www.gp32x.com/board/index.php?/topic/49415-caprice32-opengl-es-2-0/
Nov 09 06:14:32 <tmzt> and patch it to work with the old pixel format
Nov 09 06:14:56 <dtzWill> same ioctl, fails similarly, and they claim it's a mismatch between the powervr module and the driver versions
Nov 09 06:14:57 <tmzt> as long as the c++ abi matches
Nov 09 06:15:10 <destinal> dtzWill: nice find
Nov 09 06:15:32 <dtzWill> are we using that driver from 2.0.0 or from 1.4.5?
Nov 09 06:15:38 <destinal> it doesn't exist in 1.4.5
Nov 09 06:15:42 <dtzWill> by driver i mean the libar...... ah.
Nov 09 06:15:44 <dtzWill> well, okay :).
Nov 09 06:16:25 <destinal> dtzWill: if it's   usr/plugins/gfxdrivers/libqgfxegl.so
Nov 09 06:16:29 <tmzt> I had this problem too the other day.
Nov 09 06:16:29 <tmzt> It is a problem with the driver & powervr module not matching.
Nov 09 06:16:30 <tmzt> If you are using version .06 of the gles driver revert the module back to version 1397 and that should fix it, it did for m
Nov 09 06:16:32 <halfhalo> totally going to target before class tomorrow
Nov 09 06:16:49 <dtzWill> tmzt: yeppers :).
Nov 09 06:16:50 <SineOt> hah, why?
Nov 09 06:17:01 <halfhalo> http://www.target.com/Stir-Crazy-Popcorn-Popper-qt/dp/B00004RC6R/ref=sr_1_1?ie=UTF8&searchView=grid5&keywords=popcorn&fromGsearch=true&sr=1-1&qid=1289282742&rh=&searchRank=target104545&id=Stir%20Crazy%20Popcorn%20Popper%20qt&node=1038576|1287991011&searchSize=30&searchPage=1&searchNodeID=1038576|1287991011&searchBinNameList=subjectbin,price,target_com_primary_color-bin,target_com_size-bin,target_com_brand-bin&frombrowse=0 shall be mine fo
Nov 09 06:17:10 <destinal> halfhalo: lol
Nov 09 06:17:26 <destinal> dtzWill: I replaced all the gl  and es libs
Nov 09 06:17:27 <SineOt> haha
Nov 09 06:17:55 <destinal> I couldn't find that ibqgfxegl.so so it must not exist, but the others..
Nov 09 06:18:27 <destinal> s/ibqgfxegl/libqgfxegl/
Nov 09 06:18:28 <halfhalo> wait, wrong one, that one is for home
Nov 09 06:18:28 <infobot> destinal meant: I couldn't find that libqgfxegl.so so it must not exist, but the others..
Nov 09 06:19:13 <destinal> let me confirm the issue is the same before I did that :)
Nov 09 06:19:22 <dtzWill> destinal: sure thing.
Nov 09 06:20:21 <dtzWill> http://www.beagleboard.org/irclogs/index.php?date=2010-09-30 also mentions that ioctl and our issue, but don't see anything immediately useful.
Nov 09 06:20:39 <dtzWill> seems to be version mismatch as well. might be worth poking into that channel at some point, shrug
Nov 09 06:21:42 <destinal> well I may have created it by overwriting these ones:  usr/lib/libGLESv2.so usr/lib/libEGL.so usr/lib/libIMGegl.so usr/lib/egl/libeglsubdrv.so usr/lib/libpvr2d.so  but the error stayed the same, let me roll back to 2.0.0 stock libs
Nov 09 06:22:03 <dtzWill> destinal: sounds good, ty :).
Nov 09 06:27:37 <destinal> dtzWill: http://webos.pastebin.com/VaEL4mWG
Nov 09 06:27:45 <destinal> with 2.0.0 stock libs again
Nov 09 06:28:24 <dtzWill> destinal: same error, hmm.
Nov 09 06:28:57 <destinal> yeah, has to be something going on with the kernel driver being different
Nov 09 06:29:15 <destinal> I wonder if we can just grab that one module
Nov 09 06:29:31 <destinal> (if it's a module and not monolithic)
Nov 09 06:30:44 <destinal> /lib/modules/tisgxgfx/pvrsrvkm.ko
Nov 09 06:31:20 <destinal> fortunately I think all the version symbols match up
Nov 09 06:31:44 <dtzWill> destinal: if that works it'd be EPIC hahah
Nov 09 06:31:58 <dtzWill> destinal: j/w how'd you check if they matched up?
Nov 09 06:32:24 <destinal> dtzWill: actually I don't know for sure, can you check a stock pre uname -a?
Nov 09 06:32:26 <destinal> Linux roadrunner 2.6.24-palm-joplin-3430 #1 281.1.23.2 armv7l unknown
Nov 09 06:32:58 <dtzWill> errr i have
Nov 09 06:32:59 <dtzWill> Linux palm-webos-device 2.6.24-palm-joplin-3430 #2 Thu Oct 21 00:50:32 UTC 2010 armv7l unknown
Nov 09 06:33:02 <destinal> err nevermind that's the stock pre 1.4.5 kernel, but the .20 has the same name
Nov 09 06:33:09 <dtzWill> but that's using uberkernel lol :P
Nov 09 06:33:14 <dtzWill> okay.
Nov 09 06:33:26 <dtzWill> well i mean what's the worst that could happen, try it! O:)
Nov 09 06:33:31 <destinal> kk
Nov 09 06:34:18 <destinal> wonder if I should replace all 3 tisgxgfx modules
Nov 09 06:35:03 <dtzWill> not sure, i'd say just play with it :/.
Nov 09 06:35:06 * dtzWill jealous :)
Nov 09 06:36:08 <destinal> ok here goes nothing
Nov 09 06:37:00 <destinal> kernel booted at least, I can novaterm
Nov 09 06:37:12 <dtzWill> so either it rejected the module or loaded it :)
Nov 09 06:37:18 <dtzWill> quick, check which! ;)
Nov 09 06:37:38 <destinal> stopping LunaSysMgr and running by hand
Nov 09 06:38:10 <dtzWill> destinal: gl! :)
Nov 09 06:38:34 <destinal> ** (LunaSysMgr:2976): CRITICAL **: FATAL:(TextureSharedCompositingWindow::TextureSharedCompositingWindow(int, int, unsigned int, HostWindowDataOpenGLTextureShared*):66) gc
Nov 09 06:38:34 <destinal> Segmentation fault
Nov 09 06:38:44 <dtzWill> bummer.
Nov 09 06:39:59 <dtzWill> the module...loaded, then?
Nov 09 06:40:08 <destinal> nope, appears not
Nov 09 06:40:10 <destinal> open("/dev/pvrsrvkm", O_RDWR)           = -1 ENOENT (No such file or directory)
Nov 09 06:40:14 <dtzWill> ah, okay.
Nov 09 06:40:30 <dtzWill> when you try to load it, does it complain?
Nov 09 06:40:39 <dtzWill> maybe it's some VERMAGIC stuff that you can try overriding O:)
Nov 09 06:41:27 <destinal> FATAL: Error inserting pvrsrvkm (/lib/modules/2.6.24-palm-joplin-3430/tisgxgfx/pvrsrvkm.ko): Unknown symbol in module, or unknown parameter (see dmesg)
Nov 09 06:41:36 <destinal> [  291.940000] pvrsrvkm: disagrees about version of symbol dump_stack
Nov 09 06:41:51 <dtzWill> of all the symbols to have conflicts on.... lol
Nov 09 06:42:19 <destinal> [  291.940000] pvrsrvkm: Unknown symbol contig_page_data
Nov 09 06:42:19 <destinal> [  291.940000] pvrsrvkm: Unknown symbol __alloc_pages
Nov 09 06:42:28 <destinal> lots of drivers are doing that tho for some reason
Nov 09 06:43:25 <dtzWill> yeah, source sure would help here.
Nov 09 06:45:06 <destinal> no idea, there's references to __alloc_pages  in both 1.4.5 and 2.0 vers too
Nov 09 06:45:15 <dtzWill> interesting:
Nov 09 06:45:23 <dtzWill> modinfo the 2.0 has:
Nov 09 06:45:37 <dtzWill> srcversion:     533BB7E5866E52F63B9ACCB
Nov 09 06:45:40 <dtzWill> 1.4.5 has:
Nov 09 06:45:45 <dtzWill> srcversion:     480797BEFAF67A817EC3DDC
Nov 09 06:46:09 <dtzWill> http://www.google.com/search?q=533BB7E5866E52F63B9ACCB etc
Nov 09 06:46:37 <dtzWill> including sr71 haha
Nov 09 06:46:47 <destinal> ?
Nov 09 06:46:48 <dtzWill> quick someone swap out the kernel with sr71
Nov 09 06:46:56 <destinal> sr71?
Nov 09 06:46:58 <dtzWill> destinal: http://forums.precentral.net/webos-internals/257347-sr71-blackbird-f105-thunderchief-testing-kernels-20.html#post2612778
Nov 09 06:47:06 <dtzWill> one of the experimental kernels
Nov 09 06:47:32 <dtzWill> apparently backported the piece that results in a version of that module that matches the srcversion hash on the 2.0 module
Nov 09 06:47:55 <destinal> so you're saying that kernel may satisfy the version #'s it wants?
Nov 09 06:48:21 <dtzWill> destinal: or that using that kernel and the modules from it might make the driver happy.
Nov 09 06:48:26 <dtzWill> all of this is a big 'maybe'
Nov 09 06:49:00 <dtzWill> curiously that srcversion in that post is for the wlan driver however. but presumably it's a hash and can't imagine we're looking at a collision :)
Nov 09 06:50:52 <destinal> dtzWill: I only see a uImage in http://forums.precentral.net/webos-internals/257347-sr71-blackbird-f105-thunderchief-testing-kernels.html
Nov 09 06:52:47 <dtzWill> surely the source for taht is in our git somewhere.
Nov 09 06:52:53 <destinal> dtzWill: wait, could that be because it's a binary module from palm, but how could a wifi module from a while back and this pvr version be from the same tree?
Nov 09 06:53:27 <dtzWill> destinal: that kernel in particular was forward porting to 2.6.24 it looks lke
Nov 09 06:53:42 <dtzWill> oh wait, they're both 2.6.24.... hmm.
Nov 09 06:54:31 <dtzWill> destinal: I don't know, sorry.
Nov 09 06:54:38 <dtzWill> lots of thinking out loud here :)
Nov 09 06:55:32 <destinal> /home/reviewdaemon/projects/nova/oe/BUILD-roadrunner/work/tisgxgfx-roadrunner-L 23.17M3-38/submissions/38/src/eurasia_km/services4/srvkm/env/linux/kbuild/../../../../srvkm/env/linux/osfunc.c
Nov 09 06:55:49 <dtzWill> what's that from?
Nov 09 06:56:10 <destinal> strings pvrsrvkm.ko  on the 2.0 ver
Nov 09 06:57:17 <destinal> http://linux.derkeiler.com/Mailing-Lists/Kernel/2007-02/msg03381.html
Nov 09 06:57:17 <destinal> Sure, you simply remove the information from the module, then the kernel will load it.  To strip modversions, remove the __versions section.  To strip the version magic, it's easiest to remove the entire .modinfo section.  You can use objcopy for this...
Nov 09 06:57:38 <destinal> but I don't know for sure if it's failing versions, wouldn't it give an error saying that instead?
Nov 09 06:58:13 <dtzWill> destinal: it's failing versions on dump_stack, you said the other drivers had the __alloc_pages and loaded anyway?
Nov 09 06:58:24 <dtzWill> *had the same errors with unknown symbol __alloc_pages?
Nov 09 06:59:21 <destinal> how can you clear dmesg?
Nov 09 07:00:22 <destinal> I don't think the 145 version gives any errors
Nov 09 07:00:32 <destinal> just the bad one
Nov 09 07:00:39 <destinal> but other modules were giving symbol errors
Nov 09 07:01:53 <ashi_todd> I don't think you need to clear dmesg...
Nov 09 07:02:19 <destinal> ashi_todd: I checked the numeric value next to the error to make sure it wasn't a duplicate, it wasn't, just old error
Nov 09 07:02:25 <ashi_todd> ah
Nov 09 07:02:26 <destinal> so proper module is fine it seems
Nov 09 07:02:42 <ashi_todd> well, good luck, and good night
Nov 09 07:03:05 <destinal> yeah, it's version conflicts
Nov 09 07:03:12 <destinal> now if I can only figure out how to strip it
Nov 09 07:03:23 <destinal> ashi_todd: thanks
Nov 09 07:04:20 <dtzWill> destinal:
Nov 09 07:04:37 <dtzWill> arm-none-linux-gnueabi-objcopy --remove-secton __versions oldmodule.ko newmodule.ko
Nov 09 07:04:43 <dtzWill> (just tried it
Nov 09 07:04:49 <dtzWill> (made the module load on my normal pre)
Nov 09 07:04:54 <dtzWill> (gulp)
Nov 09 07:05:16 <destinal> dtzWill: nice
Nov 09 07:05:24 * destinal tries it
Nov 09 07:05:28 <dtzWill> but on my pre it broke GL as a result
Nov 09 07:05:38 <dtzWill> but i don't have the new libraries from 2.0
Nov 09 07:05:42 <dtzWill> so...cross your fingers ;)
Nov 09 07:07:52 <destinal> find / -name arm-none-linux-gnueabi-objcopy
Nov 09 07:07:54 <destinal> nothing
Nov 09 07:08:09 <dtzWill> destinal: errr that'll be objcopy on YOUR machine
Nov 09 07:08:23 <destinal> ah, but I need the toolchain
Nov 09 07:08:24 <dtzWill> presuming you have some codesourcery toolchain installed, and on your path
Nov 09 07:08:25 <destinal> sigh
Nov 09 07:08:35 <dtzWill> i'm not sure if the fileformat is such that you can't use objcopy
Nov 09 07:08:46 <destinal> can you just ompload me the one you made?  :)
Nov 09 07:08:53 <dtzWill> destinal: sure
Nov 09 07:10:21 <destinal> probably should do all three in tisgxgfx  and tar them
Nov 09 07:13:39 <dtzWill> http://ompldr.org/vNjNuYg
Nov 09 07:17:07 <destinal> seems to be running
Nov 09 07:17:13 <destinal> trying luna
Nov 09 07:17:57 <destinal> hmm, segfault but let me try one more thing
Nov 09 07:18:47 <dtzWill> bah, i was hopeful we'd at least change the error :(
Nov 09 07:18:56 <dtzWill> or cause some crazy behavior ;)
Nov 09 07:19:19 <destinal> actually, yeah, we did
Nov 09 07:19:31 <destinal> sec, will pastebin
Nov 09 07:19:38 <dtzWill> destinal: great, ty.
Nov 09 07:20:39 <destinal> weird lunaservicebus related things now
Nov 09 07:20:57 <destinal> copying across vms to pastebin, sifgh
Nov 09 07:21:52 <dtzWill> destinal: that sounds like an improvement!!!! :)
Nov 09 07:23:14 <destinal> oops, no, tired and ran it while it was already trying to start as service :(
Nov 09 07:23:16 <destinal> sec
Nov 09 07:26:41 <destinal> http://webos.pastebin.com/XAxdFdeh
Nov 09 07:26:44 <destinal> sorry about the confusion
Nov 09 07:27:11 <destinal> fcntl64(43, F_SETFD, FD_CLOEXEC)        = 0
Nov 09 07:27:11 <destinal> ioctl(43, 0xc01c670c, 0x9ed81fd4)       = -1 EFAULT (Bad address)
Nov 09 07:27:11 <destinal> close(43)
Nov 09 07:27:12 <destinal> sigh
Nov 09 07:27:17 <dtzWill>  haha dammit :)
Nov 09 07:28:52 <destinal> dtzWill: the .stripped files were the right ones, correct? definitely stripped 2.0 versions?
Nov 09 07:29:09 <destinal> I really thought this would do something -- oh well
Nov 09 07:30:39 <dtzWill> destinal: yep, definitely stripped 2.0 versions.
Nov 09 07:31:32 <destinal> I still don't get why I get this:
Nov 09 07:31:33 <destinal> [  377.690000] bridgedriver: disagrees about version of symbol dump_stack
Nov 09 07:31:42 <destinal> puff mentioned that earlier too
Nov 09 07:32:00 <destinal> wonder if it could be related to our issues
Nov 09 07:34:01 <dtzWill> misc googlings: http://webos.pastebin.com/bNXxySSJ
Nov 09 07:34:50 <dtzWill> suggests mismatch between powervr and the gles libs? which version of those are we using?
Nov 09 07:35:37 <destinal> dtzWill: can you lsmod on your pre (if it's working)
Nov 09 07:36:06 <destinal> nevermind, LOTS of modules that aren't loading on mine
Nov 09 07:36:13 <destinal> and what's up with the version number problem?
Nov 09 07:36:17 <destinal> for bridgedriver
Nov 09 07:36:51 <dtzWill> destinal: http://webos.pastebin.com/5rKBTQKq
Nov 09 07:37:02 <dtzWill> no idea re:bridgedriver
Nov 09 07:37:08 <destinal> yeah it's probably my issue
Nov 09 07:37:56 <destinal> I have an idea..
Nov 09 07:38:06 <destinal> let me get that toolchain
Nov 09 07:39:13 <destinal> I bet it's those other drivers failing
Nov 09 07:39:53 <destinal> which honestly probably means something wonky is happening with metadoctor
Nov 09 07:40:08 <destinal> and the wrong files are there
Nov 09 07:40:35 <destinal> because my lsmod is one line
Nov 09 07:40:36 <destinal> pvrsrvkm              148092  1 omaplfb
Nov 09 07:41:49 <dtzWill> destinal: [   31.340000] bridgedriver: no version for "struct_module" found: kernel tainted.
Nov 09 07:42:02 <dtzWill> destinal: from my 1.4.5 pre, fwiw (running UK, but still)
Nov 09 07:42:19 <dtzWill> destinal: well but maybe the others are loaded as it continues to boot up?
Nov 09 07:42:39 <destinal> nope, I think it's monolithic enough to get most stuff running without
Nov 09 07:42:42 <dtzWill> try modprobe'ing sd8xxx and whatnot, or are you seeing dmesg fails?
Nov 09 07:43:42 <destinal> uh it just crashed my kernel
Nov 09 07:43:44 <destinal> doing that
Nov 09 07:43:51 <dtzWill> >_> <_< wasn't me
Nov 09 07:44:02 <destinal> yeah I think /lib/modules is hosed
Nov 09 07:44:03 <destinal> rebuilding
Nov 09 07:44:08 <destinal> err copying
Nov 09 07:45:02 <destinal> I think rwhitby assumed just cause it booted modules were fine
Nov 09 07:45:06 <destinal> but it's not
Nov 09 07:49:40 <destinal> oh well, recopied it and it still fails.  let me try the modules directory from a sprint pre for the hell of it because I don't get why modprobing sd8xxx crashes my system
Nov 09 07:50:51 <dtzWill> destinal: sounds good, and gl.
Nov 09 07:50:58 <destinal> hmm, bridgedriver doesn't fail anymore though
Nov 09 07:50:58 <destinal> hmm
Nov 09 07:51:11 <dtzWill> yeah since we're running the 1.4.5 kernel, sd8xxx shouldn't crash things.... you'd think! heh
Nov 09 07:54:13 <destinal> keeps crashing it. very odd.  I could load the rest of those and no more version errors
Nov 09 07:57:11 <destinal> dtzWill: oh wait, maybe it's because the driver and the firmware file don't match
Nov 09 07:57:35 <destinal> new rootfs, new firmware
Nov 09 07:58:03 <destinal> at least I think they use one
Nov 09 07:58:23 <dtzWill> bah. the firmware of which?
Nov 09 07:58:30 <destinal> of the wifi chip
Nov 09 07:58:44 <dtzWill> oh oh. that could be. is that loaded to the device dynamically?
Nov 09 07:58:58 <destinal> I think so but don't remember for sure
Nov 09 07:59:12 <dtzWill> destinal: /lib/firmware
Nov 09 07:59:28 <destinal> ok let me try 145's /lib/firmware
Nov 09 07:59:29 <dtzWill> destinal: looks lke you're right, and the rootfs contains them.. although haven't diffed against old
Nov 09 07:59:34 <dtzWill> yeah, that seems reasonable.
Nov 09 07:59:42 <dtzWill> more like 'pre 1's firwmare vs the new hardware
Nov 09 07:59:53 <destinal> right, that thing you said
Nov 09 08:00:00 <destinal> :)
Nov 09 08:00:27 <dtzWill> destinal: fwiw they are different :)
Nov 09 08:00:35 <dtzWill> (just diffed, but no surprises there)
Nov 09 08:06:42 <destinal> sd8xxx still crashes the kernel and luna still hates compositing
Nov 09 08:06:59 <dtzWill> destinal: :(. well it was a good call though
Nov 09 08:07:00 <destinal> even after that.  but probably you need that dir for the camera anyway
Nov 09 08:07:07 <dtzWill> yeah seems like a good thing to do anyway
Nov 09 08:07:47 <destinal> I wonder if a file in rootfs defines what firmware too.. sigh
Nov 09 08:08:50 <destinal> let me try one more thing
Nov 09 08:13:56 <destinal> hmm, can't get the 2.0 stripped bc_example.ko to work, crashes.  that may be needed, as well
Nov 09 08:16:58 <dtzWill> hmm I can't get it to work either.
Nov 09 08:17:11 <dtzWill> [ 4173.110000] bc_example: Unknown symbol PVRGetBufferClassJTable
Nov 09 08:17:11 <dtzWill> [ 4193.310000] bc_example: BC_Example_ModInit: can't init device
Nov 09 08:17:15 <dtzWill> that what you're seeing?
Nov 09 08:17:26 <destinal> yeah, look for agree on version before that
Nov 09 08:17:42 <destinal> just in case
Nov 09 08:18:09 <dtzWill> destinal: nope
Nov 09 08:18:16 <destinal> ok, interesting
Nov 09 08:18:21 <destinal> so where might it exist?
Nov 09 08:21:28 <destinal> dtzWill: hmm, I don't get Unknown symbol
Nov 09 08:21:33 <destinal> dtzWill: I do get [  154.970000] bc_example: BC_Example_ModInit: can't init device
Nov 09 08:21:40 <dtzWill> interesting.
Nov 09 08:22:01 <destinal> coiuld be because i have the 2.0 pvrsrvkm and  omaplfb
Nov 09 08:22:54 <destinal> can you try removing and moving/ renaming that module to see what luna does without it?
Nov 09 08:23:09 <dtzWill> hmm /etc/init.d/rc.pvr seems interesting
Nov 09 08:24:02 <destinal> yes, that it does
Nov 09 08:25:12 <dtzWill> destinal: are we using the libpvr stuff from 1.4.5 too? O:)
Nov 09 08:25:16 <destinal> /usr/bin/pvrsrvinit  || echo failed
Nov 09 08:25:17 <destinal> ouch
Nov 09 08:25:26 <destinal> failed
Nov 09 08:26:13 <dtzWill> well pvrsrvinit changed in 1.4.5->2.0.0 as well fwiw
Nov 09 08:27:10 <destinal> can you run it on your pre to see if it only runs once per boot or something?
Nov 09 08:27:15 <destinal> check result code?
Nov 09 08:27:25 <dtzWill> yeah it fails on mine when i run it again :)
Nov 09 08:27:29 <destinal> ah ok
Nov 09 08:27:30 <dtzWill> result code 1
Nov 09 08:31:19 <[AD]Turbo> hi all
Nov 09 08:32:00 <destinal> dtzWill: wow
Nov 09 08:32:00 <destinal> -rwxr-xr-x    1 root     root        86065 Nov  9 01:25 pvrsrvinit
Nov 09 08:32:00 <destinal> -rwxr-xr-x    1 root     root         5748 Oct  9 01:29 pvrsrvinit.20
Nov 09 08:33:08 <destinal> dtzWill: and you know this system is a very complex combination of kernel and userland stuff talking, it's really easy for little things to be going wrong mixing the versions this way
Nov 09 08:33:18 <destinal> [AD]Turbo: hi
Nov 09 08:34:25 <destinal> dtzWill: tried with the 1.4.5 pvrsrvinit too, no dice
Nov 09 08:37:17 <destinal> and on that note I'm going to bed. maybe tomorrow I'll read back through these logs to see if any of this makes sense when more awake :)
Nov 09 08:37:36 <destinal> hopefully others have ideas as well and can duplicate some of this
Nov 09 08:38:01 <destinal> night
Nov 09 09:43:40 <dtzWill> destinal-sleep: sounds good and ty for your efforts :D. and for trying my ideas ;)
Nov 09 09:43:49 <dtzWill> ttyl :)
Nov 09 09:54:31 <tmzt> destinal-sleep: wifi driver sd8xxx?
Nov 09 13:19:13 <don8> I've come to the realization that you won't get people more envious than the people at precentral
Nov 09 13:19:30 <en0x> oO
Nov 09 13:19:54 <en0x> w00t w00t w00t? we?
Nov 09 13:19:54 <en0x> :S
Nov 09 13:20:45 <don8> you can't go two pages without hearing "super phone or android"
Nov 09 13:23:31 <don8> no matter how many times you tell them "I don't want to hear it" it's always the same story
Nov 09 13:28:44 <en0x> oh
Nov 09 13:28:47 <en0x> who cares :p
Nov 09 13:41:24 <don8> lol I know I just get frustrated hearing the same "experts" every day
Nov 09 15:09:52 <destinal> dtzWill: I wonder if Dr Frankenstein was ever frustrated like this trying to sew body parts together to make his creature work :)
Nov 09 15:11:54 <ka6sox> yes
Nov 09 15:14:29 <destinal> ka6sox: I was sure once we got the 2.0 pvr kernel module to work things would just start working (or at least get further)  but I wonder about that "example" module.  it's a bit more than an example methinks if the module is running on dtzWill's device, maybe it's the problem (or again maybe some userland initialization piece is)
Nov 09 15:15:58 <destinal> ka6sox: are the pvr drivers binary only?  I don't see them in Palm's kernel patch.
Nov 09 15:18:46 <destinal> I think in the spirit of the GPL binary drivers should be a license violation but I know that they're common practice these days and kind of a grey area (whether they're derivatives of the kernel or not, and even if not, whether plugging them in in-memory makes them part of the same running program as the kernel for GPL purposes)
Nov 09 15:20:45 <ka6sox> its entriely possible that there are other daemons that are broken too..and causing things like infinate loops.
Nov 09 15:34:56 <ka6sox> re: binary modules I'm against them personally but some functionality wouldn't be available without them.
Nov 09 15:57:39 <destinal> So an interesting comment on the precentral front page article is that A6 is the keyboard controller.  I want to confirm that but it makes a lot of sense.  chrisa said that windsor is dead (either a stillborn product or an internal only codename), and since we've seen broadway has a 320x400 resolution in web logs, I'd say:   mantaray = palmpad  and broadway = touchscreen phone or pda
Nov 09 15:59:55 <destinal> leaving stingray to be a next gen phone with keyboard, perhaps?
Nov 09 16:05:30 <destinal>  I suppose broadway could be a printer too but who's gonna browse the web on one?  ;)
Nov 09 16:06:06 * oil  
Nov 09 16:06:33 <destinal> heh
Nov 09 16:07:04 <drnull> destinal: but wasn't broadway supposed to be a pixi 2?  http://www.precentral.net/palm-broadway-pops-app-log-pixi-2-coming-well
Nov 09 16:08:05 <destinal> it's 320x400 which is why people say that, I don't think they know any more than that
Nov 09 16:09:20 <destinal> if A6 is a keyboard, I suppose pixi2 could have a different kind with a different controller, but that doesn't jube well with stingray having an A6 (again, *if* it is the keyboard, have to confirm that -- it's what was making me think it's touchscreen only)
Nov 09 16:09:35 <destinal> s/jube/jibe/
Nov 09 16:09:35 <infobot> destinal meant: if A6 is a keyboard, I suppose pixi2 could have a different kind with a different controller, but that doesn't jibe well with stingray having an A6 (again, *if* it is the keyboard, have to confirm that -- it's what was making me think it's touchscreen onl...
Nov 09 16:10:49 <destinal> but then again, why make a touchscreen only device with 320x400 resolution?
Nov 09 16:11:03 <destinal> doesn't seem that practical
Nov 09 16:13:46 <drnull> agreed.  I have a hard time reconciling (A6=keyboard) and (broadway=320x400) and (broadway != A6). Maybe the reported resolution was incorrect for the broadway, and it is actually a tablet.  Then the rest of this makes sense.
Nov 09 16:17:19 <destinal> drnull: true. it claims nova-blowfish and nova-barleywine in weblogs, and I don't know if we've confirmed if either one was ~1.4.5 or something larger on the 1.x.x  continuum and the other was 2.0, or if we're looking at totally internal versions that don't correspond to what we've seen.
Nov 09 16:17:55 <destinal> also barleywine is an awesome codename
Nov 09 16:20:35 <destinal> drnull: I have to think that 320x400 may have been a lie..  I don't think we'll see a pixi2 but I could be wrong
Nov 09 16:21:58 <drnull> Yeah, I wish metrix could let you break down their stats a bit more.  I show 3 downloads at webOS build 362 and 9 at webOS build 363.  And I see the following webOS version downloads: 2.0.0: 2, 2.0.1: 3, Nova-Barleywine: 6, Nova-Main: 1
Nov 09 16:22:34 <drnull> So depending on how you break that down, Barleywine is almost definitely 363 (which is the highest build number I show).
Nov 09 16:24:20 <drnull> and all my resolutions show 320x400 or 320x480, just for the record.
Nov 09 16:25:18 <destinal> yeah, since tablets are almost certainly being developed and tested now and used against real world sites for an early 2011 release, I think the resolutions are fake
Nov 09 16:25:39 <destinal> so broadway may be the first table
Nov 09 16:25:40 <destinal> t
Nov 09 16:28:51 <drnull> I've also not seen devices not named "Pixi" or "Pre" (well, other than the occasional Device/Emulator from my own testing).  So I guess they're faking those as well?
Nov 09 16:29:49 <drnull> I guess after the first slipup, they probably started faking all the meta-data that gets reported.
Nov 09 16:30:16 <destinal> heh, they used to show planned version numbers too, now codenames instead
Nov 09 16:30:33 <destinal> before you'd see 1.3 hitting you and know it was coming soon
Nov 09 16:31:03 <drnull> Well... we see 2.0.0 and 2.0.1 now, so I don't see how they've strayed from that?
Nov 09 16:31:23 <destinal> unless dev uses codenames and integration testing uses the version numbers
Nov 09 16:32:16 <destinal> probably depends on how close to a release things are too
Nov 09 16:33:00 <destinal> if it's already packaged up as 2.0.1 then that's what it'll be, but if it's purely R&D it'll be whatever name they give it
Nov 09 16:34:09 <drnull> gotcha.  And the R&D devices should really have metrix's server ip blacklisted.  :)
Nov 09 16:34:34 <destinal> sssshh ;)
Nov 09 16:43:20 <threethirty> howdy all, theoretically how hard would it be to complile all of the packages need to get a vnc server running on my pre
Nov 09 16:57:07 <dtzWill> threethirty: vnc server that.... such that you could control your phone remotely via vnc?
Nov 09 16:57:42 <dtzWill> threethirty: afaik no one's pieced that together yet, so it's definitely not 'just go here and do this'.
Nov 09 16:58:22 <dtzWill> getting a vnc server and pointing it at the device's framebuffer works (for non-3d games), and then maybe combine with the uinput stuff egaudet was doing iirc.
Nov 09 16:58:30 <dtzWill> could probably rig a vnc server that way. if you do let us know :D
Nov 09 16:59:49 <threethirty> dtzWill: will do
Nov 09 16:59:57 <threethirty> thanks
Nov 09 17:07:41 <oilsworkn> a wircer!
Nov 09 17:10:14 <destinal> dtzWill: I'm leaning towards doing vns server with framebuffer capture and a hidd plugin for input. (though it clearly neaeds more research)
Nov 09 17:10:25 <destinal> s/vns/vnc/
Nov 09 17:10:25 <infobot> destinal meant: dtzWill: I'm leaning towards doing vnc server with framebuffer capture and a hidd plugin for input. (though it clearly neaeds more research)
Nov 09 17:11:37 <destinal> maybe with hidd we could generate remote button pushes too and create a virtual control panel around the screen area
Nov 09 17:12:36 <destinal> actually function keys would be better still
Nov 09 17:12:50 <destinal> kind of like the emulator does with orientation etc
Nov 09 17:14:37 <destinal> dtzWill: are there any framebuffer capture programs that will make a movie capture of the pre's screen today?  wondering how goot the refresh rate could be
Nov 09 17:15:48 <threethirty> im glad you all like this idea because this discussion went over my head quickly :)
Nov 09 17:16:10 <oilsworkn> haha
Nov 09 17:16:15 <dtzWill> destinal: idk. I actually used a framebuffer vnc and it seemed reasonable
Nov 09 17:16:22 <destinal> threethirty: so, if you don't mind my asking, what would be the practical use of the thing?
Nov 09 17:16:23 <dtzWill> err by used i mean watched ;)
Nov 09 17:16:39 <destinal> threethirty: Ie why do you want it
Nov 09 17:18:16 <dtzWill> destinal: I don't know a way to record, but don't see why you couldn't
Nov 09 17:18:47 <threethirty> destinal: i need to keep an eye on my phone to see if my brother tries to get a hold of me durring class (he has a mental illness) but I would get in trouble for "playing with my phone in class". I have my laptop out anyway because I take notes on it (my handwritting is terrible)
Nov 09 17:19:00 <destinal> ah...
Nov 09 17:19:11 <destinal> that's kind of a cool use case
Nov 09 17:19:17 <oilsworkn> yeah it is
Nov 09 17:19:34 <oilsworkn> stealth phone usage
Nov 09 17:19:37 <oilsworkn> sneaky sneaky
Nov 09 17:22:14 <oilsworkn> you could also write a patch for the messaging app and have it forward your messages to your email and check it via the web
Nov 09 17:23:00 <oilsworkn> (though, im sure if you explained the situation to your teacher, they would let you use your phone for such purpose in class)
Nov 09 17:23:50 <oilsworkn> from my experience, not all teachers are completely unreasonable
Nov 09 17:40:33 <threethirty> oilsworkn: im in class with a lot of freshmen, i dont want to put my proffs in weird situations. kids these days..
Nov 09 17:44:55 <oilsworkn> yeah, other students doing the "well, threethirty gets to play with his phone during class!"
Nov 09 17:45:13 <threethirty> and they would :(
Nov 09 17:45:23 <oilsworkn> bastards
Nov 09 17:56:46 <dtzWill> destinal: just built ffmpeg with the patch here: http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/91238
Nov 09 17:57:13 <dtzWill> destinal: (using the first one, the followups are just arguing about licensing and a feature that never got implemented anyway)
Nov 09 17:57:21 <dtzWill> and it works! xD :D
Nov 09 18:05:21 <dtzWill> ~10fps fwiw
Nov 09 18:05:35 <dtzWill> but that's probably/possibly because it's encoding in software
Nov 09 18:05:46 <dtzWill> and so just the encoding is beasting the cpu
Nov 09 18:06:24 <dtzWill> test.mp4: http://ompldr.org/vNjN0dQ hehe
Nov 09 18:07:02 <dtzWill> hmm that didn't work :(.
Nov 09 18:07:26 <destinal> lol Sorry, we're still working on it.  Go grab some lunch or something.
Nov 09 18:07:38 <destinal> technical issues over there
Nov 09 18:08:22 <dtzWill> destinal: http://wdtz.org/files/fb_test.mp4
Nov 09 18:09:05 <dtzWill> with that, i'm off to get some work done, but i'll be in/out :)
Nov 09 18:13:12 <dtzWill> unfortunately that's not a good direction to go anyway :/ since in 2.0 they look to be using hw-accel'd qt
Nov 09 18:13:33 <dtzWill> suggesting even if we did put together a sweet vnc server, it'd be deprecated rather soon :(
Nov 09 18:13:46 * oilsworkn will hold out on 1.4.5 FOREVER
Nov 09 18:14:07 <dtzWill> yay we'll have 1 user
Nov 09 18:49:05 <ice8lue> hey erveryone
Nov 09 18:51:32 <ka6sox> Got a response from opensourcequestions@palm.com: "Thank you for reporting the error, we will correct it. There is a typographical problem, the kernel is actually 2.6.24 and available from the link immediately below."
Nov 09 18:51:40 <ka6sox> so NO 2.6.25 for Pixi.
Nov 09 18:51:56 <oilsworkn> ;(
Nov 09 18:54:13 <ice8lue> anyone knows about the state of the 'frankendoctor'?
Nov 09 18:54:41 <ice8lue> did some tryes myself but cant't go further than a loop of the palm screen...at least no crash ;)
Nov 09 18:54:52 <oilsworkn> afaik the townfolk rose up and grabbed their pitchforks and torches and chased him out of town
Nov 09 18:55:30 <halfhalo_T400> I made a killing with the stocks I bought in pitchfork manufacturers
Nov 09 18:56:01 <oilsworkn> in all seriousness, i think they hit a wall where they need the source for the 2.0 kernel to move forward
Nov 09 18:56:05 <oilsworkn> but i could be totally wrong
Nov 09 18:57:07 <ice8lue> ya that's what i expected :/
Nov 09 18:57:16 <ice8lue> well....I will keep on trying
Nov 09 18:57:29 <ka6sox> gl
Nov 09 18:57:59 <halfhalo_T400> I will give you this cookie if you suceed: o
Nov 09 18:58:17 <ice8lue> xD amazing^^ a whole cookie just for that?^^
Nov 09 18:58:48 <halfhalo_T400> yes
Nov 09 18:58:49 <oilsworkn> thats a small cookie
Nov 09 18:58:53 <oilsworkn> didn't even use teh capital o
Nov 09 18:59:08 <halfhalo_T400> no, because that makes it a deformed reject cookie
Nov 09 18:59:46 <threethirty> later on everyone class calls
Nov 09 19:03:24 <ice8lue> mh..now it crashed -.-
Nov 09 19:03:39 <halfhalo_T400> AS IT SHOULD
Nov 09 19:03:49 <ice8lue> not really...
Nov 09 19:03:50 <halfhalo_T400> Because I don't want to give up this cookie
Nov 09 19:04:06 <ice8lue> I never claimed that cookie^^
Nov 09 19:04:27 <dtzWill> ice8lue: might want to bounce through last night's logs and pastebins strace'ing luna when it dies.
Nov 09 19:04:53 <ice8lue> i'm already on, reading through them
Nov 09 19:04:57 <dtzWill> don't want to push you into the rut we might have fallen into, but on the other hand wanna make sure you have latest findings
Nov 09 19:04:59 <ice8lue> just a lot of stuff in there
Nov 09 19:05:02 <dtzWill> ice8lue: haha okay :).
Nov 09 19:05:10 <ice8lue> =) thx
Nov 09 19:05:30 * halfhalo_T400 is hungry
Nov 09 19:05:37 * halfhalo_T400 makes a samich
Nov 09 19:07:40 <halfhalo_T400> cd /usr/local/kitchen; ./configure --type="samich" --bread="white" --cheese="provalone"  --with-mustard --with-mayo --with-bacon --with-peperroni; make; make install
Nov 09 19:08:11 <dtzWill> i was with you until your kitchen installed the sandwich
Nov 09 19:08:30 <dtzWill> but that does sound delicious :)
Nov 09 19:08:58 <dtzWill> also autoconf has some really sweet features these days, i should look into upgrading...
Nov 09 19:09:04 <halfhalo_T400> the --manual-intake option is implicit
Nov 09 19:10:49 <ice8lue> replacing --with-mayo with --with-ketchup should kill some unneccessary temporary files
Nov 09 19:11:13 <ice8lue> just fills needed space on hd
Nov 09 19:15:49 <ice8lue> do the xorg port sources be open? what if we temporarely stop luna from being loaded but load xorg? should make bugtracing luna more simple
Nov 09 19:18:03 <dtzWill> ice8lue: I like the way you think. which port are you referring to?
Nov 09 19:18:21 <dtzWill> ice8lue: you should be able to run X against the fb device, assuming that's working.
Nov 09 19:18:37 <dtzWill> I believe it is, you could check with fbdraw if that's still shipped on the device
Nov 09 19:19:26 <ice8lue> mh need to search the forums... there was a port used with xterm
Nov 09 19:19:36 <ice8lue> to run a wm for example
Nov 09 19:20:00 <dtzWill> ice8lue: oh I did that :). that used SDL/GLESv2 which means you need a working luna.
Nov 09 19:20:18 <dtzWill> that made some things harder, but also meant it was nice and in a card so you dind't have to stop luna to use it (big plus)
Nov 09 19:20:57 <ice8lue> hmz...and it wouldn't bepossible (without harder work) to seperate that?
Nov 09 19:21:25 <dtzWill> ice8lue: I'd just find w/e X bit that uses the framebuffer and grab that.  There was work on that a while back I thought...
Nov 09 19:22:24 <dtzWill> ah seems that was done via debian chroot, although I can't find a directfb x (although I could've sworn that existed)
Nov 09 19:22:34 <dtzWill> but things like dfbterm should work....
Nov 09 19:22:37 <destinal> dtzWill: but would that actually test the egl stuff..
Nov 09 19:23:05 <dtzWill> destinal: not sure what you mean. I would fully expect directfb to work.
Nov 09 19:23:18 <dtzWill> destinal: if you have your franken'd phone handy can you just try scribbling the fb with fbdraw?
Nov 09 19:23:25 <destinal> dtzWill: I think he's talking about troubleshooting the issues we ran into last night
Nov 09 19:23:42 <dtzWill> don't remember the invocation offhand but you should be able to make a rectangle
Nov 09 19:23:48 <destinal> dtzWill: the shutdown screen works btw and that uses fb, right?
Nov 09 19:23:52 <dtzWill> and that'll test if the fb device 'works'
Nov 09 19:23:55 <dtzWill> oh i suppose so O:)
Nov 09 19:24:54 <destinal> do SDL games work without luna?  that would be maybe a better test
Nov 09 19:26:17 <ice8lue> thought supertux was a sdl game?
Nov 09 19:26:30 <destinal> halfhalo_T400: way to go,  now I'm hungry and it's all your fault ;)
Nov 09 19:27:04 * halfhalo_T400 has popcorn maker for making popcorn for lunch at his office
Nov 09 19:28:53 <destinal> by the way if anyone else has a victim pre (seperate from the one you actually use) I would love it if someone would sanity check / duplicate what I've done so far.   most of the hard part is already figured out except why we can't initialize a surface
Nov 09 19:30:00 <bsm> Thank you for your email. We are in the process of preparing the source code package and when ready they will be published on http://opensource.palm.com. Please watch the open source web site for availability. :)
Nov 09 19:30:32 <destinal> bsm: yeah, I'm starting to think that's a form reply ;)
Nov 09 19:30:42 <ice8lue> Oo really? without even having the doctors out? cool^^
Nov 09 19:31:00 <destinal> ice8lue: the 2.0 doctor is out for SFR, that's what we're playing with
Nov 09 19:31:29 <destinal> we're asking for the source code packages for the open source bits in it
Nov 09 19:31:52 <ice8lue> of course, but the pre/pixi kernel should be different from what is running on the pre 2, shouldnt it?
Nov 09 19:32:49 <destinal> ice8lue: pre and pre2 kernels should be very close
Nov 09 19:33:55 <ice8lue> oh, okay^^ thought there would be differences...
Nov 09 19:34:12 <ACap> my non-techie brother just got a Pixi minus (i know, right?) Any word on how soon the overclock will be ready for the public?
Nov 09 19:34:32 <destinal> pre2 is more like a small hardware upgrade for the pre, we've already loaded some of its kernel drivers successfully on the pre if that tells you anything
Nov 09 19:35:13 <destinal> think of it as (pre+)+
Nov 09 19:35:32 <ice8lue> really? which ones? ya i'm studying embedded technologys, i'm a little into that
Nov 09 19:36:24 <destinal> ice8lue: two of the three powervr modules last night when we got desperate.   :)
Nov 09 19:36:44 <destinal> (GPU)
Nov 09 19:37:19 <destinal> pre and pre2 use the same powervr integrated with their omap SoC it seems
Nov 09 19:37:56 <destinal> omap3430 abd omap3630 but GPU is apparently not one of the differences
Nov 09 19:38:00 <dtzWill> SDL shouldn't work, although again I suppose we can get a version of SDL that uses directfb.... maybe
Nov 09 19:38:03 <destinal> s/abd/and/
Nov 09 19:38:04 <infobot> destinal meant: omap3430 and omap3630 but GPU is apparently not one of the differences
Nov 09 19:38:26 <destinal> dtzWill: directfb's not the problem he's trying to troubleshoot though, right?
Nov 09 19:38:50 <dtzWill> i'm sorry I think there's a disconnect.
Nov 09 19:39:02 <dtzWill> what is the issue then?
Nov 09 19:39:10 <dtzWill> i thought we wanted to test if the fb worked at all separate from luna.
Nov 09 19:39:38 <dtzWill> :( sorry
Nov 09 19:39:41 <destinal> ah and i thought he was trying to run other egl test code
Nov 09 19:39:50 <ice8lue> will have a look on both omaps datasheets... let's see what our university got about that
Nov 09 19:40:47 <bsm> somehow I don't think thats worth it as palm HAS TO release the source some day
Nov 09 19:41:02 <ice8lue> definitely^^
Nov 09 19:41:15 <destinal> bsm: sure but the point of this effort is to get webos2.0 going before a month from now if possible
Nov 09 19:41:34 <bsm> yeah I know, I followed all threads
Nov 09 19:41:55 <destinal> bsm: and I think it's doable
Nov 09 19:42:47 <bsm> would've been nicer without driver problems
Nov 09 19:42:49 <ice8lue> destinal: me too, but the question is what to do with these efforts after it is possible via OTA update
Nov 09 19:43:14 <ice8lue> possible custom roms?
Nov 09 19:43:45 <destinal> ice8lue: you can always make your own custom doctor images with metadoctor
Nov 09 19:43:52 <bsm> yeah and roms for poor mexicans
Nov 09 19:44:38 <destinal> this specific example of such will just be on a wiki page somewhere for reference, after a 2.0 doctor comes out for castle it won't be necessary
Nov 09 19:45:59 * ice8lue hates datasheets
Nov 09 19:46:05 <destinal> also the features rwhitby has developed for making this beast work will stay in metadoctor in case we ever need to do it for an image only available for some other odd platform
Nov 09 19:46:19 <ice8lue> pages of assembler code...~~.~
Nov 09 19:46:36 <destinal> ah, yeah, SoC datasheets can be pretty huge
Nov 09 19:47:01 <ice8lue> we had one for a prism cpu....phew -.-
Nov 09 19:48:57 * halfhalo_T400 looks at asm code and goes to get the magnetized hammer
Nov 09 19:49:40 <ice8lue> halfhalo_T400: you'll love asm code once you work with it half the day^^
Nov 09 19:50:49 * halfhalo_T400 would go insame
Nov 09 19:50:55 <halfhalo_T400> INSAME!
Nov 09 19:51:54 <ice8lue> oh well...it's cool once you learn to think that way...coming from c/c++/c# didn't make it easy though.
Nov 09 19:52:42 * halfhalo_T400 goes crazy when ruby misbehaves, asm would be infinitly worse
Nov 09 19:57:20 <ice8lue> can't order the pre 2 in germany -.-
Nov 09 19:57:39 <ice8lue> hopefully O2 will get it soon...
Nov 09 19:58:11 <ice8lue> looking at the datasheets didn't really bring up something new though... will have a deeper look at that tomorrow
Nov 09 20:07:30 <destinal> well I guess frankendoctor just became redundant as of Friday
Nov 09 20:07:30 <destinal> http://www.precentral.net/palm-uk-sell-pre-2-sim-free-november-12th
Nov 09 20:08:51 <destinal> or wait, pre2, not webos2.0 for o2 pre's.  I'm obviously in need of more coffee..
Nov 09 20:09:29 <destinal> still maybe it's more likely source code will be released soon, now
Nov 09 20:11:09 <destinal> I guess o2's OTA update will be 12/6 --  http://www.precentral.net/webos-20-getting-rolled-out-uk-december-6th
Nov 09 20:12:41 <en0x> rwhitby: got pissed because some guy exchanged his bricked pre at sprint store
Nov 09 20:12:54 <en0x> heh ppl man ppl
Nov 09 20:12:55 <destinal> en0x: you did?
Nov 09 20:13:03 <en0x> nah
Nov 09 20:13:19 <en0x> and he closed the thread
Nov 09 20:13:29 <destinal> oh, he did, well the thread doesn't matter..
Nov 09 20:13:42 <en0x> but sayin
Nov 09 20:13:43 <ice8lue> but i do understand rod
Nov 09 20:13:53 <destinal> we're still working on it, I made some progress, see the logs..
Nov 09 20:14:01 <en0x> :)
Nov 09 20:14:05 <ice8lue> it was impossibly stupid by that one...
Nov 09 20:14:22 <en0x> it says clearly bla bla bla and he went and exchanged it
Nov 09 20:14:33 <en0x> that's how it is
Nov 09 20:15:18 <R_E> hi everyone
Nov 09 20:15:42 <ice8lue> yeah but it's idiotic to kill your pre KNOWING that you would without knowing how to make it back alive
Nov 09 20:15:59 <ice8lue> and than take it back to sprint, to somehow steal a new one
Nov 09 20:16:22 <en0x> PuffTheMagic: the psfreedom module works fine on uberkernel and 1.4.5 i'm using it ;D
Nov 09 20:16:41 <ice8lue> psfreedom?
Nov 09 20:16:57 <destinal> ice8lue: I wouldn't say steal a new one, he exchanged it and the old one is probably fine
Nov 09 20:17:05 <destinal> still, bad form
Nov 09 20:17:40 <en0x> ice8lue: ps3 exploit
Nov 09 20:17:50 <ice8lue> ah ok
Nov 09 20:18:06 <destinal> the thread clearly stated it would make your device unusable, don't do it if you wouldn't hit your pre with a hammer, don't go crying to palm / carrier etc
Nov 09 20:18:21 <destinal> pretty dumb of him
Nov 09 20:18:26 <ice8lue> yap...
Nov 09 20:19:09 <en0x> Nov 08 02:31:04 <webOSSUCKs> web os SUCKS
Nov 09 20:19:12 <en0x> oO
Nov 09 20:19:24 <ice8lue> troll...
Nov 09 20:19:32 * en0x is going to read this
Nov 09 20:22:39 <ice8lue> offtopic... is somebody able to help me making my outlook 2k7 stop asking me for my mail account password over and over again?
Nov 09 20:38:46 <Naomi> Can someone help a total irc noob?
Nov 09 20:39:17 <en0x> don't ask to ask just ask
Nov 09 20:39:46 <Naomi> Well, a game I play uses irc for help questions
Nov 09 20:39:54 <Naomi> they use darktree.net for their irc channel
Nov 09 20:40:17 <Naomi> I really know nothing about irc, so I'm not sure how to set it up to use on my phone, or whether it will work at all.
Nov 09 20:40:32 <Naomi> so i guess, how do I know if darktree has a qwebirc url?
Nov 09 20:43:11 <en0x> idk what qwebirc url is
Nov 09 20:43:18 <Naomi> me either
Nov 09 20:43:19 <Naomi> lol
Nov 09 20:44:13 <dtzWill> Naomi: install wirc and point it at darktree?
Nov 09 20:46:11 <Naomi> I thought that was what I had tried, but I just get errors
Nov 09 20:47:15 <Naomi> crap i'm sorry, I hit the wrong link and downloaded pound irc instead of wirc
Nov 09 20:47:18 * Naomi sighs
Nov 09 20:47:23 <dtzWill> Naomi: thought so :).
Nov 09 20:47:23 <Naomi> it's one of those days
Nov 09 20:47:48 <dtzWill> i cuoldn't remember what the other one was called but i knew it used some other way to speak to servers, required them to have some special protocol/interface
Nov 09 20:48:07 <Naomi> yeah, i'm removing it now
Nov 09 20:48:19 <Naomi> it didn't even work on the default
Nov 09 20:48:41 <Naomi> oh, what's fstab fixer?
Nov 09 20:48:50 * Naomi feels sheepishly newbish
Nov 09 20:49:07 <dtzWill> Naomi: you're fine :). fstab fixer fixes a bug that resulted in the 1.4.5 OTA update
Nov 09 20:49:18 <dtzWill> was a result /of/ that update, rather
Nov 09 20:49:27 <dtzWill> resulting /in/ applications being unable to write to the usb partition
Nov 09 20:49:30 <Naomi> so don't need it then
Nov 09 20:49:42 <dtzWill> Naomi: i have no idea, did you just get your device, and it came with 1.4.5?
Nov 09 20:49:45 <dtzWill> if so, then no.
Nov 09 20:49:53 <Naomi> bright red letters on a download page make me nervous lol
Nov 09 20:49:57 <dtzWill> haha then avoid it :).
Nov 09 20:49:58 <ice8lue> wow...got a palm boot up screen fading out and a crash afterwards^^ thats new =)
Nov 09 20:50:18 <dtzWill> Naomi: good instincts. wish others followed your lead...
Nov 09 20:50:19 <Naomi> no, I've had my palm from day one
Nov 09 20:50:28 <Naomi> I believe I downloaded 1.4.5 after the bug was caught though
Nov 09 20:50:34 <dtzWill> Naomi: it was never fixed.
Nov 09 20:50:38 <Naomi> crap
Nov 09 20:50:53 <dtzWill> Naomi: only gets 'fixed' if you doctor your phone to 1.4.5, fix it by hand, or install fstab fixer.
Nov 09 20:51:36 <Naomi> can't remember if I doctored before or after the update, so fstab fixer it is
Nov 09 20:51:44 * Naomi waits
Nov 09 20:52:24 <ice8lue> dtzWill: fb outside of luna doesn't seem to work =/
Nov 09 20:52:33 <dtzWill> ice8lue: no?? what'd you try? :(
Nov 09 20:53:34 <ice8lue> i was trying to disable luna starting at all. got it running so far i think
Nov 09 20:53:49 <ice8lue> but don't really know how to test fb
Nov 09 20:53:56 <ice8lue> at least no SDL^^
Nov 09 20:54:04 <dtzWill> ice8lue: should be a binary called 'fbdraw' on the device
Nov 09 20:54:52 <dtzWill> /usr/bin/fbdraw -fill 100 100 255 0 0 200 200
Nov 09 20:55:00 <dtzWill> should scribble a 200x200 rectangle onto your fb :)
Nov 09 20:55:29 <ice8lue> okay it does =)
Nov 09 20:55:41 <ice8lue> hrmpf crashed afterwards -.^
Nov 09 20:55:45 <Naomi> is that is?  just download fstab and luna restart? seems to easy
Nov 09 20:56:05 <dtzWill> er you need an alpha too, i just had RGB sorry
Nov 09 20:56:11 <dtzWill> but shouldn't crash :(
Nov 09 20:56:37 <ice8lue> well it did something but the screen turned of nearly immediately
Nov 09 20:57:00 <ice8lue> oh, forget about the crash..battery down
Nov 09 20:57:22 <dtzWill> oh haha
Nov 09 20:58:02 <ice8lue> well, gota load some time...
Nov 09 20:58:30 <Naomi> okay, this definitely looks more like the real deal lol
Nov 09 20:58:30 <ice8lue> i did set alpha to 255
Nov 09 20:58:36 <Naomi> thanks for the help!
Nov 09 20:58:46 <dtzWill> okay, glad you checked instead of just listening to me :D
Nov 09 20:59:13 <ice8lue> hope it will get the project further =)
Nov 09 21:01:06 <dtzWill> heh, there's a /usr/bin/luna-test that runs unit tests on luna xD
Nov 09 21:01:33 <ice8lue> ^^
Nov 09 21:03:13 <ice8lue> somehow that's like porting android to the htc diamond some time ago
Nov 09 21:04:01 <ice8lue> but without the need to use an emulator between host and guest os...
Nov 09 21:13:58 <rsanchez_> hello
Nov 09 21:14:09 <ashi_todd> hi
Nov 09 21:14:16 <rsanchez_> are there any linux gurus here?
Nov 09 21:14:42 <ashi_todd> probably... what do you need help with?
Nov 09 21:14:51 <oil> linux obviously :)
Nov 09 21:14:55 <ashi_todd> oh!
Nov 09 21:14:58 <Mercury> rsanchez_: Don't ask to ask, just ask.
Nov 09 21:15:04 <oil> lol
Nov 09 21:15:26 <rsanchez_> I accidentally did: sudo chmod 777 /
Nov 09 21:15:36 <rsanchez_> doesn't look like it did anything bad, just wondering if it is bad
Nov 09 21:16:07 <Mercury> Yes, it is bad, however you can fix it very easily since you only did a chmod without options.
Nov 09 21:16:22 <Mercury> You probably want to do 'sudo chmod 755 /'
Nov 09 21:16:51 <rsanchez_> ok
Nov 09 21:17:20 <rsanchez_> I just did a google search for it and found horror stories of systems being destroyed if you did chmod -R or something
Nov 09 21:17:25 <rsanchez_> wanted to make sure
Nov 09 21:17:47 <Mercury> Yeah, you didn't do the -R
Nov 09 21:17:48 <ashi_todd> yeah, -R will chmod everything.
Nov 09 21:17:56 <Mercury> If you had, it would be a rather different story.
Nov 09 21:17:57 * halfhalo likes the -R
Nov 09 21:18:28 <ice8lue> ya. But what you did was to 'just' set the read/write rules for all users
Nov 09 21:18:36 <ice8lue> on /
Nov 09 21:19:29 <rsanchez_> currently following setup instructions for WIDK
Nov 09 21:19:30 <Mercury> Yeah.
Nov 09 21:19:38 * halfhalo gets the popcorn
Nov 09 21:19:47 <rsanchez_> since I'm on linux can't run the official palm pdk
Nov 09 21:38:05 <destinal> rsanchez_: didn't they ever come out with a linux version?  they had promised to IIRC
Nov 09 21:38:24 <destinal> like, it's part of the linux SDK now, right?
Nov 09 21:39:20 <rsanchez_> is it part of the linux sdk??
Nov 09 21:41:01 <rsanchez_> gah, either way I'm almost finished getting the WIDK set up
Nov 09 21:41:30 <destinal> rsanchez_: hmm, maybe not.  their pdk page only mentions mac and windows (but that it's part of the sdk on those two platforms)
Nov 09 21:41:52 <rsanchez_> yeah I just noticed that too
Nov 09 21:41:56 <rsanchez_> hmm...
Nov 09 21:42:12 <destinal> that would be just silly to include a cross compilation kit for a linux target and not make it available for  linux hosts
Nov 09 21:43:44 <destinal> especially when it's now bundled with the sdk for the other two and the sdk exists for linux.  I dunno.. odd.
Nov 09 21:43:55 <destinal> I've only ever used our WIDK myself
Nov 09 21:50:03 <rsanchez_> are Palm's PDK tutorials compatible with the WIDK?
Nov 09 21:51:44 <halfhalo> oh, you know what I totally just realized?  I have an att sim in my house thats active!
Nov 09 21:52:01 <ashi_todd> handy
Nov 09 21:52:08 <halfhalo> yup
Nov 09 21:52:16 <halfhalo> has a data plan and everything
Nov 09 21:53:44 <ashi_todd> Whee, I have a procedure to write up today, for my company's ISO audit next week.
Nov 09 21:53:58 <halfhalo> fake it
Nov 09 21:54:07 <ashi_todd> heh
Nov 09 22:01:56 <rsanchez_> oh man wish I had noticed that earlier, perfect "that's what she said" moment
Nov 09 22:02:08 <rsanchez_> too late to be funny now
Nov 09 22:05:51 <rsanchez_> gah the WIDK installer is taking forever...
Nov 09 22:18:59 <destinal> rsanchez_: it downloads lots of big things
Nov 09 22:19:57 <rsanchez_> it looks like it's just installing all those things now
Nov 09 22:20:30 <destinal> rsanchez_: I imagine that it would be reasonably compatible with palm's tutorials, but I haven't looked at them.  try it and report back! :)
Nov 09 22:21:07 <destinal> first thing you should try to do when you get the WIDK going is compile and run DOOM, it really is the hello world
Nov 09 22:21:45 <destinal> http://www.webos-internals.org/wiki/Building_DOOM_with_scratchbox2  linked that bottom of   http://www.webos-internals.org/wiki/WebOS_Internals_PDK
Nov 09 22:22:43 <destinal> I probably should update the packaging part since the fix script for type: game apps isn't needed any longer
Nov 09 22:23:18 <rsanchez_> one question on doom: right now I can only test on a Pixi, not a Pre. Is it Pixi-compatible?
Nov 09 22:26:44 <rsanchez_> nice, finally done installing
Nov 09 22:28:31 <destinal> rsanchez_: hmm, all the WIDK directions have you set up for armv7 (pre)
Nov 09 22:28:53 <destinal> if SDL stuff works on pixi now then you should be able to target armv6 though
Nov 09 22:29:15 <PuffTheMagic> en0x, ping
Nov 09 22:29:26 <ajay> hey frolks
Nov 09 22:29:33 <ajay> s/r//
Nov 09 22:29:33 <infobot> ajay meant: hey folks
Nov 09 22:29:38 <ajay> haha
Nov 09 22:29:46 <destinal> rsanchez_: someone probably needs to clean up and document this stuff foe developing on pixi
Nov 09 22:30:23 <ajay> maybe i meant frolics.. oh wait, did infobot just take my sed cmd and implement it? sick!
Nov 09 22:30:52 <destinal> ajay: yup, he
Nov 09 22:31:00 <destinal> s/he/he rules/
Nov 09 22:31:00 <infobot> destinal meant: ajay: yup, he rules
Nov 09 22:31:03 <destinal> ~botsnack
Nov 09 22:31:04 <infobot> destinal: thanks
Nov 09 22:31:17 <rsanchez_> another quick question: obviously, root@pre won't work, so how should I get it on my pixi?
Nov 09 22:31:32 <ajay> haha
Nov 09 22:32:04 <destinal> rsanchez_: novaterm should work or you can set up ssh according to palm's directions
Nov 09 22:32:11 <destinal> or you can install ssh via preware
Nov 09 22:32:24 <rsanchez_> I'll do the preware route
Nov 09 22:33:59 <ajay> quick q--i removed uberkernel earlier as i was troubleshooting my phone. figured out the problem, but i cant get it back now
Nov 09 22:34:39 <Mercury> ajay: What happens when you try?  And how did you remove it and how are you trying to get it back?
Nov 09 22:34:43 <ajay> it says the kernel installed is not factory
Nov 09 22:35:38 <ajay> "refusing replacement".. how do i make it look like factory? i tried ipkg reinstall kernel but it doesn't know where to get the kernel from, i dont see it in ipkg-opt either
Nov 09 22:36:00 <Mercury> ajay: Preware lists a factory default kernel.
Nov 09 22:36:13 <ajay> oh yeah
Nov 09 22:36:47 <ajay> so turns out.. letting /opt (aka /var) fill up will bork your phone, at least temporarily. crazy huh?
Nov 09 22:36:58 <Mercury> I can't imagine why. :)
Nov 09 22:37:07 <SineOt> what filled up /opt?
Nov 09 22:37:44 <ajay> i dont know why it filled up though.. i threw like 80-100 mb in there (my bad i know) but that was weeks ago, i wonder what's filled up the rest of the space since then..
Nov 09 22:38:03 <ajay> i havent looked yet (to see what filled up opt)
Nov 09 22:38:18 <ajay> i dont understand why, when i removed uberkernel, the result wasnt a factory kernel
Nov 09 22:38:49 <ajay> or maybe i'm still running uberkernel, hrmm..
Nov 09 22:38:52 <SineOt> I think Uberkernel will fail if any number of files isn't stock, so something trivial might have slipped through. That's just my guess.
Nov 09 22:40:18 <ajay> yeah.. it mentioned the big ones though..
Nov 09 22:44:43 <ajay> oops wrong channel :)
Nov 09 22:46:00 <ajay> i wonder if that (or / being at 94% used) is related to the slowness ive been feelin
Nov 09 22:46:10 <ajay> (even post doctor)
Nov 09 22:47:23 <destinal> ajay: doctor should totally blow away those filesystems so I would doubt it
Nov 09 22:49:57 <rsanchez_> I'm having problems with this line: scp doom doom1.wad  root@pre:
Nov 09 22:50:20 <rsanchez_> installed openSSH from preware, but can't get it to work with root@pre:, root@pixi:
Nov 09 22:50:40 <rsanchez_> I get this message: ssh: Could not resolve hostname pixi: Name or service not known
Nov 09 22:54:26 <Mercury> rsanchez_: What's the IP address of your phone on your wireless network?
Nov 09 22:54:36 <Mercury> rsanchez_: Or cabled in with usbnet, either way.
Nov 09 22:54:46 <Mercury> rsanchez_: Replace pre/pixi with that IP address.
Nov 09 22:55:20 <rsanchez_> usbnet?
Nov 09 22:56:51 <Mercury> Read palm's documentation.  In very short, your phone can show up as a USB network adapter when you plug it in via USB to your computer, for a 'network' involving just the two computers.  This requires some configuration on your end, and on the whole you'll probably be happier just going over the wifi.
Nov 09 23:00:15 <rsanchez_> alright I'll check up on that
Nov 09 23:00:25 <rsanchez_> wifi is not an option since I'm using a Pixi-
Nov 09 23:01:04 <destinal> rsanchez_: usbnet would be the best way to go, or just for now do:
Nov 09 23:01:05 <destinal> novacom put file:///var/home/root/doom1.wad < doom1.wad
Nov 09 23:06:19 <destinal> rsanchez_: TBH you're the only one I've really talked to trying to build stuff for pixi with WIDK, you'll definitely want to change the target for cross compilation, probably change armv7 to armv6 everywhere in the instructions for WIDK
Nov 09 23:07:33 <destinal> others have built with it but until recently the SDL pdk stuff didn't work on pixi so there wasn't a lot of interest and the wiki pages don't really speak to it much
Nov 09 23:08:00 <rsanchez_> actually, I don't even have armv7 in /srv/preware/cross-compile/staging/, only armv6
Nov 09 23:08:10 <rsanchez_> just replaced them anyway in the docs
Nov 09 23:08:12 <destinal> cool, should work then
Nov 09 23:08:40 <rsanchez_> I should add a note in the wiki page for that then
Nov 09 23:09:13 <destinal> rsanchez_: great idea
Nov 09 23:10:14 <rwhitby> ajay: I suspect you forgot to run the prerm script for the kernel, as it seems you're running manual ipkg commands for some reason.
Nov 09 23:20:30 <rsanchez_> I got usbnet set up, it recognizes the pixi at usb0, but I can't get a connection set up
Nov 09 23:20:41 <rsanchez_> pixi has no ip address on the network
Nov 09 23:23:40 <ajay> destinal: oh i filled them up post-doctor :)
Nov 09 23:24:35 <ajay> oh you can put files with novacom, nice. i was trying to figure out the usbnet, but wasnt sure about filesystems
Nov 09 23:25:42 <ajay> rwhitby: i removed uber from preware, i started playing on the cmdline when my preware wasnt working (due to the full FS)
Nov 09 23:25:57 <destinal> you can get and put with novacom and get a terminal and other stuff
Nov 09 23:26:12 <ajay> destinal: nice
Nov 09 23:26:19 <rwhitby> did you have some other kernel or overclocking patch or script on there before uberkernel ?
Nov 09 23:26:24 <Loudergood> novabrowser is quite handy
Nov 09 23:26:43 <ajay> novabrowser?
Nov 09 23:27:35 <destinal> rsanchez_: typically you have to  ifconfig usb0 inet 192.168.0.2  or similar on your linux box.   the pixi shouldl be 192.168.0.202
Nov 09 23:27:45 <Loudergood> it's in the pdk http://developer.palm.com/index.php?option=com_content&view=article&id=1938
Nov 09 23:28:11 * rwhitby wishes someone would work out the host end of usbnet for OSX snow leopard
Nov 09 23:28:39 <destinal> rwhitby: it came stock in leopard and apple broke it in snow leopard, wasn't it?
Nov 09 23:28:44 <rwhitby> yep
Nov 09 23:28:47 <ice8lue> destinal: heyho^^ did some testing with the fb
Nov 09 23:28:57 <ice8lue> it definitely runs, quite good so far
Nov 09 23:29:37 <destinal> ice8lue: yeah. as I said the shutdown animation uses framebuffer so I figured things were fine there
Nov 09 23:29:53 <SineOt> it comes stock, but from what I remember reading about it, it uses usbnet as a serial connection and not an ethernet connection
Nov 09 23:30:00 <SineOt> so it won't let you use it as a network device
Nov 09 23:31:07 <ice8lue> one problem i got atm, i did set up a new user, but when i try to login, it gives me a bunch of errors (cannot chdir to my home dir, connot mkdir /var/luna,..)
Nov 09 23:31:11 <SineOt> but you can use it as a com port basically! :V
Nov 09 23:31:43 <dtzWill> ice8lue: what'd you do with the fb, etc? :)
Nov 09 23:32:38 <ice8lue> dtzWill: just playing arround so far. i did draw some pictures and so on, just to test it's working
Nov 09 23:38:19 <dtzWill> ice8lue: great! :)
Nov 09 23:38:30 <ice8lue> gnah....the home dir exists....why can't it cd there?
Nov 09 23:39:02 <ice8lue> these are the times when i hate unix^^
Nov 09 23:39:46 <dtzWill> ice8lue: permissions?
Nov 09 23:40:39 <ice8lue> dtzWill: well yes...but when the user is created, shouldnt he have the permissions to access his homedir?
Nov 09 23:45:06 <ice8lue> did chmod the homedir to 777 temporaryly...same result
Nov 09 23:56:23 <chrisa> This is an intervention
Nov 09 23:56:42 <halfhalo> is it the drugs?  its always the drugs...
Nov 09 23:56:46 <chrisa> The sprint pre fell apart, had issues, had a ridiculous return rate and didn't really do sprint much good
Nov 09 23:56:57 <chrisa> You should stop dreaming of a new sprint headset soon
Nov 09 23:57:36 <SineOt> For falling apart, having issues, and having a ridiculous return rate, they sure kept it around for a long time. :p
Nov 09 23:57:55 <chrisa> ROI
Nov 09 23:59:28 <SineOt> well, and on top of that-6 months later, they picked up the Pixi, so they couldn't have been hating the Pre -that- much to go and grab another webOS handset
Nov 10 00:00:01 <chrisa> These deals are made far earlier than you believe
Nov 10 00:00:10 <chrisa> Also, pixi had abysmal sales, especially on Sprint
Nov 10 00:00:38 <SineOt> yeah, they are
Nov 10 00:00:58 <SineOt> but phones that genuinely do terrible don't stick around for 522 days, and get a successor picked up 6 months later :p
Nov 10 00:01:11 <SineOt> they killed off like the Kin line, or quietly shuffled off into nowhere
Nov 10 00:01:42 <dtzWill> lol intervention
Nov 10 00:01:43 <dtzWill> :(
Nov 10 00:01:54 <dtzWill> ice8lue: depends how you create the user :)
Nov 10 00:02:00 <chrisa> Kin was a special form of market failure
Nov 10 00:02:08 <chrisa> They sold so few they knew they couldn't clear the inventory
Nov 10 00:02:09 <SineOt> Kin got a ton of advertising, and then they completely killed it within like 5 weeks of launch
Nov 10 00:03:16 <ice8lue> dtzWill: nevermind, got it working (tried to get the homedir at the wrong place...couldn't work ;) )
Nov 10 00:03:38 <dtzWill> ice8lue: great :)
Nov 10 00:05:08 * ice8lue will go to bed now, its far beyond midnight over here^^ cu
Nov 10 00:05:17 <dtzWill> ice8lue: cya :)
**** ENDING LOGGING AT Wed Nov 10 02:59:57 2010