**** BEGIN LOGGING AT Wed Aug 31 02:59:57 2011
Aug 31 03:00:55 * scoutcamper just realized he does have a 11.04 VM, goes to boot that to test with
Aug 31 03:19:29 <scoutcamper> w00t! it built
Aug 31 03:19:33 <scoutcamper> now to install
Aug 31 03:21:16 <halfhalo> hehehe: http://pastebin.com/Hn3hYVuW
Aug 31 03:21:56 <dtzWill> haha 'nodepad' ;)
Aug 31 03:22:44 * scoutcamper lols
Aug 31 03:24:13 <dtzWill> scoutcamper: built/install.....ing your upstart-fix-ified ubuntu?
Aug 31 03:25:00 <scoutcamper> yes
Aug 31 03:25:16 <dtzWill> :D
Aug 31 03:43:59 <scoutcamper> my install keeps hanging on making the webos user
Aug 31 03:44:18 * scoutcamper moves his postinst script higher in the file
Aug 31 04:08:19 <scoutcamper> W00T
Aug 31 04:08:23 <scoutcamper> rebuilt
Aug 31 04:27:25 <scoutcamper> w00t
Aug 31 04:27:30 <scoutcamper> postinst mod worked
Aug 31 04:27:42 <ka6sox> push the diff.
Aug 31 04:38:26 <scoutcamper> rwhitby, diff is in your inbox
Aug 31 06:54:19 <booto> is it preferred to actually include header files in the cross-compile git repo, or have a Makefile grab them from the src packages hosted at opensource.palm.com? I've gone with the latter method: http://booto.bur.st/webos/cross-compile/0001-adding-libsysfs-headers-to-the-build-env.patch
Aug 31 06:54:37 <rwhitby> yes, latter method
Aug 31 06:58:02 <rwhitby> booto: see /msg
Aug 31 07:02:05 <rwhitby> booto: git clone git@git.webos-internals.org:preware/cross-compile.git
Aug 31 07:02:19 <rwhitby> then please push your patch, welcome aboard.
Aug 31 07:23:04 <booto> pushed
Aug 31 08:13:28 <Xuzz> rwhitby: I added a highlighted state to the back buttonsin TouchPad preware.
Aug 31 08:14:08 <Xuzz> http://pastie.org/private/3quzl7jfuil1nwi8tussog and http://dl.dropbox.com/u/3177211/preware_back-pill-active.png
Aug 31 08:14:23 <Xuzz> I didn't do the dark theme, is that needed?
Aug 31 08:14:49 <Xuzz> (the button is simply a hue change from the other button that looks similar-ish to the built-in blue highlightes states)
Aug 31 08:16:08 <Xuzz> (yes, tiny UI things are really what I spend my time focusing on.)
Aug 31 09:03:19 <rwhitby> Xuzz: yes, all the themes are used
Aug 31 15:04:19 <Xuzz> okay, i didn't see a place to enable that on the TouchPad, maybe i just missed it
Aug 31 15:55:29 <ka6sox-zzzz> the point of KIOSK mode usually is to NOT let them out of it...
Aug 31 16:40:03 <ka6sox> scoutcamper, send me your diff and a postinst file for the chroot.
Aug 31 16:40:08 <ka6sox> and I will test
Aug 31 16:40:59 <scoutcamper> sent
Aug 31 16:41:22 <ka6sox> thanks.
Aug 31 16:41:30 <scoutcamper> anyone interested in testing my chroot upstart postinst mod /msg me your email address and ill send the diff and postinst
Aug 31 16:44:03 <ka6sox> it should be "idempotent" so that would mean even if you have it installed already it won't hurt.
Aug 31 16:47:31 <scoutcamper> correct
Aug 31 16:48:17 <scoutcamper> the fix will apply to any chroot, even those currently installed
Aug 31 16:48:27 <scoutcamper> s/any chroot/any ubuntu chroot/
Aug 31 16:49:59 <scoutcamper> brb
Aug 31 21:05:12 <scoutcamper> so, ka6sox xapp script is where?
Aug 31 21:05:32 <ka6sox> /dev/null
Aug 31 21:05:43 <ka6sox> got wiped accidentally.
Aug 31 21:06:10 <ka6sox> "one to throw away"
Aug 31 21:06:25 <scoutcamper> *facepalm
Aug 31 21:07:09 <ka6sox> having learned what I have on others that I've been working on will make it idempotent.
Aug 31 21:07:56 <scoutcamper> ka6sox, so, we are making this a seperate app or part of xecutah?
Aug 31 21:08:44 <ka6sox> xapps installs a script in the chroot we can call (or we add it as part of the chroot itself)
Aug 31 21:09:23 <scoutcamper> i meant as the install from preware
Aug 31 21:09:26 <ka6sox> that script attempts to install 1 app (previously known to work).
Aug 31 21:09:50 <ka6sox> I think this becomes a mod to the chroot installation
Aug 31 21:10:59 <scoutcamper> k
Aug 31 21:11:17 <ka6sox> similar to what you just did with the last patch
Aug 31 21:11:24 <scoutcamper> yeah
Aug 31 21:11:31 * scoutcamper goes to install xchat in his chroot
Aug 31 21:11:57 <ka6sox> no
Aug 31 21:11:58 <ka6sox> don't
Aug 31 21:12:02 <scoutcamper> ok
Aug 31 21:12:08 <ka6sox> thats what the script does.
Aug 31 21:12:13 <ka6sox> install things in the chroot.
Aug 31 21:12:24 <scoutcamper> once we right it
Aug 31 21:12:28 <scoutcamper> write*
Aug 31 21:12:40 <ka6sox> meet me in a screen session on fiona.
Aug 31 21:12:45 <ka6sox> we will "write" it
Aug 31 21:12:57 <ka6sox> I already know how.
Aug 31 21:13:20 <scoutcamper> "no session to be attached"
Aug 31 21:13:20 <ka6sox> (since I did it once already)
Aug 31 21:13:26 <ka6sox> create one
Aug 31 21:13:41 <scoutcamper> done
Aug 31 21:35:22 <ka6sox> okay so the basic idea is that we would update in the chroot, then attempt to install
Aug 31 21:36:02 <ka6sox> the postinst script waits till the installer is done or fails and then reads the error log
Aug 31 21:36:40 <scoutcamper> ok. do we plan on adding this script to the chroot before it is ipk?
Aug 31 21:37:21 <ka6sox> no, I intend for this to be part of the payload of a ubuntu/debian chroot (I don't know what arch linux does)
Aug 31 21:37:37 <scoutcamper> thats what i meant lol
Aug 31 21:37:53 <ka6sox> this could be separate however...
Aug 31 21:38:23 <ka6sox> since it uses public repos already populated in the chroot there isn't any danger of injection.
Aug 31 21:39:14 <scoutcamper> i like the idea of pre-installing it, and if someone chooses to use its functionality they download the launcher from preware
Aug 31 21:40:21 <ka6sox> so for arch or some other chroot flavour you would change the command only inside the script
Aug 31 21:40:33 <ka6sox> opkg-install or yum install or whatever.
Aug 31 21:40:40 <scoutcamper> yeah
Aug 31 21:41:40 <ka6sox> http://pastebin.com/uTUxBRAu
Aug 31 21:42:00 <ka6sox> this bit lives in the chroot
Aug 31 21:42:05 <scoutcamper> yep
Aug 31 21:42:11 <ka6sox> probably in /usr/sbin
Aug 31 21:42:33 <scoutcamper> i was going to suggest that >.<
Aug 31 21:43:04 <ka6sox> okay now get me the postinstall for ubuntu-natty-chroot and make a copy of it (.orig)
Aug 31 21:43:09 <ka6sox> then we will mod that to use this.
Aug 31 21:43:30 <scoutcamper> ka6sox, the one with my dbus mod or current from repos?
Aug 31 21:43:37 <ka6sox> repos
Aug 31 21:43:49 <ka6sox> because we are diffing against what is in the repo only
Aug 31 21:44:11 <scoutcamper> done
Aug 31 21:44:53 <ka6sox> this is going to be one of those "inside/outside" things
Aug 31 21:45:15 <ka6sox> I need to put a timeout on running the commands inside the chroot...or it could get hung.
Aug 31 21:46:21 <ka6sox> also we need to check a few things before we install as well.
Aug 31 21:46:27 <ka6sox> (like is it installed?)
Aug 31 21:48:28 <ka6sox> okay, now that I look @ this....
Aug 31 21:48:32 <ka6sox> we have 2 choices
Aug 31 21:49:01 <ka6sox> 1) have rwhitby include it in the chroot when we make it via rootstock (cp that file in)
Aug 31 21:49:26 <ka6sox> 2) figure out how "files" work so we can include it like a config.
Aug 31 21:49:41 <scoutcamper> i vote option 1
Aug 31 21:49:44 <ka6sox> and have it cp'd by ipkg installer
Aug 31 21:50:02 <ka6sox> the right way is to have the files and let the installer put it where it belongs.
Aug 31 21:50:31 <ka6sox> see: optware/ssh
Aug 31 21:50:39 <scoutcamper> true
Aug 31 21:50:53 <ka6sox> sshd_config is a "file" we supply
Aug 31 21:51:20 <scoutcamper> but, to make it easier to maintain why not have the chroot makefile git pull the file and cp it in
Aug 31 21:52:03 <ka6sox> s/config/\/opt\/etc\/sshd_config/
Aug 31 21:52:41 <ka6sox> a git pull would put it in the right place if we $origin correctly.
Aug 31 21:53:08 <ka6sox> but thats dangerous
Aug 31 21:53:32 <ka6sox> the builder does a git-clone iirc
Aug 31 21:53:37 <ka6sox> and then builds it
Aug 31 21:53:43 <scoutcamper> yes
Aug 31 21:53:57 <ka6sox> so its part of "files" in the git repo
Aug 31 21:54:19 <scoutcamper> i meant have the makefile git clone git://git.webos-internals.org/xapp/main.git and cp xapp.sh in the chroot
Aug 31 21:54:29 <scoutcamper> or something similar
Aug 31 21:54:31 <ka6sox> no main.git
Aug 31 21:54:33 <ka6sox> gone
Aug 31 21:54:50 <ka6sox> this is a mod to the chroot
Aug 31 21:54:57 <ka6sox> unless we are forced to go thatway.
Aug 31 21:55:02 <scoutcamper> kk
Aug 31 21:56:28 <ka6sox> assuming we aren't forced to go that way, then we need to mod the Makefile to cp the scripts from the git repo to the chroot rootstock before its compressed.
Aug 31 21:56:44 <scoutcamper> let me cp the makefile
Aug 31 21:56:58 <ka6sox> don't forget the .orig
Aug 31 21:57:02 <ka6sox> (make one)
Aug 31 21:57:25 <scoutcamper> done
Aug 31 21:58:10 <ka6sox> one reason I wanted it to be done on a chroot by chroot basis is to make sure it works with that chroot version/distro.
Aug 31 21:58:20 <scoutcamper> yeaj
Aug 31 21:58:23 <scoutcamper> yeah*
Aug 31 21:59:32 <ka6sox> looks like rootstock creates a .tgz
Aug 31 21:59:45 <scoutcamper> correct
Aug 31 21:59:58 <scoutcamper> and then that file is untarred on device by the postinst
Aug 31 22:00:54 <ka6sox> okay so we would have to make the xapps installer part of the payload and mv it into the chroot.
Aug 31 22:01:10 <ka6sox> during postinst
Aug 31 22:01:23 <scoutcamper> yes
Aug 31 22:02:18 <ka6sox> so figure out how the the config for optware/ssh gets put where it belongs.
Aug 31 22:03:08 <ka6sox> why did you modify the "orig"?
Aug 31 22:03:29 <scoutcamper> because that file had my old test didnt work well mod in it from trying to build on fiona
Aug 31 22:03:54 <ka6sox> okay I wanted the file from the repo.
Aug 31 22:04:04 <scoutcamper> that is the repo file now
Aug 31 22:04:05 <ka6sox> otherwise we are diffing against an unknown.
Aug 31 22:04:20 <ka6sox> a "space" or line would make this fail.
Aug 31 22:04:38 <ka6sox> we have to use what is in the repo to create the patch.
Aug 31 22:06:35 <scoutcamper> fixed
Aug 31 22:08:22 <ka6sox> okay going with the same theme that we will include this in the chroot....
Aug 31 22:08:59 <ka6sox> you need to figure out where that file is going to show up in the FS when ipkg installs it.
Aug 31 22:12:38 <ka6sox> I think it will be in: /media/crpytofs/apps/usr/palm/applications/org.webosinternals.ubuntu-natty-chroot/usr/bin
Aug 31 22:13:04 <scoutcamper>  /media/crpytofs/apps/usr/palm/applications/org.webosinternals.ubuntu-natty-chroot/bin
Aug 31 22:13:21 <ka6sox> whats in that?
Aug 31 22:13:33 <scoutcamper> thats where xterm-server goes
Aug 31 22:14:17 <scoutcamper> wait, i broke that
Aug 31 22:14:29 <scoutcamper> thats where chroot-wrapper and start-xterm go
Aug 31 22:14:34 <ka6sox> think optware
Aug 31 22:14:48 <ka6sox> not chroot_wrapper
Aug 31 22:15:01 <ka6sox> it goes in the tree where it goes in real life
Aug 31 22:15:08 * scoutcamper hasnt done much optware stuff
Aug 31 22:15:22 <scoutcamper> you are right, it would be /usr/bin
Aug 31 22:15:32 <ka6sox> mobi.optware.openssh/opt/etc/sshd_config
Aug 31 22:15:51 <ka6sox> this would be like an include
Aug 31 22:16:31 <scoutcamper> yeah
Aug 31 22:16:36 <scoutcamper> i get it
Aug 31 22:17:03 <scoutcamper> so we need to stage this now
Aug 31 22:17:10 <ka6sox> yes
Aug 31 22:17:12 <scoutcamper> s/need to/should/
Aug 31 22:18:22 <ka6sox> look @ how the git is arrainged for optware/openssh
Aug 31 22:18:54 <ka6sox> then we put the script in the right place.
Aug 31 22:22:03 <scoutcamper> interesting
Aug 31 22:22:08 <scoutcamper> openssh make borks
Aug 31 22:24:09 <scoutcamper> ka6sox, optware openssh in build needs a link update
Aug 31 22:24:31 <scoutcamper> its pointing at 5.3 and in the feeds it needs to be 5.8
Aug 31 22:26:09 <scoutcamper> more specifically 5.8p2-1
Aug 31 22:27:48 <scoutcamper> someone should fix that
Aug 31 22:46:13 <ka6sox> scoutcamper, are you someone?
Aug 31 22:46:29 <scoutcamper> someone with write access
Aug 31 22:47:18 <ka6sox> send patch
Aug 31 22:48:19 <scoutcamper> its in the screen session
Aug 31 22:49:21 <ka6sox> what is the name?
Aug 31 22:50:33 <scoutcamper> openssh.diff in /srv/preware/build
Aug 31 22:51:26 <ka6sox> email it to rwhitby
Aug 31 22:51:30 <ka6sox> its your patch.
Aug 31 22:51:33 <scoutcamper> kk
Aug 31 22:54:42 <scoutcamper> rwhitby, in your inbox
Aug 31 23:11:09 * scoutcamper|away bbiab
Sep 01 00:02:56 <ka6sox> scoutcamper, once we hear from rwhitby we can move forward :D
Sep 01 00:03:05 <ka6sox> but you sent the patch right?
Sep 01 00:03:05 <scoutcamper> indeed
Sep 01 00:03:09 <scoutcamper> yes i did
Sep 01 00:03:32 <scoutcamper> patch for openssh and patch to fix upstart
Sep 01 00:04:24 <ka6sox> okay, it doesn't matter where that eventually lands for the next phase
Sep 01 00:06:13 <ka6sox> the next phase involves: a new headless app that:
Sep 01 00:07:02 <ka6sox> calls that script upon installation as a postinst, creates teh chroot_wrapper necessary to call an the executable with the correct user (webos)
Sep 01 00:07:19 <ka6sox> inside the chroot.
Sep 01 00:08:13 <ka6sox> also we would have to change the whitelist in xecutah
Sep 01 00:12:42 <scoutcamper> k
Sep 01 00:12:49 <scoutcamper> makes sense
Sep 01 00:21:51 <NuttyBunny> Sorry for the spam, who has a PCL printer and wish to help to create a way to print on any linux compatible printer from TouchPads? http://forums.precentral.net/hp-touchpad-tips-information-resources/296118-use-any-printer-attached-your-linux-computer-pcl-printer-your-touchpad.html
Sep 01 00:46:58 <ka6sox> scoutcamper, is there a xapps repo yet?
Sep 01 00:47:16 <scoutcamper> negative
Sep 01 01:08:50 <ka6sox> okay I am not qualified to add the repo.
Sep 01 01:38:30 <NuttyBunny> nobody on the printer experiment?
Sep 01 01:39:02 <destinal_>  NuttyBunny  printer experiment?
Sep 01 01:39:15 <NuttyBunny> http://forums.precentral.net/hp-touchpad-tips-information-resources/296118-use-any-printer-attached-your-linux-computer-pcl-printer-your-touchpad.html
Sep 01 01:39:45 <NuttyBunny> I'm looking for a way of making the Touchpad recognize any linux machine as a PCL printer so it can use any printer attached to the machine
Sep 01 01:40:19 <destinal_> NuttyBunny: ah neat
Sep 01 01:40:23 <NuttyBunny> right now I'm stuck, as I don't know what does a standard PCL printer answers to the Touchpad when it gets queried :(
Sep 01 01:41:02 <NuttyBunny> I got one of those cheap HP laser printers, and it is not compatible. But I intend to keep it anyway :)
Sep 01 01:41:43 <destinal_> NuttyBunny: yeah I'd be interested too.  I want the ability to print to google cloud print and to Kinko's  (fedex office now)
Sep 01 01:42:43 <NuttyBunny> mmm... well, you would need to find a way to print a postscript file to those services, but the first step is connecting the TP to the computer and make it think it's a compatible printer :D
Sep 01 01:43:07 <destinal_> yup
Sep 01 01:43:54 <destinal_> have you looked to see what the touchpad print client stuff is?  ie is it a node.js service or a native binary or what
Sep 01 01:44:03 <destinal_> if it's readable we could learn a lot from what it's trying to do
Sep 01 01:45:20 <NuttyBunny> I just did a "ps" in the TP, but didn't find anything interesting. But yes, maybe I was doing it at the wrong moment, lemme check it while it's trying to discover the printer
Sep 01 01:45:59 <destinal_> well also can you find the html  / javascript for the button and see what service it calls natively
Sep 01 01:46:43 <NuttyBunny> good idea!! :D
Sep 01 01:46:49 * NuttyBunny looks in the apps dis
Sep 01 01:46:51 <NuttyBunny> dir
Sep 01 01:47:38 <NuttyBunny> hey rwhitby !
Sep 01 01:47:49 <NuttyBunny> by any chance you have a TouchPad compatible printer??
Sep 01 01:49:03 <rwhitby> yes
Sep 01 01:49:11 <rwhitby> HP Photosmart Premium
Sep 01 01:49:40 <NuttyBunny> Ah, could you help me? I'm trying to make a CUPS server act like a PCL compatible printer
Sep 01 01:50:42 <NuttyBunny> (see backlog), from a wireshark capture I saw that the TP queries the printer via snmp, but it doesn't like the response my printer gives (it's a cheap P1102w, not PCL)
Sep 01 01:51:39 <NuttyBunny> so I have the idea of: forging a SNMP service which would answer the queries with valid answers, and use netcat to pipe the input on port TCP 9100 , send it to gs to convert it to ps, and then feed that to any printer on the machine
Sep 01 01:51:56 <destinal_> and I think just hacking the client side will be easier :)
Sep 01 01:52:59 <NuttyBunny> yep, 2 options :)
Sep 01 01:53:52 <destinal_> rwhitby: at any rate if you could capture packets for installing a printer and printing to it and make a pcap file available it would be useful
Sep 01 01:54:10 <NuttyBunny> destinal_, I think the client option won't be as easy, we have a binary service: /bin/printmgrd
Sep 01 01:54:40 <NuttyBunny>  /usr/bin/printmgrd
Sep 01 01:57:54 <destinal_> hmm does 3.0.2 emu still have an ssh daemon and still do the port redirect thing?
Sep 01 01:58:04 <destinal_> thought it was 5022 but no answer
Sep 01 01:58:10 <NuttyBunny> no idea
Sep 01 02:00:20 <NuttyBunny> Or, maybe we could bypass the printer discovery app and just add the printer to the database
Sep 01 02:01:13 <destinal_> ah 5522 not 5022 right
Sep 01 02:05:40 <NuttyBunny> hmmm... from what I see in the binary print manager.....  it seems to check quite a few things from the printer, online status, jammed paper, low ink... argh...
Sep 01 02:06:56 * NuttyBunny mumbles .. stupid HP... "all printers made after 2006" 
Sep 01 02:07:21 <destinal_> wow when did HP start including the full putty suite in webos sdk?
Sep 01 02:07:34 <NuttyBunny> full?
Sep 01 02:08:53 <destinal_> well putty, pscp  and plink anyway
Sep 01 02:08:55 <destinal_> not quite full
Sep 01 02:21:20 <chpwn> okay, I did the dark-skin version as well now
Sep 01 02:59:38 <halfhalo> anyone know where/how a devices nuid is generated?  I need to grab it inside a command line node app
**** ENDING LOGGING AT Thu Sep 01 02:59:57 2011