**** BEGIN LOGGING AT Wed Aug 24 02:59:56 2005 **** BEGIN LOGGING AT Wed Aug 24 03:13:02 2005 Aug 24 03:16:30 RP: that cardid table can be expanded a bit more? Aug 24 03:17:23 RP: /etc/pcmcia/hostap_cs.conf have more cards iirc Aug 24 03:21:23 hrw|work: its only actually essential for Connect+ Aug 24 03:22:12 XorA: not if I have pcmciautils instead of pcmcia-cs Aug 24 03:23:47 with pcmcia-cs it is really nearly not needed as cardmgr force loading of modules. but with pcmciautils all work has to be done by hotplug which try to modprobe with "modalias" which is created from ID table in module so normal modutils is needed Aug 24 03:36:07 morning folks Aug 24 03:36:24 mickeyl: I'm seeing you online during the day more and more... how's the phd going? :) Aug 24 03:36:27 * Twiun hides Aug 24 03:38:09 hi mickeyl Aug 24 03:39:25 * mickeyl throws a stone in Twiuns direction Aug 24 03:39:35 bzzzt. wrong subject. choose another one Aug 24 03:39:55 how's, err, monotone? Aug 24 03:40:08 :D Aug 24 03:40:08 * Twiun crawls back under his rock. Aug 24 03:40:16 * mickeyl looks for another stone in his bag Aug 24 03:41:27 mickeyl: I joined few lines in OOO Aug 24 03:41:33 hrw|work: thanks Aug 24 03:41:54 mickeyl: basically you dont have to use

- wordpress add them Aug 24 03:42:26 I see. Now that I know how I can add things to not-in-menu I'll do the drafts completely in wordpress. it's too annoying to convert it later Aug 24 03:45:40 ;) Aug 24 03:46:22 mickeyl: there are plugins for wordpress which make it easier but it will require converting of all content so I dont enable them Aug 24 03:48:52 sure Aug 24 03:50:15 RP: I'm doing a 3100 image atm. Will give you the location when it's done Aug 24 04:03:49 morning mickeyl Aug 24 04:04:13 mickeyl: I'll pass the location on. I told him to post in oz-dev but he doesn't seem to have done so yet Aug 24 04:04:33 hrw: Yes, you can add deviceids at will. If you let me know of any extras, I'll push them into mainline... Aug 24 04:04:57 Should that bluetooth id be pushed btw or is someone doing that? Aug 24 04:05:10 RP: it'll be @ http://www.Vanille.de/temp/borzoi in a couple of hours Aug 24 04:05:18 RP: which bt id? Aug 24 04:06:13 the one which gets added to dtl1_cs Aug 24 04:06:18 ah Aug 24 04:06:39 btw., i got a id tuple from a usb host card which is apparantly not in yet Aug 24 04:06:55 oh and we had much fun with a SanDisk 2G sd card Aug 24 04:06:57 I don't even know who added that... Aug 24 04:07:12 it doesn't work in _any_ Linux device, not even the 2.6.13 zaurus Aug 24 04:07:14 "much fun"? Aug 24 04:07:27 hmm... Aug 24 04:07:38 Its probably too slow at some critical moment... Aug 24 04:07:48 Try increasing the value in the mmc_timeout patch Aug 24 04:08:08 hi all Aug 24 04:08:26 Gather any cardids toegther and I'll try and get them in mainline. I don't want any patches we can't help... Aug 24 04:08:53 yep Aug 24 04:09:39 i think there's a Sandisk ID missing in hostap Aug 24 04:09:43 i'll check when I'm home Aug 24 04:09:57 RP: that patch was sent to upstream Aug 24 04:09:57 it's a rare OEM card which has a Connect+ chipset, however 0 MB memory. Aug 24 04:10:01 * mickeyl shakes head Aug 24 04:10:39 hrw|work: Ok, that's fine then. It should therefore be at the top of the file ;-) Aug 24 04:10:53 RP: dtl1_cs-add-socket-revE.patch is my patch Aug 24 04:10:55 RP: ok Aug 24 04:10:57 or towards the top... Aug 24 04:14:02 hi BigAl Aug 24 04:14:04 sup Aug 24 04:14:11 hi BigAl Aug 24 04:14:44 there are some pending patches from Dominic that I want to see merged Aug 24 04:14:44 mainline first, before I am going to submit patches for the Bluetooth Aug 24 04:14:45 PCMCIA drivers. Aug 24 04:14:53 thats answer from module author Aug 24 04:15:58 hi BigAl Aug 24 04:17:52 RP: fwiw, I still have that odd idea that 2.6 doesn't shut down all devices and drains more power in standbye than 2.4. Do you have a means to measure that? Aug 24 04:17:58 Hi hrw Aug 24 04:18:13 Hi RP, mickeyl Aug 24 04:21:07 hrw|work: Ok, I'm happy just to know its all being dealt with :) Aug 24 04:21:41 mickeyl: Its entirely possible. Try the latest battery patches and see if they help as I did make a lot of changes (for the better) Aug 24 04:21:56 I should perhaps see if I can measure it somehow... Aug 24 04:22:03 ok will do. will have to upgrade my c7x0 anyway to look at the new hinge support Aug 24 04:22:25 RP: added info on top of patch Aug 24 04:22:40 the c7x0 runs without a battery once it booted. so if you have measuring equipment for AC you could compare it like that Aug 24 04:22:47 mickeyl: It could have been that the battery just doesn't charge as much under 2.6... Aug 24 04:23:10 morning Aug 24 04:23:26 mickeyl: The last patches should be a bit more active about battery charging... Aug 24 04:23:35 RP: ok, I'll try that Aug 24 04:24:01 Thinking about it, I'd better remove the debug code in the battery code... Aug 24 04:24:04 hrw|work: thanks Aug 24 04:25:02 I've updated my status page to reflect the C3000 kernel status :) Aug 24 04:25:46 cool Aug 24 04:26:31 03mickeyl 07org.oe.dev * r4f3c6ab6... 10/packages/sharp-binary-only/sharp-sdmmc-support_2.4.18-rmk7-pxa3-embedix.bb: add borzoi variant of sharp-sdmmc driver Aug 24 04:26:36 03hrw 07org.oe.dev * rb7f12e46... 10/packages/linux/linux-openzaurus-2.6.13-rc5-mm1/dtl1_cs-add-socket-revE.patch: linux-oz: added info from upstream about future of dtl1_cs-add-socket-revE.patch Aug 24 04:30:23 <__law__> here is a new compile error if i try to build an gpe-image http://pastebin.com/344902 Aug 24 04:30:35 <__law__> should i try to build xtst ? Aug 24 04:31:51 __law__: ah, I suspected, xdpyinfo should DEPEND on xtst Aug 24 04:33:35 __law__: thats pb_'s package Aug 24 04:34:21 <__law__> pb_, could you change this? Aug 24 04:35:04 <__law__> after installing xtst manually all works fine Aug 24 04:47:58 03florian 07org.oe.dev * r7a93d384... 10/conf/ (distro/maemo-preferred.inc machine/netvista.conf): Aug 24 04:47:58 Update netvista cinguration. Aug 24 04:47:58 Fix several version numbers for the maemo images. Aug 24 04:48:02 03florian 07org.oe.dev * rc5331de7... 10/packages/maemo/osso-gnome-vfs2_2.8.4.4-1.bb: Update dependencies to a fixed version of dbus. Change PROVIDES to virtual/gnome-vfs. Aug 24 04:48:06 03florian 07org.oe.dev * r52fd1ab0... 10/packages/xdpyinfo/xdpyinfo_cvs.bb: Add missing dependency to xtst. Aug 24 04:48:10 03florian 07org.oe.dev * r603d3c23... 10/packages/gpe-bluetooth/gpe-bluetooth_0.42.bb: Introduce a new version of gpe-bluetooth. Aug 24 04:48:15 03florian 07org.oe.dev * r771edb7e... 10/packages/gpe-conf/gpe-conf_0.1.25.bb: Add new version of gpe-conf. Aug 24 04:50:23 What's a good media player for Opie? Aug 24 04:53:08 Hello Aug 24 04:53:53 could anybody tell me what LEAD_SONAME is for? Aug 24 04:53:57 <__law__> mickey|lunch, i cant update my monotone repo any ideas? http://pastebin.com/344916 Aug 24 04:55:23 patw: it's needed to name the package according to debian package naming rules Aug 24 04:56:46 aaah, ok, thank you. Aug 24 05:00:12 BigAl: hard to tell - opie-mediaplayer2 should work for most Aug 24 05:02:14 hrw|work: I could never get that to work, it'd always lock up in the "scanning for files" stage. Aug 24 05:02:24 03koen 07org.oe.dev * rce4141d6... 10/conf/distro/preferred-gpe-versions-2.7.inc: GPE 2.7: use gpe-conf 0.1.25 Aug 24 05:05:22 BigAl: works for me Aug 24 05:06:51 X-Scale 1.25GHz... intel is weird Aug 24 05:09:22 hrw|work: I'm rebuilding the image now, hopefully it'll work. Aug 24 05:15:36 BigAl: Did you ever try gpe on spitz-2.6? Aug 24 05:18:43 RP: I've tried to build it, but the built kept failing. Aug 24 05:19:02 _law_: I lack the privs to change it. If you can do so, please go ahead; otherwise, please file a request in bugzilla. Aug 24 05:20:48 BigAl: It appears to set pxafb into a very strange state when the xserver starts :-/ Aug 24 05:21:37 RP: Weird. With 2.4 it'd build and run to the point where I could set a username and password, but that's it. Aug 24 05:21:54 RP: If I can get an image to build properly I'll have a poke at it Aug 24 05:26:52 03florian 07org.oe.dev * r791935fd... 10/packages/maemo/osso-gnome-vfs2_2.8.4.4-1.bb: Fix typo. Aug 24 06:06:57 BigAl: Under 2.6 it works if you disable the command line options in Xserver.... Aug 24 06:07:04 (unrotated) Aug 24 06:08:58 RP: Can the Xserver do rotation itself? Aug 24 06:12:56 BigAl: yes, although the rotation applet doesn't appear to be installed Aug 24 06:13:27 Is there a commandline to tell gpe to rotate? Aug 24 06:13:49 xrandr -o {left,right,normal,inverted} Aug 24 06:14:01 which is the regular tool from X Aug 24 06:14:06 koen: That had no effect on it... Aug 24 06:14:07 (just like on your desktop) Aug 24 06:16:39 koen: It will change the login screen. It has no effect after you've logged in - must get changed back by gpe-dm? Aug 24 06:16:58 RP: those hostap modules dont work :-( Aug 24 06:16:59 iirc it spawns a new xserver after login Aug 24 06:17:55 XorA: And the ones in kernel do? Strange... Aug 24 06:18:11 XorA: Did you check the card id was correct? Aug 24 06:18:12 RP: Its oopsed my kernel, trying to extract now Aug 24 06:18:31 XorA: Ah. Not good :-/ Aug 24 06:18:31 RP: it recognises card, but firmware wont load then it oopsed Aug 24 06:19:08 Maybe try disabling the ID in the ide module - perhaps its a resource conflict? Aug 24 06:20:02 RP: http://pastebin.ca/20976 Aug 24 06:21:31 XorA: Its not a "proper" oops - its a soft lockup Aug 24 06:21:43 RP: its taking 100% CPU load :-( Aug 24 06:22:05 XorA: Its in the IDE layer - its probably requesting ide resources which are in conflict with hostap Aug 24 06:23:08 RP: battery charging LED no longer works either Aug 24 06:23:30 XorA: It wouldn't as the charger code needs cpu time to function Aug 24 06:23:41 XorA: Disable the IDE driver od for that card for now and try that... Aug 24 06:23:51 RP: no, ive rebooted, and no LED Aug 24 06:24:39 RP: battery applet shows AC though Aug 24 06:26:20 RP: odd, rebooted, card is now functioning in both IDE and wlan mode simultaneously Aug 24 06:28:22 XorA: but it fails when not present at boot time? Aug 24 06:28:34 hrw|work: yeah Aug 24 06:28:52 ah, broken it again, it doesnt last long in when both are running :-) Aug 24 06:29:14 wireles is now running, ide is locked up Aug 24 06:31:03 RP: my Z no longer charges from USB on latest kernel Aug 24 06:34:32 XorA: Sounds like some races are present. Still, its a start :) Aug 24 06:34:39 The charging stuff has me concerned... Aug 24 06:34:49 USB doesn't change the reported AC status? Aug 24 06:35:34 Can you have a look over the logs for an AC insertion to USB and see if it notices anything at all? Aug 24 06:37:19 RP: ahah, it seems it doesnt like it when USB is inserted on boot, just unplugged, re-insterted and its lit LED Aug 24 06:37:25 RP: http://pastebin.ca/20978 - my c760 with 2.6.13-rc6-mm2 after remove/connect usb charger Aug 24 06:37:32 RP: and voltage is now going up not down Aug 24 06:38:34 XorA: Ok, that's much easier to fix :) Aug 24 06:38:43 hrw|work: Thanks, it looks fine :) Aug 24 06:38:43 I will reboot with usb charger connected Aug 24 06:39:14 hrw|work: Ok, it should do an AC check as it boots - if it doesn't I know what to fix... Aug 24 06:39:30 Z started with power led off Aug 24 06:39:38 We used to have a whole load of these bugs and I thought I'd tidied it all up! Aug 24 06:41:06 ;) Aug 24 06:42:38 RP: with reference to your previous comment to mickey|lunch my "fully" charged c7x0 under older kernel now registers as only 70% under new Aug 24 06:42:42 my Z is fully charged so charge led was off Aug 24 06:43:13 XorA: I've tweaked the charging profile quite a bit. Is it charging? Aug 24 06:43:31 RP: its gone from 70->75 Aug 24 06:43:48 and the charger is still on? Aug 24 06:43:54 RP: yes Aug 24 06:44:07 XorA: Interesting :) Aug 24 06:44:12 RP: http://pastebin.ca/20979 - kind of dmesg from charged with charger connected on boot Aug 24 06:44:12 now hovering 75-78-75-78 Aug 24 06:45:07 hrw|work: You' Aug 24 06:45:15 hrw|work: Your battery is fully charged? Aug 24 06:45:41 RP: yep Aug 24 06:45:45 hrw|work: That log suggests it is - it tried to turn the charger on and found it was full... Aug 24 06:45:52 Battery Voltage: 217 Aug 24 06:45:52 Status: 0 Aug 24 06:45:52 Percentage: 100 (at time 1084) Aug 24 06:46:15 voltage at 100% varies from 216 to 222 on my machine iirc Aug 24 06:47:04 That sounds normal. I think XorA's machine just got confused as it all appears to be doing what it should... Aug 24 06:48:42 wierd while my machine was "confused" it showed a battery percentage while on AC in GPE applet, not it is charging it shows 100% Aug 24 06:49:41 XorA: That sounds like someone got the logic inversed in the battery driver apm code ;-) Aug 24 06:51:08 I like to see infos like 'on AC, charging, battery is 70% charged now' Aug 24 06:51:38 hrw|work: We can do that but not through apm without breaking the specification Aug 24 06:51:48 just rebooted again with usb connected, and it starts with LED on, then turns it off Aug 24 06:52:01 I don't want to use apm for the battery info... Aug 24 06:52:05 Hello. Aug 24 06:52:09 XorA: Watch for it flicking on then off again Aug 24 06:52:15 hi sirfred Aug 24 06:52:23 Hi, RP Aug 24 06:52:38 seems to be pretty soliddly off Aug 24 06:53:46 RP: iirc we have all needed info now.. /sys/devices/platform/corgi-battery/battery_percentage give one info Aug 24 06:54:17 XorA: We'll have to see what dmesg says then Aug 24 06:55:45 RP: does /sys/devices/platform/corgi-battery/power/state have to give info about charging/not_charging? Aug 24 06:55:55 RP: the initial boot info is lost by the time I get far enough to get mdesg :-( Aug 24 06:56:32 hrw|work: No, that state file is something different. Adding extra information is easy though Aug 24 06:58:15 http://pastebin.ca/20980 Aug 24 07:01:26 XorA: As you say, too late in the bootup :-( Aug 24 07:09:11 <__law__> has anyone else mt pull problems? Aug 24 07:09:21 always Aug 24 07:09:28 * CosmicPenguin hugs hg Aug 24 07:10:08 __law__: are you on oe@handhelds.org ? Aug 24 07:10:21 if not, might subscribe to it Aug 24 07:10:42 <__law__> mickeyl, no why Aug 24 07:11:08 mickeyl: Hi. Aug 24 07:11:17 __law__: essential oe info Aug 24 07:11:50 morning Aug 24 07:12:01 everyone with OE core access should be on the mailing list or at least regularly read the archives Aug 24 07:12:54 morning kergoth Aug 24 07:13:06 hey kergoth_ Aug 24 07:14:12 morning kergoth Aug 24 07:14:44 * koen wonders where to drill a hole in the slug for the serial cable Aug 24 07:17:19 * wrobbie learned a few new things from OE newsletter sample session Aug 24 07:17:50 mickeyl: Hi, could you please tell me if there is an easy way to install all essential python packages at once. Thanks Aug 24 07:18:15 who defines what Aug 24 07:18:18 s essential? Aug 24 07:19:21 mickeyl: we just need an app to autogenerate python dependencies. run it against the script the user is trying to run and have it emit a list of the packages tehy should install Aug 24 07:19:27 heh Aug 24 07:19:30 mickeyl: Yes, I also found the newsletter extremely helpful - I'd missed a couple of developments :) Aug 24 07:19:48 that sounds like what I am after ;-) Aug 24 07:20:02 RP :) Aug 24 07:20:14 kergoth: it's nearly impossible to do that Aug 24 07:20:31 I sent into to one of Polish news services about newsletter Aug 24 07:20:40 you can statically look at import statements, but you won't grab all dependencies that way Aug 24 07:21:07 mickeyl: then I will write a script which intall's most of them Aug 24 07:21:46 worst case you could swipe the parser from the python interpreter's codebase or something. Aug 24 07:21:46 heh Aug 24 07:22:29 it's not a parser thing Aug 24 07:22:42 you can do fancy things @ runtime Aug 24 07:22:51 foo = "mymodule" Aug 24 07:22:54 __import__( foo ) Aug 24 07:23:54 mickeyl: the python interpreter knows what really needs to be imported given what the user is doing. worst case we'd hack up an interpreter to remove the unnecessary bits Aug 24 07:24:02 heh Aug 24 07:27:56 RP: ah, even though charge LED is off it seems unit is charging Aug 24 07:29:47 * __law__ oe pull works again ;-) Aug 24 07:31:30 XorA: So its just the LED not reflecting the state. Interesting... Aug 24 07:41:05 * RP plays with gtk+-2.8 on spitz Aug 24 07:41:58 RP: you have a .bb for that? Aug 24 07:42:37 koen: .bbs, yes Aug 24 07:42:48 I've not ported all the patches yet though Aug 24 07:43:03 I'm not ported any of the patches really, just built without them... Aug 24 07:44:10 koen: I was wondering whether to check them in with DEFAULT_PREFERENCE=-1 ? Aug 24 07:44:26 RP: no need for that Aug 24 07:44:48 preferred-versions-gpe-2.7.inc selects the right GTK and glib Aug 24 07:45:33 koen: Are you sure? I didn't have to alter that to build with the newer versions... Aug 24 07:46:02 preferred-gpe-versions-2.7.inc:PREFERRED_VERSION_gtk+ ?= "2.6.7" Aug 24 07:46:04 I am using an older oe tree so if that was a recent change, that would explain the difference... Aug 24 07:49:25 koen: I'm using an older oe and DISTRO=oz which includes preferred-gpe-versions.inc which doesn't set a version for gtk+ Aug 24 07:50:43 that's expected behaviour Aug 24 07:50:56 we only care about versioned distro's Aug 24 07:52:18 Hi koen! Aug 24 07:52:44 RP: 2.8 is ABI and API compatible with 2.6, so no need for DEFAULT_PREFERENCE Aug 24 07:52:45 hey lamikr Aug 24 07:53:25 koen: ok. I'll list the other things I had to upgrade in a minute and if its ok, I'll push Aug 24 07:53:31 We have finally submitted h6300 specific patches to bugzilla. I have tested them working on yesterday nights familiar-0.8.3 Aug 24 07:53:44 http://bugs.openembedded.org/show_bug.cgi?id=255 and http://bugs.openembedded.org/show_bug.cgi?id=256 Aug 24 07:58:10 After updating its just pango and glib missing so I should be ok to push? Aug 24 07:58:15 hi Aug 24 07:58:28 RP: glib is in already Aug 24 07:58:31 pango too Aug 24 07:58:46 and atk Aug 24 07:58:52 and cairo Aug 24 07:59:11 hi CoreDump Aug 24 07:59:26 I went to glib 2.8.1 ;-) Aug 24 07:59:34 aha Aug 24 07:59:47 I go away for a week and they release new stuff ;) Aug 24 08:00:05 and pango_10.1 ;-) Aug 24 08:00:18 I'll tidy it up and push Aug 24 08:09:24 03koen 07org.oe.dev * rf83ca42f... 10/packages/gpe-mini-browser/gpe-mini-browser_cvs.bb: Add sqlite to gpe-mini-browser depends Aug 24 08:10:06 RP: Am I able to build gpe-image with those Å„ew libs if I just change the DISTRO to pure familiar insteat of using familiar-0.8.3 Aug 24 08:12:05 lamikr: Yes. They're not in yet though... Aug 24 08:12:50 koen: gpe-mini-browser and gpe-pim use sqlite2 still? Aug 24 08:13:44 hrw|work: yes Aug 24 08:19:15 koen: any plans to move to sqlite3? Aug 24 08:20:02 hrw|work: plans yes, just no developer time available Aug 24 08:20:55 could use edb :P Aug 24 08:21:04 no Aug 24 08:21:11 that's berkely db Aug 24 08:21:13 since e depends on gpe anyway now Aug 24 08:21:17 yeah i know Aug 24 08:21:41 koen: ok Aug 24 08:22:03 curiousl why you would use sqlite for a browser tho Aug 24 08:22:13 seems a bit overkill Aug 24 08:29:07 CoreDump|home: http://www.vanille.de/temp/efl-altboot.png Aug 24 08:29:22 CoreDump|home: just a mockup :) Aug 24 08:29:27 ooh. Aug 24 08:29:31 cool :D Aug 24 08:29:40 nice Aug 24 08:29:47 mickeyl: sweetness!!!! Aug 24 08:29:51 hehe Aug 24 08:29:53 fluff Aug 24 08:29:58 unnecessary eye candy Aug 24 08:30:10 * CoreDump|home likes eyecandy Aug 24 08:30:22 I think trolltech just hired a hitman Aug 24 08:30:39 :) Aug 24 08:34:01 239x321, funky Aug 24 08:34:04 altboot using gtk/x11 will be next step Aug 24 08:34:30 nah, this will remain fb-only Aug 24 08:34:35 evasfb would be more appropriate Aug 24 08:34:43 my goal is to make it fit into the spitz flash Aug 24 08:34:50 which means 6MB is the limit Aug 24 08:34:53 or was it 5? Aug 24 08:35:27 spitz/2.6 should be able to boot stright from /dev/hda1 Aug 24 08:36:32 hrw|work: sure, but consider altboot being more than just a bootmanager, i.e. also backup functionality via network and stuff. It makes sense to have nothing running from hd in that case Aug 24 08:36:43 yep Aug 24 08:37:48 next thing I'll do will be to add tslib support to ecore-fb Aug 24 08:37:56 atm. it's hardcoded to some h3600 thingy Aug 24 08:37:59 :D Aug 24 08:38:05 monotone doesn' Aug 24 08:38:07 heh Aug 24 08:38:15 monotone doesn't have a copy command does it? Aug 24 08:38:23 RP: not yet Aug 24 08:38:23 * koen looks at the 3870 running opie on his desk Aug 24 08:38:55 so we just copy files and lose history? Aug 24 08:39:08 when you copy yes Aug 24 08:39:12 you can move instead Aug 24 08:39:26 i think this is on the TODO list Aug 24 08:39:45 np, just checking :) Aug 24 08:41:28 mickeyl: nice work on the latest newsletter, finally read it. good stuff. Aug 24 08:42:13 thanks Aug 24 08:42:34 mickeyl: iirc reenoo asked for a 'copy' command and they weren't interested Aug 24 08:43:18 koen: too bad, that's handy for projects like ours. guess that most people just rename and are not interested in having multiple files with the same history Aug 24 08:43:35 RP: I added few cards from hostap_cs.conf to hostap cardid table: http://pastebin.ca/20988 Aug 24 08:44:05 i'm going to look at getting tailor to support a monotone source, which will make it easy to test out other source control tools without losing history, or maintain gateways to other tools Aug 24 08:44:12 i think we should seriously consider svk, myself. Aug 24 08:44:21 mickeyl: how about having a netstumbler (http://www.netstumbler.com/) style 'o' in the A for angstrom? Aug 24 08:44:34 hrw|work: fine. feel freel to update the diff Aug 24 08:44:36 hostap 0.4.4 have support for pcmcia changes in 2.6.13 Aug 24 08:44:44 2005-08-21 - v0.4.4 Aug 24 08:44:44 * added support for pcmcia changes in Linux 2.6.13 Aug 24 08:44:44 * added support for creating sysfs device files (Linux 2.6.x) Aug 24 08:45:23 03rpurdie 07org.oe.dev * r8981688a... 10/packages/glib-2.0/ (glib-2.0_2.8.0.bb glib-2.0_2.8.1.bb): glib-2.0: upgrade to 2.8.1 Aug 24 08:45:25 Time to upgrade I suspect :) Aug 24 08:45:27 03rpurdie 07org.oe.dev * r0ad28d6b... 10/packages/pango/pango_1.10.0.bb: pango: add version 1.10.0 Aug 24 08:45:32 03rpurdie 07org.oe.dev * rab3bfea8... 10/packages/gtk+/ (14 files in 2 dirs): Add gtk+ 2.8.0 Aug 24 08:47:18 hmm Aug 24 08:47:52 RP: add_event.patch looks like applied, hostap_cardids.patch also Aug 24 08:47:53 kergoth, you read my paste lastnight? Aug 24 08:48:01 dont think so, what was it? Aug 24 08:48:04 hey good news for monotone: njs emte: and we have a Plan For Fixing That, hopefully will be working in a ~few months Aug 24 08:48:10 re: speed and computational req Aug 24 08:48:15 ah Aug 24 08:48:17 koen: you mean the antenna symbol? Aug 24 08:48:19 cool Aug 24 08:49:13 njs and mpm were discussing monotone, i belive mpm was getting info for his presentation at linux.conf.au Aug 24 08:49:17 hmm, i should run up to the nearby grocery store and get some soda to bring to work Aug 24 08:49:20 emte: ah Aug 24 08:49:54 lot of discussion about monotone vs bk design history Aug 24 08:49:59 whats the irc channel for hg? Aug 24 08:50:08 #mercurial Aug 24 08:50:46 very active group Aug 24 08:50:57 hrw|work: That would be what they needed to add to have it work with 2.6.13 Aug 24 08:52:01 looks like our hostap_cs.conf need cleaning and comparing with upstream one Aug 24 08:52:59 and if your curious it looks liek the newest project is the hgeditor Aug 24 08:53:45 simular to how bk worked i think ... Aug 24 08:54:35 okay back to biohazard .... time for all the mouse/touch callbacks Aug 24 08:54:53 er, fig segfault first Aug 24 08:54:56 fix* Aug 24 08:55:59 mickeyl: yes, the antenna symbol Aug 24 08:57:57 Afternoon all; I've just done a bitbake gpe-image for which I pulled last night; I have a strange error in log.do_rootfs: http://pastebin.ca/20989 Aug 24 08:58:04 Basically the error for unsatisfied deps for task-bootstrap, those modules exist as ipks, I don't know why they're not picked up Aug 24 08:59:43 Lardman I have the same error Aug 24 09:00:01 would the naming make any difference to whether they are picked up? Mine seem to have a 2.6 prefix before the kernel version: http://pastebin.ca/20990 Aug 24 09:00:37 With me task bootstrap generates an error but it is not saying what it is missing Aug 24 09:00:56 gerwinin: even in the log file Aug 24 09:00:57 ? Aug 24 09:01:26 Yes in the log file it is not saying anything about this Aug 24 09:01:36 It is saying unsatisfied depency Aug 24 09:02:25 RP: gtk 2.8.0 is so last week ;) Aug 24 09:02:33 lardman|work: http://bugs.openembedded.org/show_bug.cgi?id=51 Aug 24 09:02:52 03rpurdie 07org.oe.dev * r306834bf... 10/packages/gtk+/gtk+_2.8.0.bb: gtk-2.8.0: syntax correction Aug 24 09:03:00 RP: with 2.6.13-rc6-mm2 hostap-modules 0.4.4 compiles without utsname.patch Aug 24 09:03:12 gerwinin: are you sure the log file (not the console output) doesn't contain any detailed info? Aug 24 09:04:21 reenoo_: Thanks, I'll work through Aug 24 09:04:47 renoo_ I amm looking at log.do_rootfsdoesn't say anything interesting Aug 24 09:05:22 gerwinin: look about half way through, that's where my pastebin.ca post came from Aug 24 09:06:34 hrw|work: That's good to know Aug 24 09:09:11 Okay Aug 24 09:12:11 mickeyl: can you look at http://pastebin.ca/20993 - its hostap_cs.conf a bit changed to be more up-to-date with upstream one Aug 24 09:13:25 NOTE: package hostap-modules-0.4.4-r0: task do_build: completed Aug 24 09:13:59 * Twiun applauds hrw|work Aug 24 09:14:44 someone familiar with hostap should post our changes to hostap_cs.conf for upstream Aug 24 09:16:31 hrw|work: why do you do that? Aug 24 09:16:45 don't forget that our hostap_cs.conf is _appended_ to the upstream one, not replacing it Aug 24 09:16:47 :D Aug 24 09:16:54 aha ;) Aug 24 09:17:01 tricky, isn't it? :)) Aug 24 09:17:19 mickeyl: so it needs cleaning a bit then Aug 24 09:17:23 possible Aug 24 09:17:31 some things may have landed upstream already Aug 24 09:19:33 how's the stuff with the celf people working out? Aug 24 09:20:11 http://pastebin.ca/20994 - modinfo hostap_cs 0.4.4 Aug 24 09:22:09 hmmm i will take some time to get used to the new style aliasses Aug 24 09:22:40 "modprobe pcmcia:m*c*f*fn*pfn*pa4B801A17pb6345A0BFpcC9049A39pdC23ADC0E*" is so easy to remember... Aug 24 09:22:46 hehe Aug 24 09:23:11 pcmciautils will need modutils-modprobe for it Aug 24 09:23:17 * mickeyl wanders home Aug 24 09:23:22 the full thng? Aug 24 09:23:23 bummer Aug 24 09:23:36 mickeyl: to get module aliases Aug 24 09:28:25 cu Aug 24 09:28:36 03rw 07org.oe.dev * r67c06308... 10/packages/vlc/ (4 files in 3 dirs): vlc-gpe: make 0.8.1 work with more recent tremor API. patch courtesy of Philipp Zabel. fixes Bug #205. Aug 24 09:47:01 mickeyl: Do you have gpe with a 2.4 kernel on spitz at the moment? Aug 24 09:48:37 The config line in /etc/X11/Xserver looks wrong to me, even for 2.4. I suspect it needs 640x480@270 changed to 480x640@270 - the console might be rotated but I doubt the framebuffer is... Aug 24 09:48:59 RP: I do, yes Aug 24 09:49:28 RP: I'm the one who added that line for spitz...is it not correct? it works for me Aug 24 09:49:52 and it worked for jnc when he was testing it Aug 24 09:49:54 JustinP: Could you try changing the 640x480 to 480x640 and see if Xserver still works please? You'd better stop gpe-dm from initing on boot just in case... Aug 24 09:50:05 (and I assume it's workin on the e-image since people haven't yelled at me about it) Aug 24 09:50:18 JustinP: It might work but it looks wrong - the framebuffer is 480x640, not 640x480 Aug 24 09:50:34 It doesn't support hardware rotation Aug 24 09:50:44 RP: surely I don't need to reboot to make it take effect... Aug 24 09:51:11 JustinP: No, just restart X - if it screws up the display - reboot instantly though Aug 24 09:51:14 and I think I tried that at onc epoint but it didn't help my TS rotation problems so I left it as it is now... Aug 24 09:51:18 ok Aug 24 09:52:26 I'm seeing certain problems under 2.6 which I doubt will happen with 2.4 - 2.6 pxafb has a massive black hole in its mode handling... Aug 24 09:52:45 worke exactly the same Aug 24 09:52:56 That's what I thought would happen Aug 24 09:53:01 the only reason that's in there is to fix rotation problems Aug 24 09:53:19 the screen size works fine if that isn't there, it's just rotated incorrectly Aug 24 09:53:37 Yes, you can keep the @270 which should do that Aug 24 09:54:06 It just should be 480x640@270 and not 640x480@270 Aug 24 09:54:21 ok Aug 24 09:54:24 So I'm ok to change that in OE? Aug 24 09:54:25 I'll check in that change Aug 24 09:54:28 or you can Aug 24 09:54:29 sure Aug 24 09:54:37 no problems Aug 24 09:54:42 I'm in the file so I can do it Aug 24 09:54:47 does it fix your 2.6 problems? Aug 24 09:54:59 make sure you change the right one... Aug 24 09:55:09 Yes, 2.6 works if I fix that Aug 24 09:55:30 packages/x11/x11-common/etc/X11/Xserver, right? Aug 24 09:55:33 There is a bug here as well as 2.6 should bomb out with the setting that's there at the moment Aug 24 09:56:10 I can see that Xserver-cxk.patch in gpe-dm/files has soething to do with this... Aug 24 09:56:19 not any more Aug 24 09:56:26 look at the "remove files" patch Aug 24 09:56:34 I'm removing Xserver from gpe-dm Aug 24 09:56:39 it's in x11-common now Aug 24 09:56:47 there was a discussion on the GPE list Aug 24 09:57:07 JustinP: I'm not subbed to GPE :) Aug 24 09:57:25 I'll alter the other file then, thanks for the pointer :) Aug 24 09:57:28 I moved a bunch of scripts from gpe-dm and gpe-session-scripts to x11-common to allow for other X systems to use them (i.e. e17) Aug 24 09:57:31 np Aug 24 09:58:02 right now they're removed with makefile patches...hopefully they'll get updated soon and we can put up a tarball for x11-common somewhere..... Aug 24 09:58:45 hey JustinP Aug 24 09:59:04 emte: hey Aug 24 09:59:07 got to run, meeting Aug 24 09:59:10 I'll be back in a bit Aug 24 09:59:24 i am curious ... you are aware that e17 has its own rotation code ... Aug 24 09:59:53 night all Aug 24 10:00:29 emte: Its more just specifiying which is the default rotation for different devices... Aug 24 10:02:01 yeah i understand that Aug 24 10:02:23 i just wasnt sure if he was aware of it Aug 24 10:02:34 e doesnt care which direction it goes Aug 24 10:03:31 emte: excellent :). To complicate things the C series Zaurii have switches so the screen can auto rotate to the right orientation depending on how the device is held :) Aug 24 10:03:31 03rpurdie 07org.oe.dev * r26a17ae9... 10/packages/x11/x11-common/etc/X11/Xserver: x11-common: Correct spitz resolution - confirmed on 2.4 and fixes Xserver on 2.6 kernels Aug 24 10:03:39 i was playing with the rotation code the otehr day is why it just occured to me Aug 24 10:04:02 nah that doesnt complecate it Aug 24 10:04:11 setup a listener to auto-rotate Aug 24 10:04:19 problem solved Aug 24 10:05:04 i am discovering ecore's listeners to be very handy things Aug 24 10:14:25 emte: They sound it :). Just needs someone to write the platform support code I guess Aug 24 10:21:18 http://enlightenment.org/doxy/ecore/config__listener__example_8c-example.html Aug 24 10:21:24 RP, example for you Aug 24 10:21:29 easy peasy Aug 24 10:21:49 all that one does is alter ecore_config Aug 24 10:22:41 i'll keep in mind the zaurus-rotate tho when i setup biohazard's Aug 24 10:23:14 i wasnt aware it had a mercury/tilt switch Aug 24 10:23:58 emte: Not tilt switch - the screen has a hinge and it can change from a clamshell to a tablet Aug 24 10:24:15 There are physical switches to detect the position of the screen Aug 24 10:24:59 ah, i thought you ment it was a gravity rotate Aug 24 10:25:22 tho the code would be the same Aug 24 10:25:25 emte: No, but the principle is the same Aug 24 10:25:34 yup Aug 24 10:25:42 When you fold the screen, its nice to have it react :) Aug 24 10:25:57 yup Aug 24 10:42:59 mickeyl: is pcmcia applet included in all images? Aug 24 11:01:06 hi Aug 24 11:01:09 mickeyl: ping Aug 24 11:19:35 good evening !!! Aug 24 11:19:40 hey Aug 24 11:20:01 hi kergoth ! Aug 24 11:20:44 03hrw 07org.oe.dev * rb1c39c88... 10/packages/hostap/ (4 files in 2 dirs): added hostap modules 0.4.4 (development version which works with kernel 2.6.13) Aug 24 11:20:50 03hrw 07org.oe.dev * r6ea9d797... 10/packages/pcmciautils/ (pcmciautils_005.bb pcmciautils_007.bb): updated pcmciautils to 007 Aug 24 11:20:55 03hrw 07org.oe.dev * r5a11cecc... 10/packages/pcmciautils/ (pcmciautils_001.bb pcmciautils_002.bb pcmciautils_003.bb): removed old versions of pcmciautils Aug 24 11:27:41 mhhh question about bitbake ... what are the full list of commands ( supplied with -c ) ??? Aug 24 11:28:34 do a bitbake -c listtasks on the .bb file and look :P Aug 24 11:28:34 bitbake -b file.bb -clisttasks Aug 24 11:28:49 thanks ... Aug 24 11:32:46 I have put patches that add h6300 support today to bugzilla, is there any hope somebody could check them today? I will leave on friday for a one week trip, so if we need to make some changes I can only try on today or tomorrow. (bugs 255 and 256) Aug 24 11:33:23 kergoth_: is your monotone key in already? Aug 24 11:33:49 koen: heh, mickey did so, but then i did something stupid.. removed my .db without exporting my keys Aug 24 11:33:54 so now i need to generate a new set :) Aug 24 11:37:30 in update-alternatives higher priority value means higher priority or vice versa? in other words - does 20 is more important then 50 or 50 then 20? Aug 24 11:38:10 from the man page: Aug 24 11:38:14 " Each alternative has a priority associated with it. When a link group Aug 24 11:38:14 is in automatic mode, the alternatives pointed to by members of the Aug 24 11:38:14 group will be those which have the highest priority. Aug 24 11:38:14 " Aug 24 11:38:29 ok Aug 24 11:38:40 so i'd say 50 is more important than 20 Aug 24 11:39:07 man, these coffee shops are gonna bankrupt me. i need to buy an espresso/cappuccino machine Aug 24 11:39:41 i can't find where (specially in kernel 2.6 but also 2.4) is defined the partition scheme of the flash ... my attemps to run kernel 2.6 in h3600 fail cause a kernel panic on trying to mount root fs :( ... Aug 24 11:43:27 still trying to build gpe-image. now getting "Cannot satisfy the following dependencies for gpe-task-base: hotplug" Aug 24 11:44:17 I see build/tmp/work/hotplug-dbus-0.9-r1 Aug 24 11:44:19 Ideas? Aug 24 11:44:41 does hotplug-dbus provide 'hotplug'? Aug 24 11:44:54 mreimer: How about if you run "bitbake hotplug" and then "bitbake gpe-image"? Aug 24 11:44:55 hi koen. How do I check that? Aug 24 11:45:15 lamikr: ERROR: Nothing provides hotplug Aug 24 11:45:37 mreimer: ar x build/tmp/work/hotplug-dbus-0.9-r1.ipk and check control.tar.gz Aug 24 11:46:31 or check build/tmp/work/Packages Aug 24 11:47:18 koen: what am I looking for? Provides: hotplug? Aug 24 11:47:30 something like that Aug 24 11:47:56 mreimer: hotplug-dbus RDepend on hotplug Aug 24 11:48:31 kergoth_: so module-init-tools needs update to be higher priority then busybox Aug 24 11:48:43 hrw: yes, "Depends: hotplug" Aug 24 11:49:42 koen: linux-hotplug-doc and linux-hotplug both provide, replace, conflict with hotplug Aug 24 11:49:49 is that a problem? Aug 24 11:50:18 iirc linux-hotplug is the correct one Aug 24 11:50:31 but I confused them earlier Aug 24 12:01:51 03ph5 07org.oe.dev * r74801abf... 10/conf/bitbake.conf: mkdir the 'patches' directory before calling quilt Aug 24 12:02:39 heh, thats a quick and easy fix for that bug Aug 24 12:03:34 kergoth_: wait till you see the next one Aug 24 12:05:34 Wait - I thought we already did that Aug 24 12:05:37 a long time ago Aug 24 12:09:02 03ph5 07org.oe.dev * r65c6c7a1... 10/classes/kernel.bbclass: fix regexp to strip cvs kernel versions, too. (fixes #230) Aug 24 12:13:25 Has others expericenced that /etc/profile file build by gpe-image does not have executable bits "x" set? That's happening in familiar-0.8.3 distro. Aug 24 12:14:28 /etc/profile should not have the executable bits set Aug 24 12:16:16 lamikr: it should be sourced by the shell, not executed. Aug 24 12:18:08 what's a good machine conf file to start from for a 2.6 kernel? I want to make h2200.conf Aug 24 12:19:37 is that really needed? Aug 24 12:19:52 I dunno. What would be the right way to build a rootfs for h2200? Aug 24 12:19:53 e.g different mkfs.jffs2 params Aug 24 12:20:26 KERNEL="kernel26" in local.conf and MACHINE=h3900 Aug 24 12:21:06 mreimer was u working on kernel 2.6 for h3600 ??? Aug 24 12:21:07 that's what I'm doing right now, but I'm getting errors: Cannot satisfy the following dependencies for task-bootstrap: kernel-module-h3900-lcd-2.6 kernel-module-h5400-lcd-2.6 kernel-module-asic2-adcts-2.6 kernel-module-apm-2.6 kernel-module-h5400-battery-2.6 kernel-module-h3900-battery-2.6 kernel-module-pcmcia-core-2.6 Aug 24 12:21:11 ok, then I have a somekind of problem. We have /etc/profile.d/tslib.sh that is executed by the /etc/profile script. Tslib.sh sets two environment variables that are needed by the touschscreen driver but for some reason that script is not executed before /etc/init.d/gpe-dm is started. --> touschreen driver does not work. I have put that working simply by adding call to /etc/profile from the /etc/init.d/gpe-dm script but it seems that this is not the cor Aug 24 12:21:17 gremlin[it]: no, h2200 Aug 24 12:21:30 mreimer ops my mistake :) Aug 24 12:21:52 cu Aug 24 12:22:29 lamikr: init scripts are not run by login shells, and as such do not get the profile sourced. having the init script do so is probably the best solution, or execute gpe-dm in the script using exec -l (iirc, exec's it using a login shell) Aug 24 12:31:26 Wondering how this is working for other devices like h3600, as we are using same initscripts than other devices... We have Just added /usr/share/tslib/ts.conf-h6300 and /etc/profile.d/tslib.sh configs to use it. h2200 and h3600 uses same approach. Aug 24 12:33:08 03ph5 07org.oe.dev * r4a2c1667... 10/classes/kernel.bbclass: Aug 24 12:33:08 makes the kernel-module-af-packet package add "alias net-pf-17 af_packet" to Aug 24 12:33:08 /etc/modules.conf (fixes #233). Aug 24 12:39:26 03ph5 07org.oe.dev * r1d77ddb0... 10/packages/libdaemon/libdaemon_0.6.bb: install dexec.h into the staging area Aug 24 13:00:32 '/clear Aug 24 13:00:45 ooops - sorry :) Aug 24 13:12:23 koen: yes, it should be in all opie-images. Aug 24 13:12:26 hrw|gone: pong Aug 24 13:12:57 mickeyl: I noticed Aug 24 13:13:18 the conceptronic pcmcia card is just rubbish Aug 24 13:13:33 conceptronic? what's that doing Aug 24 13:13:42 10mbit ethernet card Aug 24 13:13:47 ah. Aug 24 13:13:50 * mickeyl hugs his belkin Aug 24 13:13:54 no Aug 24 13:13:55 billionton Aug 24 13:14:15 so it will be either wifi or usb Aug 24 13:14:42 * koen is going to setup his own 'cluster' Aug 24 13:14:45 whois ph5?` Aug 24 13:14:53 philip zabel Aug 24 13:15:08 hmm, k. Aug 24 13:15:19 * mickeyl always slightly worried when he sees a new nickname touching core files Aug 24 13:15:59 emte: I'm not sure what you mean by "e17 has its own rotation code. I have used xrandr to rotate my screen (this is what GPE uses) and it works fine with e17. Are you saying that e has a built-in way of doing this? Aug 24 13:16:20 xrandr is the X way Aug 24 13:16:21 JustinP: don't take too much attention :D Aug 24 13:16:23 emte: AFAIK the rotation is an X thing, not a WM thing. e17 can rotate its orientation but wouldn't this leave the apps rotated the wrong way? Aug 24 13:16:26 anything other is crap Aug 24 13:16:31 heh Aug 24 13:16:34 ok, thanks Aug 24 13:16:40 is e17 running in X? Aug 24 13:16:44 I really don't know anything about xrandr Aug 24 13:16:46 yes, it's in X Aug 24 13:16:49 I was under the impression it was running on the fb Aug 24 13:16:50 hence xrandr works Aug 24 13:16:51 treke: yeppa. Aug 24 13:16:55 nope Aug 24 13:16:56 treke: configurable Aug 24 13:17:01 treke: no, most e17 apps are X only yet. Aug 24 13:17:05 e17 and pretty much all the other *programs* need X Aug 24 13:17:08 ah Aug 24 13:17:16 JustinP: what's the best way to give e17 a try with OE? Aug 24 13:17:21 the libraries are set up to run on fb/qt/whatever as well, though Aug 24 13:17:29 mreimer: bitbake e-image Aug 24 13:17:32 thanks Aug 24 13:17:47 I thought I remembered something about raster doing an evas only e17 as well Aug 24 13:17:48 mreimer: although the unzipped image is >80MB, maybe over 90MB now Aug 24 13:17:48 maybe not Aug 24 13:18:04 treke: raster told me himself that e17 was X only Aug 24 13:18:19 Then I'm clearly mistaken Aug 24 13:18:32 treke: it *should* be possible...but all the apps would have to be built on evas as well. A WM is of no use without programs. Aug 24 13:18:33 what's the RAM footprint like? Aug 24 13:18:42 mreimer: ummm....let me check.... Aug 24 13:18:53 works fairly well on 64MB RAM devices Aug 24 13:19:14 "mouse" buttons being the only major thing to solve before we can release such an image Aug 24 13:19:23 oh and a lighter theme Aug 24 13:19:26 but that's minor Aug 24 13:19:57 mickeyl: load up the engage module and start xmonobut ;-) Aug 24 13:20:17 sounds like a Start Trek command Aug 24 13:20:26 "make it so" Aug 24 13:20:34 e16 did have an lcars theme Aug 24 13:20:38 Then, you should rerount the power through the phasers and cause the DNA in the gas cloud to realign Aug 24 13:20:43 'energie' Aug 24 13:20:45 * koen shrugs Aug 24 13:21:00 the german voice-overed startrek is just to scary Aug 24 13:21:17 hehe Aug 24 13:21:45 mhhh should sound good klingon in doich :) :) :) Aug 24 13:22:03 ~57MB used on my Z with 1.3MB buffers and 33MB cached Aug 24 13:22:28 so....23MB or so? Aug 24 13:23:15 I suppose that's enough for another decent app or two to run Aug 24 13:23:18 (assuming I'm reading that right) Aug 24 13:23:31 of course I also have a swap partition ;-) Aug 24 13:24:06 actually it's not on right now Aug 24 13:24:13 forgot ot turn it on after my rebuild Aug 24 13:24:29 I need to rebuild again, actually...I've still got a hybrid GPE/e17 for my main boot Aug 24 13:25:00 rebuild? how yesterday Aug 24 13:25:02 there is screenshoot of e17 ? Aug 24 13:25:16 JustinP: http://vanille.de/temp/efl-altboot.png ;) Aug 24 13:25:37 mickeyl: huh? I mean that the "image" on my zaurus was the GPE image and I installed e17 on top of it. I've been testing my e-images with image files. Aug 24 13:25:47 mickeyl: I saw that, looks sweet Aug 24 13:25:50 JustinP: i know. i was just longing for a chance to post this mockup again :)) Aug 24 13:25:54 mickeyl: too bad it's only a mockup Aug 24 13:25:59 nice! Aug 24 13:26:00 well Aug 24 13:26:02 lets say Aug 24 13:26:07 60% is mockup Aug 24 13:26:09 40% is code Aug 24 13:26:12 * mickeyl grins Aug 24 13:26:22 you actually have some done? awesome :-) Aug 24 13:26:36 gremlin[it]: http://handhelds.org/scap/ Aug 24 13:26:46 oops...damn Aug 24 13:26:46 most work is my C++ abstraction layer, because i can't stand C Aug 24 13:26:51 heh Aug 24 13:26:53 there were some e17 screenshots there a little while ago Aug 24 13:27:15 mickeyl: ah....so you're readying your C++ layer for use? Any actual altboot code done? Aug 24 13:27:24 mickeyl: hehe, i've been shamelessly posting your mockup to #elinux and places ;) Aug 24 13:27:33 kergoth: excellent :D Aug 24 13:27:51 unfortunately you can't see it's animated. the two horizontal bars are glowing Aug 24 13:28:00 oooh Aug 24 13:28:00 ooh. Aug 24 13:28:04 and to be honest... I ripped that as part of an E background Aug 24 13:28:21 btw, cairo 1.0 is out Aug 24 13:28:25 cool Aug 24 13:29:28 FYI: first dude gives actual feedback on OZ/3100. I guess it somehow works then Aug 24 13:29:39 mreimer: here you go: http://handhelds.org/scap/port.5276.png Aug 24 13:29:42 mmmm... cairo Aug 24 13:29:58 JustinP: sweet! Aug 24 13:30:43 hopefully cairo will get a little acceleration on ARM (some asm and soft-float for starters) Aug 24 13:30:47 cute Aug 24 13:31:13 * koen wants the kdrive mmx code working on pxa27x Aug 24 13:31:56 mreimer: 2 more: http://handhelds.org/scap/port.5604.png Aug 24 13:32:09 http://handhelds.org/scap/port.5655.png Aug 24 13:32:31 JustinP: you're torturing me. I wonder how it will look on 240x320? Aug 24 13:32:40 mreimer: try it and see... Aug 24 13:32:50 mreimer: I'm not sure how usable entrance will be on that size screen... Aug 24 13:32:55 JustinP: I hope to Aug 24 13:33:22 has anyone successfully built gpe-image for 2.6 recently? Aug 24 13:33:42 NOTE: package cairo-1.0.0: completed Aug 24 13:33:50 koen: you're fast :-) Aug 24 13:33:56 JustinP: E's theming is absolutely great. Making entrance work on 240x320 is just a matter of adjusting the theme. Aug 24 13:34:11 mreimer: I just copied the 0.9.2 .bb and adjusted the SRC_URI :) Aug 24 13:34:38 mickeyl: well, I know it's good, I just haven't played with the theme stuff as yet.... Aug 24 13:34:56 mickeyl: actually....e will be compiled with the lower profile for the smalle rdevices, won't it? Aug 24 13:35:12 although I'm not sure if this affects entrance.... Aug 24 13:35:15 what provides things like kernel-module-h3900-lcd-2.6? I'm getting unsatified dependencies when building gpe-image Aug 24 13:35:26 JustinP: yes, but that won't affect the entrance theme. Aug 24 13:35:39 should be pretty easy to add a post-compile bbclass which themes Aug 24 13:35:42 mreimer: I think the defconfig needs to be adjusted Aug 24 13:36:12 is that packages/linux/handhelds-pxa-2.6/defconfig ? Aug 24 13:36:26 03koen 07org.oe.dev * r889c17fb... 10/packages/cairo/cairo_1.0.0.bb: Add cairo 1.0.0 Aug 24 13:36:29 I think so Aug 24 13:36:39 but I haven't looked into kernels for a while Aug 24 13:36:59 Fatal error: Call to undefined function: get_bloginfo() in /home/groups/o/op/openzaurus/htdocs/wordpress/wp-includes/wp-db.php on line 304 Aug 24 13:37:00 eh Aug 24 13:37:02 thats no good Aug 24 13:37:31 got that too when i hit a layer 2 likn Aug 24 13:37:33 link even Aug 24 13:37:45 didn't occur when starting from the main site Aug 24 13:38:08 odd Aug 24 13:38:35 ~shoot sourceforge.net Aug 24 13:38:40 * ibot shoots sourceforge.net in the eye with a glue gun! Aug 24 13:38:56 we really should move to another provider, it's unbearably slow. guess they moved oz.org to the slowest server they have Aug 24 13:39:02 ouch Aug 24 13:40:58 hehe Aug 24 13:40:59 i still have problem with "Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(31,1)" on kernel 2.6 with h3600 :( ... Aug 24 13:41:07 go sourceforge! Aug 24 13:41:44 that's set linuxargs="mtdparts=ipaq:0x00040000(bootldr)ro,0x00FC0000(root) root=/dev/mtdblock1 init=/linuxrc noinitrd console=ttySA0 rootfstype=jffs2" Aug 24 13:41:56 but i got from kernel : Aug 24 13:42:14 SA1100 flash: CFI device at 0x00000000, 16MiB, 32-bit Aug 24 13:42:14 RedBoot partition parsing not available Aug 24 13:42:17 SA1100 flash: using static partition definition Aug 24 13:42:22 Creating 4 MTD partitions on "sa1100": Aug 24 13:42:34 ... Aug 24 13:42:40 jffs2: Too few erase blocks (2) Aug 24 13:42:40 Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(31,1) Aug 24 13:43:05 gremlin[it]: using the right updater.sh? Aug 24 13:43:16 updater.sh ?!?!? Aug 24 13:43:25 gremlin[it]: just a guess Aug 24 13:43:34 no idea what you're actually doing... Aug 24 13:43:40 h3600 (the old) ... not h6300 ... Aug 24 13:44:06 JustinP: updater.sh is a sharp only thing Aug 24 13:44:34 ahhh :) ... i thought i was doing completelly wrong procedure Aug 24 13:44:47 ah....yes, ok, sorry Aug 24 13:45:14 i built a task-image with a familiar kernel26 h3600 ... using oe :) Aug 24 13:52:11 JustinP, no it handles rotating everything Aug 24 13:52:22 wethere the apps like it or not ... Aug 24 13:52:26 weather* Aug 24 13:52:51 i cant imagine a non-e app behaving very well, you are correct Aug 24 13:53:16 a *wm* handles rotating? Aug 24 13:53:17 mmm red bean bun Aug 24 13:53:22 yup Aug 24 13:53:25 are you sure you aren't on crack? Aug 24 13:53:46 there are even examples in the cookbook Aug 24 13:53:53 sec i'll grab you a link Aug 24 13:54:40 slowly guys. Evas handles rotation - which is perfectly on spec for a canvas abstraction layer. You *could* use that on the X backend as well, but it's pretty worthless. It has been designer for the non-rotatable backends like the FB. Aug 24 13:55:03 that's what I thought Aug 24 13:55:15 http://enlightenment.org/docs/eflcookbook.html#id2460452 Aug 24 13:55:18 emte: exactly, non-EFL apps are going to go bonkers. You can go ahead and patch e17 and try it yourself, but I'm positive it's not going to work for X apps Aug 24 13:55:20 and why I said '*wm*' Aug 24 13:55:26 * mickeyl watches his daylie simspons epsiode Aug 24 13:55:30 same thing I said earlier Aug 24 13:55:49 * koen should catch up with the last two seasons Aug 24 13:56:08 nah, i dont plan on doning stuff for x-apps Aug 24 13:56:23 trying to convince some programmers now to write e-apps Aug 24 13:56:33 03pohly 07org.oe.dev * rb4bff47a... 10/conf/machine/h3900.conf: bugzilla bug 83: list all supported architectures so that tune-xscale can be used Aug 24 13:56:58 getting them to do some of the cookbook examples to get a basic handle on the code Aug 24 13:57:18 emte: I can't see an example of the wm doing rotation Aug 24 13:57:21 only evas Aug 24 13:57:51 e is one evas canvas container taht contains objects Aug 24 13:58:03 are you blind? Aug 24 13:58:15 WM WM WM WM WM WM WM Aug 24 13:58:25 if you simplify it down enough that is all it is Aug 24 13:59:10 no, it isn't Aug 24 13:59:11 then you add in all your callbacks, listeners, and handlers Aug 24 13:59:24 the wm manages X windows Aug 24 13:59:29 including the rootwin Aug 24 14:00:12 it does *NOT* paint regular apps on an evas canvas Aug 24 14:00:32 no it cant Aug 24 14:00:32 so it does *NOT* do rotation Aug 24 14:00:40 * reenoo_ kindly points out that this isn't #e or whatever the channel is called Aug 24 14:01:08 * kergoth_ not so kindly points out the same thing, wielding the cluebat menacingly Aug 24 14:01:09 ;) Aug 24 14:01:11 * koen /ignores emte again Aug 24 14:01:37 * emte shrugges i never said it could rotate anything to do with x Aug 24 14:01:49 kergoth_: hehe. I've been missing that while you were away ;) Aug 24 14:02:05 hehe Aug 24 14:02:13 my old boss told me i should tell prospective employers to google for kergoth Aug 24 14:02:26 heh Aug 24 14:02:26 i considered all the ircl ogs out there of me berating some noob, and decided against that Aug 24 14:02:27 heh Aug 24 14:02:44 Mabye you can bribe tim to do a perl s/kergoth// on the irc logs Aug 24 14:02:49 hehe Aug 24 14:03:11 "no really, i'm nice when i'm paid to be! just not on my own time!" Aug 24 14:03:37 mhhh i start to suppose kernel-2.6 don't support flash for h3600 :( :( :( ... Aug 24 14:04:14 i only suggested it as a possible option for dealing with the zaurus rotate switch for the e-image nothing more Aug 24 14:04:55 gremlin[it]: I don't think 2.6 is ready for end users on h3600 Aug 24 14:05:49 no end user ... either (beginners) developer ... if u enable mtd ipaq maping kernel don't compile ;( Aug 24 14:06:42 | drivers/built-in.o: In function `h3xxx_really_set_vpp': Aug 24 14:06:54 gremlin[it], which part of the sa-1110 did you write your code for? Aug 24 14:06:54 | freq_table.c:(.text+0x40b90): undefined reference to `assign_h3600_egpio' Aug 24 14:07:00 gremlin[it]: this really isnt the channel for this. try #handhelds.org Aug 24 14:07:17 emte ... micro Aug 24 14:07:51 kergoth i know ... but here at least i got some answer ... on #handhelds.org no one speak :( ... sorry Aug 24 14:08:16 djeez Aug 24 14:08:21 yeah, the same thing happens when i try to talk to myself like reenoo_ tells me to Aug 24 14:08:35 when will people realize that opie isn't installed on every embedded device Aug 24 14:08:38 morons Aug 24 14:10:08 * koen is getting increasingly aggrevated Aug 24 14:10:14 :( Aug 24 14:10:29 koen: relax ;) Aug 24 14:10:48 ibot! :( Aug 24 14:10:48 kergoth_: ibot quit! you can berate at will again ;) Aug 24 14:10:51 hehe Aug 24 14:10:52 yay Aug 24 14:12:41 koen: calling a script (or echo "..." > ...) in up/down is a nice hack actually. I've been thinking of that myself but the point was that it's non-standard Aug 24 14:14:07 reenoo_: I use it for my tunnels Aug 24 14:14:15 up 'script foo' Aug 24 14:14:55 same here :) Aug 24 14:15:14 quite a bit of magic in there here actually >:) Aug 24 14:15:55 * koen looks at the ubholy combination of 'ip', 'route' and ipv4 Aug 24 14:21:03 ~kergoth Aug 24 14:21:05 methinks kergoth is Chris Larson, an embedded linux guy at Texas Instruments. Email at kergoth@handhelds.org or clarson@ti.com. OpenZaurus project founder. Core dev / architect on the BitBake, OZ, and OpenEmbedded projects. Aug 24 14:21:15 heh, thats old Aug 24 14:21:29 ~cluebat Aug 24 14:21:31 *WHACK* *WHACK* *WHACK* Aug 24 14:21:35 :) Aug 24 14:22:36 ~no, kergoth is Chris Larson, an embedded linux guy currently at Astrocom Corporation (http://www.astrocorp.com/). Email at kergoth@handhelds.org or clarson@astrocorp.com. Core dev / architect on the BitBake, OZ, and OpenEmbedded projects. Maintainer of the tslib project. Aug 24 14:22:42 ibot: no, kergoth is Chris Larson, an embedded linux guy currently at Astrocom Corporation (http://www.astrocorp.com/). Email at kergoth@handhelds.org or clarson@astrocorp.com. Core dev / architect on the BitBake, OZ, and OpenEmbedded projects. Maintainer of the tslib project. Aug 24 14:22:49 bah, silly bot Aug 24 14:22:50 ibot: no, kergoth is Chris Larson, an embedded linux guy currently at Astrocom Corporation (http://www.astrocorp.com/). Email at kergoth@handhelds.org or clarson@astrocorp.com. Core dev / architect on the BitBake, OZ, and OpenEmbedded projects. Maintainer of the tslib project. Aug 24 14:22:52 kergoth_: okay Aug 24 14:24:28 grr Aug 24 14:24:57 must be a rule ... Aug 24 14:30:22 damn it! Aug 24 14:30:27 okay Aug 24 14:30:30 stupid...you should not be sleeping! Aug 24 14:30:32 * emte damns it Aug 24 14:31:15 i'll damn my segfault while i am at it Aug 24 14:32:28 hopefully that will take care of it....stupid ipaq-sleep Aug 24 14:33:17 uh-oh Aug 24 14:33:19 mickey|tv: access to branch 'org.openembedded.nslu2-linux' denied by server Aug 24 14:33:27 oh wait...perhaps it was removed Aug 24 14:33:35 that's right, they're supposed to be using org.oe.dev now Aug 24 14:35:22 all right, going to get some work done now Aug 24 14:43:20 Good night. Aug 24 14:47:32 How do I upload a new kernel to oe ? Aug 24 14:52:23 to your device or the metafile to the repo? Aug 24 14:52:59 I'm still getting errors building gpe-image: "Cannot satisfy the following dependencies for gpe-task-base: hotplug". Any more ideas? According to Packages, linux-hotplug-doc and linux-hotplug both claim to provide, replace, and conflict with hotplug Aug 24 14:53:30 mreimer: that would be the problem. Aug 24 14:53:55 pb_: I thought that might be. How can I go about solving it? Aug 24 14:54:07 The metafile Aug 24 14:54:09 both ipks are built Aug 24 14:54:24 mreimer: linux-hotplug-doc should not be providing, conflicting with, or replacing hotplug. I think the .bb file for linux-hotplug is defective. Aug 24 14:55:03 that would be linux-hotplug_20040920.bb ? Aug 24 14:55:13 mreimer: right. it sounds like you need to replace RDEPENDS with RDEPENDS_${PN}, and likewise for RCONFLICTS and RREPLACES. Aug 24 14:55:41 do you mean RPROVIDES (not RDEPENDS)? Aug 24 14:55:48 er, yeah, sorry Aug 24 14:55:51 ~lart pb_ Aug 24 14:55:52 * ibot DoSes pb_ Aug 24 14:55:56 ok, thanks Aug 24 14:57:04 gerwinin, send it to the mailing list/or the other place taht i for get Aug 24 14:59:35 gerwinin: attach the files you want to be added to a bug filed at bugs.openembedded.org Aug 24 15:03:08 odd concept, metafile additions are bugs :P Aug 24 15:03:24 most likly true for the most part Aug 24 15:03:39 bug trakcing systems are generally used for both feature requests and bugs Aug 24 15:03:54 As they should be Aug 24 15:05:04 GOT PROMPT !!! Aug 24 15:05:09 root@h3600:~# uname -a Aug 24 15:05:15 Linux h3600 2.6.12-hh2 #17 Wed Aug 24 23:45:15 CEST 2005 armv4l unknown Aug 24 15:05:47 good job gremlin[it] Aug 24 15:05:59 If you want to start a built from the beginning which bitbake command do you use ? Aug 24 15:06:21 rm -Rf tmp ? Aug 24 15:06:33 what do you mean? Aug 24 15:06:53 or rather: what do you want to do? Aug 24 15:07:07 I want to erase my previous built from opie-image Aug 24 15:07:12 And start all over again Aug 24 15:07:52 rm -rf build/tmp ; bitbake opie-image Aug 24 15:08:03 i'm happy and i go to bed ;) :) :) Aug 24 15:08:17 gerwinin: if you want to automatically purge the directory, re-extract the source, apply patches - without rebuilding everything. I do 'bitbake rebuild ' Aug 24 15:08:29 night gremlin[it] Aug 24 15:08:40 byeeeeeeeee Aug 24 15:08:50 'night gremlin[it] Aug 24 15:10:49 night gremlin Aug 24 15:15:30 pb_: what do I need to do after editing the .bb to get bitbake to see the change? I ran "parse" and then "rebuild linux-hotplug", but the resulting still has the same provides, conflicts, replaces Aug 24 15:15:55 in the bb I'm using RPROVIDES = "hotplug_${PN}", RCONFLICTS = "hotplug_${PN}", RREPLACES = "hotplug_${PN}" Aug 24 15:16:02 mreimer: did you increment PR? Aug 24 15:16:11 if not, do that Aug 24 15:16:14 nope. that's what I'm missing Aug 24 15:16:19 reparse? Aug 24 15:16:25 yeah Aug 24 15:18:03 are ipkgs like kernel-module-h3900-lcd-2.6 generated automatically by ipkg-make-kernel2.6-packages based on what the build generates? So whatever modules are created get packages up according to some munging of their names? Aug 24 15:19:40 right, although oe doesn't use the ipkg-make-kernel*-packages scripts. Aug 24 15:19:51 it has its own stuff, which lives in kernel.bbclass. Aug 24 15:20:46 the reason I'm asking is because when building gpe-image I'm getting "Cannot satisfy the following dependencies for task-bootstrap: kernel-module-h3900-lcd-2.6...". Do I just need to keep tweaking the ipaqpxa_defconfig until it properly builds these missing modules? Aug 24 15:21:06 yeah, probably Aug 24 15:21:17 ok, I'll keep at it. thanks as always Aug 24 15:24:42 damn strcpy ... Aug 24 15:26:58 mreimer: okay, cool Aug 24 15:34:05 package xdpyinfo compilation error on configure Aug 24 15:34:48 checking xext xtst package xtst was not sound :( Aug 24 15:35:10 yOyOeK: try updating; I think a fix was committed in the last day Aug 24 15:46:01 bitbake -i rocks Aug 24 16:00:17 03nail 07org.oe.dev * ra91a76a6... 10/packages/ (3 files in 3 dirs): Move qc-usb Aug 24 16:00:23 03nail 07org.oe.dev * ra6cc7e9f... 10/packages/meta/openslug-packages.bb: Re-add qc-usb-messenger to openslug feed Aug 24 16:00:27 03nail 07org.oe.dev * r617b38d3... 10/packages/ (2 files in 2 dirs): Update DESCRIPTION Aug 24 16:08:47 03nail 07org.oe.dev * rd1519ff3... 10/packages/musicpd/mpd_0.11.5.bb: Update mpd to latest version Aug 24 16:08:52 03nail 07org.oe.dev * r9fd4fd7a... 10/packages/libao/libao_0.8.6.bb: Update libao to latest version Aug 24 16:09:45 ok....if I want to run a script whenever usbd0 is started up (i.e. when I plug a mini-B connector into my zaurus and the other end into a PC) where do I put it? Aug 24 16:09:53 hotplug seems like the plae but I'm not sure where.... Aug 24 16:32:51 ::sigh:: Aug 24 16:33:20 and it would probably not work anyway...I added some code to net.agent and it works...except reloading the module causes it to be run again, of course... Aug 24 16:43:59 03rpurdie 07org.oe.dev * r8bdd9843... 10/packages/linux/linux-openzaurus_2.6.13-rc5-mm1.bb: linux-oz-2.6: general cleanup of spitz patches. Disable debugging of common power/battery code Aug 24 16:57:35 'night all **** ENDING LOGGING AT Wed Aug 24 17:00:41 2005 **** BEGIN LOGGING AT Wed Aug 24 17:00:41 2005 **** ENDING LOGGING AT Wed Aug 24 17:04:20 2005 **** BEGIN LOGGING AT Wed Aug 24 17:04:39 2005 Aug 24 17:07:57 can you get away with updating the bb's but not the db file? will the build system still work? Aug 24 17:08:50 slisk: the build has nothing to do with monotone Aug 24 17:09:33 JustinP: so you could type monotone --update int he metafile dir and build up to date packages without touching the .db file? Aug 24 17:09:48 slisk: ummmm....what? Aug 24 17:09:57 slisk: to update your bb file syou need to update your monotone db Aug 24 17:10:41 slisk: let's keep it in this channel Aug 24 17:10:54 slisk: to update your bb files you have to pull from a monotone server Aug 24 17:10:57 ok Aug 24 17:11:14 slisk: you're perfectly welcome to just download the snapshots Aug 24 17:11:22 slisk: in which case you get ~daily updates Aug 24 17:11:51 JustinP: when you update the db, doesn't it only need to process changes, not everything all over again? Aug 24 17:12:09 slisk: yes Aug 24 17:12:16 slisk: and it doesn't do everything again Aug 24 17:12:34 slisk: it all depends on how many changesets you're downloading Aug 24 17:12:44 JustinP: what constitues a "revision" as mentioned in the console output when updating the db? Aug 24 17:12:46 revisions I mean Aug 24 17:12:54 slisk: any node in the graph, I believe Aug 24 17:13:22 JustinP: so you're considering all the steps between your copy and the current, not just the end result?? Aug 24 17:14:06 slisk: of course Aug 24 17:14:10 slisk: there is no "end result" Aug 24 17:14:27 slisk: each revision is its own node and has to be verified on its own Aug 24 17:14:44 slisk: there's a "head" but to verify the head you have to verify all the steps leading up to it Aug 24 17:16:14 JustinP: is each revision a change to an individual file or possible a group? Aug 24 17:16:51 group Aug 24 17:16:53 slisk: you can change multiple files in one revision Aug 24 17:17:14 iirc certs are per file Aug 24 17:17:36 JustinP: why can't the head simply be dl'ed with a hash and verified on it's own, abandoning the old state? Aug 24 17:17:49 oops Aug 24 17:17:50 it's a set of changes to any number of files (as well as file additions/deletions/renames) Aug 24 17:18:01 slisk: because then you wouldn't have the revision history Aug 24 17:18:18 JustinP: is the revision history stored on the server? Aug 24 17:18:25 monotone's mode of working is that each DB is a full copy of the repository with all revision history Aug 24 17:18:35 there is no "server" really Aug 24 17:18:44 each DB is its own fill copy of the repo Aug 24 17:18:51 any copy may serve its contents Aug 24 17:18:53 a head without revision history is called a tarball and can be found at http://ewi546.ewi.utwente.nl/OE/ Aug 24 17:19:04 slisk: go read the monotone FAQ / manual Aug 24 17:19:53 JustinP: 1 more quick question. is there a command line option to avoid tracking the history? Aug 24 17:20:36 so that you can simple dl the difference between the local and the "head" Aug 24 17:21:11 it already does that Aug 24 17:21:37 that's what a pull does Aug 24 17:21:51 i'm assuming the longest part of the update is the revision history verification, is it possible to skip that and save time? Aug 24 17:22:02 no Aug 24 17:22:03 slisk: no, you can't just "update" your checkout in the CVS sense. At least not until you pull all of the revisions from the server Aug 24 17:22:28 monotone is a distributed system, it doesn't work that way Aug 24 17:22:36 would it be possible to just rsync a db, to skip the long verification phase? Aug 24 17:22:40 you're very welcome to set up a cron job to pull every hour or 2 Aug 24 17:22:53 mreimer: yes Aug 24 17:22:56 mreimer: that would be a horrible waste of bandwidth Aug 24 17:23:04 it takes about that long to update though Aug 24 17:23:25 JustinP: because the file changes all over, not just at, say, the end? Aug 24 17:23:30 slisk: I suggest you just update mor eoften Aug 24 17:23:43 mreimer: I have no idea, it's an sqlite database Aug 24 17:24:09 JustinP: doesn't rsync just xfer the portions of the file that changed, not the whole thing? Aug 24 17:24:30 JustinP: I don't know my machine is capable of updating & verifying fast enough to do it every hour :-( Aug 24 17:24:35 mreimer: by default it only does changes, but you can alter the behaviour Aug 24 17:24:38 mreimer: I'm not sure. It's a binary file, though, not text Aug 24 17:24:56 JustinP: rsync loves binary files too Aug 24 17:25:00 mreimer: of course it is. There aren't that many changes in 1 hour Aug 24 17:25:19 mreimer: I run monotone on my nslu2 Aug 24 17:25:28 koen: how long does it take to run? Aug 24 17:25:42 mreimer: ~1-2 minutes per revision usually Aug 24 17:25:56 mreimer: I have never seen 30+ revisions in an hour Aug 24 17:26:22 mreimer: fast enough to be usefull is as close as I can estimate it Aug 24 17:26:24 mreimer: and even if there were, you just add a little check to make sure monotone isn't running before you pull Aug 24 17:26:38 ok, I'll give that a try Aug 24 17:28:03 I think you all just need to stop being so impatient. You have to wait for it to verify, that's just how it is. It will be faster when the devs come up with a way to make it faster. Aug 24 17:28:14 run a cronjob at night and rsync the checkout to other machines Aug 24 17:28:43 that's what I do for my tinderboxes Aug 24 17:29:12 yeah, that's probably what I'll have to do. On my wee PIII 450, it means I have to plan ahead Aug 24 17:29:12 (sort off) Aug 24 17:30:01 * koen looks at intels 1.25GHz xscale announcement Aug 24 17:30:05 ehm Aug 24 17:30:08 MHz Aug 24 17:30:11 whoa Aug 24 17:30:23 how's the power consumption? Aug 24 17:30:31 might not be very useful for a handheld Aug 24 17:30:41 less W/MHz Aug 24 17:31:07 not useful for a calculator, i got PIC's faster than that Aug 24 17:31:24 it's supposed to be 30% faster than a 600MHz pxa270 Aug 24 17:31:49 wam, did you correct 1.25GHz to 1.25MHz? Aug 24 17:31:53 that's all? Aug 24 17:31:56 koen: link? Aug 24 17:32:46 http://www.tweakers.net/nieuws/38663 (dutch) Aug 24 17:32:50 heh Aug 24 17:32:52 they announced it at the IDF Aug 24 17:33:08 my father might be able to read it, but alas not me Aug 24 17:33:17 http://www.theregister.co.uk/2005/08/23/intel_demos_monahans/ Aug 24 17:57:08 JustinP, they are working on that Aug 24 17:57:25 i dunno if you missed my paste earlier to kergoth Aug 24 17:57:49 hey good news for monotone: njs emte: and we have a Plan For Fixing That, hopefully will be working in a ~few months Aug 24 17:58:17 re: speed and computational req Aug 24 17:58:52 now if i could only figure out my damn problem segfault Aug 24 18:02:49 or why the build process keeps breaking every other day Aug 24 18:05:34 you using current cvs packages? Aug 24 18:13:26 i'm using monotone and bitbake from a few weeks ago. i had this problem b4 and updated the db and bb tree to fix the problem Aug 24 18:13:37 it's happened again, i updated. but no fix Aug 24 18:14:07 don't tell me i need to be up to hour on the build programs too... Aug 24 18:14:12 slisk i'm using monotone and bitbake from a few weeks ago. slisk or why the build process keeps breaking every other day Aug 24 18:14:20 those two comments dont even match Aug 24 18:14:28 unless your breaking it yourself Aug 24 18:19:57 1. i installed the current versions of monotone and bitbake from a few weeks ago. Aug 24 18:20:27 2. between now and then i had thwe problem i'm having now. updating the db file and the bb files fixed it Aug 24 18:20:50 3. a time later (now) i have the problem again. i updated the db and bb files, but the problem remains Aug 24 18:20:52 what problem? Aug 24 18:21:10 hold on while i recreate it Aug 24 18:23:16 hmm Aug 24 18:23:34 03nail 07org.oe.dev * r01c816a6... 10/packages/syslog-ng/syslog-ng_1.6.8.bb: Add initscript. Remove/reinstall busybox syslog on install/remove. Should work entirely out of the box now Aug 24 18:23:34 wonder if raster can enlighten me on my issue ... Aug 24 18:23:56 this problem: http://pastebin.ca/21060 Aug 24 18:24:52 The SL-C3100 source code has been released... Aug 24 18:25:08 and? Aug 24 18:25:32 * arrummzen was just stateing a fact... Aug 24 18:25:37 the other sharp code wasnt great will the 3100 be better? Aug 24 18:26:01 * arrummzen thought the OE community would be interested in knowing Aug 24 18:26:19 slisk, looks like your quit version is off or the metafile needs adjusting Aug 24 18:26:43 not sure, i see a lot of cursing at sharp code :) Aug 24 18:27:13 probably help figure out some driver issues tho Aug 24 18:28:02 I am going to buy a SL-C3100, but I was waiting until the kernel code was released. Aug 24 18:28:13 out of curiosity, in what language was it released? Aug 24 18:28:36 i doubt it will be very great Aug 24 18:28:43 Well I found the code on a jappense site... Aug 24 18:28:52 for the most part all the other zaurus oem code was tossed Aug 24 18:30:08 i am not sure how far the sl-3100 device stuff is Aug 24 18:31:54 i am trying to remember who even has one here Aug 24 18:32:13 i think RP was doing some driver stuff earlier Aug 24 18:34:44 03nail 07org.oe.dev * r372eed5d... 10/packages/syslog-ng/syslog-ng_1.6.8.bb: Forgot INITSCRIPT_PARAMS. Duh. Aug 24 19:00:29 emte: yes, I saw that and mentioned to slisk that they were working on it Aug 24 19:01:24 slisk: it doesn't break very often for me. It normally works just fine. For all 3 of the images I compile. If you have a problem I suggest you file a bug so it has som visiblity. Stating it here isn't good enough unless someone is here who can fix it. Aug 24 19:02:00 arrummzen: it's already on OE and compiling. We even have testing 3100 images. I don't think any devs have one, but it's basically the same as the 3000 anyway. Aug 24 19:11:35 should mention it to raster later Aug 24 19:11:47 i think he is still using the other method Aug 24 19:11:52 for his 3100 Aug 24 19:12:19 he has a 3100? I thought he had a 3000.... Aug 24 19:12:27 "older method"? he's using pdaXrom Aug 24 19:12:45 IIRC Aug 24 20:09:05 03nail 07org.oe.dev * rce3e6a98... 10/packages/syslog-ng/syslog-ng_1.6.8.bb: Add -f to update-rc.d Aug 24 20:09:09 03nail 07org.oe.dev * r3aec9f1e... 10/packages/syslog-ng/syslog-ng_1.6.8.bb: Bump PR *sigh* Aug 24 20:28:17 well he has a couple zauri i thought he bought a 3100 Aug 24 20:28:29 hey raster, you have a 3100? Aug 24 20:28:37 btw i have a problem Aug 24 20:28:54 no Aug 24 20:28:56 c830 Aug 24 20:28:57 :) Aug 24 20:29:29 btw, i dont suppose youve saw any cheap flybooks around ... or u71P Aug 24 20:29:59 http://www.labotomy.net/cgi-bin/biohazard/?cmd=file;file=configHandler.c;filenode=177a7157c999cc0ff5cdb066ccd9ae058968a66b Aug 24 20:30:02 line 38 Aug 24 20:30:29 its giving me a segfault Aug 24 21:26:10 en fin! "NOTE: package gpe-image-1.0: completed" Aug 24 21:26:25 thanks all of you who patiently helped Aug 24 21:26:39 tu hablas español :D Aug 24 21:32:56 no hablo :-( Aug 24 21:33:04 tu parle francais? Aug 24 21:33:09 mo hinga Sango? Aug 24 21:33:10 nu, :( Aug 24 22:49:45 http://dfbcar.sourceforge.net/ who was looking for a lightweight media palyer? This was developed on an epia-m 10000 ezra. Aug 25 00:37:22 morning all. Aug 25 00:37:36 Does anybody have a clue what I can do about this : | /home/fille/oe/build/tmp/work/handhelds-sa-2.4.19-rmk6-pxa1-hh37.4-r6/temp/run.do_patchcleancmd.13700: line 457: syntax error near unexpected token `'INHIBIT_PACKAGE_STRIP',' ? Aug 25 00:37:42 thanks! Aug 25 00:38:05 morning Aug 25 00:38:33 morning XorA ;) Aug 25 00:39:33 morning XorA ;) Aug 25 00:48:51 aaaargh, roll on Angstrom Aug 25 00:49:47 shit and I broke monotone bigtime :-( Aug 25 00:55:02 morning all Aug 25 00:55:53 RP: morning Aug 25 01:30:52 hi Aug 25 01:31:08 NAiLzZz: can you add package name into commit logs? Aug 25 01:31:36 RP: looks like I have to update my 2.6.13-rc6-mm2 with your new patches Aug 25 01:41:54 hrw|work: i didn't touch the i2c code... Aug 25 01:42:12 re Aug 25 01:42:24 hi zecke_ Aug 25 01:47:09 NOTE: package linux-openzaurus-2.6.13-rc6-mm2-r12: task do_unpack: started Aug 25 01:51:03 RP: all patches apply Aug 25 01:52:05 hrw|work: excellent :) Aug 25 01:52:34 hrw|work: It was mostly a re arrange of them to try and make them easier to manage - I need to prepare them for upstream... Aug 25 01:54:23 RP: you are kernel mastah, I only test newer version ;) Aug 25 01:58:25 hi pb Aug 25 02:00:22 is there a timer with a nanosecond or µsecond resolution? (linux kernel) Aug 25 02:04:31 hi hrw|work Aug 25 02:04:37 dont think so - at least not on all arms. on pxa i'd try the cccr register Aug 25 02:04:37 hail zecke_ Aug 25 02:05:06 zecke_: depends what you mean by a timer. gettimeofday() will give you microsecond resolution on almost all platforms. Aug 25 02:06:27 sorry s/cccr/oscr/ Aug 25 02:06:40 pb_: okay, but that sounds like polling in a kernel thread Aug 25 02:06:51 * XorA phews that koen exports an oe db as he manages to recover from corrupt db Aug 25 02:07:06 zecke_: yeah. if you want an interrupt after a certain time, you will need to use machine-specific code. Aug 25 02:07:30 RP: NOTE: package linux-openzaurus-2.6.13-rc6-mm2-r12: task do_build: completed Aug 25 02:07:32 some machines can't do that at all, but a significant number (probably most nowadays) can. Aug 25 02:07:55 stupid battery... anyway Aug 25 02:08:19 'night Aug 25 02:23:50 pb_: thanks master, my first public glib code gone public :} Aug 25 02:32:31 zecke_: very good Aug 25 02:36:52 pb_: hehe Aug 25 02:40:48 hey zecke_ Aug 25 02:43:34 koen_: hey Aug 25 02:44:05 koen_: somebody send bogus data to the tinderbox... I#m going to rewrite the tinderbox... Aug 25 02:44:29 ok Aug 25 02:44:47 it can not be that a server application stops working when one client is broken... Aug 25 02:45:01 koen_: temporarily you could delete the data directory Aug 25 02:46:14 done Aug 25 02:54:22 Building an opie-image for Spitz is failing for me on the do_rootfs; is this a know issue? The problem seems to be "| ERROR: Cannot satisfy the following dependencies for task-bootstrap:" (with nothing afterwards. Aug 25 02:57:40 Check further up in the log. You might just be seeing the output from the crazy logfile grepping thing. Aug 25 02:58:27 evening All ;) Aug 25 02:58:38 hey ZonE-HaKin9| **** ENDING LOGGING AT Thu Aug 25 02:59:56 2005