**** BEGIN LOGGING AT Sat Feb 19 23:59:56 2005 Feb 20 00:02:31 where can I get unslung-able ?? Feb 20 00:03:38 you have to build it Feb 20 00:03:55 sorry mis read what it does Feb 20 00:04:12 I need support for the multible luns Feb 20 00:05:27 the hd enclose I have supports 4 hd on 2 usb ports so 2 hds per port Feb 20 00:05:41 any idea how I can have them register seprattely? Feb 20 00:06:01 it seems to be running on both like they are 1 drive Feb 20 00:30:53 03jp30 * 10unslung/ (8 files in 2 dirs): initial checking of apache Feb 20 00:39:19 03jp30 * 10unslung/ (4 files in 2 dirs): initial checkin of php Feb 20 00:50:56 03jp30 * 10unslung/make/ (apr-util.mk apache.mk): oops: missed dependencies for apr-util and apache Feb 20 00:59:33 anyone here ever setup raid ? Feb 20 00:59:39 or more then 2 usb drives? Feb 20 01:11:20 not that I've heard Feb 20 01:13:29 jp30: we only have libdb 4.2 in unslung, and it seems apache is looking for 4.3 Feb 20 01:14:12 rwhitby-away, does it actually fail to configure? Feb 20 01:14:15 v1s: you will need to run OpenSlug to be able to access more than two drives, raided or not Feb 20 01:14:36 jp30: home/unslung/packages/toolchain/armv5b-softfloat-linux/gcc-3.3.5-glibc-2.2.5/lib/gcc-lib/armv5b-softfloat-linux/3.3.5/../../../../armv5b-softfloat-linux/bin/ld: cannot find -ldb-4.3 Feb 20 01:14:50 building htpasswd Feb 20 01:19:21 is there firmware I can just load on to it ? Feb 20 01:20:06 rwhitby-away, can you grep for UTIL_LDFLAGS in apache's config.status and see if that's where the -ldb-4.3 is coming from? Feb 20 01:20:34 jp30, is there openslug firmware that can download ? Feb 20 01:21:33 v1s, no, you need to build it yourself Feb 20 01:22:22 where do Iget the files to build ? Feb 20 01:23:28 v1s, you need to use bitkeeper. read the wiki Feb 20 02:02:27 rwhitby@nomic:/home/unslung/packages$ grep UTIL_LDFLAGS builds/apache/config.status Feb 20 02:02:27 s,@UTIL_LDFLAGS@,,;t t Feb 20 02:06:18 rwhitby-away, i've committed some changes to apr-util that I think will fix the problem Feb 20 02:37:34 jp30: apache built Feb 20 02:38:10 let me know when you think it's ready for the feeds Feb 20 03:08:33 03rwhitby * 10unslung/Makefile: Promoted apache and php. Feb 20 03:19:25 anyone know offhand what the apache memory footprint is on the nslu2? Feb 20 03:19:44 I've been wanting to resetup my personal cgiirc gateway. Feb 20 03:22:28 ipkg install and see :-) Feb 20 03:24:30 03rwhitby * 10unslung/Makefile: Promoted apr and apr-util Feb 20 03:30:21 got the apache test page on prodslug :-) Feb 20 03:30:55 3,3MB per instance Feb 20 03:31:01 gah Feb 20 03:31:13 ~seen siddy Feb 20 03:31:14 siddy <~sid@p509263BE.dip.t-dialin.net> was last seen on IRC in channel #nslu2-linux, 1d 15h 28m 20s ago, saying: '[g2], yeah, fortunately netconsole.c is quite short, but I can't see where it could cause an alignment-error.'. Feb 20 03:31:23 well, i'll wait Feb 20 03:33:18 dyoung: let me know when you've got CGI::IRC working ;-) Feb 20 03:48:15 Okay it works Feb 20 03:48:28 its really damn slow though. Feb 20 03:50:42 how slow is slow - unbearable? Feb 20 03:51:04 I hve a unslubng 3.x fix. Feb 20 03:51:12 heh Feb 20 03:51:20 * dyoung hides Feb 20 03:51:25 dyoung-slugweb, what app is the slugweb? Feb 20 03:51:36 apache Feb 20 03:51:42 03rwhitby * 10unslung/Makefile: Move apr to cross only. Feb 20 03:51:43 ow ow owwwww. Feb 20 03:51:45 Apache runs as nobody; but /tmp and /var/tmp are permisstion 755 Feb 20 03:51:46 it's an ipk now (last half hour) Feb 20 03:52:26 Lemme use it from here for a bit Feb 20 03:52:33 I guess once it starts up its okay Feb 20 03:52:39 the starting is pretty slow though Feb 20 03:52:41 ok, that needs to be fixed in the posting (like openssl or dropbear does) Feb 20 03:53:06 dyoung-slugweb: what does your swap look like? Feb 20 03:53:11 Swap? Feb 20 03:53:18 This one dont got no swap. ;-) Feb 20 03:53:40 ka6sox-away: 3.3MB per apache instance it looks like Feb 20 03:54:22 its just straight apache...no modperl/ssl etc..? Feb 20 03:54:52 dunno, I just build 'em Feb 20 03:55:00 jp30 contributed it Feb 20 03:55:02 my httpds are 2600 Feb 20 03:55:08 the cgiirc is 3.5 Feb 20 03:55:30 dyoung: cgiirc is same here. Feb 20 03:55:32 or close Feb 20 03:55:34 3824 to be exact. Feb 20 03:56:08 I guess its usable. Feb 20 03:56:14 It doesnt completely suck. Feb 20 03:59:10 can't be worse than a trans-pacific gateway ... Feb 20 03:59:30 brb Feb 20 03:59:59 dyoung: can you put up a howto for CGI::IRC ? Feb 20 04:02:54 okay its 4am here...I need sleep Feb 20 04:04:41 hmm Feb 20 04:04:47 well, it worked once. Feb 20 04:05:00 crap I type there now doesnt come out. Feb 20 04:05:29 but stuff from here shows up there. Feb 20 04:10:28 03rwhitby * 10unslung/Makefile: Cleaned up the package categorisation - now there is just cross and native. New packages default to cross. Feb 20 04:13:33 i'll put up a wiki page when i dtermine why it only work ssometimes Feb 20 04:16:25 hm, with unslung, why doesn't my usb drive in port2 get detected? Feb 20 04:17:06 is it formatted vfat? Feb 20 04:17:26 yeah Feb 20 04:17:48 but it doesn't get detected in the sense that the usb storage doesn't see it Feb 20 04:17:51 no sdb at all Feb 20 04:18:34 http://www.nslu2-linux.org/wiki/HowTo/TroubleshootADriveNotFormattedCondition Feb 20 04:19:04 got the right devices and *_conn entries? Feb 20 04:19:57 dunno, but that's not even the problem Feb 20 04:20:11 it doesn't get detected in dmesg Feb 20 04:20:19 and it also isn't in /proc/bus/usb/devices Feb 20 04:20:30 dunno then Feb 20 04:21:07 what brand is it, and does it only have one storage device LUN in it? Feb 20 04:21:25 WD Feb 20 04:21:28 probably Feb 20 04:21:31 scratch the second question - irrelevant if it's not detected at all Feb 20 04:21:41 removing and reinserting the drive doesn't give any output in dmesg Feb 20 04:24:43 gotta prob peeps Feb 20 04:24:55 its about CUPS Feb 20 04:25:17 peeps? Feb 20 04:25:39 people Feb 20 04:25:43 hey whitby Feb 20 04:25:52 you installed cups? Feb 20 04:25:57 nope Feb 20 04:26:02 do you know how? Feb 20 04:26:11 ipkg install cups Feb 20 04:26:24 yes but where do you put the cups thing or whatever Feb 20 04:26:31 im a noob with linux ive gotta say Feb 20 04:26:38 in the cups holder? Feb 20 04:26:48 had redhat and they had RPMs but I dunno about the rest Feb 20 04:26:58 very funny Feb 20 04:27:14 dunno - never used cups Feb 20 04:27:24 do you have a printer on urs? Feb 20 04:27:39 have you checked the cups doco on the net? Just add /opt to most paths ... Feb 20 04:28:10 no i havent Feb 20 04:28:35 and I assume you've read the CUPS howto on the wiki, or else you wouldn't be here, right ;-) Feb 20 04:28:51 yeah Feb 20 04:30:46 how would you compile cups in windows? Feb 20 04:31:18 CoLinux Feb 20 04:31:31 ta] Feb 20 04:33:37 ok Feb 20 04:33:38 weird Feb 20 04:33:54 boot slug, kill the USB_Detects, plug in drive, slug is locked up Feb 20 04:34:14 weird Feb 20 04:34:22 you haven't got serial yet, have you ? Feb 20 04:34:29 it isn't my own slug Feb 20 04:34:36 i'm gonna try it in the first port Feb 20 04:35:01 disk1 will try to mount it with three ext3 partitions Feb 20 04:35:20 also if i kill the USB_detects first? Feb 20 04:35:26 no Feb 20 04:35:40 it will only try to mount it if it gets detected at all Feb 20 04:35:43 make sure rc.bootbin doesn't run either Feb 20 04:35:46 that's the first problem now Feb 20 04:36:00 right Feb 20 04:36:09 i don't care if mounting fails, i want it to show up in dmesg Feb 20 04:36:24 if i just plug it in, it won't go fdisking/formatting on its own, rihgt? Feb 20 04:37:26 no Feb 20 04:37:51 ok, let's try that then Feb 20 04:38:16 does a hard disk work in Disk2? maybe it's a hardware problem Feb 20 04:38:34 okay, as disk1 it gets detected Feb 20 04:39:05 and i can mount it and stuff Feb 20 04:39:17 smells like a hardware problem with disk2 port Feb 20 04:39:29 i mean, a hdd should work on port 2 Feb 20 04:39:37 you can hook up 2 hdds Feb 20 04:39:40 y dont you rip it out and put a serial port in? Feb 20 04:39:57 Gorx: it isn't his slug Feb 20 04:40:08 o lol Feb 20 04:40:09 maybe the hardware gets confused because i had a flash stick in port 1 :) Feb 20 04:40:16 but still, imo it should matter Feb 20 04:40:20 giel: does a hard disk work in disk2? Feb 20 04:40:46 in general or on this specific slug? Feb 20 04:42:45 okay, booting up without devices, pluggin in hdd in port2 and then flash in port1 seems to work Feb 20 04:42:50 after killing usb_detect Feb 20 04:43:48 hey, this is weird Feb 20 04:44:14 hdd in port2 plugged in first gets assigned sda1, and then the flash gets assigned sda1 sda2 sda3 Feb 20 04:44:18 giel: on that specific slug - it should work in general Feb 20 04:44:36 anything in port2 should be assigned sdb Feb 20 04:46:17 that is weird'cause it isn't Feb 20 04:46:24 it was assigned sda just now Feb 20 04:46:46 that slug smells funny Feb 20 04:46:47 looks like a problem in the usb storage driver or something? Feb 20 04:52:13 is the usbstorage driver patched to force the sda/sdb assignment? Feb 20 04:52:34 'cause normally mass storage devices get assigned the first free one Feb 20 04:53:05 yeah, see scsi_scan.c Feb 20 04:53:24 The MSC driver in the stock Linksys firmware is hacked to assign port1 to sda and port2 to sdb Feb 20 04:53:48 ok Feb 20 04:54:07 then i've got a really weird hardware problem, or something is wrong with my firmware Feb 20 04:58:01 can someone mail me the scsi_scan.c of unslung 3.18? (no bk access atm) Feb 20 04:59:47 03rwhitby * 10unslung/Makefile: Promoted e2fsprogs parted proftpd recordext x11 xau xdmcp xdpyinfo xext xextensions xproto xtrans xtst ice sm xt xmu xauth to cross. Feb 20 05:00:06 giel: i can Feb 20 05:00:18 msg me your email address Feb 20 05:01:54 hmm - something is screwy with my connection to gmail - maybe I can't at the moment. Feb 20 05:02:09 no hurries Feb 20 05:03:19 Just did a pull, but the build failed - anyone here that can help me out? Feb 20 05:04:38 Error was: Feb 20 05:04:39 slug@ali build $ bb unslung-image Feb 20 05:04:41 NOTE: Using cache in '/home/slug/build/tmp/cache' Feb 20 05:04:42 NOTE: Parsing finished. 0 cached, 2069 parsed, 26 skipped, 0 masked. Feb 20 05:04:44 NOTE: Building provider hash: [####################] (100%) Feb 20 05:04:45 NOTE: build 200502201355: started Feb 20 05:04:47 Traceback (most recent call last): Feb 20 05:04:49 File "/home/slug/bitbake/bin/bitbake", line 706, in ? Feb 20 05:04:50 bb.event.fire(bb.event.BuildStarted(buildname, pkgs_to_build, make.cfg)) Feb 20 05:04:52 File "/home/slug/bitbake/lib/bb/event.py", line 49, in fire Feb 20 05:04:53 if tmpHandler(event) == Handled: Feb 20 05:04:55 File "tmpHandler(e)", line 30, in tmpHandler Feb 20 05:04:57 NameError: global name 'bb' is not defined Feb 20 05:05:24 rwhitby: i'm afk for a while, can you msg me if you've succeeded in mailing? Feb 20 05:05:36 bob_tm, you need to get a newer bitbake Feb 20 05:05:43 svn up Feb 20 05:05:43 Thanks Feb 20 05:05:57 giel: perhaps dyoung can help - my mail is offline atm Feb 20 05:07:00 dyoung: I don't need to pull again? Just build? Feb 20 05:08:13 03rwhitby * 10unslung/Makefile: Promoted perl-digest-hmac perl-digest-sha1 perl-html-parser perl-html-tagset perl-mime-base64 perl-net-dns perl-net-ident perl-storable perl-time-hires to native. Feb 20 05:08:48 bob_tm: you might want to try nslu2-package-symlinks too - parses 338 files instead of 2069 Feb 20 05:09:13 rwhitby: I will. Feb 20 05:10:16 you may or may not have to re-setup bitbake. But you probably dont need to p ull. Feb 20 05:10:40 I run bitbake in place. Saves installing it. Feb 20 05:10:54 (heh thats why its may or may not....) Feb 20 05:13:12 dyoung: can you mail me scsi_scan.c of unslung 3.18? pretty plz? ;) Feb 20 05:13:21 giel: sent Feb 20 05:13:23 I'm looking at it now Feb 20 05:13:30 debating if I know your email addy. Feb 20 05:13:32 dyoung: oh nm, rwhitby sent it already Feb 20 05:13:40 ok Feb 20 05:13:51 dyoung: it is in the maintainer of the unslung glibc ipk :) Feb 20 05:14:15 I havnt looked at the unslung package repo since last year. Feb 20 05:15:11 But I'll try to remember that. Feb 20 05:16:42 :) Feb 20 05:17:16 I wasnt aware that Linksys removed telnetd from R29 Feb 20 05:17:55 yep Feb 20 05:18:13 telnetd and telnet.cgi gone Feb 20 05:18:26 but we can put them back :-) Feb 20 05:18:59 I wonder if that was our fault. Feb 20 05:19:08 no other changes as far as I can see in the rootfs (unless the binaries have changed). Feb 20 05:19:25 buzbee's :-) Feb 20 05:19:58 In R25 they removed indexing in the Management dir, but buzbee got around that, so they went further in R29 Feb 20 05:20:10 oh yeah. hes not here. We can blame him. ;-) Feb 20 05:20:20 the irony is that vi is still there, but there is no way to get to it :-) Feb 20 05:20:30 serial? Feb 20 05:20:30 I think a maxtor onetouch util is also missing Feb 20 05:20:45 Tiersten: yeah, good point Feb 20 05:20:58 many binaries differ but that probably just recompiling Feb 20 05:21:11 yeah, that's what I figured Feb 20 05:21:16 I it still glibc 2.2.5 ? Feb 20 05:21:44 hmm - good question. Feb 20 05:21:55 let me see if I've still got it around anywhere ..... Feb 20 05:21:57 still no sign of the source package Feb 20 05:22:45 Any suggestions on replacing syslogd/klogd? It is in inittab, so it can't be diverted directly. Would it be an idea to divert it indirectly, like replacing syslogd and klogd by a script which in turn checks the unslung dirs. That way we could just kill syslogd/klogd once the disks are mounted and get the replacements restarted? Can that be done? Feb 20 05:24:41 you could have a script on boot that moves the old syslogd aside and puts in a link to the new one, then kills the process to restart it. you'd need to be careful on package removal though ... Feb 20 05:27:05 rwhitby: Yes, but what if the slug crashes - then it will fail to load Linksys syslogd on restart. Feb 20 05:27:25 since the link will still be in place Feb 20 05:28:17 true Feb 20 05:29:27 will an external disk be available when things in inittab are first run? Feb 20 05:29:47 Nope Feb 20 05:30:08 So it will always start Linksys logds first Feb 20 05:31:13 I will try it out. I am very happy with syslog-ng - I log all my machines to the slug now. Feb 20 05:31:33 ... since it's my only 24/7 machine Feb 20 05:31:53 I even log my routers to it. Feb 20 05:32:42 so the package will replace /sbin/syslogd and /sbin/klogs, or will the package modify /etc/inittab? Feb 20 05:33:33 I thought of modifying /etc/inittab and redirect the entries to /unslung which will check the disk diversion dirs before starting orginal syslogd/klogd Feb 20 05:34:13 packages are not allowed to touch /unslung Feb 20 05:34:40 Ok - any suggestions on where to place things like this on the flash? Feb 20 05:35:01 I'm not sure this is something a package should do - it would probably be manual config Feb 20 05:35:15 manual config is fine Feb 20 05:36:15 BTW - checked install. If we go for the one in coreutils - it build nicely on Gentoo and on the slug, so it will be the same both in cross and native Feb 20 05:37:20 That was no welformed sentence :-) Ok - I found no existing dir in which to place it, so maybe a new tools dir in the toolchain and define $(INSTALL) macro in the make system. Feb 20 05:37:44 agreed Feb 20 05:37:58 like we do for ipk-utils Feb 20 05:38:01 Yes Feb 20 05:38:15 good plan Feb 20 05:38:28 We could put more tools in there if problems arise with portability of standard unix tools Feb 20 05:38:49 What should I do, practically? Feb 20 05:39:23 03rwhitby * 10unslung/Makefile: Demoted postfix and cyrus-imapd. Feb 20 05:41:11 bob_tm-away: same as for IPKG_BUILD Feb 20 05:43:53 03rwhitby * 10unslung/make/torrent.mk: Fixed the dependencies so it wouldn't keep rebuilding. Feb 20 05:51:32 03rwhitby * 10unslung/sources/perl-net-ident/control: The contents were completely wrong. Feb 20 05:54:37 03rwhitby 07 * r1.2817.1.1 10openembedded/packages/busybox/switchbox-1.00/linuxrc: Silenced the stderr output from all the calls to dd in switchbox linuxrc Feb 20 05:54:43 dyoung-zzzz: with all the package maintenance, that's all I got to do on switchbox tonight Feb 20 06:28:26 Good morning. Anyone have a moment to help me with an ipkg problem? Feb 20 07:22:28 JWynen: will try it, so please follow me to #nslu2-general Feb 20 07:23:09 thanks caplink811, I'm a noob on irc, so excuse me if i do something wrong Feb 20 08:29:12 rwhitby-away: I did a pull 4 hours ago and built the image. .unslung on the rootfs says "V2.3R25-uNSLUng-standard-4.2-alpha", but defconfig in standard kernel does not have devfs enabled. Is that correct? Feb 20 08:32:25 Iow - has unslung-able replaced unslung-standard as kernel image for unslung in the new version? Feb 20 08:55:28 Sorry for the newbie question, but how to you get bitbake to only re-build the kernel after I make changes? Feb 20 08:55:49 Remove the relevant stamps Feb 20 08:56:04 You know how? Feb 20 08:56:21 Nope; I'm assuming they're in the build/tmp directory... Feb 20 08:57:07 Yes - build/tmp/stamps - You get the drift - there is a stamp for each stage in the process for each package Feb 20 08:57:57 Just rm each one from package-do_compile and onwards with regards to file creation date Feb 20 08:58:00 I thought I'd tried that, but it seemed to continue to re-use the built components... Feb 20 08:58:39 Hmm... I usually clean the kernel too Feb 20 08:58:57 but I'm "Newbie light" when it comes to this make system. Feb 20 08:59:00 Ahh, but package-do_compile is a sub of stamps, isn't it... Feb 20 08:59:09 No Feb 20 08:59:20 It is just one file Feb 20 08:59:28 ....as I recalll Feb 20 08:59:32 ok; I'll double check what's in it Feb 20 08:59:48 Why isn't there a "clean" target like make usually has? Feb 20 09:00:07 It probably is somewhere - I just don't know where. Feb 20 09:00:11 Anyone else? Feb 20 09:00:28 does bitbake -c clean work in this situation? Feb 20 09:00:46 jp30: How do you specify package to clean? Feb 20 09:00:54 with the -b option Feb 20 09:01:09 Ok - it will not fecth the sources again then and overwrite changes? Feb 20 09:01:39 I'm not sure - also a bitbake noob here, I'm afraid... Feb 20 09:02:00 We all are :-) I've used it for months, but just used it. Feb 20 09:02:27 I've just started with it, but it doesn't seem to be as "open" as a standard GNU make... Feb 20 09:03:18 <[g2]> rm -f tmp/stamps/openslug-kernel-2.6.9-r1.do_compile Feb 20 09:03:28 <[g2]> http://www.nslu2-linux.org/wiki/HowTo/TweakKernelOrPackagesInOE Feb 20 09:04:22 Ahh... I'll double check that howto... but again, is there a reason why I can't target it? Feb 20 09:05:13 g2: Do you know the answer to my query about the 4.2 kernel and devfs? Feb 20 09:08:43 <[g2]> bob_tm-away, sorry, but no, I'm an OpenSlug guy and I'm not following Unslung very much Feb 20 09:09:19 Thanks - I'll wait for rwhitby-awake :-) Feb 20 09:09:34 <[g2]> bob_tm-away, you don't know anything about hotplug-ng do you ? Feb 20 09:09:36 <[g2]> :) Feb 20 09:10:02 * [g2] continues reading the code Feb 20 09:10:38 * bob_tm-away looks away ashamed of his ignorance Feb 20 09:10:55 <[g2]> no need to be ashamed Feb 20 09:11:27 <[g2]> except if you can't admit ignorance and ignore it :) Feb 20 09:11:40 It's kind'a new to me, and it's not really new. Feb 20 09:11:52 Haven't been keeping up. Feb 20 09:12:17 <[g2]-bbiab> well it's a week old :) Feb 20 09:12:34 Here, yes, but in general? Feb 20 09:13:40 I'm doing house cleaning and throwing away schematics and code from 1999. Feb 20 09:19:17 Here's another newbie question: I read the FAQ on integrating a Bluetooth usb device, but it was for OpenSlug; is there any reason why it wouldn't work on unslung? Feb 20 09:20:16 What were the needed drivers? Feb 20 09:20:27 Could be a problem with devfs Feb 20 09:20:44 ...since unslung doesn't have it until 4.x Feb 20 09:21:00 hci_usb, bluetooth Feb 20 09:22:06 I'll look through the code to see if there are any references to devfs; anything I should look for in particular? Feb 20 09:22:06 Nah, no devfs. Uses ordinary register. Feb 20 09:22:27 Just a sec... Feb 20 09:22:48 I have a binary ready for unslung. Feb 20 09:23:06 But only the bluetooth Feb 20 09:23:09 one Feb 20 09:23:22 What does hci_usb do? Feb 20 09:24:15 it runs the low level Host Controller Interface Feb 20 09:24:50 Ok - they should both compile - just enable the modules in defconfig Feb 20 09:25:24 That's what I kind of thought as well... but was wondering if there was any reason why OpenSlug might be preferred. Feb 20 09:25:33 I change both defconfig and linux-version/arch/arm/defconfig, but one should be enough I think Feb 20 09:26:00 Probably because it came from an openslugger? Feb 20 09:26:40 * bob_tm-away found a complete set of Geode family reference manuals and is heading for the recycle bin in the back yard. Feb 20 09:47:08 03jp30 * 10unslung/ (sources/php/php.ini make/php.mk): distribute a php.ini with php Feb 20 10:01:56 <[g2]> bob_tm-away, are you an EE Feb 20 10:02:29 03jp30 * 10unslung/ (make/php.mk sources/apache/apxs.patch): apache: clean up apxs behaviour when staging an apache module; php: take advantage of this Feb 20 10:02:35 Well, no - between the rock and the hard place - drivers & uCode at some point. Feb 20 10:03:09 <[g2]> uCode on processors ? Feb 20 10:03:31 Jepp - or rather processor cards (long time ago) Feb 20 10:03:46 mini machines. Feb 20 10:03:51 <[g2]> cool Feb 20 10:03:59 <[g2]> ot Feb 20 10:04:07 <[g2]> it's all merging anyway Feb 20 10:04:34 Eh... it had its day. Not many places today can you get a request for an instuction and have it ready the week after :-) Feb 20 10:04:58 but it ends in patch hell - the ucode was loadable from disk. Feb 20 10:05:26 <[g2]> softcores start the process all over again :) Feb 20 10:05:39 <[g2]> just a little faster load from disk :) Feb 20 10:06:30 I always loved the error message "No such CPU" - it was for real - the bootstrap/io processor reported it if the main processor was not found. Feb 20 10:08:15 The only thing I really miss was the ability to add and remove RAM while running. Feb 20 10:09:31 ...welll sometimes it crashed, but that feature was done by another department, so I didn't have to care. Feb 20 10:09:44 <[g2]> sounds fun Feb 20 10:10:27 Yeah, well. Wasted lots of perfectly good party-years in air-conditioned work spaces. Feb 20 10:11:15 When I reflect on all the cool music I could have listened to and all the nice sex I could have had, I wouldn't have chosen the same thing over. Feb 20 10:17:07 * bob_tm-away goes to the kitchen to make a cup of espresso and forget about lost youth. Speedy J keeps the beat up and facilitates dumping of tons of old manuals. Feb 20 11:59:51 03jp30 * 10unslung/make/apache.mk: apache: enable all modules; patch httpd.conf sources, rather than distributing our own static httpd.conf Feb 20 11:59:57 03jp30 * 10unslung/sources/apache/ (httpd-std.conf.in httpd.conf): apache: enable all modules; patch httpd.conf sources, rather than distributing our own static httpd.conf Feb 20 12:06:51 jp30: \o/ Hurray Feb 20 12:07:58 yeah, looks like its all go - only problem is rwhitby rushed the version I committed last night into the feed. now he will have to rebuild them all over again Feb 20 12:08:46 Well, worse things have happened. I'm much happier. Feb 20 12:09:03 Does it build cross? Feb 20 12:09:24 oh, yes... it builds cross. actually i'd better check that native works... Feb 20 12:09:47 * bob_tm-away is frantically running cvs update Feb 20 12:10:38 I haven't check that ssl works... and ldap support is missing at present. will have to get openldap to build cross. Feb 20 12:10:51 <[g2]> jp30, so it (apache/php) builds cross with dynamic libs now ? Feb 20 12:10:58 LDAP is not the end of the world. SSL is. Feb 20 12:11:22 [g2], yeah, absolutely. I put some notes on how its done on the wiki Feb 20 12:11:55 oh, PEAR is also missing. it will only build native Feb 20 12:12:32 ssl *should* work. it's just that I haven't tried it. Feb 20 12:12:52 Ok - time will tell Feb 20 12:12:55 it's not enabled by default at least... Feb 20 12:13:13 ...will check whether mod_ssl is actually being built... Feb 20 12:26:37 Ok - just leave a note - I'm back and forth Feb 20 12:33:17 is there a way to use a different portnumber with libbt C BitTorrent library? Feb 20 12:39:25 wow what great client Feb 20 12:39:29 i love java Feb 20 12:39:35 and i love all of you Feb 20 12:39:39 pretty l'll people Feb 20 12:39:57 hugs and kisses don't go changin for me! Feb 20 12:51:37 03jp30 * 10unslung/make/apache.mk: apache: enable many more modules, including ssl Feb 20 12:52:23 jp30 is on fire. Feb 20 12:52:40 thank you :) Feb 20 12:53:13 no, thank *you*! Feb 20 12:53:22 ~hail jp30 Feb 20 12:53:24 * jbot bows down to jp30 and chants, "I'M NOT WORTHY!!" Feb 20 12:54:30 next stop, ldap. Is the maintainer of openldap in the house? Feb 20 12:55:27 looks like it's caplink811-away Feb 20 12:56:11 yeah.. Feb 20 12:57:19 jp30: what's the matter? Feb 20 12:57:30 how close is openldap to building cross? it looks like i need some ldap libraries for apr-util to depend on - would openldap provide those? Feb 20 12:58:35 didn't try it yet, i build only native, cause i think ,openldap could also be building cross Feb 20 12:59:22 would try tomorrow build ldap cross Feb 20 12:59:54 hmm... ok I'll experiment. if I can get it to work, is it OK by you if I commit changes to openldap.mk? Feb 20 13:00:18 jp30: yes, sure, no problem Feb 20 13:01:06 caplink811-away, thanks, i'll see what i can do Feb 20 13:19:30 03rwhitby 07 * r1.2826 10openembedded/packages/ (3 files in 3 dirs): Enabled DEVFS for Unslung standard - 4.3-alpha Feb 20 13:24:28 jp30: unslung package feed is now updated with latest apache and php Feb 20 13:24:46 bob_tm-away: unslung standard now has DEVFS - thanks for reminding me Feb 20 13:24:57 * dyoung-zzzz revalidating switchbox build area Feb 20 13:25:19 ~worship jp30 for getting apache cross compiled Feb 20 13:25:21 * jbot prostrates himself before jp30 for getting apache cross compiled. "I'm not worthy!" Feb 20 13:25:23 <[g2]> rwhitby, I had another idea about migration to OpenSlug Feb 20 13:25:42 morning all! Feb 20 13:25:56 [g2]: I'm all ears :-) Feb 20 13:25:59 <[g2]> we could start my migrating the make scripts to use the OE tools and build all with OE created toolchains Feb 20 13:26:25 <[g2]> and phase out packages as they are OE buildable Feb 20 13:26:39 rwhitby, i may have a new apache for you in a couple of hours :/ if i can get openldap to cross-build Feb 20 13:26:49 <[g2]> kergoth would probably sh**-a-brick over the idea Feb 20 13:27:28 <[g2]> anybody know if jbowler ever is in here ? Feb 20 13:27:33 [g2]: what about the init scripts stuff that OE puts in the packages that Unslung can't handle? Feb 20 13:27:51 <[g2]> I wasn't talking for Unslung :( Feb 20 13:27:55 or are you talking about different makefile targets for openslug packages Feb 20 13:28:06 <[g2]> It was for OpenSlug Feb 20 13:28:06 same CVS repository, yes? Feb 20 13:28:42 <[g2]> I don't know how much in the make files will have to be tweaked Feb 20 13:28:55 sounds pointless to me Feb 20 13:29:10 <[g2]> kergoth, I thought you'd see it that way Feb 20 13:30:06 [g2]: why wouldn't we just have an OE package which uploads all the ipks to a feed (I presume there's something like that already) and just have that package depend on all the OE packages that have been verified to work with OE, and then take any Unslung packages that are not already in OE and add them to OE ? Feb 20 13:30:26 ... verified to work with OpenSlug .... Feb 20 13:30:54 kergoth: which page did you want me to update with init script stuff ? Feb 20 13:31:13 there's no point in building openslug packages outside of oe. it gains you nothing Feb 20 13:31:19 rwhitby: that wiki page on making unslung packages work in oe Feb 20 13:31:32 i'm working my way through the items on that page Feb 20 13:31:50 ok, will do that now Feb 20 13:33:11 <[g2]> rwhitby, the issue is twofold, 1) getting stuff built for OpenSlug, 2) Updating the packages in OE to work for OpenSlug (and all others) Feb 20 13:33:29 if apache in oe doesnt build, fix the build. Feb 20 13:33:32 this isnt that complecated. Feb 20 13:34:54 <[g2]> kergoth I agree with you, but there are a ton of devs out there that haven't made the transition and don't see it that way Feb 20 13:35:23 <[g2]> the approach you mention is skip 1) and focus on 2) Feb 20 13:35:40 <[g2]> but everyone isn't on the OE train Feb 20 13:35:43 you're trying to leverage unslung developers for openslug, instead of leveraging the countless folks in #oe that hack on many distributions Feb 20 13:35:47 its pointless Feb 20 13:35:50 a waste of your time and theirs Feb 20 13:36:58 <[g2]> possibly, I was thinking it's a bridge getting the unslung developers eased into being OE developers Feb 20 13:37:39 you're going to try to do that when oe's packages dont even work on unslung? Feb 20 13:37:42 good luck. Feb 20 13:38:26 I agree that getting OE packages to work with Unslung is the important first step. Feb 20 13:40:51 kergoth: Info added to the end of http://www.nslu2-linux.org/wiki/Unslung/OEPackagesForUnslung Feb 20 13:40:58 cool, thanks Feb 20 13:41:16 let me know if there is anything else you need over that - I have some time right now to add stuff, but that window will close in 10 minutes or so Feb 20 13:42:08 <[g2]> kergoth, I was going to push hotplug-ng use into OpenSlug Feb 20 13:42:39 <[g2]> I've been testing and things seems to be working much better with hotplug-ng Feb 20 13:42:54 i've been testing it a bit on my zaurus too Feb 20 13:42:59 i'm going to create a hotplug-common package Feb 20 13:43:11 with some of the agents and things needed by both Feb 20 13:43:21 (i.e. the net agent that calls ifup) Feb 20 13:43:41 <[g2]> kergoth ifup is getting called properly on OpenSlug Feb 20 13:43:59 hotplug-ng doesnt include any scripts to handle network events, so i dont see how thats possible. Feb 20 13:44:01 <[g2]> both the usbnet and ixp425 driver bring up the iface Feb 20 13:44:19 it includes scripts for pci, usb, and scsi only Feb 20 13:44:22 <[g2]> I just needed to add an auto eth0 Feb 20 13:44:27 gah Feb 20 13:44:36 so you're bringing them up only on boot. Feb 20 13:44:41 rather than by hotplug at insertion time Feb 20 13:45:11 <[g2]> I'm pretty sure it some up hotplug with the auto in there Feb 20 13:45:19 <[g2]> I can double check Feb 20 13:45:28 there's no net hotplug agent in hotplug-ng. Feb 20 13:45:36 therefore, hotplug runs no scripts that call ifup. Feb 20 13:45:42