**** BEGIN LOGGING AT Sat Mar 17 02:59:57 2007 Mar 17 06:05:05 the OpenPBX.org package awaits review http://bugs.openembedded.org/show_bug.cgi?id=2011 Mar 17 09:58:33 can anyone offer some help installing debian usb pen drive on the NSLU2 ? Mar 17 10:35:27 hillct: Will look into it shortly :) Mar 17 10:36:08 having problems installing debian to usb pen drive, can anyone help? Mar 17 10:56:15 hillct: ping? Mar 17 10:56:25 --with-perl-shebang='#! /usr/bin/perl' <-- absolute path. Bad. Mar 17 10:56:50 /usr/share/openpbx.org/ogi/*" <-- absolute path Mar 17 10:57:43 also, the postinst_prepend has a bunch of absolute paths Mar 17 10:58:20 It'll work with 3.10-beta, but it's not going into HEAD until that is fixed Mar 17 11:39:28 hello! Mar 17 11:40:15 I've build the slugosle-image, and would like to flash it to the slug, how to proceed? Mar 17 14:27:46 NAiL: the guys in #oe has some other comments too. I'll try and get them integrated over the next few days (today if I get a chance) Mar 17 14:28:32 hillct: great :) Mar 17 14:29:01 now that this part is almost done, comes the fun pert - coming up with a dialplan Mar 17 14:29:30 also, what are your thoughts on this? http://www.fivn.com/index.php?page=software&menu=asterisk Mar 17 14:29:56 I should be able to use such a script as-is although I haven't so much as looked at that code yet Mar 17 14:30:26 as in, in that OpenPBX is backward compatible with Asterisk when it comes to config files Mar 17 14:40:47 NAiL: the version of populate_volitiles.sh in 3.10-beta seems to be out of date in that it doesn't support creation of links Mar 17 14:58:49 hillct: hmm. I'll take a look at it. I'm too sleepy right now to do any work ;) Mar 17 14:59:29 the internet will never be able to escape the evils of timezones Mar 17 15:04:33 Well, it's 16:04 here, so I should really be awake now. Mar 17 20:53:55 huhu Mar 17 20:54:43 guys.. maybe someone has an idea regarding this; the problem is on the wl500g/oleg firmware... try/catch for C++ exceptions does not work at all on uclibc++? Mar 17 21:03:43 hi Mar 17 21:04:58 hey Mar 17 22:30:23 * Hamish feels like chanting....ah, the bliss of vmware.... Mar 17 22:30:30 anyone in here? Mar 17 22:31:14 I have a question regarding compilation from source for my slug. Mar 17 22:32:16 don't ask to ask, just ask Mar 17 22:32:35 okido. Mar 17 22:33:02 being a relative newbie: How do I use the cross-compile env. to compile a program from source? Mar 17 22:33:59 I have a dir, containing the source, I want to compile that source to be run on the slug. How to proceed? Mar 17 22:34:41 hmmmm, rwhitby has this makefile thingy that does all the setting of for you Mar 17 22:35:01 okay, where to get that makefile thingy? Mar 17 22:35:53 www.nslu2-linux.org/Makefile Mar 17 22:37:23 Would it help mentioning I have succesfully built e.g. the slugosle & be firmware? Mar 17 22:37:54 make setup, make update, make slugosle etc etc work, and builds. Mar 17 22:38:46 How do I use the Makefile, NAiL linked? Mar 17 22:39:07 Hamish: which distro are you going to build for? Mar 17 22:39:46 not quite sure yet, Mar 17 22:39:50 either is fine, Mar 17 22:40:04 slugos, optware? Mar 17 22:40:17 slugos is fine. Mar 17 22:40:24 make setup-openembedded Mar 17 22:41:22 *done* Mar 17 22:41:39 ok, are you building for littleendian or bigendian? Mar 17 22:41:51 I can still choose that. Mar 17 22:42:13 if I want to run debian later, which endianness will that have/require? Mar 17 22:42:38 is that little endian? Mar 17 22:42:50 yes Mar 17 22:42:58 littleendian is slow network-wise though Mar 17 22:43:03 little endian is it then. Mar 17 22:43:06 since a network is bigendian Mar 17 22:43:28 I would prefer stability over performance. Mar 17 22:43:41 ah, ok. Mar 17 22:44:05 is the difference big? Mar 17 22:44:16 used to be ~30 Mar 17 22:44:19 % Mar 17 22:44:26 I can live with that Mar 17 22:44:28 dunno what the numbers are with the new network driver Mar 17 22:44:39 it will only get better, I imagine. Mar 17 22:45:08 I won't be taxing the network much at first. Mar 17 22:45:16 ok Mar 17 22:45:23 then do "make slugosle-image" Mar 17 22:45:44 That'll build a toolchain and such Mar 17 22:46:14 Ok, I have done that already. Mar 17 22:46:33 it will report a succesfull build in a minute. Mar 17 22:46:41 uh, that takes a little while Mar 17 22:47:00 NOTE: package slugos-image-1.0: completed Mar 17 22:47:00 NOTE: build 200703172340: completed Mar 17 22:47:13 oookaay Mar 17 22:47:19 fast machine? Mar 17 22:47:22 it took hours before Mar 17 22:47:25 Hamish: you have a supercomputer :) Mar 17 22:47:32 not really, hehe Mar 17 22:47:42 oh, you have already done that before? Mar 17 22:47:43 it has everything cached I suppose. Mar 17 22:47:46 yes. Mar 17 22:47:50 to test the env. Mar 17 22:47:56 well then you already have a toolchain Mar 17 22:48:01 Hamish: you probably want to know how to work with .bb files Mar 17 22:48:03 I suppose I do, yes. Mar 17 22:48:34 I just don't know how to compile with the ARm as a target yet. Mar 17 22:48:59 if it's a very simple app Mar 17 22:49:06 it is. Mar 17 22:49:20 CC = /path/to/arm-linux-gcc make Mar 17 22:49:29 it's in tmp/cross/bin Mar 17 22:49:39 okay, let's give that a go. Mar 17 22:49:41 if it doesn't even have a makefile, /path/to/arm-linux-gcc -o outfile file.c Mar 17 22:49:50 thanks a million so far. Mar 17 22:50:10 and suggestions where I generally should place the sources? Mar 17 22:50:20 is there some kind of convention for that? Mar 17 22:50:36 e.g. usr/local/src/**appname**/ Mar 17 22:51:10 just trying to keep things a bit tidy. it's complex for me as it is. Mar 17 22:51:32 see "Making a package using the template" here http://www.nslu2-linux.org/wiki/Optware/AddAPackageToOptware if you want to make a package Mar 17 22:52:08 pierrelux: it's not optware Mar 17 22:53:52 I'll fetch the source of the app from svn, edit the makefile and give the compile a try. Mar 17 22:54:03 (this seems like a logical course of steps, right?) Mar 17 22:54:07 yeah Mar 17 22:55:26 i need to compile a custom driver for slugosbe. Any tips about that ? it's a one-makefile, one-source file project. I managed to get the image compiled using monotone Mar 17 22:56:06 pierrelux: look at one of the other similar out-of-kernel modules in OE Mar 17 22:56:38 eg. zd1211 Mar 17 22:56:45 or spca5xx Mar 17 22:57:10 ah, hello Rod. Mar 17 22:57:24 g'day Mar 17 22:57:39 * CoreDump|home point out that spca5xx doesn't compile for kernels > 2.6.11 Mar 17 22:57:48 rob >I am taking your advice to heart Mar 17 22:58:09 rwhitby: I see something like /tmp/work/ixp4xxbe-linux/ixp4xx-kernel-2.6.20-r3.768/install/kernel-module-usbserial. Could be ok ... Mar 17 22:58:36 pierrelux: that is an in-kernel-source module, not an out-of-kernel-source module Mar 17 22:58:40 which do you need? Mar 17 23:00:56 rwhitby: it's called ibot.c ... it's not on internet. an ov511 modified driver to be non-v4l Mar 17 23:02:31 ok, so out of kernel source module Mar 17 23:03:20 rwhitby: yes it is out of kernel Mar 17 23:04:33 so NAiL gave you an example package Mar 17 23:06:23 ok, checking that out... Mar 17 23:08:18 ok.. I've editted the Makefile, 'make' now tells me it cannot find usb.h, and libptp2/ptp.h Mar 17 23:08:25 cd .. Mar 17 23:10:02 i've downloaded the source for both. both come with a configure script. Mar 17 23:10:10 I don't know how to handle these. Mar 17 23:10:44 (configuring libptp & libusb I mean) Mar 17 23:12:38 libusb is in oe already Mar 17 23:14:03 ok, that is a step in the right direction, how do I go about using it? Mar 17 23:14:29 cd slugosle, source setup-env, bb libusb Mar 17 23:14:37 that should stage it in tmp/stage/ somewhere Mar 17 23:18:55 ok.. bb libusb completed. Mar 17 23:21:25 I have found these: ~/slug/slugosle/tmp/staging/arm-linux/lib$ ls -al |grep libusb Mar 17 23:21:25 lrwxrwxrwx 1 nslu2 nslu2 19 Mar 18 00:14 libusb-0.1.so.4 -> libusb-0.1.so.4.4.4 Mar 17 23:21:25 -rwxr-xr-x 1 nslu2 nslu2 37622 Mar 18 00:14 libusb-0.1.so.4.4.4 Mar 17 23:23:15 Do I need to edit the Makefile of the application to point to these? Mar 17 23:52:51 Hamish: have you read the openembedded manual? Mar 18 00:07:35 not in it's entirety, rod Mar 18 00:20:46 I am fairly cluesless as to where openembedded fits in the picture. Mar 18 00:35:46 NAiL: how big was that C++ fis executable? Mar 18 00:35:48 can kernel options be set via installable kernel modules? Mar 18 00:36:35 rwhitby: 888kb, but it didn't execute Mar 18 00:36:41 that is statically linked Mar 18 00:37:40 how about dynamically linked? Mar 18 00:38:16 ~90kb IIRC Mar 18 00:38:33 requires a bunch of libraries though Mar 18 00:39:06 yeah Mar 18 00:39:15 * rwhitby writes an email to phil Mar 18 00:39:28 why? Mar 18 00:40:30 k Mar 18 00:43:05 rwhitby: forwarded a mail to you Mar 18 00:46:00 ah, ok Mar 18 00:46:47 imho you came out a bit harsh there Mar 18 00:47:12 > Do you know how much work it'd take to port this utility to plain C? Mar 18 00:47:13 Not much.  Mainly the C++ stuff is just "convenience wrappers" around C Mar 18 00:47:13 system calls, which would need to be written out in full. Mar 18 00:50:39 NAiL: follow-up sent Mar 18 00:54:48 NAiL: the boost dependency is still a killer for build time. I thought it would be a five minute build, but all the build depencies has pushed that out to almost an hour. Mar 18 00:55:41 yeah, but if ported to plain c, that dep would disappear Mar 18 01:00:46 (And the size would probably be closer to 40kb shared) Mar 18 01:41:04 NAiL: you may want to check out http://bugs.openembedded.org/show_bug.cgi?id=2011 for the path updates you asked for (among other things) Mar 18 02:48:02 Cool, 2 prolific usb -> RS232 adapters for $21 shipped... Mar 18 02:52:03 joshin: if you make your own electronic stuff, you could get your hand on ft232 (5$). i used it for the slug: http://tech.ph.groups.yahoo.com/group/nslu2-linux/photos/browse/6ab5 Mar 18 02:53:13 pierrelux: yeah, I have a Max3232 chip right in front of me but I wanted a few more serial ports. Mar 18 02:53:47 I want to see if the slug has enough oomph for a couple ham radio things (like PSK31 with an USB audio device) Mar 18 02:54:25 let us know about it... looks interesting Mar 18 02:55:34 Could be. I expect it to be a bit too slow since it is lacking a FPU Mar 18 02:56:01 But I hope to be surprised. :) **** ENDING LOGGING AT Sun Mar 18 02:59:56 2007