**** BEGIN LOGGING AT Tue Feb 05 02:59:57 2008 Feb 05 05:05:28 nbd * r10391 /trunk/package/uci/Makefile: fix another crash Feb 05 05:09:12 bah! you always wait until I've started another build :-P Feb 05 05:56:07 git log Feb 05 10:50:26 nbd: can you upload the dtc-1.1.0.tgz to the openwrt mirror? Feb 05 13:36:33 blogic * r10392 /trunk/package/madwifi/files/lib/wifi/madwifi.sh: fix wpa_supplicant for madwifi Feb 05 14:34:48 hi, question. where do i have to look if I want to change the rootfs on the wg302v2-2 to NFS ? it seems hard coded to always be build to use jffs2 Feb 05 15:13:37 i dont know how it is done on other architectures, but on adm5120 you can change rootfs location in: make kernel_menuconfig, and there in "kernel hacking" you can change "default kernel command string" Feb 05 15:14:20 in it - could be setted rootfs on nfs Feb 05 15:21:02 thanks Feb 05 15:39:06 florian * r10393 /packages/net/natpmp/Makefile: Cleanup natpmp Makefile Feb 05 15:39:25 florian * r10394 /packages/net/weechat/Makefile: Make sure lua will get built Feb 05 15:40:45 florian * r10395 /packages/libs/libnetfilter-queue/ (. Makefile patches/ patches/001-no_utils.patch): Add libnetfilter-queue (#2870) Feb 05 15:43:35 florian * r10396 /trunk/package/comgt/files/3g.sh: Allow more ttyUSB devices to use the 3g script (#3098) Feb 05 15:58:16 florian * r10397 /packages/net/nufw/ (6 files in 2 dirs): Add nufw Feb 05 19:40:02 ralph * r10398 /trunk/tools/ (Makefile bison/ bison/Makefile): Device Tree Compiler require more recent bison ... Feb 05 20:22:58 nbd * r10399 /trunk/package/uci/files/uci/lib/config/uci.sh: fix a stupid copy&paste mistake Feb 05 20:30:47 nbd * r10400 /trunk/package/uci/Makefile: update uci - fixes a bug in parse_uci_tuple Feb 05 21:21:49 nbd: I'm having issues with my wl500gP wireless again. Bit different than the radar problem Feb 05 21:21:59 what kind of issues? Feb 05 21:22:07 the network is visible, but I can't connect. It just times out Feb 05 21:22:18 (And windows is ofcourse ultra-helpful on "why") Feb 05 21:22:24 hm, strange Feb 05 21:22:34 The router looks a-ok when I connect via wired Feb 05 21:22:51 except possibly this: Feb 05 21:22:51 11a or 11g? Feb 05 21:22:51 Mode:Master Frequency:5.3 GHz Access Point: 00:16:E3:83:F6:CE Feb 05 21:22:52 Bit Rate:0 kb/s Tx-Power:17 dBm Sensitivity=1/1 Feb 05 21:23:01 shouldn't bitrate be =! 0? Feb 05 21:23:10 no, it's ok if no client is connected Feb 05 21:23:17 it's automatic Feb 05 21:23:20 ah, ok Feb 05 21:23:44 there's nothing in dmesg at all Feb 05 21:24:31 According to Windows, the signal is nothing short of excellent Feb 05 21:24:41 but I can't connect Feb 05 21:25:59 does it happen in 11g as well? Feb 05 21:26:13 now windows auto-connected all of a sudden... Feb 05 21:26:26 I haven't tried using 11g Feb 05 21:28:15 btw, do you know what the difference between the 11a channels are? Feb 05 21:28:34 Channel 48 (5240 MHz) Max Tx Power 17 dBm (hw limited) [17 hw 20 reg] Flags CF_OFDM CF_5GHZ PF_4MS_LIMIT Feb 05 21:28:37 Channel 52 (5260 MHz) Max Tx Power 17 dBm (hw limited) [17 hw 20 reg] Flags CF_OFDM CF_5GHZ CF_PASSIVE_SCAN_ONLY PF_DFS_REQUIRED PF_4MS_LIMIT Feb 05 21:29:38 not really Feb 05 21:30:31 btw, it still says bitrate == 0, even though I'm now connected via wifi Feb 05 21:31:24 probably still normal Feb 05 21:31:29 ok Feb 05 21:32:04 I suspected wpa-supplicant in the beginning. I kinda thought it'd run as a daemon or such Feb 05 21:32:10 but that doesn't appear to be the case? Feb 05 21:34:12 Anyway, this whole disconnection-thing happens on a daily basis. Feb 05 21:41:12 does kamikaze do parallell builds btw? Feb 05 21:41:21 ie, make -J? Feb 05 21:43:44 yes Feb 05 21:43:49 advanced options Feb 05 21:44:17 nbd, whats the word on .24 for broadcom? Feb 05 21:45:12 Weedy: last I heard he's working on other things first. the code accepted upstream will require some work to accomodate. Feb 05 21:45:47 Weedy: where in advanced options? Feb 05 21:46:36 NAiL: the option was removed from menuconfig around 1000 revisions ago, you have to run `make -j#' (where # is the job number) Feb 05 21:47:05 Bartman007: ah, ok :) Feb 05 21:47:41 oops, ~2150 revisions ago (r8249) Feb 05 21:48:21 the exact amount of revisions aren't that important to me ;) Feb 05 21:48:48 yeah, but I thought the actual revision might be of historical use. Feb 05 21:49:15 ah, yeah Feb 05 21:49:26 well, just adding it on the commandline is good enough for me Feb 05 21:49:29 oops, r8237 actually killed it, r8249 just removed a leftover option. Feb 05 21:49:35 hahah :P Feb 05 21:51:33 Just want to speed up compiles on the box now that "the box" is actually dual-core with 6gb ram Feb 05 21:53:41 I've found the extra cpu power is better spent performing concurrent builds. I can build 3 arches in about the same time as it takes one on a quadcore box. Feb 05 21:54:35 ah, yes, but I'm only building for my own (mipsel?) router Feb 05 21:55:49 and anyway it's only using one core Feb 05 21:56:19 some packages are forced to compile on a single core Feb 05 21:56:21 Bartman007, it was? Feb 05 21:56:25 awesome Feb 05 22:01:04 Weedy: maybe I misspoke: It is my understanding that the code accepted upstream is different then ours, our kernel hackers will have to restructure our code around the new upstream code. Feb 05 22:01:39 hmm Feb 05 22:02:20 forward-port the patches? Feb 05 22:02:39 but nbd's main priority at the moment appears to be getting the new uci code stabilized Feb 05 22:02:45 the code upstream is from debian people, mostly for WGT634U. If I read that correctly Feb 05 22:02:51 yes Feb 05 22:03:02 * Weedy is annoyed UCI isnt in svn Feb 05 22:03:06 uci is in svn Feb 05 22:03:14 where? Feb 05 22:03:20 there. Feb 05 22:03:26 im diffing tarballs ATM Feb 05 22:03:34 a tarball referenced from an openwrt build system Feb 05 22:03:39 as for the actual repo, it's git Feb 05 22:03:43 yeah Feb 05 22:03:44 git://nbd.name/uci.git Feb 05 22:03:47 i ment repo Feb 05 22:03:58 http://nbd.name/gitweb.cgi?p=uci.git;a=summary Feb 05 22:04:02 easy to browse as well Feb 05 22:04:13 with tagged releases Feb 05 22:07:08 Weedy: the hope is that the uci format will be adopted by other projects and become the primary config format for some programs. IMO keeping the code directly in the openwrt codebase would hinder the potential for adoption Feb 05 22:07:59 besides, git is fun! Feb 05 22:08:08 heh Feb 05 22:08:42 Bartman007: you have a f*cked-up definition of "fun" ;) Feb 05 22:08:47 one thing that would be useful for git is the reverse of git-svn - an svn server that serves from a git repo Feb 05 22:09:01 nbd: there is work on that already. Feb 05 22:09:12 oh, cool Feb 05 22:09:18 looking forward to that Feb 05 22:09:28 I haven't looked at the progress in at least a month but was quite excited when I heard about it Feb 05 22:10:49 Weedy: If hg or bzr or foo was able to work as well (and fast) w/ svn as git does I'd probably say they were fun instead. Feb 05 22:10:55 btw, regarding 2.6.24... is updating from .23.1 to .24 a lot of work, or mainly quilt fingerwork? Feb 05 22:11:31 git-svn was the one thing that made me go with git instead of hg Feb 05 22:11:43 im assuming lots of work since it isnt done already Feb 05 22:22:07 crappy wireless! Feb 05 22:22:37 nbd: what was that trick to switch off radar detection again? Feb 05 22:22:58 iwpriv ath0 doth 0 Feb 05 22:23:16 thank you Feb 05 22:23:23 madwifi doth protest too much? Feb 05 22:23:39 radar misdetects happen Feb 05 22:23:42 it's a known issue Feb 05 22:23:57 i think i'll just add that command by default in openwrt Feb 05 22:24:20 * Bartman007 is sitting outside of NAiL's building with a radar gun. Feb 05 22:24:44 nbd: where's the best place to add it? Feb 05 22:24:54 Bartman007: rather wide frequency radar ;) Feb 05 22:25:39 nbd * r10401 /trunk/package/madwifi/files/lib/wifi/madwifi.sh: disable radar detection by default until it is fixed Feb 05 22:25:45 ^^ Feb 05 22:25:48 :) Feb 05 22:47:57 what is the point of the radar detection anyhoo? Feb 05 22:52:40 NAiL: from what I've seen DFS is a European restriction being pushed so wifi doesn't interfere with other devices using the spectrum Feb 05 22:53:28 when it detects anohter device it is supposed to send a packet to all connected clients that tells them what channel to switch to, then switches itself. Feb 05 22:54:37 I tend to think of DFS as a bunch of teenagers partying while the parents are away; when the parents come back they scatter and meet up somewhere else. Feb 05 23:05:44 nice in theory Feb 05 23:06:34 I've gotta say I find it very unlikely that my router is interfering with anything at all. It can hardly push the signal all the way to my bedroom Feb 05 23:08:55 the thing I find interesting is that it is supposed to switch when it detects other wifi networks also. Feb 05 23:09:20 I can envision chaos as the number of networks increased, esepcailly if all of them support DFS. Feb 05 23:09:29 s/increased/increases/ **** ENDING LOGGING AT Wed Feb 06 02:59:57 2008