**** BEGIN LOGGING AT Thu Apr 01 02:59:57 2010
Apr 01 03:17:10 <cshore> is it possible to use a fallback to git (instead of URL) for source code?
Apr 01 05:19:56 <builder> build #54 of ppc44x is complete: Failure [failed compile_6]  Build details are at http://tksite.gotdns.org:8010/builders/ppc44x/builds/54
Apr 01 05:29:51 <{Nico}> cshore: fallback ?
Apr 01 05:30:19 <cshore> {Nico}: i.e. if the tarball at the URL isn't there use git
Apr 01 05:32:44 <{Nico}> humm, no
Apr 01 05:33:57 <cshore> {Nico}: what about SVN?
Apr 01 05:34:56 <{Nico}> cshore: we try to mirror tarballs, instead of fallback'ing
Apr 01 05:36:16 <cshore> ah....I was trying to deal with freeswitch-latest which only keeps one or two days worth of builds (it's a personal repository, not in tree until it becomes a release version)
Apr 01 05:45:12 <CIA-39> nico * r20643 /packages/net/ctorrent/Makefile: [packages] ctorrent: fix build failure on avr32
Apr 01 05:57:22 <CIA-39> nico * r20645 /packages/net/hping3/patches/ (100-makefile.patch 101-endianness.patch 102-lib_rename.patch): [packages] hping3: refresh patches
Apr 01 06:00:58 <CIA-39> nico * r20646 /packages/net/hping3/patches/101-endianness.patch: [packages] hping3: fix build failure on avr32 (closes: #7006)
Apr 01 06:38:13 <builder> build #31 of uml is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/uml/builds/31
Apr 01 07:18:01 <builder> build #53 of atheros is complete: Failure [failed compile_11]  Build details are at http://tksite.gotdns.org:8010/builders/atheros/builds/53
Apr 01 07:48:51 <acoul1> any rb600 owner out there willing to test this: http://dev.ozo.com/openwrt/mpc83xx/openwrt-mpc83xx-rb600.elf ?
Apr 01 08:01:14 <Lalloso> cshore:
Apr 01 08:01:30 <Lalloso> after updating and installing your freeswitch-latest feed
Apr 01 08:01:37 <Lalloso> should it appear under net in make menuconfig?
Apr 01 08:03:13 <Lalloso> maybe there is a problem with your domain?
Apr 01 08:13:21 <cshore> it may need updating again
Apr 01 08:14:11 <cshore> in some dir try git pull http://cshore.is-a-geek.com/openwrt-repo/feed
Apr 01 08:14:27 <cshore> sorry git clone http://cshore.is-a-geek.com/openwrt-repo/feed
Apr 01 08:15:25 <cshore> if that works, then the problem is that the feed needs to be updates (tarballs are built daily and are only kept for a day or two by the freeswitch team)
Apr 01 08:15:54 <cshore> or else I got the md5sum wrong
Apr 01 08:16:49 <cshore> so pls let me know if git clone works....if it doesn't then I have to fix my site, otherwise it's the package
Apr 01 08:22:21 <xMff> last time feeds complained that dumb http transport does not support --depth
Apr 01 08:22:32 <xMff> ... which is used by feeds
Apr 01 08:27:52 <dkostousov> Hi. Which is svn-revision number for 10.03-rc2
Apr 01 08:28:21 <dkostousov> ?
Apr 01 08:29:17 <xMff> take alook at the OpenWrt.config
Apr 01 08:30:09 <xMff> or see here https://dev.openwrt.org/milestone/Backfire%2010.03-rc2  ("revisions used")
Apr 01 08:30:24 <dkostousov> thx
Apr 01 08:32:22 <cshore> xMff: oh yeah...feeds has to be edited to allow that
Apr 01 08:32:35 <cshore> xMff: what does --depth do anyway?
Apr 01 08:33:02 <xMff> cshore: make the checkout smaller by skippping all the history information
Apr 01 08:33:49 <cshore> xMff:  maybe we should have a fallback so that if with depth fails we do a full checkout
Apr 01 09:13:38 <CIA-39> nico * r20647 /packages/utils/upx/patches/101-avr32.patch: [packages] upx: add support for avr32 (closes: #7009)
Apr 01 09:18:18 <Lalloso> cshore: your site works well
Apr 01 09:19:02 <Lalloso> i think there is some problem with the package
Apr 01 09:19:30 <Lalloso> https://dev.openwrt.org/wiki/GetSource
Apr 01 09:19:48 <Lalloso> i don't understand if this co of packages is needed or not?
Apr 01 09:24:47 <Lalloso> what is the difference between
Apr 01 09:24:48 <Lalloso> scripts/feed update -a
Apr 01 09:25:21 <dkostousov> Is it way to update feeds to specific revision with "feeds" script?
Apr 01 09:25:31 <dkostousov> *"feed"
Apr 01 09:26:04 <Lalloso> and svn co svn://svn.openwrt.org/openwrt/packages ?
Apr 01 09:26:44 <xMff> Lalloso: feeds will always check out the latest version and maintain its own internal workcopy
Apr 01 09:27:07 <xMff> Lalloso: while a manual checkout allows you to use specific version
Apr 01 09:27:29 <xMff> Lalloso: the manual checkout can be used by editing feeds.conf to use src-link instead of src-svn
Apr 01 09:27:39 <Lalloso> ah i see
Apr 01 09:27:46 <Lalloso> so i can stick to feeds for my need
Apr 01 09:28:24 <Lalloso> cshore i think your freeswitch-latest package somehow conflicts with freeswitch 1.0.4 now i'm redoing a fresh checkout
Apr 01 09:29:55 <Lalloso> besides I think your package lacks mod-lua :-)
Apr 01 09:49:42 <noz> What's the preferred method for -rc2 installation reports and issues? Trac?
Apr 01 09:50:01 <cshore> Lalloso: I work on that....and on getting a git daemon going
Apr 01 09:50:25 <Lalloso> it is going now cshore
Apr 01 09:50:30 <Lalloso> don't misunderstand me
Apr 01 09:50:37 <Lalloso> i had a problem with my corporate proxies at usual
Apr 01 09:50:48 <Lalloso> I think it now your git repo works well
Apr 01 09:51:06 <xMff> noz: trac is fine
Apr 01 09:51:10 <Lalloso> what is the time there?
Apr 01 09:51:23 <cshore> 5:51 am
Apr 01 09:54:35 <Lalloso> you never sleep
Apr 01 09:54:51 <cshore> no, just not the same time every day
Apr 01 09:55:09 <Lalloso> when you woke up?
Apr 01 09:55:48 <cshore> around 1 or 2pm....
Apr 01 09:56:35 <Acinonyx> hello
Apr 01 10:03:30 <cshore> hello
Apr 01 10:09:20 <noz> FYI, I've just added "Backfire 10.03 RC2" to the list of versions in Trac
Apr 01 10:52:40 <xMff> noz: do you have referenced wlan0 in /e/c/network by any chance?
Apr 01 11:19:41 <noz> xMff: Yes, I have a "config 'interface' 'wlan'" section, with the line
Apr 01 11:19:41 <noz> 	option 'ifname' 'wlan0'
Apr 01 11:19:41 <noz> in it
Apr 01 11:20:19 <xMff> this is the reason for you race condition most likely
Apr 01 11:20:36 <noz> I migrated forward to Backfire-RC2 from an ancient self-built trunk (round about r12000)
Apr 01 11:20:42 <noz> but preserved config
Apr 01 11:20:58 <noz> xMff: What's the solution?
Apr 01 11:21:10 <xMff> you should use option network wifi in /e/c/wireless instead and leave ifname blank in /e/c/network
Apr 01 11:21:24 <noz> OK. I'll try that. Many thanks
Apr 01 11:21:26 <xMff> this ensures that the wifi is added to wlan once it is ready
Apr 01 11:21:40 <xMff> erm option network wlan I meant
Apr 01 11:22:53 <noz> Question is: dare I brave another reboot (and ensuing chaos)? :-)
Apr 01 12:05:29 <CIA-39> kaloz * r20648 /trunk/ (30 files in 5 dirs): [x86]: add a subtarget and drivers for the Intel Tolapai SoC
Apr 01 14:27:19 <OutBackDingo> nbd: ping
Apr 01 14:47:35 <nbd> OutBackDingo: pong
Apr 01 14:48:02 <OutBackDingo> nbd: ok need ur brain on madwifi
Apr 01 14:48:22 <OutBackDingo> if channel set to auto and adhoc mode with SSID for interface
Apr 01 14:48:34 <OutBackDingo> boot radio should scan for SSID and connect correct ?
Apr 01 14:48:49 <OutBackDingo> though it seems in 5gh it always seems to default to chan 36
Apr 01 14:49:07 <OutBackDingo> atheros 2k device
Apr 01 14:49:13 <nbd> no idea. i never worked on autochannel for ad-hoc
Apr 01 14:49:21 <OutBackDingo> whos the autochannel brain ?
Apr 01 14:49:28 <OutBackDingo> because it doesnt appear to function
Apr 01 14:49:30 <nbd> nobody
Apr 01 14:49:34 <OutBackDingo> LOL
Apr 01 14:49:40 <OutBackDingo> great
Apr 01 14:49:44 <nbd> i figured most people doing mesh stuff should avoid autochannel anyways
Apr 01 14:50:02 <OutBackDingo> nbd: yeah well im smarter then the marketing team
Apr 01 14:50:44 <nbd> static bssid is also a good idea for mesh
Apr 01 14:50:47 <OutBackDingo> someone wanted to find a smart intelligent way to recover orphaned nodes due to a channel change
Apr 01 14:50:52 <nbd> makes the whole thing more reliable
Apr 01 14:51:02 <OutBackDingo> which is what ive tried to explain
Apr 01 14:52:06 <OutBackDingo> i spoke to xMff about iwinfo to see if there was really a smart way to accomplish a radio searching for its adhoc mesh neighbors and switching its channel to rejoin the mesh
Apr 01 14:52:20 <OutBackDingo> to me, the logic really make no sense
Apr 01 14:52:44 <nbd> it certainly makes no sense to abuse autochannel for that
Apr 01 14:52:52 <OutBackDingo> without having to loop through every channel
Apr 01 14:52:52 <nbd> since that's just as likely to hit the home channel
Apr 01 14:53:09 <OutBackDingo> nbd: any thoughts on a better logic ?
Apr 01 14:53:14 <nbd> regular scanning code
Apr 01 14:53:18 <OutBackDingo> other then dont do it :)
Apr 01 14:53:31 <nbd> make sure it does not treat networks on different channels as duplicate
Apr 01 14:53:52 <nbd> that way it's much more targetted and comes with less downtime for the node on the main channel
Apr 01 14:54:01 <OutBackDingo> i looked at iwlist ath0 scan but the is not adhoc BSSID in the output
Apr 01 14:54:12 <nbd> either way, i won't have time to look at madwifi for a few weeks
Apr 01 14:54:19 <nbd> since i'm still doing contract work on ath9k
Apr 01 14:54:31 <OutBackDingo> no worries, just wanted to pick your brain
Apr 01 14:55:01 <jow_laptop> OutBackDingo: "is no adhoc BSSID in the output", cannot confirm that
Apr 01 14:55:21 <jow_laptop> [...] Cell 01 - Address: 02:CA:FF:EE:BA:BE [...]
Apr 01 14:56:08 <OutBackDingo> jow_laptop: ODD
Apr 01 14:56:21 <OutBackDingo> your using iwlist iface scan ?
Apr 01 14:56:24 <jow_laptop> yea
Apr 01 14:56:29 <OutBackDingo> in trunk ?
Apr 01 14:56:36 <jow_laptop> no, 8.09
Apr 01 14:56:41 <OutBackDingo> ahhhahhh
Apr 01 14:56:46 <OutBackDingo> im in trunk
Apr 01 14:56:52 <OutBackDingo> and there is no output
Apr 01 14:56:54 <jow_laptop> paste the output somewhere
Apr 01 14:57:16 <jow_laptop> well if there is no output then try to repeat the scan multiple times
Apr 01 14:58:10 <OutBackDingo> http://pastebin.com/kcNAraA2
Apr 01 14:58:15 <OutBackDingo> jow_laptop: dont that
Apr 01 14:58:18 <OutBackDingo> done that
Apr 01 14:59:36 <jow_laptop> but you do use ad-hoc and not ahdemo or something?
Apr 01 14:59:50 <OutBackDingo> interface is in ahdemo
Apr 01 15:00:10 <OutBackDingo> on 5ghz 802.11a
Apr 01 15:00:16 <jow_laptop> wasn't the whole point of ahdemo to not broadcast beacons?
Apr 01 15:00:21 <jow_laptop> no wonder you cannot scan it
Apr 01 15:00:27 <OutBackDingo> crap, your right
Apr 01 15:01:24 <jow_laptop> ahdemo even more relies on correct static config since there is no way to find the cell (apart from probing all channels)
Apr 01 15:04:49 <OutBackDingo> jow_laptop: yeah we use a static bssid
Apr 01 15:08:16 <OutBackDingo> jow_laptop: sooo now i wonder if channel "auto" would find the other node and connect
Apr 01 15:08:25 <OutBackDingo> though i really dont agree with their logic
Apr 01 15:11:25 <jow_laptop> no, channel auto won't find the node and does not load to a reconnect
Apr 01 15:11:44 <jow_laptop> *lead
Apr 01 15:19:47 <OutBackDingo> so a process to do iwlist grep for bssid, get channel iwconfig interface, guess iwlist or iwinfo
Apr 01 15:22:15 <CIA-39> acinonyx * r20649 /trunk/package/base-files/files/sbin/hotplug-call: [package] base-files: Cancel firmware loading if file doesn't exist
Apr 01 15:39:26 <CIA-39> juhosg * r20650 /trunk/package/mac80211/ (2 files in 2 dirs):
Apr 01 15:39:26 <CIA-39> mac80211: remove MCU requests for SoC platforms in the rt2x00 driver
Apr 01 15:39:26 <CIA-39> The ralink SoC platforms do not have an MCU.
Apr 01 15:39:26 <CIA-39> Signed-off-by: Luis Correia <luis.f.correia@gmail.com>
Apr 01 15:43:38 <OutBackDingo> xMff: did you say stunnel with uhttpd ?
Apr 01 15:44:52 <thepeople> OutBackDingo: uhttpd support ssl with cyassl
Apr 01 15:45:08 <OutBackDingo> hrmmm
Apr 01 15:46:00 <OutBackDingo> thepeople: thanks
Apr 01 15:49:41 <OutBackDingo> thepeople: so i have a non-ssl daemon and the only way to secure it i guess is with stunnel ? which used openssl any thoughts on using cyassl to accomplish this ?
Apr 01 15:50:08 <OutBackDingo> guess i could try to create a stunnel-cyassl package :PP
Apr 01 15:50:32 <jow_laptop> depending on the complexity
Apr 01 15:50:51 <OutBackDingo> jow_laptop: true
Apr 01 15:50:52 <jow_laptop> cyassl is meant to be a drop-in replacement but it only covers a subset of the openssl api
Apr 01 15:51:27 <jow_laptop> can't you jsut add ssl to your daemon? Its easy
Apr 01 15:52:10 <OutBackDingo> jow_laptop: maybe for you :P LOL
Apr 01 15:52:13 <thepeople> OutBackDingo: stunnel won't work on cyassl, at least when I last tried it, it would take a bit of modification
Apr 01 16:11:55 <OutBackDingo> man
Apr 01 16:12:16 <OutBackDingo> i dont really want two separate ssl libs and i dont want the fat of openssl
Apr 01 17:28:02 <CIA-39> hauke * r20651 /trunk/target/linux/brcm-2.4/files/arch/mips/bcm947xx/prom.c:
Apr 01 17:28:02 <CIA-39> brcm-2.4: fix commit r18413 "128MB ram problem"
Apr 01 17:28:02 <CIA-39> This was missing in commit r20072
Apr 01 17:59:16 <dogge10> xMff, https://forum.openwrt.org/viewtopic.php?pid=105829#p105829
Apr 01 19:53:56 <[florian]> cshore: the imagetag segfault disappeared, going to test on bcm6338 now
Apr 01 20:39:05 <builder> build #44 of octeon is complete: Success [build successful]  Build details are at http://tksite.gotdns.org:8010/builders/octeon/builds/44
Apr 01 21:20:18 <[florian]> cshore: everything worked ok on bcm6338, bcm6348 (Gw6k, gw6k2)
Apr 01 21:20:47 <CIA-39> florian * r20652 /trunk/ (5 files in 4 dirs):
Apr 01 21:20:47 <CIA-39> [brcm63xx] flashmap and image generation: reduced union bcm_tag to a single struct
Apr 01 21:20:48 <CIA-39> combining the elements so that it is no longer necessary to create an openwrt-only
Apr 01 21:20:48 <CIA-39> tagid and tagcrc, and elimate the tagid detection and switch statements which
Apr 01 21:20:48 <CIA-39> made dealing with imagetags overly complicated, especially since the logic would
Apr 01 21:20:48 <CIA-39> need analogs in all code that touched the imagetag. Patch from cshore.
Apr 01 22:29:38 <blogic> rtz2: ping
**** ENDING LOGGING AT Fri Apr 02 02:59:56 2010