**** BEGIN LOGGING AT Thu Oct 22 02:59:57 2009
Oct 22 08:46:58 <CIA-94> jow * r18119 /trunk/package/opkg/ (5 files in 2 dirs): (log message trimmed)
Oct 22 08:46:58 <CIA-94> [package] opkg:
Oct 22 08:46:58 <CIA-94> - replace package list parser by something that works with fixed buffers
Oct 22 08:46:58 <CIA-94> to avoid slurping the whole file into memory during parsing
Oct 22 08:46:58 <CIA-94> - fix a possible null pointer dereference in opkg_remove_cmd()
Oct 22 08:46:59 <CIA-94> - get rid of upgrade, it cannot upgrade single packages and kills
Oct 22 08:47:01 <CIA-94> the system in most cases
Oct 22 08:48:13 <loswillios> nice
Oct 22 08:48:37 <xMff> :)
Oct 22 08:54:21 <xMff> still the opkg codebase is an unbelievable mess
Oct 22 08:55:03 <xMff> ok, time to go to work. bbl
Oct 22 08:56:54 <blogic> yay, ethernet working on ramips
Oct 22 08:57:11 <blogic> need to make aplatform drievr etc etc
Oct 22 08:57:16 <blogic> but its sort of working now
Oct 22 13:54:14 <netprince> xMff: I use opkg upgrade <package> all the time, was it really broken? :(
Oct 22 13:55:18 <netprince> I have all my important scripts in a package, and I can upgrade each router via opkg upgrade <scriptpackage>
Oct 22 13:58:21 <xMff> netprince: problem is that it is dangerous, will probably re-add it but only allow upgrades for single packages
Oct 22 14:02:25 <nbd> netprince, xMff: shouldn't opkg install <package> work as well?
Oct 22 14:02:34 <nbd> instead of upgrade
Oct 22 14:03:17 <xMff> nbd: just checked, upgrade may do install but install does not do upgrade
Oct 22 14:03:28 <nbd> oh
Oct 22 14:04:15 <xMff> everything in this code is inconsistent, the function names, the error handling, the indentination, the logic ...
Oct 22 14:05:02 <xMff> and the usage text does not even mention that upgrade may take args
Oct 22 14:06:45 <Lalloso> {Nico}: hello {Nico} are you still working on freeswitch packages?
Oct 22 15:10:45 <{Nico}> Lalloso: not really, just waiting for feedback :)
Oct 22 15:11:26 <Lalloso> from me?
Oct 22 15:11:45 <Lalloso> i mean packages are ready to be tested? :-)
Oct 22 15:15:21 <CIA-94> jow * r18120 /trunk/package/opkg/patches/ (5 files):
Oct 22 15:15:21 <CIA-94> [package] opkg:
Oct 22 15:15:21 <CIA-94> - re-enable upgrade and restrict it to signle packages, fix usage text
Oct 22 15:15:21 <CIA-94> - only read package descriptions if they're actually needed (almost never),
Oct 22 15:15:21 <CIA-94> saves even more space when parsing package lists
Oct 22 15:15:22 <CIA-94> - refresh patches
Oct 22 15:16:01 <xMff> netprince: ^^^
Oct 22 15:16:06 <netprince> :)
Oct 22 15:16:07 <netprince> thanks!
Oct 22 15:16:15 <{Nico}> Lalloso: yep
Oct 22 15:16:49 <Lalloso> wow
Oct 22 15:17:00 <Lalloso> _trine: have you heard?
Oct 22 15:17:08 <Lalloso> we must test now!!!
Oct 22 15:17:23 <Lalloso> how many flavours have you done {Nico} ?
Oct 22 15:17:55 <Lalloso> could you explain me a little bit how one could decide the list of freeswitch modules to build with the package?
Oct 22 15:23:38 <Lalloso> is there a virtual machine system which works well with openwrt?
Oct 22 15:36:48 <Lalloso> {Nico}: wonderful, i've seen you have done all the freeswitch packages for the different modules also
Oct 22 15:37:30 <Lalloso> you know what could be missing maybe
Oct 22 15:37:35 <Lalloso> a freeswitch-core
Oct 22 15:37:37 <Lalloso> package
Oct 22 15:37:45 <Lalloso> so that one could install just that core and some modules
Oct 22 15:38:02 <Lalloso> i guess that freeswitch-minimal already contains such modules like mod_console
Oct 22 15:57:57 <{Nico}> Lalloso: i made it fully modular, so there's freeswitch main package and all freeswitch-mod-* modules
Oct 22 15:58:46 <Lalloso> so the freeswitch main package does not contain any module?
Oct 22 15:58:54 <{Nico}> Lalloso: then there are freeswitch-default (default upstream config with a common set of modules) and freeswitch-minimal (a minimal setup, the one we talked about)
Oct 22 15:59:01 <{Nico}> Lalloso: nope
Oct 22 15:59:07 <Lalloso> thx wonderful than
Oct 22 15:59:25 <Lalloso> maybe you could call it freeswitch-core just to clarify but maybe it's better just freeswitch
Oct 22 16:00:06 <Lalloso> thx a lot for your work {Nico} when _trine we'll be back online we'll test and give feedbacks
Oct 22 16:00:25 <{Nico}> cool
Oct 22 16:01:18 <{Nico}> we should add a minimal default config with exactly what's required for the freeswitch-minimal to work
Oct 22 16:05:15 <Lallobot> that's not hard
Oct 22 16:05:28 <Lallobot> oh sorry (wrong irc client)
Oct 22 16:26:13 <CIA-94> nbd * r18121 /trunk/rules.mk: add some extra stripping options for kernel modules to reduce size even more (thx, matteo)
Oct 22 16:39:40 <CIA-94> nbd * r18122 /trunk/rules.mk: fix a typo
Oct 22 16:54:20 <CIA-94> nbd * r18123 /trunk/tools/mklibs/patches/008-gc_sections.patch: mklibs: use -Wl,--gc-sections when relinking - currently only saves a few bytes, maybe more in the future
Oct 22 17:16:15 <CIA-94> jow * r18124 /trunk/package/opkg/patches/008-fix_parsing_insanity.patch: [package] opkg: ignore Section, Source and Maintainer as well when ignoring descriptions, saves a few hundred KB memory
Oct 22 17:32:31 <OutBackDingo> nbd: whoa dude....  nice
Oct 22 17:33:14 <OutBackDingo> saved 100k in the image 2359704 and thats with luci web ui include
Oct 22 17:46:47 <netprince> my build r18112 on broadcom 2.4 seems to be rebooting every few mins, strange
Oct 22 17:50:17 <netprince> no kernel panic, just: CFE version 1.0.37 for BCM947XX (32bit,SP,LE) ...
Oct 22 17:50:47 <xMff> netprince: do you see spurious sysrq messages in dmesg?
Oct 22 17:52:46 <netprince> I'm interfaced w/ serial port, watching the messages that way, haven't seen anything strange.  Should I check dmesg also?
Oct 22 17:52:53 <xMff> netprince: yes
Oct 22 17:53:00 <netprince> ok, will do
Oct 22 17:53:18 <xMff> however, I experienced sometimes that a shaky serial triggered sysrq
Oct 22 17:53:57 <xMff> and you see absolutely nothing before the reboot? No watchdog stuff etc.?
Oct 22 17:54:01 <netprince> I was wondering about that, will try without and see
Oct 22 17:54:36 <netprince> no, I scrolled back to check
Oct 22 17:54:41 <netprince> will look again
Oct 22 17:56:32 <netprince> xMff: here is one reboot: http://openwrt.pastebin.ca/1637902
Oct 22 17:57:11 <xMff> netprince: what device is that exactly?
Oct 22 17:57:19 <netprince> wl-520gu
Oct 22 17:57:25 <xMff> hm
Oct 22 17:58:13 <xMff> well, you should check whether this happens without serial cable too
Oct 22 17:58:21 <netprince> yes, will try that next
Oct 22 17:58:39 <xMff> sometimes connecting/touching the wrong pin will instantly reboot the device
Oct 22 18:13:44 <netprince> xMff: I bet that was the problem, seems fine now, and I was getting some serial jibberish earlier
Oct 22 18:25:39 <OutBackDingo> xMff: sup, question can we not run luci ui on its own now without lucid ?
Oct 22 18:26:05 <xMff> OutBackDingo: should work fine if you use the "sgi-cgi"
Oct 22 18:26:58 <OutBackDingo> xMff: ahhh ok ll look at it
Oct 22 18:29:05 <OutBackDingo> xMff: ever heard of flashsys2 ?
Oct 22 18:29:31 <xMff> OutBackDingo: yes, even looked at the code. Shiny ui and crappy backend
Oct 22 18:29:51 <OutBackDingo> hahaha shiny ui that doesnt seem to be able to add / remove interfaces
Oct 22 18:30:35 <OutBackDingo> i did a build on trunk of it, and couldnt even configure VAPs or add / remove wireless interfaces
Oct 22 18:32:15 <xMff> well, the usual way to achieve apparent simplicity is cutting down "complex" features
Oct 22 18:34:33 <OutBackDingo> xMff: wasnt found of needing to use hostapd either, though they claim they support VAPs there didnt appear to be anywhere to configure them
Oct 22 18:34:40 <OutBackDingo> gui was clean though
Oct 22 18:36:24 <xMff> no wonder if it is restricted to a few common use cases :)
Oct 22 18:36:37 <xMff> need to run now, will be back later
Oct 22 19:19:57 <blogic> OutBackDingo: did you get lucid running ?
Oct 22 19:22:29 <OutBackDingo> blogic: yeah seems xMff also mentions i can run stand alone with sgi-cgi
**** ENDING LOGGING AT Fri Oct 23 02:59:57 2009