**** BEGIN LOGGING AT Fri Sep 05 02:59:56 2008
Sep 05 03:29:42 <frogonwheels> I'm getting an issue compiling asterisk1.4 because my autoconf on my computer is earlier than the toolchain one!
Sep 05 03:29:49 <frogonwheels> It looks like the toolchain one is being used - but using the libs from the host one!
Sep 05 06:53:15 <CIA-12> juhosg * r12536 /trunk/target/linux/generic-2.6/patches-2.6.26/980-backport_gpio_sysfs_support.patch: [kernel] 2.6.26: fix build error caused by the gpio sysfs stuff
Sep 05 08:23:11 <frogonwheels> ok luci  svn3104 is still not working for me: dispatcher.lua:123 attempt to index field 'main' ( a nil value)
Sep 05 08:28:42 <KanjiMonster> hi, I have a dev-question: how do I check whether a configuration variable is set (a boolean) in a Makefile? I want to do something depending whether its a BE or LE system (and consequently want to check whether BIG_ENDIAN is set or not)
Sep 05 08:34:05 <frogonwheels> look at the asterisk config - think that has some ifdefs in it.
Sep 05 08:34:51 <KanjiMonster> okay, thanks
Sep 05 08:35:37 <KanjiMonster> (my target is to work around psybnc's broken endianess detection)
Sep 05 09:16:16 <exovii> Kaloz: here ?
Sep 05 09:39:28 <frogonwheels> luci  svn3104 is still not working for me: dispatcher.lua:123 attempt to index field 'main' ( a nil value)
Sep 05 09:46:40 <KanjiMonster> frogonwheels: just a shot in the dark, but I had a similiar problem some revisions ago after upgrading uci and forgetting to also upgrade libuci
Sep 05 09:47:27 <KanjiMonster> (and libuci-lua)
Sep 05 09:47:51 <frogonwheels> ah
Sep 05 09:49:28 <frogonwheels> that's probably it.
Sep 05 09:49:30 <frogonwheels> ta
Sep 05 09:49:40 <frogonwheels> possibly building them woudl help too.
Sep 05 12:29:05 <frogonwheels> Well that was fun
Sep 05 12:29:35 <frogonwheels> luci managed to temporarily brick my router.
Sep 05 12:43:02 <xMff> frogonwheels: where you able to determine why?
Sep 05 12:43:20 <frogonwheels> xMff: nope. had to reload a trx
Sep 05 12:43:30 <xMff> damn
Sep 05 12:43:37 <frogonwheels> xMff: probably due to pressing save on the ssl page.
Sep 05 12:43:43 <frogonwheels> ssl keys
Sep 05 12:43:48 <frogonwheels> that's the only thing I did
Sep 05 12:44:02 <frogonwheels> ssh sorry
Sep 05 12:44:07 <xMff> hmm
Sep 05 12:44:11 <frogonwheels> (hoping it would restart dropbear)
Sep 05 12:44:18 <frogonwheels> then booom
Sep 05 12:44:27 <xMff> so Dropbear was gone then?
Sep 05 12:44:35 <xMff> permanently
Sep 05 12:44:43 <frogonwheels> well that's not unusual -
Sep 05 12:44:51 <frogonwheels> if I load up the system, something often drops out.
Sep 05 12:45:00 <frogonwheels> and the interface certainly loaded up the sytem.
Sep 05 12:45:02 <xMff> hmm perhaps had fubar'd the ssh config
Sep 05 12:45:15 <frogonwheels> xMff: yeah - but everything was screwed.
Sep 05 12:45:23 <xMff> really strange
Sep 05 12:45:29 <xMff> what platform?=
Sep 05 12:45:37 <frogonwheels> mipsel
Sep 05 12:45:59 <xMff> okay, I noticed strange behaviour on a wl500gp yesterday too
Sep 05 12:46:12 <frogonwheels> wl500gp for me too
Sep 05 12:46:15 <xMff> sometimes you had to reboot twice to get the system up and running
Sep 05 12:46:29 <frogonwheels> twice, thrice, even rescue mode failed.
Sep 05 12:46:34 <xMff> will investigate this weekend
Sep 05 12:46:37 <frogonwheels> went into rescue mode, but refused to be rescued.
Sep 05 12:47:19 <frogonwheels> had to use Rescue.exe
Sep 05 12:47:21 <frogonwheels> :)
Sep 05 12:47:24 <xMff> hehe
Sep 05 12:48:29 <xMff> If you try again sometime it would be nice if you can run "top" in background to see whether unusual things happen
Sep 05 12:48:51 <xMff> luci sometimes restarts several services on the fly
Sep 05 12:49:09 <xMff> i think something is wrong there under certains circumstances
Sep 05 13:02:48 <frogonwheels> oh F$ uci isn't working very nicely
Sep 05 13:05:27 <nbd> 15:02 < frogonwheels> oh F$ uci isn't working very nicely
Sep 05 13:05:28 <nbd> details?
Sep 05 13:05:46 <frogonwheels> i've just rebooted.
Sep 05 13:05:49 <frogonwheels> wait a sec
Sep 05 13:06:53 <frogonwheels> nbd: I'm doing a uci export asterisk
Sep 05 13:07:05 <frogonwheels> and getting: uci: Parse error (option/list command found before the first section) at line 159, byte 1
Sep 05 13:07:21 <nbd> show me the contents of the asterisk config
Sep 05 13:07:28 <frogonwheels> 	option 'message' 'Q IMp'
Sep 05 13:07:28 <frogonwheels> config 'voicezone' 'military'
Sep 05 13:07:28 <frogonwheels> 	option 'zone' 'Zulu'
Sep 05 13:07:39 <frogonwheels> hmm. I'll pastebin the whole lot
Sep 05 13:07:41 <frogonwheels> sorry
Sep 05 13:10:55 <frogonwheels> http://pastebin.com/d541d3957
Sep 05 13:13:32 <nbd> ok, the error message for this problem could be improved
Sep 05 13:13:35 <frogonwheels> nbd The strangest thing is that when using -S .. the line  option 'sip
Sep 05 13:13:43 <nbd> but the problem is that there are multiple sections with the same name
Sep 05 13:14:00 <frogonwheels> ?
Sep 05 13:14:27 <nbd> config 'voice' 'general'
Sep 05 13:14:30 <frogonwheels> which bit is the name.
Sep 05 13:14:31 <nbd> config 'incoming' 'general'
Sep 05 13:15:01 <nbd> 'general' in this case is the name
Sep 05 13:15:19 <nbd> it must be unique or left out
Sep 05 13:15:33 <frogonwheels> you mean it has to be unique for the whole config!?
Sep 05 13:15:47 <nbd> yes
Sep 05 13:16:00 <nbd> because it's used by the shell code as pointer to the section
Sep 05 13:16:38 <frogonwheels> well that rather screws me over.
Sep 05 13:16:46 <frogonwheels> was working in a previous version :(
Sep 05 13:17:17 <nbd> it could not have worked properly
Sep 05 13:17:28 <nbd> because in shell you reference config sections by the name only
Sep 05 13:17:30 <frogonwheels> ah well it did. possibly much older script.
Sep 05 13:17:48 <nbd> maybe it was just hidden
Sep 05 13:17:50 <frogonwheels> ah - I use config_load
Sep 05 13:18:04 <nbd> because previously it did not complain about something that is problematic
Sep 05 13:21:53 <frogonwheels> nbd: ok - rather bizarre that it doesn't use both the section and name for the key..
Sep 05 13:22:19 <frogonwheels> hey how do you match blank in a case) in shell?
Sep 05 13:22:41 <nbd> the first parameter is not the section itself
Sep 05 13:22:44 <nbd> the first parameter is the type
Sep 05 13:23:05 <frogonwheels> oh.
Sep 05 13:24:30 <frogonwheels> nbd: can you have non alpha chars in the type?
Sep 05 13:35:59 <nbd> some, yes
Sep 05 13:36:05 <nbd> e.g. "-"
Sep 05 13:37:25 <frogonwheels> and _
Sep 05 13:37:30 <frogonwheels> but that seems about it.
Sep 05 13:37:34 <frogonwheels> - could work for me.
Sep 05 13:40:00 <frogonwheels> nbd: an error when the type changes would be good too
Sep 05 13:40:17 <nbd> i'll simply add a duplicate section error message
Sep 05 13:40:21 <nbd> doesn't have to mention the type
Sep 05 13:40:28 <nbd> since it's not the type change that's problematic
Sep 05 13:40:44 <frogonwheels> yeah fair enough. both are a bad.
Sep 05 13:41:03 <frogonwheels> in this case it _is_ the type change that's problematic.
Sep 05 13:41:06 <frogonwheels> .. but I agree
Sep 05 13:49:46 <frogonwheels> doesn't handle -
Sep 05 15:14:42 <CIA-12> lars * r12537 /packages/XOrg/lib/pixman/Makefile: explicitly disable gtk tests.
Sep 05 17:55:51 <Yanira> http://openwrt.pastebin.com/d1bba409 -- and now, how to print that url on the web page?
Sep 05 18:12:29 <Yanira> yeah, all this hotplug stuff is still broken...
Sep 05 18:12:34 <Yanira> ups
Sep 05 18:12:40 <Yanira> http://openwrt.pastebin.com/d1bba409 -- and now, how to print that url on the web page?
Sep 05 18:13:26 <xMff> Yanira: do you mean how to embed the address in a template?
Sep 05 18:13:41 <Yanira> yep :)
Sep 05 18:13:44 <xMff> ahh, mom
Sep 05 18:14:36 <Yanira> xMff: i have the address hardcoded in m = Map("uvc_streamer", translate("uvc_streamer"), translate("uvc_streamer_desc")) uvc_streamer_desc atm
Sep 05 18:14:45 <Yanira> and i like to make it dynamic
Sep 05 18:16:09 <CIA-12> nbd * r12538 /packages/net/openntpd/files/ntpd.hotplug: fix openntpd hotplug script (based on patch from #3027)
Sep 05 18:19:12 <xMff> Yanira: so you want to include the address within the translation of "uvc_streamer_desc".. right?
Sep 05 18:19:29 <Yanira> exactly
Sep 05 18:19:50 <xMff> Yanira: there is a shortcut named "translatef()" (like sprintf)
Sep 05 18:20:12 <xMff> translatef("uvc_streamer_desc", "1.2.3.4")
Sep 05 18:20:34 <xMff> in the language file you have to put a "%s" where the parameter should included
Sep 05 18:21:10 <blogic> Yanira: which cam are you using ?
Sep 05 18:21:25 <blogic> i mean, can you tell me a uvc capable cam that i can get off the shelf ?
Sep 05 18:21:29 <Yanira> blogic: quickcam pro 2007 for notebooks
Sep 05 18:21:53 * blogic orders one online
Sep 05 18:22:34 <Yanira> i'd like to see mjpg-streamer config moved to uci
Sep 05 18:22:42 <CIA-12> nbd * r12539 /packages/net/openntpd/files/ntpd.init: only start ntpd from hotplug
Sep 05 18:22:54 <Yanira> xMff: i'll give it a try
Sep 05 18:23:49 <xMff> Yanira: my code was slightly wrong, because translatef() still takes a default translation you have to "pad" the params, e.g.: translatef("foo", nil, "param1", "param2", ...)
Sep 05 18:24:11 <xMff> then it should work
Sep 05 18:33:53 <Yanira> xMff: http://openwrt.pastebin.com/d6c3af8a8 -- to what should i change that two files to get it working?
Sep 05 18:37:59 <xMff> Yanira: http://openwrt.pastebin.com/m3bc4492b - but it's to better to make the "%s" change in the xml language file, because the lua translation file is generated (and overwritten) from the xml translation when you do make
Sep 05 18:38:28 <xMff> so don't change *.en.lua but *.en.xml instead
Sep 05 18:40:50 <Yanira> ok. so no i'll try to get uvc_streamer.@uvc_streamer[0].port
Sep 05 18:40:56 <Yanira> s/no/now/
Sep 05 18:41:37 <xMff> Yanira: mom
Sep 05 18:44:36 * Yanira is waiting...
Sep 05 18:53:32 <xMff> Yanira: retrieving values from anonymous sections is a bit trickier with the lua-uci binding
Sep 05 18:53:37 <xMff> Yanira: http://openwrt.pastebin.com/d76a15cd0
Sep 05 18:56:42 <Yanira> nice, nice. but difficult
Sep 05 18:57:04 <Yanira> and now i'd like to have that url as a linke
Sep 05 18:57:39 <xMff> Yanira: I think I will implement the @sectiontype[index] format in our uci:get()
Sep 05 18:57:53 <xMff> but will take some time
Sep 05 18:58:56 <Yanira> hmm, if i change the port and hit save & apply the url is only updated after reloading the page
Sep 05 19:06:43 <xMff> Yanira: this is a limitation of the form generation cycle, forms are processed before the uci data is written, so the vars aren't updated after the first reload
Sep 05 19:07:01 <Yanira> ah. ok
Sep 05 19:07:03 <xMff> anyway here is the version with a link: http://openwrt.pastebin.com/d76a15cd0
Sep 05 19:08:18 <xMff> ah sorry, this one: http://openwrt.pastebin.com/d7acf9e24
Sep 05 19:10:51 <Yanira> nice. works :)
Sep 05 19:11:00 <Yanira> can you commit?
Sep 05 19:11:21 <xMff> sure
Sep 05 19:14:13 <xMff> Yanira: it's okay to give credits to you? The uvc stuff is still missing file headers
Sep 05 19:14:57 <Yanira> xMff: ok
Sep 05 19:27:02 <xMff> Yanira: it's committed
Sep 05 19:28:03 <Yanira> okie :)
Sep 05 23:27:30 <CIA-12> nbd * r12540 /trunk/package/madwifi/Makefile: madwifi: select the right hal binary for EABI on ARM
Sep 05 23:27:41 <CIA-12> nbd * r12541 /trunk/package/madwifi/patches/102-multicall_binary.patch: madwifi: add missing include file for the tools multicall binary
**** ENDING LOGGING AT Sat Sep 06 02:59:57 2008