[07:11] <achim71000> i have an buildroot related question, guess thats the right place
[13:20] <CIA-4> nico * r3978 /branches/buildroot-ng/openwrt/package/linux-atm/ (Makefile patches/300-no_autotools.patch): prevent autoheader invocation, clean autotools patch.
[14:39] <CIA-4> nbd * r3979 /branches/buildroot-ng/openwrt/package/madwifi/ipkg/: remove old ipkg directory
[14:40] <CIA-4> nbd * r3980 /branches/buildroot-ng/openwrt/package/base-files/rb532-2.6/ (. etc/ etc/config/ etc/config/network): add missing base-files for rb532
[15:05] <CIA-4> nico * r3981 /branches/buildroot-ng/openwrt/package/diag/ (Makefile diag_led.c src/ src/Makefile src/diag.c): move source files to src/ and let the kernel tree build the module.
[15:09] <CIA-4> nbd * r3982 /branches/buildroot-ng/openwrt/package/rules.mk: add support for conffiles,preinst,... through templates
[15:09] <nbd> heh... using template in make makes things really weird
[15:09] <nbd> for example:
[15:09] <CIA-4> nico * r3983 /branches/buildroot-ng/openwrt/package/diag/src/diag.c: suppress warnings.
[15:09] <nbd> +COMMANDS += if [ -n "$$$$$$$$pkg_$(subst -,_,$(1))_$(2)" ]; then echo "$$$$$$$$pkg_$(subst -,_,$(1))_$(2)" > $(2); fi;
[15:10] <CIA-4> nbd * r3984 /branches/buildroot-ng/openwrt/package/iptables/Makefile: add conffiles to iptables
[15:11] <nbd> {Nico}: you can enter the contents of conffiles, preinst, etc. without any extra separators now
[15:14] <{Nico}> nbd: great
[15:15] <nbd> what i didn't try is if it can handle " in strings
[15:15] <nbd> but i believe it should work
[15:33] <CIA-4> nbd * r3985 /branches/buildroot-ng/openwrt/package/rules.mk: fix extra control files for packages containing "." and allow different conffiles for multiple packages built from the same directory
[15:38] <CIA-4> nbd * r3986 /branches/buildroot-ng/openwrt/package/base-files/Makefile: add conffiles for base-files
[16:18] <achim71000> mad
[16:33] <CIA-4> kaloz * r3987 /branches/buildroot-ng/openwrt/target/linux/ (15 files in 12 dirs): upgrade to 2.6.17, probably break some stuff temporary
[17:15] <CIA-4> kaloz * r3988 /branches/buildroot-ng/openwrt/target/linux/generic-2.6/patches/201-multiple_default_gateways.patch: fix the multiple gateway patch, too
[19:09] <CIA-4> nico * r3989 /branches/buildroot-ng/openwrt/package/alsa/ (Makefile ipkg/): remove ipkg directory, normalize Makefile.
[19:46] <CIA-4> nico * r3990 /branches/buildroot-ng/openwrt/package/openssl/ (Config.in Makefile ipkg/): remove ipkg directory, normalize Makefile.
[20:03] <CIA-4> nico * r3991 /branches/buildroot-ng/openwrt/package/mtd/ (Makefile mtd.c src/ src/Makefile src/mtd.c): move source file to src/.
[20:04] <CIA-4> nbd * r3992 /branches/buildroot-ng/openwrt/package/wlc/ (18 files in 4 dirs): add wlc utility
[20:05] <CIA-4> nico * r3993 /branches/buildroot-ng/openwrt/package/robocfg/ (7 files in 2 dirs): move source files to src/.
[20:10] <CIA-4> nico * r3994 /branches/buildroot-ng/openwrt/package/hostapd/ (Makefile ipkg/): remove ipkg directory, add conffiles, normalize Makefile.
[20:22] <CIA-4> nico * r3995 /branches/buildroot-ng/openwrt/package/dropbear/ (Makefile files/S50dropbear files/dropbear.init):
[20:22] <CIA-4> remove unneeded autoconf call and flags, rename S50dropbear to dropbear.init,
[20:22] <CIA-4> normalize Makefile:
[20:22] <CIA-4> - indent configure invocation,
[20:22] <CIA-4> - use "install -d" instead of "mkdir -p".
[20:26] <CIA-4> nico * r3996 /branches/buildroot-ng/openwrt/package/linux-atm/ (Makefile files/br2684.hotplug files/br2684.init): rename br2684.init to br2684.hotplug, normalize Makefile.
[20:27] <CIA-4> nico * r3997 /branches/buildroot-ng/openwrt/package/dnsmasq/Makefile: add conffiles, use proper optimization flags when building.
[20:30] <CIA-4> nico * r3998 /branches/buildroot-ng/openwrt/package/ (19 files in 19 dirs): normalize Makefiles.
[21:16] <CIA-4> nbd * r3999 /branches/buildroot-ng/openwrt/package/madwifi/Makefile: override LDOPTS to remove madwifi compile warning
[21:16] <CIA-4> nbd * r4000 /branches/buildroot-ng/openwrt/package/util-linux/ (. Makefile patches/ patches/501-mconfig.patch): add util-linux
[21:16] <nbd> yay... r4000 :)
[22:25] <CIA-4> kaloz * r4001 /branches/buildroot-ng/openwrt/target/linux/netfilter.mk: follow the xtables changes in 2.6.17
[22:35] <[mbm]> so .. rc6
[22:35] <nbd> rc6...
[22:35] <{Nico}> hmm, rc6
[22:35] <Kaloz> huh, another rc? ;)
[22:35] <[mbm]> I still have plans of cleaning up diag and doing some webif changes before rc6
[22:35] <nbd> yeah
[22:35] <nbd> a few things need to be done
[22:36] <[mbm]> how many boards actually have the nvram setting for reset_gpio (or any other gpio mappings?)
[22:36] <nbd> i don't think there's many of them
[22:36] <nbd> and i don't know if we can even trust these settings :)
[22:37] <[mbm]> hrm, was hoping that we could cut down on the number of magic values to identify boards and just use nvram
[22:37] <[mbm]> so we wouldn't have to constantly update for every new board revision
[22:39] <nbd> i don't think we can do that
[22:39] <[mbm]> hrm.. suppose not
[22:40] <nbd> what i would like to get in rc6 is my jffs2 patch
[22:40] <nbd> so that we can get rid of the unnecessary reboot
[22:40] <[mbm]> yep
[22:40] <[mbm]> someone was asking about that in #openwrt last week
[22:40] <[mbm]> I don't think they liked my response
[22:40] <nbd> and wbx seems to have 'fixed' this by adding an autoreboot
[22:40] <nbd> which i think is a bad idea
[22:40] <[mbm]> ditto
[22:41] <[mbm]> that was the suggestion that had been made in #openwrt
[22:41] <nbd> :)
[22:41] <[mbm]> and I said I had other plans
[22:41] <nbd> anyway, the patch works well on the routerboard
[22:41] <nbd> haven't hit any weird sideeffects yet
[22:41] <[mbm]> looked pretty simple
[22:41] <[mbm]> so I can't imagine there would be any odd cases
[22:42] <{Nico}> nbd: do you have anything we can play with?
[22:42] <nbd> {Nico}: the patch still needs to be backported
[22:42] <nbd> {Nico}: but for 2.6 it's already in svn
[22:42] <nbd> although not yet in the generic patch directory
[22:42] <[mbm]> as for the webif stuff, I want to abstract the whole lan/wan interfaces so you can setup as many ("ifup groups"?) as you want
[22:43] <nbd> i think we should do that for the /etc/config based stuff
[22:43] <nbd> otherwise we have to do the work twice
[22:43] <[mbm]> I think I'll end up doing a version for rc6
[22:44] <nbd> ok
[22:44] <CIA-4> nico * r4002 /branches/buildroot-ng/openwrt/ (10 files in 5 dirs): move wlcompat from target/linux/package/ to package/.
[22:45] <nbd> so do we all agree on releasing rc6 as 0.9?
[22:45] <{Nico}> so, that's it, there's no more packages anymore in target/linux/package/ in buildroot-ng
[22:45] <nbd> when no major issues come up?
[22:46] <[mbm]> right
[22:46] <{Nico}> ok
[22:46] <[mbm]> btw does anyone have any ideas about what to do for failsafe mode on unknown boards?
[22:47] <nbd> bridge all interfaces?
[22:47] <[mbm]> need to address the problem of "openwrt works on ___!" and then a week later "how come I can't get into failsafe on ___?"
[22:47] <nbd> better failsafe access on wan as well than no access at all
[22:47] <[mbm]> well, the problem is really that we don't have any reset button to get into failsafe on those boards
[22:48] <nbd> and we don't even know which board is unknown
[22:48] <[mbm]> the only ideas that I can come up with is that if diag doesn't explicitly know were the reset button is that it forces a bootup into failsafe mode
[22:49] <nbd> yeah
[22:49] <[mbm]> or some way of watching all the gpios for a change and assuming that the change means the reset was pressed
[22:49] <nbd> hmm... yeah
[22:52] <[mbm]> btw we're still getting some reports of the kmod-switch and b44 acting strangely on some boards
[22:52] <[mbm]> but I haven't been able to get any information to actually track down the issue
[22:52] <nbd> which boards?
[22:52] <[mbm]> the wrt54g v1.1 was one of them
[22:52] <nbd> define 'acting strangely'
[22:53] <[mbm]> from the sounds of it, absolutely no ethernet
[22:53] <nbd> i will test on my 1.1 before rc6
[22:54] <[mbm]> there was another board which was somewhat odd, I'd have to go to the logs and look
[22:54] <[mbm]> it had a vlan switch but for some reason it was still using eth0/eth1 as lan/wan
[22:55] <malbon> some switches do that. ie ks8995. You can have seperate macs talking to the switch, and phy 5 indipendently
[22:56] <[mbm]> but it was shoing up as /proc/switch which means it was one of the kmod-switch (adm or robo) which don't do that
[22:58] <nbd> [mbm]: sometimes it's also the issue of dd-wrt screwing with the nvram settings
[22:58] <[mbm]> hmm
[22:58] <nbd> [mbm]: because it often completely ignores the previous settings and just sets its own
[22:58] <nbd> it tries to correct the nvram rather than deal with the quirks
[22:58] <nbd> which is more work and more unreliable
[23:01] <[mbm]> ok, so for rc6 we have
[23:01] <[mbm]> - webif updates
[23:01] <[mbm]> - diag updates (how to deal with failsafe on unknown boards?)
[23:01] <[mbm]> - confirm wrt54g v1.1
[23:01] <[mbm]> anything else?
[23:02] <Kaloz> nothing else comes into mind
[23:02] <nbd> i think there was something else, i just don't remember it right now
[23:02] <h3sp4wn> usb2 fix (i.e so it can be used on wl500-gd stable)
[23:02] <nbd> yeah, usb2
[23:02] <nbd> for the gd and the premium
[23:02] <[mbm]> ok
[23:02] <nbd> maybe the premium fix will also fix the gd bugs
[23:03] <nbd> it seems to be a backport of several 2.6 things
[23:03] <[mbm]> and after the rc6 release comes the kamilaze/whiterussian merger for 1.0
[23:03] <nbd> yeah, after 0.9
[23:03] <[mbm]> well, I plan on starting that prett much immediately after rc6 hits the front page
[23:03] <malbon> is wgt634u a target for rc6?
[23:03] <nbd> yeah
[23:03] <nbd> no
[23:04] <malbon> k just checking.
[23:04] <[mbm]> so there will basically be a 0.9 and a 1.0 in parallel
[23:04] <nbd> wgt634u has no usable nvram
[23:04] <nbd> makes sense
[23:04] <nbd> but before that happens, we need to fold back buildroot-ng into trunk
[23:04] <[mbm]> yeah, that happens now while we're working on rc6
[23:05] <nbd> right
[23:05] <{Nico}> [mbm]: how will 1.0 appear? in a new branch?
[23:05] <[mbm]> the usb stuff might need to wait until the 1.0 tree has started so we can update some drivers
[23:05] <nbd> i might have an usb fix soon
[23:06] <nbd> asus has the mentioned backports in their tree
[23:06] <[mbm]> {Nico}: think that branches/whiterussian will just be overwritten with the new 1.0 branch
[23:06] <nbd> (they finally got around to releasing the wl-500gP gpl code)
[23:06] <[mbm]> and 0.9 will work out of tags/0.9
[23:06] <nbd> yeah
[23:06] <{Nico}> [mbm]: ok
[23:07] <[mbm]> how's the packages tree coming along?
[23:07] <nbd> doesn't have many packages yet
[23:07] <nbd> but most of the stuff in the buildroot-ng branch is sorted out
[23:07] <nbd> but there are a few things that really need to get fixed soon
[23:07] <[mbm]> hmm different structure than I expected
[23:07] <nbd> what did you expect?
[23:08] <[mbm]> well, I didn't expect to see the libs/ directory
[23:08] <nbd> i thought we were going to make a hierarchical structure in packages/
[23:08] <[mbm]> figured it would just be something that you could dump directly on packages/ in buildroot-ng
[23:08] <[mbm]> not complaining
[23:09] <nbd> in the future, i would like to write some scripts that automatically resolve dependencies and copy over a certain package to the buildroot from the packages directory
[23:09] <[mbm]> ok
[23:09] <nbd> so you say copy /path/to/packages/pkgname and it pulls all the libraries along as well
[23:10] <[mbm]> no idea how you intend to pull that off, but I'll wait until I see an implementation
[23:10] <nbd> easy: go into that directory, do cd .. until you find the packages root
[23:10] <[mbm]> well, that covers everything I wanted to discuss
[23:10] <nbd> then look for a cached .pkginfo
[23:11] <nbd> [mbm]: how's the verbose stuff for buildroot-ng coming along?
[23:11] <nbd> [mbm]: got any new ideas on how to implement it properly?
[23:11] <nbd> you don't need to do the implementation alone, i just don't know where to start yet
[23:11] <nbd> i want to avoid the previous mess
[23:12] <[mbm]> yeah, I still haven't come up with any good ideas
[23:12] <[mbm]> so I've been working on the web stuff
[23:12] <nbd> k
[23:13] <[mbm]> if make wasn't so subborn I'd just redefine $MAKE to be a macro
[23:13] <nbd> yeah, make is a bitch
[23:13] <nbd> some things are just weird
[23:13] <[mbm]> might end up putting something in the scripts directory and setting $MAKE to that
[23:14] <nbd> it took me quite a while until i came up wiath a way to even pass a multiline string to the shell
[23:14] <nbd> the make developers probably never thought of templates as being used for anything other than make code
[23:14] <[mbm]> yeah, we're really abusing poor make
[23:15] <nbd> still better than writing a bloated python monster :)
[23:15] <[mbm]> well, our scripts directory means that the build system requires the perl monster
[23:15] <florian__> re
[23:15] <nbd> hey
[23:16] <[mbm]> wrt: stfu.
[23:16] <nbd> hehe... i'm faster than wrt :)
[23:16] <florian__> sorry for being late
[23:16] <florian__> I was in the train from Nantes to Paris
[23:16] <nbd> no problem
[23:16] <[mbm]> actually I put a delay into wrt
[23:16] <nbd> ah
[23:16] <[mbm]> (types too fast, must be a bot ..)
[23:16] <nbd> :)
[23:17] <nbd> the delay works sometimes
[23:17] <[mbm]> yeah, it's funny to see people talk to wrt
[23:17] <[mbm]> and even funner is when he seems to respond properly
[23:17] <malbon> wrt is a good conversaionalist. ;)
[23:17] <nbd> [mbm]: wombat in #madwifi is also quite good
[23:17] <nbd> seems to use something like megahal
[23:17] <Kaloz> hehe, yeah
[23:18] <[mbm]> heh, I'll have to look at that channel
[23:19] <nbd> [mbm]: if you talk to it, it'll respond with some random bot-talk
[23:19] <nbd> and it also responds to some trigger words
[23:19] <[mbm]> yeah I've seen megahal's before
[23:19] <[mbm]> they're usually very easy to spot since nothing they say makes any sense
[23:19] <nbd> but when it does, it's really funny
[23:22] <[mbm]> well, unless anyone else has any topics I think the meeting is over
[23:23] <[mbm]> ...
[23:24] <{Nico}> nbd, [mbm]: can you detail the changes you want in webif?
[23:24] <nbd> changing the structure of that confusing firewall page
[23:24] <[mbm]> yeah the firewall page really needs to be changed
[23:24] <nbd> btw. {Nico}, florian__ - could you come to this event? https://snr.freifunk.net/trac/wcc/wiki
[23:24] <nbd> it would be ideal for a developer meeting
[23:25] <[mbm]> some of the wireless settings are missing and other wireless settings are grouped in strange ways
[23:25] <{Nico}> nbd: i'll try, but i'm not sure i can stay the whole days
[23:25] <[mbm]> like the wds is mixed with the mac filtering settings
[23:25] <[mbm]> and it's somewhat confusing
[23:26] <{Nico}> [mbm]: you want a separate page for wds?
[23:26] <[mbm]> no, I mean the page looks like this
[23:26] <[mbm]> [wds setting]
[23:26] <[mbm]> [mac setting]
[23:26] <[mbm]> [wds setting]
[23:26] <[mbm]> [mac setting]
[23:27] <[mbm]> but it should be like this:
[23:27] <[mbm]> [wds settings]
[23:27] <[mbm]> ...
[23:27] <[mbm]> [mac settings]
[23:27] <[mbm]> ...
[23:28] <[mbm]> in other words there should almost be boxes of content on the page
[23:28] <[mbm]> one box for all the wds settings
[23:28] <[mbm]> and one box for all the mac settings
[23:28] <[mbm]> so that they're clearly grouped
[23:28] <nbd> right
[23:30] <{Nico}> ok, i'll try to fix it but i have to dive into webif first
[23:31] <[mbm]> heh good luck, webif is almost it's own undocumented programming language
[23:31] <{Nico}> what are the missing wireless settings? (might be easier to start with webif)
[23:31] <nbd> heh, yeah
[23:31] <[mbm]> hmm I don't have a list of them but it shouldn't be hard to look at the factory page and see
[23:32] <[mbm]> I know that the long/short preamble is missing
[23:32] <[mbm]> as are some of the rate controls
[23:32] <[mbm]> isolation should be a feature
[23:32] <[mbm]> ...
[23:33] <[mbm]> basically our advanced wireless page should be much much longer than it is
[23:33] <{Nico}> can ayone post a screenshot of the linksys page (just for inspiration)?
[23:34] <alxhh> hmmm, powerlevel
[23:34] Action: [mbm] has to go .. bb(much)l
[23:35] <h3sp4wn> Are the packages which were previously in florians repository going to be integrated at any point ?
[23:35] <nbd> powerlevel.. that'll encourage all the OMGWTF-251-mW-ZOMG-LOL1!!1 people
[23:35] <{Nico}> h3sp4wn: what are you missing?
[23:36] Action: RoDent loves the new wlc
[23:36] <h3sp4wn> ncftp specifically - I could built it myself but have not done so because of the fact he has already done it
[23:36] <{Nico}> h3sp4wn: anything else?
[23:37] <Kaloz> h3sp4wn: bug florian, he promised to integrate everything :)
[23:37] <h3sp4wn> It had a massive amount of stuff never really looked at most of it
[23:38] <{Nico}> Kaloz: i thing he won't have much time until july (exams iirc)
[23:39] Action: {Nico} has to eat, food is ready... bbl
[23:39] Action: nbd goes to the fridge to hunt down some food
[23:43] <h3sp4wn> Do people request packages they want building to your ticketting system (I am interested in learning how the new package building system works even if its not something I want to use or need)
[23:43] <nbd> h3sp4wn: some do
[23:43] <nbd> h3sp4wn: if you want to know how the build system works, just do a checkout of the buildroot-ng branch and try to build it
[23:44] <nbd> h3sp4wn: there's no documentation yet, but the format itself is very easy to read and understand
[23:45] <h3sp4wn> nbd: Is there any particular package which you would conside to be none in an ideal way
[23:45] <h3sp4wn> done in an ideal way sorry
[23:47] <nbd> actually most packages are in good shape
[23:49] <h3sp4wn> thx I will build a few simple things until I understand it then maybe try something more complicated
[23:50] <h3sp4wn> Should I just use branches/buildroot-ng or the packages one you have been talking about ?
[23:51] <malbon> night
[23:53] <nbd> branches/buildroot-ng is the new build system
[23:54] <h3sp4wn> I thought the packages were going to be kept in a seperate branch ?
[23:54] <nbd> only the non-essential packages
[23:54] <nbd> the core packages are in branches/buildroot-ng
[23:54] <nbd> not much to see in /packages yet
[23:57] <h3sp4wn> If I put a package into packages/blah on built root ng and run make will it be built automatically of do I have to edit the menu somehow ?
[23:57] <nbd> you have to run make menuconfig
[23:57] <nbd> to select it
[23:58] <nbd> but it will find the package automatically
[23:58] <h3sp4wn> That is truly impressive stuff thanks
[00:00] --- Mon Jun 19 2006
[13:20] <CIA-4> nico * r3978 /branches/buildroot-ng/openwrt/package/linux-atm/ (Makefile patches/300-no_autotools.patch): prevent autoheader invocation, clean autotools patch.
[14:39] <CIA-4> nbd * r3979 /branches/buildroot-ng/openwrt/package/madwifi/ipkg/: remove old ipkg directory
[14:40] <CIA-4> nbd * r3980 /branches/buildroot-ng/openwrt/package/base-files/rb532-2.6/ (. etc/ etc/config/ etc/config/network): add missing base-files for rb532
[15:05] <CIA-4> nico * r3981 /branches/buildroot-ng/openwrt/package/diag/ (Makefile diag_led.c src/ src/Makefile src/diag.c): move source files to src/ and let the kernel tree build the module.
[15:09] <CIA-4> nbd * r3982 /branches/buildroot-ng/openwrt/package/rules.mk: add support for conffiles,preinst,... through templates
[15:09] <nbd> heh... using template in make makes things really weird
[15:09] <nbd> for example:
[15:09] <CIA-4> nico * r3983 /branches/buildroot-ng/openwrt/package/diag/src/diag.c: suppress warnings.
[15:09] <nbd> +COMMANDS += if [ -n "$$$$$$$$pkg_$(subst -,_,$(1))_$(2)" ]; then echo "$$$$$$$$pkg_$(subst -,_,$(1))_$(2)" > $(2); fi;
[15:10] <CIA-4> nbd * r3984 /branches/buildroot-ng/openwrt/package/iptables/Makefile: add conffiles to iptables
[15:11] <nbd> {Nico}: you can enter the contents of conffiles, preinst, etc. without any extra separators now
[15:14] <{Nico}> nbd: great
[15:15] <nbd> what i didn't try is if it can handle " in strings
[15:15] <nbd> but i believe it should work
[15:33] <CIA-4> nbd * r3985 /branches/buildroot-ng/openwrt/package/rules.mk: fix extra control files for packages containing "." and allow different conffiles for multiple packages built from the same directory
[15:38] <CIA-4> nbd * r3986 /branches/buildroot-ng/openwrt/package/base-files/Makefile: add conffiles for base-files
[16:18] <achim71000> mad
[16:33] <CIA-4> kaloz * r3987 /branches/buildroot-ng/openwrt/target/linux/ (15 files in 12 dirs): upgrade to 2.6.17, probably break some stuff temporary
[17:15] <CIA-4> kaloz * r3988 /branches/buildroot-ng/openwrt/target/linux/generic-2.6/patches/201-multiple_default_gateways.patch: fix the multiple gateway patch, too
[19:09] <CIA-4> nico * r3989 /branches/buildroot-ng/openwrt/package/alsa/ (Makefile ipkg/): remove ipkg directory, normalize Makefile.
[19:46] <CIA-4> nico * r3990 /branches/buildroot-ng/openwrt/package/openssl/ (Config.in Makefile ipkg/): remove ipkg directory, normalize Makefile.
[20:03] <CIA-4> nico * r3991 /branches/buildroot-ng/openwrt/package/mtd/ (Makefile mtd.c src/ src/Makefile src/mtd.c): move source file to src/.
[20:04] <CIA-4> nbd * r3992 /branches/buildroot-ng/openwrt/package/wlc/ (18 files in 4 dirs): add wlc utility
[20:05] <CIA-4> nico * r3993 /branches/buildroot-ng/openwrt/package/robocfg/ (7 files in 2 dirs): move source files to src/.
[20:10] <CIA-4> nico * r3994 /branches/buildroot-ng/openwrt/package/hostapd/ (Makefile ipkg/): remove ipkg directory, add conffiles, normalize Makefile.
[20:22] <CIA-4> nico * r3995 /branches/buildroot-ng/openwrt/package/dropbear/ (Makefile files/S50dropbear files/dropbear.init):
[20:22] <CIA-4> remove unneeded autoconf call and flags, rename S50dropbear to dropbear.init,
[20:22] <CIA-4> normalize Makefile:
[20:22] <CIA-4> - indent configure invocation,
[20:22] <CIA-4> - use "install -d" instead of "mkdir -p".
[20:26] <CIA-4> nico * r3996 /branches/buildroot-ng/openwrt/package/linux-atm/ (Makefile files/br2684.hotplug files/br2684.init): rename br2684.init to br2684.hotplug, normalize Makefile.
[20:27] <CIA-4> nico * r3997 /branches/buildroot-ng/openwrt/package/dnsmasq/Makefile: add conffiles, use proper optimization flags when building.
[20:30] <CIA-4> nico * r3998 /branches/buildroot-ng/openwrt/package/ (19 files in 19 dirs): normalize Makefiles.
[21:16] <CIA-4> nbd * r3999 /branches/buildroot-ng/openwrt/package/madwifi/Makefile: override LDOPTS to remove madwifi compile warning
[21:16] <CIA-4> nbd * r4000 /branches/buildroot-ng/openwrt/package/util-linux/ (. Makefile patches/ patches/501-mconfig.patch): add util-linux
[21:16] <nbd> yay... r4000 :)
[22:25] <CIA-4> kaloz * r4001 /branches/buildroot-ng/openwrt/target/linux/netfilter.mk: follow the xtables changes in 2.6.17
[22:35] <[mbm]> so .. rc6
[22:35] <nbd> rc6...
[22:35] <{Nico}> hmm, rc6
[22:35] <Kaloz> huh, another rc? ;)
[22:35] <[mbm]> I still have plans of cleaning up diag and doing some webif changes before rc6
[22:35] <nbd> yeah
[22:35] <nbd> a few things need to be done
[22:36] <[mbm]> how many boards actually have the nvram setting for reset_gpio (or any other gpio mappings?)
[22:36] <nbd> i don't think there's many of them
[22:36] <nbd> and i don't know if we can even trust these settings :)
[22:37] <[mbm]> hrm, was hoping that we could cut down on the number of magic values to identify boards and just use nvram
[22:37] <[mbm]> so we wouldn't have to constantly update for every new board revision
[22:39] <nbd> i don't think we can do that
[22:39] <[mbm]> hrm.. suppose not
[22:40] <nbd> what i would like to get in rc6 is my jffs2 patch
[22:40] <nbd> so that we can get rid of the unnecessary reboot
[22:40] <[mbm]> yep
[22:40] <[mbm]> someone was asking about that in #openwrt last week
[22:40] <[mbm]> I don't think they liked my response
[22:40] <nbd> and wbx seems to have 'fixed' this by adding an autoreboot
[22:40] <nbd> which i think is a bad idea
[22:40] <[mbm]> ditto
[22:41] <[mbm]> that was the suggestion that had been made in #openwrt
[22:41] <nbd> :)
[22:41] <[mbm]> and I said I had other plans
[22:41] <nbd> anyway, the patch works well on the routerboard
[22:41] <nbd> haven't hit any weird sideeffects yet
[22:41] <[mbm]> looked pretty simple
[22:41] <[mbm]> so I can't imagine there would be any odd cases
[22:42] <{Nico}> nbd: do you have anything we can play with?
[22:42] <nbd> {Nico}: the patch still needs to be backported
[22:42] <nbd> {Nico}: but for 2.6 it's already in svn
[22:42] <nbd> although not yet in the generic patch directory
[22:42] <[mbm]> as for the webif stuff, I want to abstract the whole lan/wan interfaces so you can setup as many ("ifup groups"?) as you want
[22:43] <nbd> i think we should do that for the /etc/config based stuff
[22:43] <nbd> otherwise we have to do the work twice
[22:43] <[mbm]> I think I'll end up doing a version for rc6
[22:44] <nbd> ok
[22:44] <CIA-4> nico * r4002 /branches/buildroot-ng/openwrt/ (10 files in 5 dirs): move wlcompat from target/linux/package/ to package/.
[22:45] <nbd> so do we all agree on releasing rc6 as 0.9?
[22:45] <{Nico}> so, that's it, there's no more packages anymore in target/linux/package/ in buildroot-ng
[22:45] <nbd> when no major issues come up?
[22:46] <[mbm]> right
[22:46] <{Nico}> ok
[22:46] <[mbm]> btw does anyone have any ideas about what to do for failsafe mode on unknown boards?
[22:47] <nbd> bridge all interfaces?
[22:47] <[mbm]> need to address the problem of "openwrt works on ___!" and then a week later "how come I can't get into failsafe on ___?"
[22:47] <nbd> better failsafe access on wan as well than no access at all
[22:47] <[mbm]> well, the problem is really that we don't have any reset button to get into failsafe on those boards
[22:48] <nbd> and we don't even know which board is unknown
[22:48] <[mbm]> the only ideas that I can come up with is that if diag doesn't explicitly know were the reset button is that it forces a bootup into failsafe mode
[22:49] <nbd> yeah
[22:49] <[mbm]> or some way of watching all the gpios for a change and assuming that the change means the reset was pressed
[22:49] <nbd> hmm... yeah
[22:52] <[mbm]> btw we're still getting some reports of the kmod-switch and b44 acting strangely on some boards
[22:52] <[mbm]> but I haven't been able to get any information to actually track down the issue
[22:52] <nbd> which boards?
[22:52] <[mbm]> the wrt54g v1.1 was one of them
[22:52] <nbd> define 'acting strangely'
[22:53] <[mbm]> from the sounds of it, absolutely no ethernet
[22:53] <nbd> i will test on my 1.1 before rc6
[22:54] <[mbm]> there was another board which was somewhat odd, I'd have to go to the logs and look
[22:54] <[mbm]> it had a vlan switch but for some reason it was still using eth0/eth1 as lan/wan
[22:55] <malbon> some switches do that. ie ks8995. You can have seperate macs talking to the switch, and phy 5 indipendently
[22:56] <[mbm]> but it was shoing up as /proc/switch which means it was one of the kmod-switch (adm or robo) which don't do that
[22:58] <nbd> [mbm]: sometimes it's also the issue of dd-wrt screwing with the nvram settings
[22:58] <[mbm]> hmm
[22:58] <nbd> [mbm]: because it often completely ignores the previous settings and just sets its own
[22:58] <nbd> it tries to correct the nvram rather than deal with the quirks
[22:58] <nbd> which is more work and more unreliable
[23:01] <[mbm]> ok, so for rc6 we have
[23:01] <[mbm]> - webif updates
[23:01] <[mbm]> - diag updates (how to deal with failsafe on unknown boards?)
[23:01] <[mbm]> - confirm wrt54g v1.1
[23:01] <[mbm]> anything else?
[23:02] <Kaloz> nothing else comes into mind
[23:02] <nbd> i think there was something else, i just don't remember it right now
[23:02] <h3sp4wn> usb2 fix (i.e so it can be used on wl500-gd stable)
[23:02] <nbd> yeah, usb2
[23:02] <nbd> for the gd and the premium
[23:02] <[mbm]> ok
[23:02] <nbd> maybe the premium fix will also fix the gd bugs
[23:03] <nbd> it seems to be a backport of several 2.6 things
[23:03] <[mbm]> and after the rc6 release comes the kamilaze/whiterussian merger for 1.0
[23:03] <nbd> yeah, after 0.9
[23:03] <[mbm]> well, I plan on starting that prett much immediately after rc6 hits the front page
[23:03] <malbon> is wgt634u a target for rc6?
[23:03] <nbd> yeah
[23:03] <nbd> no
[23:04] <malbon> k just checking.
[23:04] <[mbm]> so there will basically be a 0.9 and a 1.0 in parallel
[23:04] <nbd> wgt634u has no usable nvram
[23:04] <nbd> makes sense
[23:04] <nbd> but before that happens, we need to fold back buildroot-ng into trunk
[23:04] <[mbm]> yeah, that happens now while we're working on rc6
[23:05] <nbd> right
[23:05] <{Nico}> [mbm]: how will 1.0 appear? in a new branch?
[23:05] <[mbm]> the usb stuff might need to wait until the 1.0 tree has started so we can update some drivers
[23:05] <nbd> i might have an usb fix soon
[23:06] <nbd> asus has the mentioned backports in their tree
[23:06] <[mbm]> {Nico}: think that branches/whiterussian will just be overwritten with the new 1.0 branch
[23:06] <nbd> (they finally got around to releasing the wl-500gP gpl code)
[23:06] <[mbm]> and 0.9 will work out of tags/0.9
[23:06] <nbd> yeah
[23:06] <{Nico}> [mbm]: ok
[23:07] <[mbm]> how's the packages tree coming along?
[23:07] <nbd> doesn't have many packages yet
[23:07] <nbd> but most of the stuff in the buildroot-ng branch is sorted out
[23:07] <nbd> but there are a few things that really need to get fixed soon
[23:07] <[mbm]> hmm different structure than I expected
[23:07] <nbd> what did you expect?
[23:08] <[mbm]> well, I didn't expect to see the libs/ directory
[23:08] <nbd> i thought we were going to make a hierarchical structure in packages/
[23:08] <[mbm]> figured it would just be something that you could dump directly on packages/ in buildroot-ng
[23:08] <[mbm]> not complaining
[23:09] <nbd> in the future, i would like to write some scripts that automatically resolve dependencies and copy over a certain package to the buildroot from the packages directory
[23:09] <[mbm]> ok
[23:09] <nbd> so you say copy /path/to/packages/pkgname and it pulls all the libraries along as well
[23:10] <[mbm]> no idea how you intend to pull that off, but I'll wait until I see an implementation
[23:10] <nbd> easy: go into that directory, do cd .. until you find the packages root
[23:10] <[mbm]> well, that covers everything I wanted to discuss
[23:10] <nbd> then look for a cached .pkginfo
[23:11] <nbd> [mbm]: how's the verbose stuff for buildroot-ng coming along?
[23:11] <nbd> [mbm]: got any new ideas on how to implement it properly?
[23:11] <nbd> you don't need to do the implementation alone, i just don't know where to start yet
[23:11] <nbd> i want to avoid the previous mess
[23:12] <[mbm]> yeah, I still haven't come up with any good ideas
[23:12] <[mbm]> so I've been working on the web stuff
[23:12] <nbd> k
[23:13] <[mbm]> if make wasn't so subborn I'd just redefine $MAKE to be a macro
[23:13] <nbd> yeah, make is a bitch
[23:13] <nbd> some things are just weird
[23:13] <[mbm]> might end up putting something in the scripts directory and setting $MAKE to that
[23:14] <nbd> it took me quite a while until i came up wiath a way to even pass a multiline string to the shell
[23:14] <nbd> the make developers probably never thought of templates as being used for anything other than make code
[23:14] <[mbm]> yeah, we're really abusing poor make
[23:15] <nbd> still better than writing a bloated python monster :)
[23:15] <[mbm]> well, our scripts directory means that the build system requires the perl monster
[23:15] <florian__> re
[23:15] <nbd> hey
[23:16] <[mbm]> wrt: stfu.
[23:16] <nbd> hehe... i'm faster than wrt :)
[23:16] <florian__> sorry for being late
[23:16] <florian__> I was in the train from Nantes to Paris
[23:16] <nbd> no problem
[23:16] <[mbm]> actually I put a delay into wrt
[23:16] <nbd> ah
[23:16] <[mbm]> (types too fast, must be a bot ..)
[23:16] <nbd> :)
[23:17] <nbd> the delay works sometimes
[23:17] <[mbm]> yeah, it's funny to see people talk to wrt
[23:17] <[mbm]> and even funner is when he seems to respond properly
[23:17] <malbon> wrt is a good conversaionalist. ;)
[23:17] <nbd> [mbm]: wombat in #madwifi is also quite good
[23:17] <nbd> seems to use something like megahal
[23:17] <Kaloz> hehe, yeah
[23:18] <[mbm]> heh, I'll have to look at that channel
[23:19] <nbd> [mbm]: if you talk to it, it'll respond with some random bot-talk
[23:19] <nbd> and it also responds to some trigger words
[23:19] <[mbm]> yeah I've seen megahal's before
[23:19] <[mbm]> they're usually very easy to spot since nothing they say makes any sense
[23:19] <nbd> but when it does, it's really funny
[23:22] <[mbm]> well, unless anyone else has any topics I think the meeting is over
[23:23] <[mbm]> ...
[23:24] <{Nico}> nbd, [mbm]: can you detail the changes you want in webif?
[23:24] <nbd> changing the structure of that confusing firewall page
[23:24] <[mbm]> yeah the firewall page really needs to be changed
[23:24] <nbd> btw. {Nico}, florian__ - could you come to this event? https://snr.freifunk.net/trac/wcc/wiki
[23:24] <nbd> it would be ideal for a developer meeting
[23:25] <[mbm]> some of the wireless settings are missing and other wireless settings are grouped in strange ways
[23:25] <{Nico}> nbd: i'll try, but i'm not sure i can stay the whole days
[23:25] <[mbm]> like the wds is mixed with the mac filtering settings
[23:25] <[mbm]> and it's somewhat confusing
[23:26] <{Nico}> [mbm]: you want a separate page for wds?
[23:26] <[mbm]> no, I mean the page looks like this
[23:26] <[mbm]> [wds setting]
[23:26] <[mbm]> [mac setting]
[23:26] <[mbm]> [wds setting]
[23:26] <[mbm]> [mac setting]
[23:27] <[mbm]> but it should be like this:
[23:27] <[mbm]> [wds settings]
[23:27] <[mbm]> ...
[23:27] <[mbm]> [mac settings]
[23:27] <[mbm]> ...
[23:28] <[mbm]> in other words there should almost be boxes of content on the page
[23:28] <[mbm]> one box for all the wds settings
[23:28] <[mbm]> and one box for all the mac settings
[23:28] <[mbm]> so that they're clearly grouped
[23:28] <nbd> right
[23:30] <{Nico}> ok, i'll try to fix it but i have to dive into webif first
[23:31] <[mbm]> heh good luck, webif is almost it's own undocumented programming language
[23:31] <{Nico}> what are the missing wireless settings? (might be easier to start with webif)
[23:31] <nbd> heh, yeah
[23:31] <[mbm]> hmm I don't have a list of them but it shouldn't be hard to look at the factory page and see
[23:32] <[mbm]> I know that the long/short preamble is missing
[23:32] <[mbm]> as are some of the rate controls
[23:32] <[mbm]> isolation should be a feature
[23:32] <[mbm]> ...
[23:33] <[mbm]> basically our advanced wireless page should be much much longer than it is
[23:33] <{Nico}> can ayone post a screenshot of the linksys page (just for inspiration)?
[23:34] <alxhh> hmmm, powerlevel
[23:34] Action: [mbm] has to go .. bb(much)l
[23:35] <h3sp4wn> Are the packages which were previously in florians repository going to be integrated at any point ?
[23:35] <nbd> powerlevel.. that'll encourage all the OMGWTF-251-mW-ZOMG-LOL1!!1 people
[23:35] <{Nico}> h3sp4wn: what are you missing?
[23:36] Action: RoDent loves the new wlc
[23:36] <h3sp4wn> ncftp specifically - I could built it myself but have not done so because of the fact he has already done it
[23:36] <{Nico}> h3sp4wn: anything else?
[23:37] <Kaloz> h3sp4wn: bug florian, he promised to integrate everything :)
[23:37] <h3sp4wn> It had a massive amount of stuff never really looked at most of it
[23:38] <{Nico}> Kaloz: i thing he won't have much time until july (exams iirc)
[23:39] Action: {Nico} has to eat, food is ready... bbl
[23:39] Action: nbd goes to the fridge to hunt down some food
[23:43] <h3sp4wn> Do people request packages they want building to your ticketting system (I am interested in learning how the new package building system works even if its not something I want to use or need)
[23:43] <nbd> h3sp4wn: some do
[23:43] <nbd> h3sp4wn: if you want to know how the build system works, just do a checkout of the buildroot-ng branch and try to build it
[23:44] <nbd> h3sp4wn: there's no documentation yet, but the format itself is very easy to read and understand
[23:45] <h3sp4wn> nbd: Is there any particular package which you would conside to be none in an ideal way
[23:45] <h3sp4wn> done in an ideal way sorry
[23:47] <nbd> actually most packages are in good shape
[23:49] <h3sp4wn> thx I will build a few simple things until I understand it then maybe try something more complicated
[23:50] <h3sp4wn> Should I just use branches/buildroot-ng or the packages one you have been talking about ?
[23:51] <malbon> night
[23:53] <nbd> branches/buildroot-ng is the new build system
[23:54] <h3sp4wn> I thought the packages were going to be kept in a seperate branch ?
[23:54] <nbd> only the non-essential packages
[23:54] <nbd> the core packages are in branches/buildroot-ng
[23:54] <nbd> not much to see in /packages yet
[23:57] <h3sp4wn> If I put a package into packages/blah on built root ng and run make will it be built automatically of do I have to edit the menu somehow ?
[23:57] <nbd> you have to run make menuconfig
[23:57] <nbd> to select it
[23:58] <nbd> but it will find the package automatically
[23:58] <h3sp4wn> That is truly impressive stuff thanks
[00:00] --- Mon Jun 19 2006