[00:00] <{Nico}> nbd: i have busybox-1.Ã2.1 pending
[00:00] <nbd> you need a busybox .config file and an unpacked busbox tree
[00:01] <{Nico}> nbd: and what it builds a config tree, right?
[00:02] <nbd> yes
[00:02] <nbd> you need to be in an empty directory
[00:03] <nbd> and do gen_busybox_menuconfig /path/to/busybox/tree /path/to/.config
[00:03] <nbd> make sure you diff the new tree against the old one to see if it screwed up somewhere
[00:03] <nbd> i don't know if it'll handle the new version correctly
[00:11] <common> nbd wrv54g uses some different image packing, did Kaloz already take care of that?
[00:17] <nbd> i think the wrv54g is not supported yet
[01:02] <h3sp4wn> make[3]: *** [install] Segmentation fault (is that usual when building the xscale port ?)
[01:11] <[mbm]> huh? make segfaulted?
[01:16] <h3sp4wn> It has happened multiple times - only when building the xscale-2.6 port
[01:17] <h3sp4wn> when it gets to the generation of the 128k jffs2 image (the 64k one seems to be generated fine)
[01:18] <h3sp4wn> I suppose I could try it from a 32bit chroot (there is nothing particularly strange about the box other than it is running 64 bit ubuntu)
[02:03] <common> nbd: got any idea how to add zLoader http://openixp.devel.phj.hu/cgi-bin/trac.cgi/browser/target/linux/image/wrv54g/kernel-loader/Makefile to openwrt?
[02:03] <common> where to place it
[02:05] <groz_> h3, i've found a lot of things, especially filesystem related, abend on 64bit
[02:05] <groz_> utilities that _assume_ int=32bit and parse out file system structures
[02:05] <groz_> in particular
[02:06] <common> [mbm]: can you help me with the zLoader?
[02:06] <common> once i have it i can try booting on wrv54g
[02:43] <{Nico}> nbd: why did you put libjpeg static?
[02:45] <{Nico}> and why did you rename from jpeg to libjpeg?
[02:45] <{Nico}> i thought we choose to use the source name as the package name
[02:45] <nbd> because it vehemently refused to build the shared library on my system and because it's used by only a few things
[02:46] <nbd> so it might actually save some space
[02:46] <nbd> and i renamed it to make it more consistent with the other libraries
[02:46] <nbd> and i think i found a few other
[02:46] <nbd> libraries, where the name is also lib<name>
[02:49] <{Nico}> then the source name is lib<name>, but libpcre is in pcre, libopenssl is in openssl, etc...
[02:49] <{Nico}> i think we should stay as much consistent as possible
[02:49] <nbd> ok
[02:50] <{Nico}> what was your problem with libjeg, do you have a build.log?
[02:50] <nbd> it built the static library, but not the .so
[02:50] <nbd> so the package install failed
[02:52] <{Nico}> some packages don't link against the static libjpeg
[02:52] <common> nbd, the kernel hangs with "not syncing: VFS: Unable to mount root fs on unknown block(0,0)
[02:52] <common> but it boots upto that point
[02:52] <groz> you are either missing the file system, or, the driver to get at it
[02:53] <common> the redboot partition is not found
[02:53] <common> i guess as the wrv54g uses openrg
[02:53] <nbd> i have a board with openrg as well
[02:53] <nbd> i will work on it soon
[02:54] <groz> are you using 2.6 ?
[02:54] <common> yes
[02:54] <groz> if so, do you have initramfs ?
[02:54] <common> no idea
[02:54] <groz> that's the easy way to 'get up and going' without worrying about mapping file systems
[02:55] <common> Kaloz: i owe you a beer
[02:56] <common> nbd
[02:56] <common> from what i can say it did not create any partitions too
[02:57] <nbd> yeah, probably needs a different flash map driver
[02:57] <common> http://openixp.devel.phj.hu/cgi-bin/trac.cgi/browser/target/linux/linux-2.6/wrv54g/patches/wrv54g/040-gtwx5715-mtd-parts.patch
[02:57] <common> this one?
[02:58] <nbd> don't like it the way it is
[02:59] <nbd> looks like it won't produce something that works with both redboot and openrg
[02:59] <common> yep
[02:59] <nbd> i'll change a few things
[02:59] <groz> famous last words
[02:59] <nbd> but not today
[02:59] <groz> then it never works again :)
[02:59] <common> np
[02:59] <nbd> nah, it's different for me
[02:59] <common> you may want to add
[02:59] <nbd> i change a few things, i break them
[02:59] <nbd> and when i'm done a while later, it works :)
[02:59] <common> wrv-bin to target/utils
[03:00] <groz> well i've spent most of the day doing cleanup on a client project
[03:00] <groz> shortly, i've got to go dig into brcm driver stuff
[03:00] <groz> have a client wants an eval of the brcm open source stuff
[03:00] <groz> ie a quick 'what works, what doens't work, what is reasonable for me to take on as fix list in it'
[03:01] <groz> they would like to see broadcom mini pci up and running on thier i386 platform
[03:01] <groz> as an access point
[03:04] <common> may be possible, but not worth the time
[03:04] <groz> why would you say that ?
[03:04] <common> the brcm driver ap mode is not defined as working/stable yet
[03:05] <groz> and that's the whole point of it
[03:05] <groz> they are considering tasking me with changing that
[03:05] <groz> so they want an eval of how much work it'll be
[03:05] <common> a lot
[03:05] <groz> if it's somethign that could be done in a month, they will probably say go
[03:05] <common> i'm on the brcm list, as my notebook features this fuckedup chipset
[03:06] <common> and there was a patch some days ago
[03:06] <groz> well, i have the chipset in over a thousand ap's hosted on a wrt54g, and, it works pretty good
[03:06] <groz> so it's just driver code away from working just as well on i386
[03:07] <common> you got the src code of the driver?
[03:07] <groz> not the wrt one, but my point is
[03:07] <groz> the chipset and it's on board firmware, work
[03:07] <groz> it's just the underlying linux driver layer that needs the work
[03:08] <groz> and i have a client prepared to pay to get that work done
[03:08] <common> are we both talking about the brcm 43xx driver?
[03:08] <groz> yup
[03:08] <common> and you know about linvilles wireless tree?
[03:08] <groz> same thing that is in thousands of notebooks out there
[03:09] <groz> i'm aware of numerous different directions in different trees
[03:09] <groz> trick is, pick the one with the most promise
[03:09] <groz> and work from there
[03:09] <common> linvilles dscape tree
[03:10] <groz> but for now, the task at hand, get it actually running on a notebook
[03:10] <groz> then see what does, and doesn't, work
[03:10] <CIA-17> nico * r4513 /packages/utils/deco/Makefile: change dependency on libncurses (from strict depend to select)
[03:11] <common> for me it fails with dhclient as dhclient takes the interface down, and on upping the interface again the wifi settings are lost
[03:11] <common> did not try ap mode
[03:11] <groz> i haven't even got as far as trying that yet, i gotta sit down and figure out the issues of actually getting it to work first
[03:12] <common> there are no issues, install the driver, install the firmware, works
[03:12] <groz> ya, issue number one
[03:12] <groz> find the right firmware to install
[03:13] <groz> fwcutter supposed to work with wl.o
[03:13] <groz> but it wont touch any of the wl.o variants i have around here
[03:13] <common> firmware depends on the chip revision
[03:13] <common> from what i can say
[03:13] <groz> just bitches about not able to find the stuff inside it
[03:13] <common> they ship a list with firmware urls and versions
[03:13] <CIA-17> nico * r4514 /packages/ (4 files in 4 dirs): rename glib package from liblib1 to glib1 and fix dependencies on it
[03:14] <groz> yes, and, i fetched them all, and, the fwcutter dowloanded from sources
[03:14] <groz> wont touch any of them
[03:14] <groz> so i ggotta figure out why
[03:14] <groz> will get to it
[03:14] <groz> before the day is done
[03:16] <CIA-17> nico * r4515 /packages/ (3 files in 3 dirs): change dependencies (from strict depend to select)
[03:22] <CIA-17> nico * r4516 /packages/admin/osiris/Makefile: fix download url, reorg variables, indent
[03:26] <{Nico}> nbd: it seems menuconfig is bugging when i symlink all packages into buildroot-ng
[03:26] <{Nico}> the Network category won't open
[03:26] <{Nico}> until i remove some packaes
[03:26] <{Nico}> s/packaes/links/
[03:44] <groz> common, i found a link to one on a ubuntu help page
[03:44] <groz> it cut just fine
[03:45] <groz> machine is rebooting, i may actually have it up shortly
[03:45] <groz> if this 'just works' for client mode
[03:45] <groz> i'm gonna be a happy campter
[03:45] <common> if you use dhclient
[03:45] <groz> and then start looking at ap mode
[03:45] <common> do this:
[03:46] <common> term #1: dhclient device
[03:46] <groz> hehe, i will use iwconfig and ifconfig till i see things running
[03:46] <common> term #2: iwconfig device essid fox mode managed
[03:46] <groz> or add it as a post-up in the networks file
[03:48] <groz> hawt diggety, it's working
[03:48] <groz> and dhclient didn't seem to need that kludge
[03:48] <groz> altho, i alreadhy had it up
[03:49] <groz> when i ran dhclient
[03:49] <groz> i'm using the latest dev release of ubuntu here
[03:49] <groz> just installed it on this laptop couple days ago
[03:55] <common> debian unstable
[03:56] <common> now, you may want to try ap mode
[03:56] <common> seems like the diffs got merged
[03:57] <common> [PATCH] AP (master) mode fixed (resubmit) 18.6.2006 on the brcm lists as reference
[03:57] <groz> i'm just about to try ap
[03:57] <groz> got distracted with a phone call
[03:57] <groz> but the dhcp worked out ok
[09:55] <nbd> gah, stupid menuconfig only has a static 4k buffer for building menus
[09:55] <nbd> no wonder it breaks with too many packages
[09:57] <groz> isn't it amazing how many things you gotta fix along the way ?
[09:58] <CIA-17> nbd * r4517 /branches/buildroot-ng/openwrt/scripts/config/mconf.c: increase the size of a few static buffers in menuconfig - fixes display of menus with lots of packages
[09:59] <nbd> yeah :)
[10:00] <groz> holy cow
[10:00] <groz> all of a sudden i got a lot more menus in menuconfig
[10:00] <groz> its magic
[10:00] <nbd> hehe
[10:01] <groz> am i correct in assuming you are not doing much with i386 these days nbd ?
[10:01] <groz> i've been slowly getting a bunch of kmod stuff for i386 sorted away when i got a chance
[10:01] <nbd> yeah, although in the next few days, i got a reason for checking out the i386 port of openwrt
[10:01] <groz> but i haven't committed anything back yet
[10:01] <groz> so let me know before you dive into ti
[10:02] <groz> it'll commit a bunch
[10:02] <nbd> i want to use lockdep on madwifi and that only works on i386
[10:02] <nbd> and my laptop is ppc :)
[10:02] <groz> hah, you know, i did some math today
[10:02] <groz> i want a new puter and a new monitor for my desktop
[10:02] <groz> i want one of those new 1600x1050 flat panels
[10:03] <groz> so i started comparing the price, of piece together a box with a dual core, and that monitor
[10:03] <groz> to just buying an imac
[10:03] <groz> and there's less than $100 difference in the end
[10:03] <groz> then i start looking at what that'll do to clean up the cables mess here
[10:04] <nbd> :)
[10:04] <groz> but those aren't ppc anymore
[10:04] <groz> but i think it may be fun to play with osx too
[10:05] <nbd> yeah, i like it
[10:05] <groz> kinda my 'intro to the bsd world' in a way i guess
[10:11] <Kaloz> intro to the bsd world for dummies
[10:11] <Kaloz> :P
[10:12] Action: groz pleads guilty, am a dummy
[10:12] <Kaloz> whatever, i got sick of bsd (and yes, i've tried most of them(
[10:12] <groz> i haven't played with them at all
[10:12] <groz> but the reality of the world is
[10:12] <groz> the vast majority of development these days, is going into linux et al
[10:13] <Kaloz> true. and the future is about embedded systems
[10:13] <Kaloz> and bsd has no chance there
[10:13] <groz> oh, the present is about embedded
[10:13] <groz> the future is about 'deeply embedded'
[10:13] <groz> the world is going back to a model where the software is an afterthought
[10:13] <groz> and the hardware holds the value
[10:14] <groz> the wrt54 is a prime example
[10:16] <Kaloz> nah, the world is going back to the old and nice client-server infrastucture
[10:16] <Kaloz> gmail, gdriver, gwhatever
[10:16] <groz> that's consumer stuff, for end lusers at home
[10:16] <Kaloz> people will be able to rech their stuff through terminals through vpn
[10:16] <groz> it's not being adopted in the workplace
[10:17] <Kaloz> well, the client-server model will work everywhere
[10:17] <Kaloz> desktop stuff currently are not feasable at all
[10:17] <groz> the workplace is moving to 'purpose built devices' for everything
[10:17] <groz> and migrating away from general purpose 'desktop' computers
[10:18] <Kaloz> sure. but interestingly handhelds are dieing... notebooks taking away market share from both them and desktop
[10:18] <x-db90h> it's hard to say what the future holds, but that's one possibility
[10:18] <groz> kaloz, you ever worked with a prism54 card ?
[10:18] <groz> is it one that 'just works' in ap mode ?
[10:18] <Kaloz> yeah, but it was long ago
[10:19] <groz> i need to find a mini pci that 'just works' in ap mode
[10:19] <Kaloz> if you need something that "just works" go with prism :P if you need "just works most of the time" or 802.11a/g, go with atheros
[10:20] <groz> i need g mode
[10:20] <Kaloz> go with atheros then
[10:20] <groz> and i've had considerable stability issues with madwifi over time
[10:20] <groz> gunshy of even trying that for thsi project
[10:20] <groz> and, since it's not i386 based
[10:20] <groz> i dunno
[10:50] <CIA-17> florian * r4518 /packages/lang/sablevm-classpath/ (. Config.in Makefile ipkg/): Port sablevm-classpath to -ng
[10:51] <CIA-17> florian * r4519 /packages/lang/sablevm-classpath/Makefile: Remove references to old IDIR_*
[11:05] <CIA-17> florian * r4520 /packages/lang/sablevm/ (. Config.in Makefile ipkg/): Port sablevm to -ng
[11:22] <CIA-17> florian * r4521 /packages/net/xsupplicant/ (. Config.in Makefile ipkg/): Port xsupplicant to -ng
[11:30] <CIA-17> florian * r4522 /packages/utils/sane-backends/ (. Config.in Makefile ipkg/): Port sane-backends to -ng
[11:33] <CIA-17> florian * r4523 /packages/net/scanlogd/ (. Config.in Makefile ipkg/): Port scanlogd to -ng
[11:34] <CIA-17> florian * r4524 /packages/net/scdp/ (. Config.in Makefile ipkg/): Port scdp to -ng
[11:38] <CIA-17> florian * r4525 /packages/utils/screen/ (. Makefile): Port screen to -ng
[11:42] <CIA-17> florian * r4526 /packages/net/ser2net/ (. Config.in Makefile ipkg/): Port ser2net to -ng
[11:45] <malbon> groz: avoid prism54 unless you can get the fullmac version.
[11:46] <malbon> groz: the softmac drivers just don't work for me.
[12:24] <CIA-17> nico * r4527 /branches/buildroot-ng/openwrt/Config.in: enable ext2 rootfs by default when target is uml-2.6
[12:27] <CIA-17> nico * r4528 /branches/buildroot-ng/openwrt/package/busybox/ (28 files in 19 dirs): update busybox to v1.2.1
[12:28] <groz> nico ping
[12:32] <CIA-17> nico * r4529 /packages/net/lighttpd/ (9 files in 4 dirs): port lighttpd to buildroot-ng
[12:34] <{Nico}> groz: pong
[12:35] <groz> hey, i see you fussing with uml stuff, have you actually been using it at all ?
[12:37] <{Nico}> groz: yeah, but i'm using an old rev - recent changes broke it
[12:37] <{Nico}> still not played with networking
[12:37] <groz> i'll be back into it later in the week, i'll fix it again
[12:39] <{Nico}> it should be related to the recent TARGET_{CC,PATH} changes for ccache
[12:40] <groz> probably
[12:51] <{Nico}> groz: does it make sense to enable usb support in uml-2.6 ?
[12:54] <groz> good question on a lot of things
[12:54] <groz> one of the things that uml is good for, is developing a driver
[12:55] <groz> you can pull it down when things go wrong, and start over
[12:55] <groz> so if the host has no drivers already on the specific hardware
[12:55] <groz> a uml can talk to it directly
[14:25] <CIA-17> florian * r4530 /packages/utils/setpwc/ (. Config.in Config.in Makefile Makefile ipkg/ ipkg/): Port setpwc to -ng
[14:28] <CIA-17> florian * r4531 /packages/net/shat/ (. Config.in Makefile ipkg/): Port shat to -ng
[14:33] <CIA-17> florian * r4532 /packages/net/sipp/ (. Config.in Makefile ipkg/): Port sipp to -ng
[14:41] <CIA-17> florian * r4533 /packages/net/siproxd/ (. Config.in Makefile ipkg/): Port siproxd to -ng
[14:42] <CIA-17> florian * r4534 /packages/net/sipsak/ (. Config.in Makefile ipkg/): Port sipsak to -ng
[16:42] <CIA-17> florian * r4535 /packages/net/slurm/ (. Config.in Makefile ipkg/): Port slurm to -ng
[16:42] <CIA-17> florian * r4536 /packages/net/socat/ (. Config.in Makefile ipkg/): Port socat to -ng
[16:42] <CIA-17> florian * r4537 /packages/net/srelay/ (. Config.in Makefile ipkg/): Port srelay to -ng
[16:43] <CIA-17> florian * r4538 /packages/net/ssltunnel/ (. Config.in Makefile ipkg/): Port ssltunnel to -ng
[16:43] <CIA-17> florian * r4539 /packages/net/ssmtp/ (. Config.in Makefile ipkg/): Port ssmtp to -ng
[16:46] <CIA-17> nico * r4540 /branches/buildroot-ng/openwrt/package/zd1211/Makefile: disable on uml-2.6 (lacking wireless), standardize
[16:56] <CIA-17> florian * r4541 /packages/admin/syslog-ng/ (. Config.in Makefile ipkg/): Port syslog-ng to -ng
[16:58] <CIA-17> florian * r4542 /packages/net/tinc/ (. Config.in Makefile ipkg/): Port tinc to -ng
[17:05] <CIA-17> nico * r4543 /packages/net/rarpd/Makefile: fix rarpd: revert to plain old ./configure call, Build/Configure/Default does not support overloading CFLAGS (yet)
[17:37] <CIA-17> nico * r4544 /packages/lang/sablevm/Makefile: fix sablevm, standardize, indent
[18:42] <common> groz: i got a prism54 minipci card
[22:15] <CIA-17> mbm * r4545 /branches/buildroot-ng/openwrt/package/dnsmasq/files/dnsmasq.conf: fix error if dhcp is disabled
[23:54] <{Nico}> .
[23:55] <malbon> .
[23:56] <[mbm]> so, a quick recap of the agenda .. rc6 needs to be released asap .. some nagging issues with the webif
[23:56] <[mbm]> so the buggy firewall page will get dropped and we'll add some more nvram toggles to the pages
[23:56] <[mbm]> I really want to see every nvram variable used by rc6 in webif
[23:57] <{Nico}> [mbm]: can we all aggree on a release date?
[23:57] <[mbm]> unfortunately all of the devels are pretty busy so it's hard to find time to fix the old rc6 bugs
[23:58] <[mbm]> we have to fix the bugs and we have to do a resonable breakin/quality assurance
[23:58] <[mbm]> so the absolute minimum to a release would be 2 weeks from now
[23:59] <[mbm]> oh .. something else I should add to that
[23:59] <[mbm]> we should add an stunnel package and move webif to https
[00:00] --- Wed Aug 9 2006
[00:00] <nbd> you need a busybox .config file and an unpacked busbox tree
[00:01] <{Nico}> nbd: and what it builds a config tree, right?
[00:02] <nbd> yes
[00:02] <nbd> you need to be in an empty directory
[00:03] <nbd> and do gen_busybox_menuconfig /path/to/busybox/tree /path/to/.config
[00:03] <nbd> make sure you diff the new tree against the old one to see if it screwed up somewhere
[00:03] <nbd> i don't know if it'll handle the new version correctly
[00:11] <common> nbd wrv54g uses some different image packing, did Kaloz already take care of that?
[00:17] <nbd> i think the wrv54g is not supported yet
[01:02] <h3sp4wn> make[3]: *** [install] Segmentation fault (is that usual when building the xscale port ?)
[01:11] <[mbm]> huh? make segfaulted?
[01:16] <h3sp4wn> It has happened multiple times - only when building the xscale-2.6 port
[01:17] <h3sp4wn> when it gets to the generation of the 128k jffs2 image (the 64k one seems to be generated fine)
[01:18] <h3sp4wn> I suppose I could try it from a 32bit chroot (there is nothing particularly strange about the box other than it is running 64 bit ubuntu)
[02:03] <common> nbd: got any idea how to add zLoader http://openixp.devel.phj.hu/cgi-bin/trac.cgi/browser/target/linux/image/wrv54g/kernel-loader/Makefile to openwrt?
[02:03] <common> where to place it
[02:05] <groz_> h3, i've found a lot of things, especially filesystem related, abend on 64bit
[02:05] <groz_> utilities that _assume_ int=32bit and parse out file system structures
[02:05] <groz_> in particular
[02:06] <common> [mbm]: can you help me with the zLoader?
[02:06] <common> once i have it i can try booting on wrv54g
[02:43] <{Nico}> nbd: why did you put libjpeg static?
[02:45] <{Nico}> and why did you rename from jpeg to libjpeg?
[02:45] <{Nico}> i thought we choose to use the source name as the package name
[02:45] <nbd> because it vehemently refused to build the shared library on my system and because it's used by only a few things
[02:46] <nbd> so it might actually save some space
[02:46] <nbd> and i renamed it to make it more consistent with the other libraries
[02:46] <nbd> and i think i found a few other
[02:46] <nbd> libraries, where the name is also lib<name>
[02:49] <{Nico}> then the source name is lib<name>, but libpcre is in pcre, libopenssl is in openssl, etc...
[02:49] <{Nico}> i think we should stay as much consistent as possible
[02:49] <nbd> ok
[02:50] <{Nico}> what was your problem with libjeg, do you have a build.log?
[02:50] <nbd> it built the static library, but not the .so
[02:50] <nbd> so the package install failed
[02:52] <{Nico}> some packages don't link against the static libjpeg
[02:52] <common> nbd, the kernel hangs with "not syncing: VFS: Unable to mount root fs on unknown block(0,0)
[02:52] <common> but it boots upto that point
[02:52] <groz> you are either missing the file system, or, the driver to get at it
[02:53] <common> the redboot partition is not found
[02:53] <common> i guess as the wrv54g uses openrg
[02:53] <nbd> i have a board with openrg as well
[02:53] <nbd> i will work on it soon
[02:54] <groz> are you using 2.6 ?
[02:54] <common> yes
[02:54] <groz> if so, do you have initramfs ?
[02:54] <common> no idea
[02:54] <groz> that's the easy way to 'get up and going' without worrying about mapping file systems
[02:55] <common> Kaloz: i owe you a beer
[02:56] <common> nbd
[02:56] <common> from what i can say it did not create any partitions too
[02:57] <nbd> yeah, probably needs a different flash map driver
[02:57] <common> http://openixp.devel.phj.hu/cgi-bin/trac.cgi/browser/target/linux/linux-2.6/wrv54g/patches/wrv54g/040-gtwx5715-mtd-parts.patch
[02:57] <common> this one?
[02:58] <nbd> don't like it the way it is
[02:59] <nbd> looks like it won't produce something that works with both redboot and openrg
[02:59] <common> yep
[02:59] <nbd> i'll change a few things
[02:59] <groz> famous last words
[02:59] <nbd> but not today
[02:59] <groz> then it never works again :)
[02:59] <common> np
[02:59] <nbd> nah, it's different for me
[02:59] <common> you may want to add
[02:59] <nbd> i change a few things, i break them
[02:59] <nbd> and when i'm done a while later, it works :)
[02:59] <common> wrv-bin to target/utils
[03:00] <groz> well i've spent most of the day doing cleanup on a client project
[03:00] <groz> shortly, i've got to go dig into brcm driver stuff
[03:00] <groz> have a client wants an eval of the brcm open source stuff
[03:00] <groz> ie a quick 'what works, what doens't work, what is reasonable for me to take on as fix list in it'
[03:01] <groz> they would like to see broadcom mini pci up and running on thier i386 platform
[03:01] <groz> as an access point
[03:04] <common> may be possible, but not worth the time
[03:04] <groz> why would you say that ?
[03:04] <common> the brcm driver ap mode is not defined as working/stable yet
[03:05] <groz> and that's the whole point of it
[03:05] <groz> they are considering tasking me with changing that
[03:05] <groz> so they want an eval of how much work it'll be
[03:05] <common> a lot
[03:05] <groz> if it's somethign that could be done in a month, they will probably say go
[03:05] <common> i'm on the brcm list, as my notebook features this fuckedup chipset
[03:06] <common> and there was a patch some days ago
[03:06] <groz> well, i have the chipset in over a thousand ap's hosted on a wrt54g, and, it works pretty good
[03:06] <groz> so it's just driver code away from working just as well on i386
[03:07] <common> you got the src code of the driver?
[03:07] <groz> not the wrt one, but my point is
[03:07] <groz> the chipset and it's on board firmware, work
[03:07] <groz> it's just the underlying linux driver layer that needs the work
[03:08] <groz> and i have a client prepared to pay to get that work done
[03:08] <common> are we both talking about the brcm 43xx driver?
[03:08] <groz> yup
[03:08] <common> and you know about linvilles wireless tree?
[03:08] <groz> same thing that is in thousands of notebooks out there
[03:09] <groz> i'm aware of numerous different directions in different trees
[03:09] <groz> trick is, pick the one with the most promise
[03:09] <groz> and work from there
[03:09] <common> linvilles dscape tree
[03:10] <groz> but for now, the task at hand, get it actually running on a notebook
[03:10] <groz> then see what does, and doesn't, work
[03:10] <CIA-17> nico * r4513 /packages/utils/deco/Makefile: change dependency on libncurses (from strict depend to select)
[03:11] <common> for me it fails with dhclient as dhclient takes the interface down, and on upping the interface again the wifi settings are lost
[03:11] <common> did not try ap mode
[03:11] <groz> i haven't even got as far as trying that yet, i gotta sit down and figure out the issues of actually getting it to work first
[03:12] <common> there are no issues, install the driver, install the firmware, works
[03:12] <groz> ya, issue number one
[03:12] <groz> find the right firmware to install
[03:13] <groz> fwcutter supposed to work with wl.o
[03:13] <groz> but it wont touch any of the wl.o variants i have around here
[03:13] <common> firmware depends on the chip revision
[03:13] <common> from what i can say
[03:13] <groz> just bitches about not able to find the stuff inside it
[03:13] <common> they ship a list with firmware urls and versions
[03:13] <CIA-17> nico * r4514 /packages/ (4 files in 4 dirs): rename glib package from liblib1 to glib1 and fix dependencies on it
[03:14] <groz> yes, and, i fetched them all, and, the fwcutter dowloanded from sources
[03:14] <groz> wont touch any of them
[03:14] <groz> so i ggotta figure out why
[03:14] <groz> will get to it
[03:14] <groz> before the day is done
[03:16] <CIA-17> nico * r4515 /packages/ (3 files in 3 dirs): change dependencies (from strict depend to select)
[03:22] <CIA-17> nico * r4516 /packages/admin/osiris/Makefile: fix download url, reorg variables, indent
[03:26] <{Nico}> nbd: it seems menuconfig is bugging when i symlink all packages into buildroot-ng
[03:26] <{Nico}> the Network category won't open
[03:26] <{Nico}> until i remove some packaes
[03:26] <{Nico}> s/packaes/links/
[03:44] <groz> common, i found a link to one on a ubuntu help page
[03:44] <groz> it cut just fine
[03:45] <groz> machine is rebooting, i may actually have it up shortly
[03:45] <groz> if this 'just works' for client mode
[03:45] <groz> i'm gonna be a happy campter
[03:45] <common> if you use dhclient
[03:45] <groz> and then start looking at ap mode
[03:45] <common> do this:
[03:46] <common> term #1: dhclient device
[03:46] <groz> hehe, i will use iwconfig and ifconfig till i see things running
[03:46] <common> term #2: iwconfig device essid fox mode managed
[03:46] <groz> or add it as a post-up in the networks file
[03:48] <groz> hawt diggety, it's working
[03:48] <groz> and dhclient didn't seem to need that kludge
[03:48] <groz> altho, i alreadhy had it up
[03:49] <groz> when i ran dhclient
[03:49] <groz> i'm using the latest dev release of ubuntu here
[03:49] <groz> just installed it on this laptop couple days ago
[03:55] <common> debian unstable
[03:56] <common> now, you may want to try ap mode
[03:56] <common> seems like the diffs got merged
[03:57] <common> [PATCH] AP (master) mode fixed (resubmit) 18.6.2006 on the brcm lists as reference
[03:57] <groz> i'm just about to try ap
[03:57] <groz> got distracted with a phone call
[03:57] <groz> but the dhcp worked out ok
[09:55] <nbd> gah, stupid menuconfig only has a static 4k buffer for building menus
[09:55] <nbd> no wonder it breaks with too many packages
[09:57] <groz> isn't it amazing how many things you gotta fix along the way ?
[09:58] <CIA-17> nbd * r4517 /branches/buildroot-ng/openwrt/scripts/config/mconf.c: increase the size of a few static buffers in menuconfig - fixes display of menus with lots of packages
[09:59] <nbd> yeah :)
[10:00] <groz> holy cow
[10:00] <groz> all of a sudden i got a lot more menus in menuconfig
[10:00] <groz> its magic
[10:00] <nbd> hehe
[10:01] <groz> am i correct in assuming you are not doing much with i386 these days nbd ?
[10:01] <groz> i've been slowly getting a bunch of kmod stuff for i386 sorted away when i got a chance
[10:01] <nbd> yeah, although in the next few days, i got a reason for checking out the i386 port of openwrt
[10:01] <groz> but i haven't committed anything back yet
[10:01] <groz> so let me know before you dive into ti
[10:02] <groz> it'll commit a bunch
[10:02] <nbd> i want to use lockdep on madwifi and that only works on i386
[10:02] <nbd> and my laptop is ppc :)
[10:02] <groz> hah, you know, i did some math today
[10:02] <groz> i want a new puter and a new monitor for my desktop
[10:02] <groz> i want one of those new 1600x1050 flat panels
[10:03] <groz> so i started comparing the price, of piece together a box with a dual core, and that monitor
[10:03] <groz> to just buying an imac
[10:03] <groz> and there's less than $100 difference in the end
[10:03] <groz> then i start looking at what that'll do to clean up the cables mess here
[10:04] <nbd> :)
[10:04] <groz> but those aren't ppc anymore
[10:04] <groz> but i think it may be fun to play with osx too
[10:05] <nbd> yeah, i like it
[10:05] <groz> kinda my 'intro to the bsd world' in a way i guess
[10:11] <Kaloz> intro to the bsd world for dummies
[10:11] <Kaloz> :P
[10:12] Action: groz pleads guilty, am a dummy
[10:12] <Kaloz> whatever, i got sick of bsd (and yes, i've tried most of them(
[10:12] <groz> i haven't played with them at all
[10:12] <groz> but the reality of the world is
[10:12] <groz> the vast majority of development these days, is going into linux et al
[10:13] <Kaloz> true. and the future is about embedded systems
[10:13] <Kaloz> and bsd has no chance there
[10:13] <groz> oh, the present is about embedded
[10:13] <groz> the future is about 'deeply embedded'
[10:13] <groz> the world is going back to a model where the software is an afterthought
[10:13] <groz> and the hardware holds the value
[10:14] <groz> the wrt54 is a prime example
[10:16] <Kaloz> nah, the world is going back to the old and nice client-server infrastucture
[10:16] <Kaloz> gmail, gdriver, gwhatever
[10:16] <groz> that's consumer stuff, for end lusers at home
[10:16] <Kaloz> people will be able to rech their stuff through terminals through vpn
[10:16] <groz> it's not being adopted in the workplace
[10:17] <Kaloz> well, the client-server model will work everywhere
[10:17] <Kaloz> desktop stuff currently are not feasable at all
[10:17] <groz> the workplace is moving to 'purpose built devices' for everything
[10:17] <groz> and migrating away from general purpose 'desktop' computers
[10:18] <Kaloz> sure. but interestingly handhelds are dieing... notebooks taking away market share from both them and desktop
[10:18] <x-db90h> it's hard to say what the future holds, but that's one possibility
[10:18] <groz> kaloz, you ever worked with a prism54 card ?
[10:18] <groz> is it one that 'just works' in ap mode ?
[10:18] <Kaloz> yeah, but it was long ago
[10:19] <groz> i need to find a mini pci that 'just works' in ap mode
[10:19] <Kaloz> if you need something that "just works" go with prism :P if you need "just works most of the time" or 802.11a/g, go with atheros
[10:20] <groz> i need g mode
[10:20] <Kaloz> go with atheros then
[10:20] <groz> and i've had considerable stability issues with madwifi over time
[10:20] <groz> gunshy of even trying that for thsi project
[10:20] <groz> and, since it's not i386 based
[10:20] <groz> i dunno
[10:50] <CIA-17> florian * r4518 /packages/lang/sablevm-classpath/ (. Config.in Makefile ipkg/): Port sablevm-classpath to -ng
[10:51] <CIA-17> florian * r4519 /packages/lang/sablevm-classpath/Makefile: Remove references to old IDIR_*
[11:05] <CIA-17> florian * r4520 /packages/lang/sablevm/ (. Config.in Makefile ipkg/): Port sablevm to -ng
[11:22] <CIA-17> florian * r4521 /packages/net/xsupplicant/ (. Config.in Makefile ipkg/): Port xsupplicant to -ng
[11:30] <CIA-17> florian * r4522 /packages/utils/sane-backends/ (. Config.in Makefile ipkg/): Port sane-backends to -ng
[11:33] <CIA-17> florian * r4523 /packages/net/scanlogd/ (. Config.in Makefile ipkg/): Port scanlogd to -ng
[11:34] <CIA-17> florian * r4524 /packages/net/scdp/ (. Config.in Makefile ipkg/): Port scdp to -ng
[11:38] <CIA-17> florian * r4525 /packages/utils/screen/ (. Makefile): Port screen to -ng
[11:42] <CIA-17> florian * r4526 /packages/net/ser2net/ (. Config.in Makefile ipkg/): Port ser2net to -ng
[11:45] <malbon> groz: avoid prism54 unless you can get the fullmac version.
[11:46] <malbon> groz: the softmac drivers just don't work for me.
[12:24] <CIA-17> nico * r4527 /branches/buildroot-ng/openwrt/Config.in: enable ext2 rootfs by default when target is uml-2.6
[12:27] <CIA-17> nico * r4528 /branches/buildroot-ng/openwrt/package/busybox/ (28 files in 19 dirs): update busybox to v1.2.1
[12:28] <groz> nico ping
[12:32] <CIA-17> nico * r4529 /packages/net/lighttpd/ (9 files in 4 dirs): port lighttpd to buildroot-ng
[12:34] <{Nico}> groz: pong
[12:35] <groz> hey, i see you fussing with uml stuff, have you actually been using it at all ?
[12:37] <{Nico}> groz: yeah, but i'm using an old rev - recent changes broke it
[12:37] <{Nico}> still not played with networking
[12:37] <groz> i'll be back into it later in the week, i'll fix it again
[12:39] <{Nico}> it should be related to the recent TARGET_{CC,PATH} changes for ccache
[12:40] <groz> probably
[12:51] <{Nico}> groz: does it make sense to enable usb support in uml-2.6 ?
[12:54] <groz> good question on a lot of things
[12:54] <groz> one of the things that uml is good for, is developing a driver
[12:55] <groz> you can pull it down when things go wrong, and start over
[12:55] <groz> so if the host has no drivers already on the specific hardware
[12:55] <groz> a uml can talk to it directly
[14:25] <CIA-17> florian * r4530 /packages/utils/setpwc/ (. Config.in Config.in Makefile Makefile ipkg/ ipkg/): Port setpwc to -ng
[14:28] <CIA-17> florian * r4531 /packages/net/shat/ (. Config.in Makefile ipkg/): Port shat to -ng
[14:33] <CIA-17> florian * r4532 /packages/net/sipp/ (. Config.in Makefile ipkg/): Port sipp to -ng
[14:41] <CIA-17> florian * r4533 /packages/net/siproxd/ (. Config.in Makefile ipkg/): Port siproxd to -ng
[14:42] <CIA-17> florian * r4534 /packages/net/sipsak/ (. Config.in Makefile ipkg/): Port sipsak to -ng
[16:42] <CIA-17> florian * r4535 /packages/net/slurm/ (. Config.in Makefile ipkg/): Port slurm to -ng
[16:42] <CIA-17> florian * r4536 /packages/net/socat/ (. Config.in Makefile ipkg/): Port socat to -ng
[16:42] <CIA-17> florian * r4537 /packages/net/srelay/ (. Config.in Makefile ipkg/): Port srelay to -ng
[16:43] <CIA-17> florian * r4538 /packages/net/ssltunnel/ (. Config.in Makefile ipkg/): Port ssltunnel to -ng
[16:43] <CIA-17> florian * r4539 /packages/net/ssmtp/ (. Config.in Makefile ipkg/): Port ssmtp to -ng
[16:46] <CIA-17> nico * r4540 /branches/buildroot-ng/openwrt/package/zd1211/Makefile: disable on uml-2.6 (lacking wireless), standardize
[16:56] <CIA-17> florian * r4541 /packages/admin/syslog-ng/ (. Config.in Makefile ipkg/): Port syslog-ng to -ng
[16:58] <CIA-17> florian * r4542 /packages/net/tinc/ (. Config.in Makefile ipkg/): Port tinc to -ng
[17:05] <CIA-17> nico * r4543 /packages/net/rarpd/Makefile: fix rarpd: revert to plain old ./configure call, Build/Configure/Default does not support overloading CFLAGS (yet)
[17:37] <CIA-17> nico * r4544 /packages/lang/sablevm/Makefile: fix sablevm, standardize, indent
[18:42] <common> groz: i got a prism54 minipci card
[22:15] <CIA-17> mbm * r4545 /branches/buildroot-ng/openwrt/package/dnsmasq/files/dnsmasq.conf: fix error if dhcp is disabled
[23:54] <{Nico}> .
[23:55] <malbon> .
[23:56] <[mbm]> so, a quick recap of the agenda .. rc6 needs to be released asap .. some nagging issues with the webif
[23:56] <[mbm]> so the buggy firewall page will get dropped and we'll add some more nvram toggles to the pages
[23:56] <[mbm]> I really want to see every nvram variable used by rc6 in webif
[23:57] <{Nico}> [mbm]: can we all aggree on a release date?
[23:57] <[mbm]> unfortunately all of the devels are pretty busy so it's hard to find time to fix the old rc6 bugs
[23:58] <[mbm]> we have to fix the bugs and we have to do a resonable breakin/quality assurance
[23:58] <[mbm]> so the absolute minimum to a release would be 2 weeks from now
[23:59] <[mbm]> oh .. something else I should add to that
[23:59] <[mbm]> we should add an stunnel package and move webif to https
[00:00] --- Wed Aug 9 2006