**** BEGIN LOGGING AT Tue Oct 25 02:59:57 2011
Oct 25 06:39:32 <builder_> build #67 of rdc is complete: Failure [failed compile_7]  Build details are at http://buildbot.openwrt.org:8010/builders/rdc/builds/67
Oct 25 08:10:35 <dape_at_work> hello, what should i do? http://pastebin.com/ay6NLS66
Oct 25 08:12:47 <[florian]> resolve the conflict
Oct 25 08:15:10 <dape_at_work> ;-)
Oct 25 08:16:55 <[florian]> that's more a svn workflow question than something specific to openwrt
Oct 25 08:19:32 <dape_at_work> meh, np rm it and svn update again
Oct 25 08:35:46 <dape_at_work> ok so if i understand correctly openwrt devs have tweaked kernel config related to the exact machine selected in make menuconfig
Oct 25 08:36:23 <dape_at_work> in my case i selected tp-link wr703n but in make kernel_menuconfig i can see hardware and software related options selected that arent used at all
Oct 25 08:36:34 <dape_at_work> this could lead to a bigger kernel and for the 4 mb flash thats an issue
Oct 25 08:38:39 <dape_at_work> like CONFIG_NET_DSA and the two marvel chips selected...
Oct 25 08:45:05 <[florian]> don't fool yourself, a mips compressed kernel fits within a few hundreds KBytes
Oct 25 08:45:18 <[florian]> removing some random options, will, at best make you win 100KB
Oct 25 08:45:43 <dape_at_work> enough for 3g support and maybe ipv6
Oct 25 08:45:49 <dape_at_work> i really really dont want luci out
Oct 25 08:46:17 <[florian]> so you should be able to make this fit within 3MB of flash
Oct 25 08:47:04 <dape_at_work> im recompiling right now a customized image for wr703n
Oct 25 09:06:08 <dape_at_work> uhm, opkg error.. http://pastebin.com/TfbRhkk7
Oct 25 10:22:24 <builder_> build #80 of avr32 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/80
Oct 25 11:31:02 <CIA-62> jow * r28573 /trunk/package/mac80211/Makefile: [package] mac80211: allow building ath9k_htc without PCI support, based on patch by Helmut Schaa <helmut.schaa@googlemail.com>
Oct 25 11:39:39 <CIA-62> jow * r28574 /packages/net/shorewall-lite/ (Makefile patches/110-MODULESDIR.patch):
Oct 25 11:39:39 <CIA-62> shorewall-lite: update to 4.4.24
Oct 25 11:39:39 <CIA-62> This patch updates shorewall-lite to current stable release 4.4.24
Oct 25 11:39:39 <CIA-62> Signed-off-by: Edy Corak info@loenshotel.de
Oct 25 12:42:56 <CIA-62> marek * r28575 /packages/net/batman-adv/ (5 files in 2 dirs):
Oct 25 12:42:56 <CIA-62> batman-adv: bugfixes for the 2011.3.1 release
Oct 25 12:42:56 <CIA-62> Signed-off-by: Marek Lindner <lindner_marek@yahoo.de>
Oct 25 12:45:53 <CIA-62> jow * r28576 /packages/net/shorewall-lite/Makefile: [packages] shorewall-lite: remove obsolete dependency on iptables-utils (#10217)
Oct 25 12:47:15 <CIA-62> jow * r28577 /packages/net/multiwan/Makefile: [packages] multiwan: remove obsolete dependency on iptables-utils (#10217)
Oct 25 12:52:52 <dape_at_work> uhm, anyone wants to tell me if CONFIG_PACKAGE_luci-sgi-uhttpd is critical for running luci with uhttpd? if i select luci in menuconfig this option is not auto-selected..
Oct 25 13:04:08 <jow_laptop> no itsn ot
Oct 25 13:07:24 <dape_at_work> so i leave it disabled
Oct 25 13:07:58 <dape_at_work> thanks jow_laptop !
Oct 25 13:08:29 <CIA-62> nbd * r28578 /trunk/package/busybox/patches/920-portability.patch: busybox: fix a compile error on mac os x
Oct 25 13:11:49 <CIA-62> jow * r28579 /packages/libs/firewall-mod-save/Makefile: [packages] firewall-mod-save: remove obsolete dependency on iptables-utils (#10217)
Oct 25 13:19:11 <dape_at_work> uhm, if my wr703n has u-boot then i suppose i can opt out from the kernel at mtd config redboot and myloader ..
Oct 25 13:19:28 <[florian]> correct
Oct 25 13:19:58 <dape_at_work> thanks, im trying to profit from flash at best
Oct 25 13:26:37 <CIA-62> jow * r28580 /branches/packages_10.03.1/net/shorewall-lite/Makefile: [backfire/packages] shorewall-lite: merge r28576
Oct 25 13:27:18 <CIA-62> jow * r28581 /branches/packages_10.03.1/net/multiwan/Makefile: [backfire/packages] multiwan: merge r28577
Oct 25 13:27:59 <CIA-62> jow * r28582 /branches/packages_10.03.1/libs/firewall-mod-save/Makefile: [backfire/packages] firewall-mod-save: merge r28579
Oct 25 13:32:15 <CIA-62> jow * r28583 /branches/packages_10.03.1/net/wget/Makefile: [backfire/packages] wget: merge r28519 (CVE-2010-2252)
Oct 25 14:08:49 <dape_at_work> uhm, if i want to buid (sometimes) two separate kernel and packages config, what are the config files that i must save or symlink before each compile?
Oct 25 14:16:02 <[florian]> are you using the env script? found in trunk/scripts/env?
Oct 25 14:16:49 <dape_at_work> no, hmm, just ran it, very nice, so thats the way to go?
Oct 25 14:34:13 <[florian]> dape_at_work: it's pretty self explanatory
Oct 25 14:34:18 <[florian]> the usual workflow with this script is:
Oct 25 14:34:25 <[florian]> ./scripts env new foo
Oct 25 14:34:27 <[florian]> make menuconfig
Oct 25 14:34:31 <[florian]> save your options
Oct 25 14:34:37 <[florian]> ./scripts env new bar
Oct 25 14:34:42 <[florian]> save the changes to foo
Oct 25 14:34:44 <[florian]> and switch to bar
Oct 25 14:34:57 <[florian]> then ./scripts/env switch foo (or bar) when you need to
Oct 25 14:38:45 <dape_at_work> yes, ive seen the help, its very cool, thanks!
Oct 25 15:17:26 <dape_at_work> aight, fingers crossed, compiling pppoe,3g support, ipv6 and luci in this 4 mb flash
Oct 25 15:36:43 <dape_at_work> uhm, is CONFIG_PACKAGE_kmod-textsearch=y related to some iptables string match feature or something? i would like to skip this package but its selected always..
Oct 25 15:37:49 <jow_laptop> yes
Oct 25 15:37:59 <jow_laptop> layer7 and xt_string depend on it
Oct 25 15:38:38 <jow_laptop> instead of chopping tiny bits off the kernel, did you include other obvious candidates like opkg?
Oct 25 15:38:41 <dape_at_work> hmm, wont use those so soon, anyway i can skip them?
Oct 25 15:38:48 <jow_laptop> s/include/exclude/
Oct 25 15:38:50 <dape_at_work> yes, opkg is out
Oct 25 15:39:05 <jow_laptop> did you enable the luci src diet feature?
Oct 25 15:39:06 <dape_at_work> will be a 3g and pppoe mini router with multiwan
Oct 25 15:39:11 <dape_at_work> NO, where? :)
Oct 25 15:39:27 <jow_laptop> its in luci-lib-core submenu
Oct 25 15:39:36 <dape_at_work> im on it, thanks for the tip
Oct 25 15:39:40 <jow_laptop> LuCI -> Libraries -> luci-lib-core -> ...
Oct 25 15:40:21 <dape_at_work> Build Target ? cause under luci-lib-core there's no submenu
Oct 25 15:40:25 <jow_laptop> yes
Oct 25 15:40:33 <jow_laptop> Buid Target Compressed Source
Oct 25 15:40:51 <dape_at_work> thanks!
Oct 25 15:41:35 <jow_laptop> not sure how much it saves after compression
Oct 25 15:42:09 <dape_at_work> lets see, so far i got 290k free, no multiwan installed yet and only english support for luci
Oct 25 15:42:12 <jow_laptop> but it made the lua quote quite compact in my tests
Oct 25 15:42:19 <jow_laptop> *lua code
Oct 25 15:42:22 <dape_at_work> tell you in few minutes
Oct 25 15:43:47 <jow_laptop> mhei_: I wonder whether we should make php5 depend on at least zoneinfo-core now
Oct 25 15:44:19 <jow_laptop> mhei_: because right now a standalone php5 installation will complain about a corrupted TZ database since it cannot find something for "UTC"
Oct 25 15:47:43 <jow_laptop> mhei_: also it looks like the php developers took away the ability specify a timezone with "TZ" and the heuristics to find a matching one (http://svn.php.net/viewvc/php/php-src/trunk/ext/date/php_date.c?r1=315779&r2=316640)
Oct 25 15:48:01 <jow_laptop> so more patches might be needed in case we update php5
Oct 25 15:49:14 <jow_laptop> I'd propose to make it read /etc/localtime and fall back to UTC
Oct 25 15:53:53 <karlp> if I just continuously overwrite a file on "disk" (not /tmp) will the linux drivers magically take care of wear levelling for me?
Oct 25 15:54:32 <jow_laptop> jffs2 will, but it has only so many blocks to work with
Oct 25 15:56:30 <karlp> if I keep writing just a few bytes, it will take a while to run out of even just one block though right?
Oct 25 15:56:58 <jow_laptop> nope, in order to write a few bytes, a whole block has to be erased and written again
Oct 25 15:57:07 <karlp> youch
Oct 25 15:57:19 <jow_laptop> which is usually 64k or even bigger
Oct 25 15:57:25 <karlp> I guess I'm too used to eeprom for this.
Oct 25 15:57:59 <dape_at_work> jow_laptop,  sorry for interruption, luci compressed source saved few insignifiant Ks.
Oct 25 15:58:08 <jow_laptop> dape_at_work: ok
Oct 25 15:58:10 <karlp> is there anyway to find out the block size?
Oct 25 15:58:21 <jow_laptop> karlp: cat /proc/mtd, the eraseblock column
Oct 25 15:59:07 <karlp> rootfs_data I presume? that's where /overlay seems to go
Oct 25 15:59:15 <jow_laptop> yes
Oct 25 16:23:51 <CIA-62> jow * r28584 /trunk/target/linux/ (37 files in 22 dirs): target: globally disable BSD process accounting
Oct 25 16:24:28 <dape_at_work> cool :) i wanted to ask if this is really used by any program
Oct 25 16:25:05 <CIA-62> jow * r28585 /trunk/target/linux/brcm47xx/config-3.0: brcm47xx: revert accidentially added EARLY_PRINTK symbol from previous commit
Oct 25 16:45:18 <CIA-62> jow * r28586 /trunk/target/linux/generic/ (9 files): target: remove CONFIG_IDE_PROC_FS support as well
Oct 25 16:51:02 <builder_> build #100 of s3c24xx is complete: Failure [failed compile_10]  Build details are at http://buildbot.openwrt.org:8010/builders/s3c24xx/builds/100
Oct 25 17:09:23 <karlp> if jffs wear levelling isn't a very good way of saving persistent data, are there any other tricks commonly used?
Oct 25 17:09:59 <karlp> there's never going to be enough time to detect the power condition and save to flash then, I would imagine
Oct 25 17:23:36 <builder_> build #92 of pxcab is complete: Failure [failed compile_10]  Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/92
Oct 25 17:52:30 <dape_at_home> guys sorry if i ask this here and not on the main #openwrt channel, not sure, if i play with multiwan do i still need in kernel CONFIG_IP_ADVANCED_ROUTER and CONFIG_IP_MULTIPLE_TABLES ?
Oct 25 17:52:45 <jow_laptop> yes
Oct 25 17:52:53 <dape_at_home> darn, so it uses iproute
Oct 25 17:53:13 <jow_laptop> you can try to use busybox ip
Oct 25 17:53:19 <jow_laptop> and ditch the full iproute2
Oct 25 17:53:24 <dape_at_home> i saw that has some iproute functionality in
Oct 25 17:54:35 <dape_at_home> ah, and another question, if this wr703n has only one ethernet port, do i still need CONFIG_VLAN_8021Q ?
Oct 25 17:54:54 <dape_at_home> dont plan to use trunk on that port, just pure access
Oct 25 17:55:25 <jow_laptop> no I don't think so
Oct 25 17:55:42 <jow_laptop> in this case you can also disable vconfig in busybox
Oct 25 17:55:47 <dape_at_home> should i try? dont think i can solder serial on this, its too small device :)
Oct 25 17:56:06 <dape_at_home> ah, hmm, nice idea, so i can choose the build options in busybox? hmm
Oct 25 17:56:08 <jow_laptop> well can you recover it?
Oct 25 17:56:33 <dape_at_home> dont think so, openwrt recovery will still use the kernel without vlan so..
Oct 25 17:56:55 <jow_laptop> so you have no bootloader access
Oct 25 17:57:00 <dape_at_home> i mean heck, i could still ask dad to solder the serial but his eyes could miss the pads :)
Oct 25 17:57:45 <dape_at_home> logic would tell that with only one port there's no need for vlan, but what could still use this..
Oct 25 17:58:01 <jow_laptop> wel lthere's strange hardware out there
Oct 25 17:58:26 <jow_laptop> there were definitely brcm47xx units with only one ethernet port but they had a switch like a wrt54gl
Oct 25 17:58:36 <jow_laptop> with all but one external port unconnected
Oct 25 17:58:49 <jow_laptop> one needs vlans then
Oct 25 17:59:47 <dape_at_home> swconfig dev ag71xx show doesnt work
Oct 25 18:03:25 <KanjiMonster> dape_at_home: it would be swconfig dev switch0 (or eth0)
Oct 25 18:05:30 <dape_at_home> Failed to connect to the switch to both
Oct 25 18:05:50 <KanjiMonster> then there is none
Oct 25 18:06:47 <dape_at_home> if this would be true than i do not need swconfig package
Oct 25 18:07:01 <dape_at_home> let me check things out deeper
Oct 25 18:08:34 <jow_laptop> but I would not remove 802.1q, it could become useful, especially on a single port device
Oct 25 18:08:46 <jow_laptop> I suppose removing swconfig gains you more
Oct 25 18:08:56 <dape_at_home> ook
Oct 25 18:14:15 <KanjiMonster> you can also remove several phy drivers (e.g. ar8216, rtl8366rb/s)
Oct 25 18:14:50 <dape_at_home> did it :)
Oct 25 18:15:48 <dape_at_home> guys look, checked with my dlink 825 /etc/config/network.. on this 703n there is only config interface loopback and config interface lan in /etc/config/network , thats it, from the generic trunk image installation for tp-link 703n
Oct 25 18:16:01 <dape_at_home> so this surely means swconfig ain't used, right?
Oct 25 18:18:48 <jow_laptop> yes
Oct 25 18:29:10 <dape_at_home> oki, well, i am at SPI support, i see Atheros AP83, my Atheros AR71xx controller and Atheros PB44, should i disable the other too?
Oct 25 18:30:34 <dape_at_home> the other two i meant, in dmesg i see only "Atheros AR71xx SPI Controller driver version 0.2.4"
Oct 25 18:34:50 <dape_at_home> also i'm wondering about CONFIG_JFFS2_FS_XATTR too :)
Oct 25 18:34:57 <jow_laptop> xattr is madatory
Oct 25 18:35:04 <dape_at_home> oki
Oct 25 20:35:25 <dape_at_home> uhm, i get "uci: Parse error (invalid command) at line 5, byte 0" on "uci commit" ..
Oct 25 20:35:56 <dape_at_home> any way i can increase verbosity?
Oct 25 20:37:01 <dape_at_home> meh, got it, last part of a comment got in the way on the next line
Oct 25 20:58:35 <CIA-62> nico * r28587 /packages/lang/perl-net-telnet/ (. Makefile): packages: add perl-net-telnet (closes: #10277)
Oct 25 20:58:36 <CIA-62> nico * r28588 /packages/multimedia/ffmpeg/Makefile: packages/ffmpeg: fix pkg-config detection
Oct 25 20:58:39 <CIA-62> nico * r28589 /packages/multimedia/ffmpeg/Makefile:
Oct 25 20:58:39 <CIA-62> packages/ffmpeg: update to 0.8.5
Oct 25 20:58:39 <CIA-62> This patch updates FFmpeg to 0.8.5.
Oct 25 20:58:39 <CIA-62> Signed-off-by: Ian Leonard <antonlacon@gmail.com>
Oct 25 20:58:40 <CIA-62> nico * r28590 /packages/multimedia/minidlna/Makefile:
Oct 25 20:58:40 <CIA-62> packages/minidlna: update to 1.0.22
Oct 25 20:58:40 <CIA-62> This patch updates Minidlna to 1.0.22.
Oct 25 20:58:40 <CIA-62> Signed-off-by: Ian Leonard <antonlacon@gmail.com>
Oct 25 21:15:52 <CIA-62> nico * r28591 /trunk/package/zlib/Makefile:
Oct 25 21:15:52 <CIA-62> package/zlib: include zlib.pc (pkconfig)
Oct 25 21:15:52 <CIA-62> Compiling zlib produces zlib.pc (pkgconfig), but it is not currently
Oct 25 21:15:52 <CIA-62> installed in staging_dir/usr/lib/pkgconfig. If it was included other
Oct 25 21:15:52 <CIA-62> packages depending on zlib could easier be compiled.
Oct 25 21:15:52 <CIA-62> Signed-off-by: Espen Jürgensen <espenjurgensen@gmail.com>
Oct 25 21:54:43 <builder_> build #100 of at91 is complete: Failure [failed compile_4]  Build details are at http://buildbot.openwrt.org:8010/builders/at91/builds/100
Oct 25 21:55:25 <builder_> build #98 of ubicom32 is complete: Failure [failed compile_3]  Build details are at http://buildbot.openwrt.org:8010/builders/ubicom32/builds/98
Oct 25 22:10:20 <CIA-62> nico * r28592 /packages/sound/listener/ (. Makefile patches/ patches/100-makefile.patch): packages: add listener (thanks to Raphaël Huck)
Oct 25 23:15:28 <CIA-62> jow * r28593 /packages/net/netatalk/Makefile: [packages] netatalk: use -std=gnu99 instead of -std=c99, fixes unreachable cnid_dbd - thanks dirtyfreebooter for spotting this
Oct 26 00:47:51 <CIA-62> nico * r28594 /packages/sound/listener/ (Makefile files/ files/listener.conf):
Oct 26 00:47:51 <CIA-62> packages/listener: add a config file, fix config file handling
Oct 26 00:47:51 <CIA-62> Add a config file based on:
Oct 26 00:47:51 <CIA-62> http://projects.qi-hardware.com/index.php/p/openwrt-packages/source/tree/master/listener/files/listener.conf
Oct 26 00:47:51 <CIA-62> Fix the location where listener is looking for its configuration file
Oct 26 00:47:52 <CIA-62> (thanks to Xiangfu Liu for pointing these)
Oct 26 00:48:50 <xiangfu> you are welcome :)
Oct 26 00:54:22 <CIA-62> nico * r28595 /trunk/include/package-ipkg.mk:
Oct 26 00:54:22 <CIA-62> include/package-ipkg.mk: fix typo
Oct 26 00:54:22 <CIA-62> Fix a typo that caused packaging of an empty ./lib/upgrade/keep.d
Oct 26 00:54:22 <CIA-62> directory when the package supplied conffiles
**** ENDING LOGGING AT Wed Oct 26 02:59:57 2011