**** BEGIN LOGGING AT Sun Mar 06 02:59:57 2011 Mar 06 03:13:00 build #95 of brcm47xx is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/brcm47xx/builds/95 Mar 06 09:55:02 gmorning Mar 06 10:41:18 mb * r25882 /trunk/target/linux/omap24xx/ (base-files/etc/init.d/watchdog profiles/100-n810.mk): omap24xx: Start watchdog with RT priority Mar 06 10:46:37 mb * r25883 /trunk/target/linux/omap24xx/base-files/etc/init.d/watchdog: omap24xx: Set hardware watchdog timeout to max Mar 06 11:22:41 philipp64|laptop: the change it, its a wiki Mar 06 11:41:10 jow * r25884 /packages/libs/libsdl_mixer/Makefile: [packages] libsdl-mixer: don't use host sdl-config Mar 06 11:54:53 xMff: how would I use './scripts/feeds install' with 'src-link packages /home/jan/packages' in feeds.conf.default? Mar 06 11:55:11 it doesn't seem to symlink the installed package Mar 06 11:56:01 ? Mar 06 11:56:05 works just okay here Mar 06 12:02:31 ah, pebcak :/ Mar 06 12:18:09 build #79 of ar7 is complete: Failure [failed compile_7] Build details are at http://tksite.gotdns.org:8010/builders/ar7/builds/79 Mar 06 12:20:59 [florian]: ping Mar 06 13:16:25 build #79 of sibyte is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/sibyte/builds/79 Mar 06 13:22:25 mb * r25885 /packages/utils/bemused/ (Makefile patches/200-cxx-compile-fixes.patch): bemused: Compile fixes Mar 06 13:24:43 Does anyone here know the hardware specs of a Coby 7" Android Tablet (http://shopping.dallasnews.com/ROP/ads.aspx?advid=2519&adid=10661789&subid=33305673&type=)? I am hoping OpenWRT can be ported to this device. Mar 06 13:26:10 mb * r25886 /packages/utils/bemused/Makefile: bemused: Don't need explicit iconv flags. nls.mk does that for us Mar 06 13:29:39 updated openwrt/upstream, https://home.comcast.net/~sdwalker/uscan/index.html Mar 06 13:29:57 mazilo: http://forum.xda-developers.com/showthread.php?t=880545 Mar 06 13:30:00 build #72 of etrax is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/etrax/builds/72 Mar 06 13:51:46 build #90 of ppc40x is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/ppc40x/builds/90 Mar 06 14:40:15 cshore * r25887 /packages/net/miniupnpd/ (Makefile patches/ patches/010-genconfigsh_for_openwrt.patch): Mar 06 14:40:15 [packages] miniupnpd: Updated miniupnpd to 1.5.20110302. Mar 06 14:40:15 Changed from using static config.h.openwrt from upstream tarball to using genconfig.sh from upstream config (with patches to genconfig.sh to generate the correction config.h for openwrt). closes: 8934 Mar 06 15:15:49 cshore * r25888 /packages/net/miniupnpd/ (Makefile patches/100-use_linux_makefile_for_openwrt.patch): Mar 06 15:15:50 [packages] miniupnpd: Changed from using upstream Makefile.openwrt to using Mar 06 15:15:50 upstream's Makefile.linux (after suitably patching it so it works with openwrt Mar 06 15:45:23 build #78 of x86 is complete: Failure [failed shell_9] Build details are at http://tksite.gotdns.org:8010/builders/x86/builds/78 Mar 06 15:51:08 build #85 of ramips is complete: Failure [failed shell_9] Build details are at http://tksite.gotdns.org:8010/builders/ramips/builds/85 Mar 06 16:01:26 swalker * r25889 /packages/libs/polarssl/Makefile: [packages] polarssl: update to 0.14.2, fix & add descriptions Mar 06 16:40:30 build #74 of xburst is complete: Success [build successful] Build details are at http://tksite.gotdns.org:8010/builders/xburst/builds/74 Mar 06 17:11:39 Hi ppl! Mar 06 17:28:04 jow * r25890 /packages/utils/triggerhappy/ (Makefile patches/ patches/100-fix-LINUX_INPUT_H.patch): [packages] triggerhappy: fix determination of input.h, remove explicit override (#8990) Mar 06 17:34:44 jow * r25891 /trunk/package/dropbear/patches/200-lcrypt_bsdfix.patch: [package] dropbear: fix missing -lcrypt on BSDish host systems (#8971) Mar 06 17:35:26 mb * r25892 /trunk/target/linux/omap24xx/config-2.6.38: omap24xx: Update kconfig Mar 06 17:38:20 jow * r25893 /branches/backfire/ (3 files in 3 dirs): [backfire] package iptables hashlimit match (#8846) Mar 06 17:40:02 jow * r25894 /branches/backfire/package/kernel/modules/netdevices.mk: [backfire] package IFB kernel module (#8849) Mar 06 18:24:45 jow * r25895 /packages/net/ipcad/Makefile: [packages] ipcad: fix compilation against librpc Mar 06 18:25:43 ping xMff Mar 06 18:25:50 'lo cshore Mar 06 18:29:10 Just thought I'd let you know I fixed miniupnpd. I got bit by believing a patch instead of being anal and testing it mysef. Then I got sick. Now I'm going to be completely anal and ignore requests to submit patches I haven't tested myself (at least unless it's from someone with a good track record, like KanjiMonster/Jonas). I should have known better, but there was a bug I wanted fix, and dammit I want to be able to trust people to Mar 06 18:29:59 what was it? Mar 06 18:31:08 the patch was the update to 1.5. Apparently it broke because the version number wasn't updated in the openwrt file that upstream had, and that's what was being used by the miniupnpd makefile Mar 06 18:33:23 I see Mar 06 18:33:54 Someone calculated the md5sum and submitted a patch without testing and I didn't think anyone would do that. Too trusting I guess. Mar 06 18:37:51 :) Mar 06 18:39:14 People wonder why we don't apply patches without testing them ourselves and complain...it's frustrating Mar 06 18:41:35 ah well Mar 06 18:41:39 how are you anyway? Mar 06 18:42:22 oh I don't know if you saw last week I did a couple of the things we talked about for extroot Mar 06 18:42:29 I still have some todo's there though Mar 06 18:52:43 cshore: I'd still prefer if you test my patches, or at least look at them to see if I missed something obvious ;) Mar 06 18:53:17 ok, I admit I prefer that reassurance myself too Mar 06 18:58:12 cshore * r25896 /packages/net/miniupnpd/files/miniupnpd.init: [packages] miniupnpd: Add uci option for manually specified external IP Mar 06 18:58:19 cshore * r25897 /packages/net/miniupnpd/files/miniupnpd.init: [packages] miniupnpd: Added option to change port. Default set to 5000 as before, port 'auto' set to result in a random port being used. Mar 06 18:58:20 cshore * r25898 /packages/net/miniupnpd/files/miniupnpd.init: [packages] miniupnpd: Added uci option use_system_uptime which defaults to true to maintain current behaviour, which is to use system update instead of miniupnpd daemon uptime. If false miniupnpd daemon uptime is used Mar 06 18:58:25 cshore * r25899 /packages/net/miniupnpd/files/miniupnpd.init: Mar 06 18:58:25 [packages] miniupnpd: Added uci option config_file, which allows the user to Mar 06 18:58:25 specify that miniupnpd should use the specified file as the configuration file Mar 06 18:58:35 cshore * r25900 /packages/net/miniupnpd/files/miniupnpd.init: [net] miniupnpd: Added model_number and serial_number uci options for the root desc xml Mar 06 18:58:38 cshore * r25901 /packages/net/miniupnpd/files/miniupnpd.init: Mar 06 18:58:38 [net] miniupnpd: Added uci configuration option to set uuid. A uuid is Mar 06 18:58:38 automatically generated and added to the configuration file on the first boot of Mar 06 19:00:49 cshore: I also have a bad habit of doing some last minute cosmetic fixes then breaking something subtle in the progress (and not testing the modified patch), or testing it on top of other local patches Mar 06 19:01:37 ok, I'll keep that in mind. I can understand the desire for speed and thus failure to be anal Mar 06 19:01:56 There's only so many hours in a day Mar 06 19:04:21 Even jow sometimes makes mistakes due to a fix that's supposed to 'just work', but he usually catches it pretty quickly, because he online with the person who is helping discover the original problem. Mar 06 19:04:36 yeah Mar 06 19:05:03 I usually see my mistake about ~3 seconds after the patch got accepted and committed ;) Mar 06 19:05:23 heh Mar 06 19:15:29 cshore * r25902 /packages/net/miniupnpd/files/miniupnpd.init: [net] miniupnpd: Added uci option notify_interval and presentation_url Mar 06 19:53:42 xMff: just a note, you comment on r25891, "fix missing -lcrypt on BSDish host systems", its not a fix for a missing -lcrypt, its that -lcrypt was put in the list of objects and gnumake on darwin tries to build -lcrypt target Mar 06 19:54:31 dirtyfreebooter: ... and therefore it was missing ;) Mar 06 19:54:38 dirtyfreebooter: I understood the patch Mar 06 19:54:52 oh ok, i just wanted to make sure Mar 06 19:56:00 the dropbear author, matt is going rework the makefile to be less hackish upstream, apparently the -lcrypt in the objects list was because he was having trouble linking crypt in static binary version Mar 06 19:56:22 well... I've seen worse stuff Mar 06 19:56:43 yea ;-) Mar 06 19:58:39 cshore: while you're hacking on miniupnpd, someone mentioned that it is possible to restrict possible mappings in the native config Mar 06 19:58:48 cshore: support for that in uci would be useful Mar 06 19:59:00 :) Was planning on it Mar 06 19:59:02 cshore: but that would require generating a temporary config Mar 06 19:59:49 okay, tell me if you're done so I can sync luci Mar 06 20:01:54 right Mar 06 20:02:04 so, i had no luck with openwrt Mar 06 20:02:10 one problem after another Mar 06 20:02:35 gave up... eventually -_- Mar 06 20:02:43 yep....if no config_file is specified and use_tmp_config_file is true (default true), then generate a config in /tmp/miniunpd.conf based on rule stanzas. One thing the order matters. Is it now possible to specify the order a set of stanza will be processed in ? Mar 06 20:03:13 cshore: no, thats up to the user to put a correct order Mar 06 20:03:59 and what if config_file is false and use_tmp_config_file is false? Mar 06 20:04:41 also to follow the other packages it should be /var/run or /var/etc, not /tmp Mar 06 20:06:28 Then /etc/miniupnpd.conf is used if present, otherwise no config file is used Mar 06 20:06:54 ok, /var/etc then Mar 06 20:06:57 too much options imo Mar 06 20:07:42 it's a consequence of what miniupnpd does.....miniupnpd uses a config file if present, other it just goes by command line Mar 06 20:07:46 at least use_tmp_config_file is overengineering Mar 06 20:08:10 either a user specifies a custom config or does not Mar 06 20:08:29 if he does not then he has to accept whatever the init is doing Mar 06 20:09:27 ok Mar 06 20:10:08 that'd be in line with e.g. olsr or radvd where specifying a config file bypasses uci completely Mar 06 20:10:19 or openvpn Mar 06 20:11:41 hmmm.....ok, so you want the config_file being specified to mean don't use command-line parameters? Mar 06 20:12:04 i.e. bypass default uci behaviour? Mar 06 20:12:14 I mean if "option config_file /etc/foobar.conf" is given then miniupnpd is started with that (-f ?) Mar 06 20:12:20 if not, the init script does uci Mar 06 20:12:35 ok, that's what I meant Mar 06 20:13:16 and since you're going to need a temp config for the acl stuff anyway you can also add the yet-on-commandline parameters in there Mar 06 20:13:57 yet-on-commandline (you mean not-yet, or just skip the command line altogether?) Mar 06 20:14:36 skip commandline Mar 06 20:23:51 build #94 of ubicom32 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ubicom32/builds/94 Mar 06 20:24:35 oh back to the order of stanzas....how do you recommend allowing the user to specify the order....you're most likely to be the one implementing it in LuCI so it's really up to you what UI you want to use. Mar 06 20:25:11 (basically for allow/deny it's like firewall rules, and order we add them to the config file matters) Mar 06 20:27:19 I thought you added ordering to firewall rules, or am I smoking something there? Mar 06 20:28:05 no, ordering is already implemented in luci Mar 06 20:28:16 just add .sortable = true to a tblsection Mar 06 20:28:25 I was referring to users of the uci cli Mar 06 20:28:30 they can use uci reorder Mar 06 20:28:37 and with an editor its obvious anyway Mar 06 20:29:05 okay, so the order the entries are in the file is the order they are processed with config_foreach ? Mar 06 20:29:06 therefore I believe its well covered Mar 06 20:29:17 correct Mar 06 20:29:24 excellent Mar 06 21:37:21 mb * r25903 /trunk/package/kernel/modules/usb.mk: Load MUSB before gadget and host drivers are loaded Mar 06 21:44:08 jow * r25904 /trunk/package/ (5 files in 5 dirs): [package] change PKG_FIXUP:=libtool to PKG_FIXUP:=autoreconf Mar 06 21:45:53 mb * r25905 /trunk/package/kernel/modules/usb.mk: Fix loading of ethernet gadget Mar 06 21:56:05 jow * r25906 /branches/backfire/package/ (4 files in 4 dirs): [backfire] backport r25904 Mar 06 22:04:28 mb * r25907 /trunk/target/linux/omap24xx/patches-2.6.38/850-musb-tusb-modular-fixes.patch: omap24xx: Fix modular compilation of MUSB Mar 06 22:13:22 build #75 of kirkwood is complete: Failure [failed shell_4] Build details are at http://tksite.gotdns.org:8010/builders/kirkwood/builds/75 Mar 06 22:20:44 mb * r25908 /trunk/target/linux/omap24xx/patches-2.6.38/590-cbus-tahvo-usb-fixes.patch: n810: Partially fix tahvo USB. Mar 06 22:21:01 jow * r25909 /branches/backfire/package/6in4/files/6in4.sh: [package] 6in4: add default route workarounds for Kernel 2.4, thanks to Paul Fertser for reporting Mar 06 22:29:32 mb * r25910 /trunk/target/linux/omap24xx/profiles/100-n810.mk: n810: Update profile Mar 06 22:38:01 nbd * r25911 /trunk/package/base-files/Makefile: base-files: fix Build/Compile override handling, fixes ar7 compile Mar 06 22:55:50 nbd * r25912 /trunk/toolchain/gcc/patches/linaro/850-use_shared_libgcc.patch: gcc: include assembler functions in the shared libgcc_s library, fixes missing symbol issues on powerpc Mar 06 23:00:08 mb * r25913 /trunk/target/linux/omap24xx/patches-2.6.38/590-cbus-tahvo-usb-fixes.patch: tahvo-usb: Fix NULL ptr deref in OTR irq handler Mar 06 23:02:45 build #98 of atheros is complete: Failure [failed compile_4] Build details are at http://tksite.gotdns.org:8010/builders/atheros/builds/98 Mar 06 23:12:17 nbd * r25914 /trunk/target/linux/atheros/patches-2.6.37/ (4 files): atheros: add gpio_set_debounce(), refresh patches Mar 06 23:24:25 finally got USB network working on the n810 (with about 22 MBit throughput). This is a major step forward for people without access to the serial console (most people). Mar 06 23:27:32 cool Mar 06 23:30:55 do files under /etc/config need be listed as conffile ? miniupnpd seems to be generate an upnpd-opkg becaseu there is already a upnpd under /etc/config from a previous run Mar 06 23:31:15 during build into image I mean Mar 06 23:31:54 yes Mar 06 23:33:23 I guess I'm puzzle because I'm geting that upnpd-opkg file that I don't see happening with other packages (when generation image, not when installing onto jffs2 after the fact) Mar 06 23:33:35 *generating Mar 06 23:34:27 I think I'll patch opkg right now to not do that anymore Mar 06 23:34:36 its not terribly useful on the target either Mar 06 23:46:22 cshore: I found a less intrusive solution Mar 06 23:46:45 xMff: to opkg? Mar 06 23:46:59 yes Mar 06 23:47:25 http://paste.openwrt.org/d3e2e397e Mar 06 23:48:18 cool. it's basically already been done. Mar 06 23:50:18 testing now Mar 06 23:52:10 it worked Mar 06 23:53:28 there was no upnpd-opkg afterwards Mar 06 23:53:38 jow * r25915 /trunk/include/package-ipkg.mk: [include] deploy .ipk archives with --force-maintainer to ensure that configs are kept up-to-date in case they exist already for various reasons Mar 06 23:54:06 jow * r25916 /branches/backfire/include/package-ipkg.mk: [backfire] merge r25915 Mar 06 23:54:48 do you have two packages providing a /e/c/upnpd ? Mar 06 23:55:08 personally I never managed to have prexisting conffiles in the opkg install stage Mar 06 23:56:22 I only have miniupnpd....but maybe in a sec when I post my patch for the changes we talked about you'll see something I missed Mar 06 23:57:20 could it just not be clean target/linux? Mar 06 23:57:28 I am not sure Mar 06 23:57:45 however we just made it more rebust so I don't really care Mar 06 23:59:50 xMff: http://paste.openwrt.org/m3dcefa70 Mar 07 00:00:46 cshore: hm, I'd combine *_start and *_int Mar 07 00:00:55 erm *_start and *_end Mar 07 00:01:23 using the firewall portrange syntax (x, x-y or x:y allowed) Mar 07 00:01:38 the reason there separate is that sometimes you want only one (i.e. not range but one port) Mar 07 00:02:40 oh I see Mar 07 00:03:03 hmmm....let me see if I can find where the firewall code handles the port ranges Mar 07 00:04:08 https://dev.openwrt.org/browser/trunk/package/firewall/files/lib/fw.sh#L181 Mar 07 00:11:14 mb * r25917 /trunk/target/linux/omap24xx/base-files/etc/config/network: n810: Add usb networking to default net config Mar 07 00:12:41 hmmm....I think that's not quite what i want - : isn't an allowed delimiter in the miniupnpd.conf Mar 07 00:13:05 so I would change the %- to %: ? Mar 07 00:14:02 http://paste.openwrt.org/d5ccdc11c Mar 07 00:23:44 cshore: would that work for you? Mar 07 00:24:15 yes, thanks....I'm putting it in to try it out now Mar 07 00:25:37 shortens the rules nicely Mar 07 00:27:49 ip_mask applies to clients? Mar 07 00:28:14 hosts or subnet, must always bet n.n.n.n/nn Mar 07 00:28:31 it defines the allowed forwarding destination? Mar 07 00:28:32 and yes clients Mar 07 00:28:35 yes Mar 07 00:28:44 not who is allowed to frward? Mar 07 00:28:55 no Mar 07 00:28:58 ok Mar 07 00:29:15 but in secure mode that's the same thing Mar 07 00:29:23 because you can only redirect to yourself Mar 07 00:29:56 I'd call them ext_port, int_port and int_addr then Mar 07 00:31:10 I used the plural, but yeah....will change the ip_mask Mar 07 00:32:13 are all optional? Mar 07 00:32:30 so that you can have a rule with e.g. only action deny and an ipaddr Mar 07 00:32:37 no they are all required Mar 07 00:32:40 with ports set to 0-65535 Mar 07 00:32:46 but there are no defaults Mar 07 00:33:52 mb__: was it intentional that you removed the ip address of the loopback interface? Mar 07 00:34:19 yeah it defaults to that Mar 07 00:34:43 ah okay Mar 07 00:34:53 could probably remove the netmask as well Mar 07 00:40:31 mb * r25918 /packages/utils/maemo-kexec/files/sbin/bootmaemo: maemo-kexec: Kick the watchdog before kexec-ing to make sure it doesn't expire before the new OS kernel was loaded. Mar 07 00:45:56 xMff: http://paste.openwrt.org/m35a35b7c Mar 07 00:46:16 oh, wait, I was going to test that first...sorry Mar 07 00:47:01 haha Mar 07 00:51:17 I'd propose this: "http://paste.openwrt.org/d3a1fbe0" Mar 07 00:51:41 this makes all vars optional Mar 07 00:52:53 I think the default action should be deny, but ok Mar 07 00:53:15 yes, default deny Mar 07 01:13:08 xMff: it doesn't like upnp_get_port_range...it's returning blank....should there be an eval in there somewhere to turn Mar 07 01:13:08 "${_var}_start=$val into (e.g) int_start=333 ? Mar 07 01:15:40 no I misspelled it Mar 07 01:15:44 val -> _val Mar 07 01:16:06 in the export statements Mar 07 01:16:07 doh, I didn't catch it either Mar 07 01:18:19 and yet another typo Mar 07 01:18:39 "${_var}_end=${val##[:-]}" -> "${_var}_end=${_val##*[:-]}" Mar 07 01:18:51 was missing the asterisk in the substitution pattern Mar 07 01:19:40 no that was my fault.. Mar 07 01:20:33 it was right in what you posted Mar 07 01:20:50 ok Mar 07 01:25:13 build #84 of ps3 is complete: Failure [failed compile_3] Build details are at http://tksite.gotdns.org:8010/builders/ps3/builds/84 Mar 07 01:35:35 it lives...let me post what we've got now Mar 07 01:36:50 xMff: http://paste.openwrt.org/m3d192ae8 Mar 07 01:39:13 I also turned NAT-PMP and secure mode on by default Mar 07 01:39:30 this looks fine so far Mar 07 01:39:44 now one thing that bothers me is the four different choices to disable it :) Mar 07 01:40:15 ok three Mar 07 01:40:28 there is 1) init script disable 2) option enabled 3) both upnp and natpmp disabled Mar 07 01:40:51 I think we should nuke 2 in favor to 3 Mar 07 01:41:58 ok, so make natmp and upnp both disabled by default and eliminate option disable, or have them enabled by default (and still nuke option disable) ? Mar 07 01:42:09 enabled + nuked Mar 07 01:42:14 ok Mar 07 01:42:44 the default config file is deny everything anyway, so it should be okay Mar 07 01:42:50 I bet an attempt to start the deamon with both disabled would result in an runtime error anyway Mar 07 01:43:00 probably Mar 07 01:46:03 this will eliminate one of the most common bugs you get about luci :) Mar 07 01:46:41 I'm no fan of those extra enable/disable flags anyway Mar 07 01:47:18 I mean the global ones that basically mimic /e/i/foo disable Mar 07 01:47:36 yeah Mar 07 01:48:13 The only one that really needs it is ddns-scripts and that's because it doesn't have an init.d Mar 07 01:53:58 http://paste.openwrt.org/d181e9f0 Mar 07 01:54:08 no more global disable Mar 07 01:56:39 okay, I have some more remarks Mar 07 01:56:45 ok Mar 07 01:57:11 hm or maybe its not worth it Mar 07 01:57:27 what's that? Mar 07 01:57:51 was about to suggest to put the if foo = 0/1 then echo yes or echo no stuff into a procedure Mar 07 01:58:40 yeah, I wasn't happy about that, but I don't know how much you'd actually save Mar 07 02:00:37 there's four instances so it' s borderline I think Mar 07 02:01:32 ah, what the heck I'll do it Mar 07 02:02:33 http://paste.openwrt.org/d22afd5f4 Mar 07 02:03:34 maybe with a third optional arg to override the config option name Mar 07 02:04:09 like this: http://paste.openwrt.org/d69026bef Mar 07 02:05:28 you want to try to get system uptime too? Mar 07 02:05:41 default 1 instead of 0 Mar 07 02:05:51 hm? Mar 07 02:06:36 http://paste.openwrt.org/d31b71ec3 is in there too Mar 07 02:07:00 maybe you should call the uci var like this Mar 07 02:07:08 option system_uptime Mar 07 02:07:17 okay Mar 07 02:10:54 can you do variable inside variable ${x:-${y}} ? Mar 07 02:11:07 nbd * r25919 /trunk/package/ppp/patches/ (5 files): ppp: add a few patches to reduce the package size Mar 07 02:11:07 yeah Mar 07 02:11:48 while that work with msh? Mar 07 02:11:57 or do we care? Mar 07 02:12:05 no we don't :) Mar 07 02:16:37 what about http://paste.openwrt.org/d64af3590 ? Mar 07 02:17:22 ah reverse logic? Mar 07 02:17:30 why is that needed? Mar 07 02:17:57 for system_uptime Mar 07 02:18:05 it defaults to 1 Mar 07 02:18:14 1 = 1 -> yes Mar 07 02:18:28 yes Mar 07 02:19:05 and "$defval" is always 1 as far as I can see Mar 07 02:19:26 0 unless other wise specifiec Mar 07 02:19:38 ${4:-0} Mar 07 02:20:18 yeah but I don't get it Mar 07 02:20:33 or do you mean I've change the other ones to default to one? Mar 07 02:20:39 for some reason I thought they were 0 Mar 07 02:20:39 no Mar 07 02:21:12 option foo 1 -> config_get_bool foo config foo 0 -> 1 -> yes Mar 07 02:21:25 option foo 0 -> config_get_bool foo config foo 0 -> 0 -> no Mar 07 02:21:51 same with 1 as default Mar 07 02:22:04 the default is used if the option is not set at all Mar 07 02:22:58 only system uptime isn't Mar 07 02:23:02 oh I see how I confused myself Mar 07 02:23:14 -gt 0 vs -eq 1 Mar 07 02:23:17 yes Mar 07 02:23:22 essentially the same Mar 07 02:23:43 right Mar 07 02:50:29 oh, and should we change the default download/upload to big numbers instead of usually too small? Mar 07 02:50:44 nbd * r25920 /trunk/ (5 files in 4 dirs): hotplug2: replace udevtrigger from the udev-106 source with a much smaller self-contained, stripped-down version Mar 07 02:53:23 xMff: http://paste.openwrt.org/dd072405 works with your function now. Any other comments? **** ENDING LOGGING AT Mon Mar 07 02:59:57 2011