**** BEGIN LOGGING AT Tue Aug 14 02:59:58 2012 Aug 14 03:31:27 build #53 of x86 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/53 Aug 14 05:51:26 build #61 of atheros is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/atheros/builds/61 Aug 14 05:55:40 build #58 of pxcab is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/pxcab/builds/58 Aug 14 06:02:18 build #59 of brcm47xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/brcm47xx/builds/59 Aug 14 06:05:43 build #57 of rb532 is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/rb532/builds/57 Aug 14 06:09:24 build #57 of ppc44x is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/ppc44x/builds/57 Aug 14 06:16:50 build #54 of x86 is complete: Failure [failed compile_1] Build details are at http://buildbot.openwrt.org:8010/builders/x86/builds/54 Aug 14 06:31:21 build #63 of orion is complete: Failure [failed compile_10] Build details are at http://buildbot.openwrt.org:8010/builders/orion/builds/63 Aug 14 06:35:31 build #48 of sibyte is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/sibyte/builds/48 Aug 14 08:03:22 ping nbd Aug 14 08:03:45 regarding antenna configuration profiles Aug 14 08:04:10 and how to improve that... Aug 14 08:22:15 ping dangole Aug 14 08:34:02 pong jow_laptop Aug 14 08:43:00 can you please try this patch instead? http://luci.subsignal.org/~jow/0001-package-iwinfo-transparently-handle-radioX-names-in-.patch Aug 14 08:43:35 I know it is not much different but a little bit cleaner Aug 14 08:47:32 reading it, won't static char phyname[IFNAMSIZ] be only reserved while in function madwifi_phyname? returning a pointer to a local variable is dangerous afaik, that's why i declared it in the calling function rather than in unify_ifname/madwifi_phyname Aug 14 08:47:48 thats why its static :) Aug 14 08:48:13 so it's reserved newly every time madwifi_phyname is called?! Aug 14 08:48:33 no, its one buffer, allocated once, which gets reused Aug 14 08:48:52 ok, so as long as we are not using iwinfo multi-threaeded, we are good with that Aug 14 08:49:08 it is not threadsafe :) Aug 14 08:49:20 I mean large parts of the rest of the library aren't either Aug 14 08:49:53 i'll give it a try and tell you how it goes in a few minutes Aug 14 08:53:30 (compiling the image...) Aug 14 08:54:49 blogic * r33178 /packages/multimedia/imagemagick/Makefile: [packages] fixes imagemagick Aug 14 09:07:32 (flashing image...) Aug 14 09:19:15 jow_laptop: kinda works. i can edit wifi interfaces, and scan/join in LuCI Aug 14 09:19:20 "add" doesn't work Aug 14 09:19:20 /usr/lib/lua/luci/dispatcher.lua:447: Failed to execute arcombine dispatcher target for entry '/admin/network/wireless/radio1.network2'. Aug 14 09:19:21 The called action terminated with an exception: Aug 14 09:19:21 /usr/lib/lua/luci/model/cbi/admin_network/wifi.lua:169: bad argument #1 to 'ipairs' (table expected, got nil) Aug 14 09:19:21 stack traceback: Aug 14 09:19:22 [C]: in function 'assert' Aug 14 09:19:24 /usr/lib/lua/luci/dispatcher.lua:447: in function 'dispatch' Aug 14 09:19:28 /usr/lib/lua/luci/dispatcher.lua:195: in function Aug 14 09:20:56 after Save & Apply it works and i can edit both vif's of radio1 Aug 14 09:37:57 build #49 of xburst is complete: Failure [failed compile_2] Build details are at http://buildbot.openwrt.org:8010/builders/xburst/builds/49 Aug 14 09:51:23 dangole: thats not due to the patch I suppose Aug 14 09:51:38 we rather need to patch the freqlist operation too to accept radioX ifaces Aug 14 10:13:40 blogic * r33179 /trunk/package/arptables/patches/ (. 100-always_optimize.patch): [arptables] fix breakage when compiling for armv6 Aug 14 10:23:24 <_trine> Jow thinking more on rssileds ,,, it would be a great idea to include in the LUCI gui a graphical signal strength meter Aug 14 10:33:42 jow * r33180 /trunk/package/iwinfo/src/iwinfo_madwifi.c: [package] iwinfo: transparently handle radioX names in madwifi backend Aug 14 10:33:44 jow * r33181 /trunk/tools/wrt350nv2-builder/ (Makefile src/ioapi.h): [tools] wrt350nv2-builder: don't use internal zlib macro OF() in ioapi.h, it breaks with new zlib releases Aug 14 10:48:11 jow_laptop: http://patchwork.openwrt.org/patch/2519/ and http://patchwork.openwrt.org/patch/2521/ are obsoleted by your work now Aug 14 11:13:16 dangole: is lua -liwinfo -e 'print(iwinfo.madwifi.freqlist("radio0"))' working for you if the wifi is completely unconfigured? Aug 14 11:13:59 BasicXP: I've too a look at your translation patch, most of it is fine but there are a few minor issues Aug 14 11:14:28 BasicXP: for example I really do not like separating stuff like KBit/s into Kbit and s Aug 14 11:14:44 BasicXP: it should remain as one atomic string which can be translÃated as a whole Aug 14 11:14:59 okay Aug 14 11:15:22 i'll fix it Aug 14 11:17:53 and I am not sure whether we should translate things like "B" Aug 14 11:18:22 even if it would need an e.g. cyrillic sign, the char before it would still be latin Aug 14 11:18:51 (referring to the %.2mB pattern which you turned into %.2m<%:B%> Aug 14 11:26:49 hm Aug 14 11:27:43 so format("%.2mB", 1024) -> 1MB Aug 14 11:27:51 is there no way to make those M,K,G translatable? or you think it's better to just keep the latin? Aug 14 11:27:52 so format("%.2mB", 1024*1024) -> 1GB Aug 14 11:27:54 etc. Aug 14 11:28:15 not easily, I'd prefer to keep them in latin for now Aug 14 11:31:04 and what's your opinion about modules/admin-full/luasrc/view/admin_status/bandwidth.htm? Aug 14 11:31:12 there's a return String.format("%f %s/s%s(%f %s/s)", Aug 14 11:31:21 here the seconds are separate Aug 14 11:31:41 that's why I thought of having kbit/mbit and seconds separate everywhere else Aug 14 11:31:49 so there are less "duplicate" strings Aug 14 11:31:57 i.e. that basically mean the same Aug 14 11:32:42 in bandwidth.htm I'd put the /s next to the kbyte/kbit unit Aug 14 11:32:53 <%:kB%> -> <%:kB/s%> Aug 14 11:33:20 and remove the /s from the format pattern further down Aug 14 11:34:11 alright, i'll fix that as well Aug 14 11:34:17 anything else i should consider? Aug 14 11:34:25 stills crolling through Aug 14 11:35:17 translating "MCS" is debatable, its a proper noun Aug 14 11:35:54 the "Nslookup" translation from diagnostics.htm has a slight syntax error (missing %) Aug 14 11:37:49 the rest looks fine, good work Aug 14 11:38:06 thank you Aug 14 11:38:14 ah, maybe you should treat "SNR %d dBm" as one term Aug 14 11:38:17 i'll fix all this stuff and send you the link when i'm done Aug 14 11:38:36 In german it would read like "%d dB Signalabstand" Aug 14 11:39:02 well, SNR is an abbreviation, just like MCS Aug 14 11:39:13 maybe there's no need to make them translatable at all then? Aug 14 11:39:18 maybe Aug 14 11:39:30 so jsut leave SNR and MCS Aug 14 11:39:43 * BasicXP nnods Aug 14 11:39:46 * nods Aug 14 11:43:00 jow_laptop: is it correct to write "20MHz" or "20 MHz"? Aug 14 11:43:13 good question Aug 14 11:43:26 or GHz, whatever Aug 14 11:43:32 I prefer the compatc form, but that must not necessarily be the right one Aug 14 11:43:38 i think it should be separated by a space Aug 14 11:43:41 just like with dBm Aug 14 11:43:48 or kbit/s Aug 14 11:44:00 but that's my opinion Aug 14 11:44:42 ah yes, one reason I avoided spaces was to prevent line breaks Aug 14 11:45:03 i'll keep it as is for now Aug 14 11:45:07 in tight table cells the browser might decide to break up the amount and the unit Aug 14 11:45:09 looks like it's more correct anyway Aug 14 11:47:27 build #62 of brcm63xx is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/brcm63xx/builds/62 Aug 14 11:48:17 should we translate "WEP"? Aug 14 11:48:21 no Aug 14 11:48:34 same for WPA and WPA2 actually Aug 14 11:48:46 there is "mixed WPA/WPA2" Aug 14 11:48:47 only "mixed WPA/WPA2" Aug 14 11:48:50 that's the only one Aug 14 11:48:50 yup Aug 14 11:48:57 but for the sake of consistency I simply translated all Aug 14 11:49:17 it is no biggy. translators can just copy the input string or leave it empty Aug 14 11:49:26 the template engine will fall back to the input string then Aug 14 11:49:47 I suppose WPA is always called WPA Aug 14 11:49:51 but then it's not consistent with the fact that we can't translate MCS or SNR at all Aug 14 11:50:27 jow_laptop: gives me table: 0x94a1f8 Aug 14 11:50:35 dangole: odd Aug 14 11:50:42 iwinfo radio0 f Aug 14 11:50:45 dangole: the luci crash you've seen happens if it returns nil Aug 14 11:50:46 is alright Aug 14 11:51:38 jow_laptop: it only happens if i use the "add" button in wifi_overview Aug 14 11:51:45 hmm Aug 14 11:51:48 probably because the vif isn't created yet Aug 14 11:51:59 yeah, only staged Aug 14 11:52:04 Failed to execute arcombine dispatcher target for entry '/admin/network/wireless/radio0.network2' Aug 14 11:52:05 what is the url after add? Aug 14 11:52:08 ok Aug 14 11:52:29 http://192.168.1.1/cgi-bin/luci/;stok=2607923708d8a65a93b21934dc9b4ffe/admin/network/wireless/radio0.network2 Aug 14 11:52:59 once i got back, press Save & Apply, the same URL works afterwards Aug 14 11:53:35 i guess that the admin/network/wireless uses a new uci instance instead of reusing the ucis previously used to add the new vif Aug 14 11:54:16 yes Aug 14 11:54:34 but the new cursor should pick up the save but uncommitted values Aug 14 11:55:16 is there any unsaved changes when navigating back fro mthe crashed page? Aug 14 11:57:09 yup, the whole adding of the interface Aug 14 11:57:19 on ath9k it works...? Aug 14 11:57:22 yes Aug 14 11:59:02 what is reported by lua -lwinfo -e 'print(iwinfo.type("radio0"))' during this state? Aug 14 12:00:36 the magic happens here: http://luci.subsignal.org/trac/browser/luci/trunk/libs/sys/luasrc/sys.lua#L831 Aug 14 12:00:47 it gets called with "radio0.network2" as ifname Aug 14 12:01:05 it should end up in line 845 where it falls back to "radio0" Aug 14 12:01:29 then continue to line 860 where it returns a wrapped iwinfo object for the type determined for "radio0" Aug 14 12:06:44 florian * r33182 /packages/net/gatling/Makefile: [package] gatling: fix build failure on ARM platforms Aug 14 12:06:48 <[florian]> "/win 12 Aug 14 12:06:49 <[florian]> doh Aug 14 12:08:16 jow_laptop: modules/admin-full/luasrc/view/admin_network/wifi_join.htm, line 106 Aug 14 12:08:24 shouldn't it be "dBm" instead of "dB"? Aug 14 12:09:33 jow_laptop: says in line 110: local iw = luci.sys.wifi.getiwinfo(arg[1]) Aug 14 12:09:40 and arg[1] is radio0.network2 Aug 14 12:09:50 it's reception signal level as far as i understand Aug 14 12:14:22 figured out, nvm Aug 14 12:17:23 jow_laptop Aug 14 12:17:24 : Aug 14 12:17:25 what about Aug 14 12:17:26 diff --git a/modules/admin-full/luasrc/model/cbi/admin_network/wifi.lua b/modules/admin-full/luasrc/model/cbi/admin_network/wifi.lua Aug 14 12:17:26 index b8e240e..93c53d9 100644 Aug 14 12:17:27 --- a/modules/admin-full/luasrc/model/cbi/admin_network/wifi.lua Aug 14 12:17:29 +++ b/modules/admin-full/luasrc/model/cbi/admin_network/wifi.lua Aug 14 12:17:31 @@ -107,7 +107,7 @@ local function txpower_current(pwr, list) Aug 14 12:17:33 return (list[#list] and list[#list].driver_dbm) or pwr or 0 Aug 14 12:17:35 end Aug 14 12:17:37 Aug 14 12:17:39 -local iw = luci.sys.wifi.getiwinfo(arg[1]) Aug 14 12:17:43 +local iw = luci.sys.wifi.getiwinfo(wdev:name()) Aug 14 12:17:45 local hw_modes = iw.hwmodelist or { } Aug 14 12:17:47 local tx_power_list = txpower_list(iw) Aug 14 12:17:49 local tx_power_cur = txpower_current(wdev:get("txpower"), tx_power_list) Aug 14 12:17:51 makes it work :0 Aug 14 12:17:57 dangole: pastebin, please Aug 14 12:18:05 (next time) Aug 14 12:18:15 KanjiMonser: are you logging all this?! Aug 14 12:18:25 i.e. am i eating your hdd space? Aug 14 12:18:51 never mind, i'll be less lazy next time Aug 14 12:19:11 dangole: that's not the point - pasting multiple lines is just a bit rude and should be avoided Aug 14 12:19:17 dangole: yes I know that it is "radio0.network2" and thgis is perfectly fine and anticipated Aug 14 12:19:56 if wdev:name() makes it work it means that luci.sys.wifi.getiwinfo() is broken and should be fixed Aug 14 12:20:03 how should it get iwinfo for radio0.network2 which doesn't exist yet? Aug 14 12:20:24 where does iwinfo take care of stripping whatever comes after the dot? Aug 14 12:20:43 because it does essentially the same; parse radio0.network2 into d = "radio0" and n = 2 Aug 14 12:20:54 ah ok Aug 14 12:21:00 then try to find wifi-iface #2 on device radio0 Aug 14 12:21:08 and try to read the ifname state var from it Aug 14 12:21:16 if there is no ifname state var then fall back to "radio0" Aug 14 12:21:23 and this fails Aug 14 12:21:36 I think I know why Aug 14 12:23:49 sigh Aug 14 12:23:52 yeah Aug 14 12:24:01 getiwinfo needs two uci states Aug 14 12:24:31 one with savedir set to /tmp/.uci so normal cursor Aug 14 12:24:39 one with savedir set to /var/state Aug 14 12:24:50 then iterate with the normal cursor and read the ifname with the state one Aug 14 12:24:59 I hate state vars Aug 14 12:25:51 there's an easier fix though Aug 14 12:26:36 jow_laptop: here are the updated patches https://basicxp.ru/files/luci Aug 14 12:26:49 http://pastebin.com/NzNzQybd Aug 14 12:26:58 if there's anything else you'll notice, will be glad to fix Aug 14 12:32:13 blogic * r33183 /trunk/target/linux/generic/files/drivers/char/gpio_dev.c: Aug 14 12:32:13 [kernel] fixes gpio_dev un/reload balancing Aug 14 12:32:13 Signed-off-by: SHI Bin Aug 14 12:33:59 BasicXP: translate("None") -> "%s" % translate("none") Aug 14 12:34:53 same for translate("hidden") Aug 14 12:35:02 where are those? Aug 14 12:35:06 what files? Aug 14 12:35:09 wifi_join.htm Aug 14 12:36:19 actually remove the else \n return transalte("None") entriely Aug 14 12:36:51 hm now I am confused... wait Aug 14 12:37:04 maybe it should say Open instead? Aug 14 12:37:37 no, the entire format_wifi_encryption() function is unused Aug 14 12:38:10 remove it? Aug 14 12:39:12 that would make the crypto untranslatable Aug 14 12:39:21 as the entire string is provided by iwinfo Aug 14 12:39:36 it returns something like "mixed WPA/WPA2 (TKIP, CCMP)" Aug 14 12:39:43 but you said it's not used Aug 14 12:39:52 if it's not used, how can it affect anything? Aug 14 12:40:02 i possibly misunderstood you? Aug 14 12:40:09 I am thinking about using it again Aug 14 12:42:02 i'll keep it as is Aug 14 12:42:05 you decide what to do Aug 14 12:42:56 the only thing i'm concerned about is this: Aug 14 12:42:57 <%=net.encryption.description or translate("Open")%> Aug 14 12:43:03 here it says "Open" Aug 14 12:43:08 in the function it says "None" Aug 14 12:43:21 not consistent Aug 14 12:43:27 you choose :) Aug 14 12:45:15 http://luci.subsignal.org/trac/changeset/9012 Aug 14 12:48:01 jow_laptop: if I do svn up at this point, will it overwrite the changes i've made to the files? Aug 14 12:48:50 it attempts to merge them Aug 14 12:49:13 or it will add conflict markers and let you sort it out Aug 14 12:54:13 updated the files, same place Aug 14 12:55:30 looks good now Aug 14 12:55:49 oh joy :) Aug 14 12:56:07 i hope it won't break anything Aug 14 12:56:54 we'll see Aug 14 12:57:18 there is another issue i'd like to report Aug 14 12:57:22 or re-report, actually Aug 14 12:57:34 DHCP and DNS doesn't work in LuCI for me Aug 14 12:57:44 i posted it once here, it was ignored Aug 14 12:58:32 error: http://pastebin.com/9wrwE6vW Aug 14 13:00:42 dhcp config: http://pastebin.com/AvmRtp0M Aug 14 13:01:14 same thing for "Hostnames" tab as well Aug 14 13:01:41 using revision 8897 of LuCI Aug 14 13:02:49 haven't seen fixes for this in any later revision Aug 14 13:03:01 i'll update if you think it's necessary Aug 14 13:05:41 florian * r33184 /packages/net/vtun/Makefile: [package] vtun: fix menuconfig warning Aug 14 13:07:45 <_trine> jow_laptop if you do include rssileds into LUCI could you also think about including a software attenuator which would help with a finer antenna alignment Aug 14 13:09:29 BasicXP: that means /etc/config/dhcp is broken Aug 14 13:09:37 i pasted the config Aug 14 13:09:39 nothing I can fix in luci Aug 14 13:09:50 uci show dhcp >/dev/null Aug 14 13:09:53 look for errors Aug 14 13:10:21 line 32 is the culÃprit most likely Aug 14 13:10:31 dashes are illegal in section names Aug 14 13:10:44 heh, true Aug 14 13:11:32 works now Aug 14 13:11:33 ok, I'm going to commit pootle->svn now Aug 14 13:11:40 then sync pootle from svn Aug 14 13:12:32 ok Aug 14 13:13:18 another thing, "DHCP Leases" table needs a fix for dnsmasq-dhcpv6 Aug 14 13:13:29 IPv6 address entried are wrongly displayed Aug 14 13:13:48 the MAC-Address field contains either a 0 or a number Aug 14 13:13:55 it should show a DUID instead Aug 14 13:14:25 this can be fixed by getting a MAC address from the last column of the /tmp/dhcp.leases file Aug 14 13:16:03 for IPv4 it will still show a MAC address Aug 14 13:16:09 for IPv6 it will show a DUID Aug 14 13:30:35 blogic * r33185 /packages/net/openswan/patches/100-cross_compile.patch: [packages] fix openswan cross compile Aug 14 13:33:36 jow_laptop: are you done with pootle<->svn sync? Aug 14 13:37:38 not yet Aug 14 13:38:02 I rewroked my scripts to use svn info to decide what needs to be committed or updated Aug 14 13:38:18 ah good Aug 14 13:38:20 could you please poke me when you are? thank you Aug 14 13:38:35 PM or whatever Aug 14 13:43:44 BasicXP: done Aug 14 13:45:20 okay Aug 14 13:45:31 i'll begin with the actual translation now Aug 14 13:45:38 thank you very much for help with all this Aug 14 13:48:13 yw Aug 14 13:54:14 jow_laptop: i don't see strings like "GHz" in base.po in pootle, weird Aug 14 13:54:23 I see them here Aug 14 13:54:27 fer german at least Aug 14 13:54:32 *for Aug 14 13:54:49 i might be blind or something Aug 14 13:54:51 crap Aug 14 13:54:57 i downloaded the .pos in a zip Aug 14 13:55:01 not there either Aug 14 13:55:05 i just hit ctrl-c in the wrong erminal and aborted my x86 build Aug 14 13:57:16 BasicXP: now it is there, I manually synced it again Aug 14 13:59:03 mhm, works Aug 14 13:59:13 i mean, it indeed is there now Aug 14 13:59:41 and my translations in pootle got reset again Aug 14 13:59:42 duh Aug 14 13:59:49 whatever, i'll redo them Aug 14 14:00:00 just a couple of strings Aug 14 14:00:04 bad merge logic Aug 14 14:00:07 still wanted to do that anyway Aug 14 14:00:09 yeah Aug 14 14:00:31 somehow russian was the only language that wasn't auto-updated Aug 14 14:00:37 idk why Aug 14 14:01:55 * BasicXP shrugs Aug 14 14:01:59 mystery :) Aug 14 15:05:53 jow_laptop: found a typo, patch: http://pastebin.com/jP0nvLsv Aug 14 15:16:19 committed Aug 14 15:17:15 thanks Aug 14 15:17:21 i'll look for more Aug 14 15:17:42 after that, i'm afraid, translations will have to be resynced again Aug 14 15:28:26 blogic * r33187 /packages/net/kamailio3/patches/ (. 100-cross_compile.patch): [packages] kamailio3, fix cross compile Aug 14 15:40:54 build #49 of avr32 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org:8010/builders/avr32/builds/49 Aug 14 16:19:32 florian * r33188 /trunk/target/linux/mcs814x/ (2 files in 2 dirs): Aug 14 16:19:32 [mcs814x] MCS8140 actually selects CPU_ARM926T Aug 14 16:19:32 Should make inclusion for other MCS814x SoCs easier Aug 14 16:35:46 jow_laptop: in iptables in luci, in phrases such as: "Chain nat_reflection_fwd (References: 1)", what does a reference number mean? Aug 14 16:37:11 how often other rules point to this chain Aug 14 16:37:33 every iptables ... -j chain increases the refcount of "chain" by 1 Aug 14 16:38:16 except for terminal ones like ACCEPT, REJECT or DROP Aug 14 16:40:47 thanks Aug 14 16:41:45 should I send base.po to you when it's done? or should I try n submit it to pootle? Aug 14 16:42:11 you translate it on your pc right now Aug 14 16:42:15 ? Aug 14 16:42:18 yep, in poedit Aug 14 16:42:31 then let me quickly push my base.po syncs after the typo fix Aug 14 16:42:39 before your upload it Aug 14 16:42:52 i'll make a quick fix in the file itself Aug 14 16:42:56 it's just one byte heh Aug 14 16:43:39 i'm not finished yet myself, so don't be in a hurry Aug 14 16:44:34 did the same locally Aug 14 16:44:54 I fixed the orig strings inplace with sed, to preserve the existing translations (which do not have this typo) Aug 14 16:45:42 great Aug 14 17:13:03 hi Aug 14 17:13:24 i removed my stagingdir and builddir and now i get this error when i try to update my feeds Aug 14 17:13:27 /bin/sh: /mnt/stuff/Openwrt/trunk_git/staging_dir/host/bin/find: No such file or directory Aug 14 17:13:34 WARNING: No feed for package 'libc' found, maybe it's already part of the standard packages? Aug 14 17:13:40 and errors like this Aug 14 17:13:43 how can i fix this? Aug 14 17:14:00 ahh rm everything in tmp - fixed Aug 14 17:16:58 build #41 of iop32x is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/iop32x/builds/41 Aug 14 17:19:19 build #41 of etrax is complete: Failure [failed shell_11] Build details are at http://buildbot.openwrt.org:8010/builders/etrax/builds/41 Aug 14 17:20:54 dwmw2_gone: if we had known you are coming round tonight, we would have make some tea Aug 14 17:20:58 ;) Aug 14 17:21:04 s/make/made/ Aug 14 17:51:32 jow_laptop: did as much as i could (2 untranslated, 14 fuzzy) https://basicxp.ru/files/luci/base.po Aug 14 17:51:39 should i upload it to pootle? Aug 14 17:51:58 other files will follow sooner or later Aug 14 17:52:06 as soon as i'm done with them Aug 14 17:58:24 yes Aug 14 18:00:24 it doesn't merge properly Aug 14 18:00:32 it uses old translations Aug 14 18:00:35 grr Aug 14 18:06:42 jow_laptop: http://translate.sourceforge.net/wiki/pootle/offline_translation#merging_methods Aug 14 18:06:48 there is an overwrite method in pootle Aug 14 18:07:00 but it is either not enabled or appeared in a newer version Aug 14 18:08:05 as i can see it's possible to disable it in user permissions Aug 14 18:08:06 please retry Aug 14 18:08:14 I granted overwrite permissions to registered users Aug 14 18:09:16 do i have to relogin? Aug 14 18:09:22 not sure Aug 14 18:09:27 because i don't see a third radiobutton Aug 14 18:09:36 sigh, I need to do it for each language Aug 14 18:09:57 ah okay Aug 14 18:09:57 retry Aug 14 18:10:03 enabled it for ru Aug 14 18:10:25 maybe it's per-project set as well? Aug 14 18:10:33 because there is no third radiobutton still Aug 14 18:11:11 it appeared Aug 14 18:11:19 let me check one thing... Aug 14 18:11:40 hah Aug 14 18:11:55 if you set web interface language to, say, Russian, the third radio button disappears Aug 14 18:12:01 once you set it back to english, it works fine Aug 14 18:12:14 maybe I need to restart apache Aug 14 18:12:22 a lot of stuff is cached in mod_wsgi Aug 14 18:12:27 and it's gone again Aug 14 18:12:31 what in the world? o.o Aug 14 18:12:47 even when changed back to english Aug 14 18:12:53 maybe update to pootle 2.1.6? Aug 14 18:13:04 thats easier said than done :) Aug 14 18:13:15 I restarted apache, see if something changes Aug 14 18:13:37 overwrote! yay! Aug 14 18:13:39 where do you upload anyway? I didn't find it Aug 14 18:14:05 http://luci.subsignal.org/pootle/ru/luci/base.po Aug 14 18:14:06 as pootle admin I feel a bit betrayed :) Aug 14 18:14:23 don't see any upload option Aug 14 18:14:34 ah, you mean the upload field Aug 14 18:14:43 http://luci.subsignal.org/pootle/ru/luci/ Aug 14 18:14:45 scroll down Aug 14 18:14:48 "Upload file" Aug 14 18:14:58 ah... sigh Aug 14 18:16:04 i think enabling overwriting for all registered users is a bit dangerous though Aug 14 18:16:12 translations could be erased Aug 14 18:16:19 or damaged Aug 14 18:16:53 I granted the permissions only to you now Aug 14 18:17:34 the pootle wokrflow is highly confusing Aug 14 18:17:41 unfortunately there seems to be no real alternative Aug 14 18:18:08 well, it's better than nothing anyway hehe Aug 14 18:18:16 and thank you Aug 14 18:23:32 blogic * r33189 /packages/net/freeswitch/patches/010-ppc_asm.patch: [packages] freeswitch - diable asm instructions on ppc as they dont compile Aug 14 18:33:08 asterisk.po has to be removed from the translation system, as it's empty Aug 14 18:33:38 yes, will look into that later Aug 14 18:34:53 well, there are some #~ translations in that file Aug 14 18:49:22 blogic * r33190 /packages/multimedia/imagemagick/Makefile: [packages] fix imagemagick when xorg feed is staged Aug 14 19:09:24 blogic * r33191 /packages/libs/libdmapsharing/patches/ (. 001-diable_pixbuf.patch): [packages] disable gdk_pixbuf when building libdmapsharing Aug 14 19:15:00 <_trine> http://pastebin.com/fuuGUESk Aug 14 19:15:02 <_trine> http://pastebin.com/z9CwraDS Aug 14 19:15:13 <_trine> my nano is half crashed Aug 14 19:15:44 <_trine> I could do with finding out what has eaten the ram Aug 14 19:16:26 <_trine> used 16400 free 13080 buffers 1851 buffers 1852 Aug 14 19:16:52 <_trine> total 29480 Aug 14 19:17:14 <_trine> but not much will run Aug 14 19:22:20 anyone know why I need rpath? http://pastebin.com/nghw2dBc Aug 14 19:22:32 I'm trying to crosscompile useing the toolchain built in staging, Aug 14 19:22:58 it works if I add -lcrypto and -lssl to my own makefile, but I was hoping that would be pulled in by the -lmosquitto (which depends on openssl) Aug 14 19:23:25 is this a problem with my build of the libmosquitto binary? (I've just been updating the package for openwrt, and it adds ssl) Aug 14 19:29:48 or, is the package built ok, and I should just be using LD_LIBRARY_PATH to help indicate where the extra libraries might be? Aug 14 19:31:45 is the pkconfig file staged ? Aug 14 19:31:51 of libmosquito Aug 14 19:33:46 karlp: you could just add EXTRA_LDFLAGS+= -Wl,-rpath-link,$(STAGING_DIR)/usr/lib Aug 14 19:34:22 and then make sure the deps are set properly otherwise owrt will refuse to package the binaries Aug 14 19:35:03 deps were already doing the DEPENDS=+libopenssl, so that worked. Aug 14 19:35:10 let me see if there even is a pkgconfig Aug 14 19:35:24 this package didn't have any dependencies before Aug 14 19:36:06 right, no pkgconfig, Aug 14 19:36:19