Thursday, 2010-11-11

akozakthat is awesome!00:00
akozakgridworks was fun to work with00:00
akozaki was hoping that theyd use their freebase acquisition well00:01
akozakJED3, http://code.google.com/p/google-refine/wiki/RecipeApacheLogs00:02
JED3akozak: trying this recipe out now00:03
akozakwowww this looks cool http://code.google.com/p/google-refine/wiki/FetchingURLsFromWebServices00:04
akozaki mean, nothing you couldnt do in a few lines of python00:04
akozakbut still cool00:04
*** kreynen has joined #cc00:05
*** akozak has quit IRC00:37
*** kreynen has quit IRC00:37
*** bassel has joined #cc00:37
*** kreynen has joined #cc00:42
*** JoiIto has quit IRC00:44
*** JoiIto has joined #cc00:44
*** wormsxulla has quit IRC01:18
*** oshani has joined #cc01:25
*** wormsxulla has joined #cc01:32
*** kreynen has quit IRC01:38
*** CC_Hudson has quit IRC01:38
*** CC_Hudson has joined #cc01:43
*** CC_Hudson has quit IRC01:46
*** CC_Hudson has joined #cc01:57
*** JED3 has quit IRC01:58
*** paulproteus has quit IRC02:07
*** pyrak_ has quit IRC02:08
*** pyrak has joined #cc02:08
*** paulproteus has joined #cc02:08
*** mralex has quit IRC02:10
*** mralex has joined #cc02:11
*** oshani has quit IRC02:25
*** nkinkade has quit IRC03:01
*** erlehmann has quit IRC03:51
*** erlehmann has joined #cc03:52
*** erlehmann has quit IRC04:09
*** kreynen has joined #cc04:13
*** DNS777 has quit IRC04:14
*** DNS777 has joined #cc04:14
*** kreynen has quit IRC04:32
*** TDJACR has quit IRC07:20
*** TDJACR has joined #cc07:26
*** thisisparker has joined #cc08:30
*** DNS777 has quit IRC09:00
*** DNS777 has joined #cc09:15
*** DNS777 has quit IRC09:15
*** DNS777 has joined #cc09:15
*** xelo has joined #cc10:55
*** xelo has left #cc10:55
*** pmiller has joined #cc12:31
*** pmiller has left #cc12:31
*** tvol has joined #cc13:12
*** kreynen has joined #cc13:23
*** tvol has quit IRC13:37
*** wormsxulla has quit IRC14:13
*** ghelleks has joined #cc14:19
*** DNS777 has quit IRC14:20
*** DNS777 has joined #cc14:21
*** papy|afk has quit IRC14:47
*** nkinkade has joined #cc15:02
*** erlehmann has joined #cc15:38
*** jgay has joined #cc15:42
*** wormsxulla has joined #cc15:56
*** oshani has joined #cc16:24
*** isforinsects has quit IRC16:25
*** isforinsects has joined #cc16:25
paroneayeabuilding out cc.engine on live16:28
paroneayeain case you notice a momentary downtime.16:28
*** akozak has joined #cc16:46
nyerglerparoneayea: just saw the ticket comment re: wsgi_cache17:01
nyerglerare we leaving the cron job in place for a while to make sure no more zero size files are written?17:01
paroneayeayes17:02
paroneayeawe'll determine our success based on whether or not that thing complains17:03
paroneayeanyergler: ^17:03
nyerglerthanks17:03
paroneayeayurp17:03
*** thisisparker has quit IRC17:05
*** kreynen has quit IRC17:07
*** ghelleks has quit IRC17:18
nyergler:)17:19
paroneayea:O17:19
paroneayeaughhhhhh17:21
paroneayealooks like it didn't fix it17:21
paroneayea.17:21
paroneayea:(17:21
nkinkadeparoneayea: Did you have a chance to look at that PO I sent you?17:24
nkinkadeIn terms of using it with sync.17:24
*** ghelleks has joined #cc17:24
paroneayeaI looked at it but didn't try using it with sync :)17:26
paroneayeaI'll try now17:26
*** kreynen has joined #cc17:39
*** akozak has quit IRC17:39
*** akozak_ has joined #cc17:39
*** akozak_ is now known as akozak17:39
*** JED3 has joined #cc17:42
*** thisisparker has joined #cc17:45
*** mralex has joined #cc17:50
CC_HudsonProject cc.api build (18): UNSTABLE in 16 min: http://code.creativecommons.org/hudson/job/cc.api/18/17:52
CC_Hudson* JED3: Added docstrings to the support functions17:52
CC_Hudson* JED3: Moving PDM support into the API17:52
CC_HudsonYippie, build fixed!18:07
CC_HudsonProject cc.api build (19): FIXED in 15 min: http://code.creativecommons.org/hudson/job/cc.api/19/18:07
CC_Hudson* JED3: Fix POST details call param passing18:07
CC_Hudson* JED3: Include deprecation information in license results18:07
CC_Hudson* JED3: Update docs to include deprecation responses18:07
CC_Hudson* JED3: Added docstrings to the support functions18:07
CC_Hudson* JED3: Moving PDM support into the API18:07
JED3mralex: ping18:15
mralexJED3: 'ola18:15
JED3mralex: how difficult would it be to wire GA to record when users select the copy-and-paste attribution text on the deeds?18:16
mralexJED3: if we can capture the keypress, then not hard at all18:17
JED3mralex: would that capture right-click menu "Copy" actions?18:17
mralexJED3: i'm not sure if there's a javascript event for that18:18
JED3i think capturing that.click would be enough18:18
mralexwe can definitely capture the textarea onclick18:18
mralexand/or onfocus18:18
JED3and its fairly simple to do?18:18
mralexyeah18:18
mralexi have a similar thing set up already to capture clicks to search, support, and wiki that are otherwise crossdomain18:19
JED3okay, i guess we could add it to the deedscraper's deed.js file18:20
akozakyou might be able to approximate copies by measuring select click events and then whether they exited the page or then deselected the text18:20
akozaklike if they selected and left vs. selected and then deselected18:21
akozakor something18:21
akozaki don't know18:21
mralexwe can easily handle onfocus and onblur18:21
*** DNS777 has quit IRC18:31
*** DNS777 has joined #cc18:31
mralexJED3: FWIW, webkit has some support for copy and paste JS events http://developer.apple.com/library/safari/#documentation/AppleApplications/Conceptual/SafariJSProgTopics/Tasks/CopyAndPaste.html18:45
JED3mralex: o neat, i think i'm just going to listen for onclick though :/18:53
JED3mralex: do you think its fair to assume that those click, are probably copy-and-pasting its contents?18:53
JED3maybe onselect is better18:53
JED3onclick="this.select()" onselect="log dis event" ?18:54
akozakJED3, why not log it all?18:56
mralexJED3: it's probably fair to assume a click event will result in a copy, so just handle everything in the onclick18:57
akozakyou could probably calculate the percentage of clicks are a copy event. since webkit supports copy events, just log all onclicks and see if you've also logged a copy18:58
akozakyou might find e.g. that 70% of clicks also occured with a copy18:58
akozakand then you could apply that data to non-webkit18:59
akozakbecause youd have a reasonable approximation of how many people are copying  the text when they click18:59
JED3akozak: thats also much more coding and points of failure :P19:00
akozakJED3, true19:00
akozakbut if you just log everything, the data would exist19:00
mralexJED3: it's an extra line or two in deed.js, not really a failure point. wouldn't hurt to log it.19:03
JED3mralex: have a sec?19:20
JED3mralex: nm, found what i was looking for19:22
paroneayeankinkade: btw, you were right19:44
paroneayea./bin/sync doesn't fix the problem19:44
nkinkadeparoneayea: Did it seem to just overwrite the whole file to you?19:44
nkinkadeIt did to me.19:44
paroneayeayup19:44
nkinkadeWhat astounded me about that was: where in heck did it get the old file?19:44
nkinkadeDoes sync, pull down the version controlled version or something?19:44
paroneayeait doesn't!19:45
nkinkadeSo if we overwrote the old PO file, then were did it get all the old translations from?19:45
paroneayeaI have no idea.19:45
nkinkadehaha19:45
nkinkadeMan, that's comforting! :-)19:45
nkinkadeparoneayea: So I think my route is going to be msgmerge, and then trying to just replace the existing file with the merged one, or something like that.19:46
paroneayeaok sounds good19:46
paroneayeaI'm pretty astounded by this19:46
paroneayeawhen you said it was doing that19:47
paroneayeaI thought it must have been a mangled version19:48
paroneayeaI have no idea how this could "end up" as the same file19:48
paroneayeaunless it's just dumping all their changes...19:48
*** jgay has quit IRC19:53
paroneayeaoh hey, we're featured frontpage on http://www.transifex.net20:39
akozakesperanto? ಠ_ಠ21:24
paroneayeayup ;)21:48
nyerglernkinkade: paroneayea: not sure it's relevant, but sync keeps a copy of the old master.po file around so it can match up strings where the english text changes21:48
nyergler(since it converts from po to "cc" style)21:48
nyerglernkinkade: can you reset my passwd on a6? i think this is left-over from the great-reset-of-2010, but i have no idea what it is right now21:49
nkinkadenyergler:  Done.21:50
nkinkadenyergler: The odd thing about what sync seemed to do was that msgstrs that were no longer in the file were put back in there, as if it had a backup of the es/cc_org.po file somewhere.21:51
nyerglernkinkade: well it has two different copies -- one with CC-style keys, one with "po-style" keys21:51
nyerglerso they may be coming from there21:51
nkinkadeAh.  That was it, then.21:51
*** oshani has quit IRC21:51
nkinkadeSo what I forgot is that sync works from cc-style -> english-key, not the other way around.21:52
nkinkadeWe have an english-key PO file and want to merge it with the cc-style one.  nyergler, paroneayea: is there any conceivable way to make that happen?21:53
nyerglernkinkade: has the cc-style one been changed, too?21:53
nyerglerwhy not just do ./bin/cc2po (iirc) and then use msgmerge on the resulting english key files21:54
nyerglerthen ./bin/po2cc21:54
*** ghelleks has quit IRC21:54
paroneayeawe could try that though21:54
paroneayeathis looks like the right kind of file I think...21:55
nkinkadeparoneayea: Are you testing that?21:55
paroneayeankinkade: what were some of the strings that changed?21:55
nkinkadeparoneayea: If you find the string "Recono" anywhere in the resulting PO file then it's not right.21:56
paroneayeaI'll try it but21:57
paroneayeamsgctxt "util.copyright_date"21:57
paroneayeamsgid "copyright date"21:57
paroneayeamsgstr "fecha de los derechos de autor"21:57
paroneayeaeg, it's the same way for that as in the file we're trying to copy over21:57
paroneayeaas with the "new" file21:57
nkinkadeparoneayea: The big thing to look for is that the word Reconocimiento is replaced by Atribución in the license names and elsewhere.21:58
paroneayeakk21:59
*** ghelleks has joined #cc22:01
*** MarkDude has joined #cc22:08
* paroneayea deflates22:12
paroneayeanyergler: nkinkade: not working... only about 5 messages were "converted"... as I thought, this was mostly already in the right format to begin with22:13
paroneayeastill not sure what was wrong22:13
paroneayeaI wonder if in hand-editing it they broke the file?22:13
nyergleris this the combined spanish file that's been detached for weeks?22:14
nkinkadeHmm.  Let me poke around a bit.  I don't want to have to ask the CC Chile people to retranslate everything again, if possible.22:14
paroneayeanyergler: yes22:14
nkinkadenyergler: Yeah, my guess is the PO file they have is from a couple months ago.22:14
nkinkade(just a guess)22:14
nkinkadeApparently they had some issue with Transifex, didn't mention it to me, and just downloaded the PO and did it offline, and are only now, quite some time later giving it back to me.22:15
nyergleri see22:15
nyerglerso we're trying to make sure the new strings get added to it, and the appropriate things are marked as fuzzy?22:15
nyergler(ie, what happened when you uploade to Tx?)22:16
nyerglersorry, don't want to dig into this too deep, just noticing the IRC conversation goes back almost 3 hours22:16
paroneayeanyergler: yes22:18
paroneayeatrying to merge their changes in22:18
nkinkadenyergler: I'm going to try to upload it to Tx again.22:18
nkinkadeBefore, I got an error about DOS-style newlines, so I just need to fix that and try again, which I'd hadn't got around to again since yesterday.22:18
*** erlehmann has quit IRC22:19
nyergleroh, so we don't actually know what Tx will do with it?22:19
paroneayeanorp22:21
nkinkadeUgh: From transifex: "Commit to the repository failed. Details from the VCS backend: ''."22:22
nkinkadeparoneayea: ^^22:22
paroneayea.__.22:22
paroneayeais that in relation to your uploading the file?22:22
nkinkadeparoneayea: Yeah.22:23
nkinkadePresumably Tx merged it, then tried to push to git.22:23
paroneayeawell transifex has been committing all day...22:23
nkinkadeCan we make git be more chatty?22:23
paroneayeanot on our end22:23
paroneayeaafaik22:23
nkinkadeApparently git said: ''22:23
nkinkadeWhich isn't very useful.22:23
paroneayeagit says '… wtf is this'22:24
paroneayea'get out of my repository'22:24
nkinkadeparoneayea: http://code.creativecommons.org/viewgit/i18n.git/commit/?id=544ef6de5d4ad8b9889f7a41f694be8bbff2afaa22:29
paroneayeankinkade: success? ;o22:30
nkinkadeThe git error I got before was because I was trying to upload the old PO file.22:30
paroneayeaoh, this is the msgmerged one?22:30
nkinkadeThis time I uploaded the PO I created with msgmerge and it worked.22:30
paroneayeahorray!22:30
nkinkadeBut only after I removed the DOS newlines from a few places in the file.22:30
*** figaroo has joined #cc22:31
*** figaroo has left #cc22:31
paroneayea     *22:35
paroneayea    /\22:35
paroneayea   /o \      PARTY HATS22:35
paroneayea  /o  o\22:35
paroneayea ========_22:35
paroneayea  \____.--'22:35
paroneayeaI wonder what their problem with transifex was22:36
paroneayeamaybe shitty git merge issues! ;D22:36
paroneayeaIt's goomba time22:38
nkinkadenyergler: What was opened.CC.org about?  Was opened the old discovered?22:40
nyerglerno, it was a wiki run under cclearn22:40
*** bassel has quit IRC22:41
paroneayea     ___            ___            ___22:41
paroneayea   _/   \_        _/   \_        _/   \_        ,/-\,22:41
paroneayea   /\___/\        /\___/\        /\___/\       / o_||22:41
paroneayea  |  | |  |      |  | |  |      |  | |  |     | .,  '_.----,22:41
paroneayea  \ ^___^ /      \ ^___^ /      \ ^___^ /      \// /// \,/ \\22:41
paroneayea   ------__       ------__       ------__        \||/\,/ \.//.22:41
paroneayea   _(___(__)      _(___(__)      _(___(__)        \\./_\_/_//22:41
paroneayea  (__)           (__)           (__)             /_'-------L\22:41
paroneayea                      IT'S GOOMBA TIME22:42
paroneayeanyergler: so I'm thinking for this;22:44
paroneayeahttp://code.creativecommons.org/issues/issue66722:45
nkinkadenyergler: Do we care to keep redirecting it to wiki:/OER ...  I imagine so, but just wondering where to drop the vhost for that redirect.22:45
paroneayeaI'll just have it generate a csv in cc/i18n/statistics.csv22:45
nkinkadeI can just pick something.  It's not that important.22:45
paroneayeaand do that at tarball-time22:45
nyerglernkinkade: yes, keep redirecting22:45
nyerglerno opinion re: location22:45
nkinkadeYeah, I'll just pick something reasonable.22:46
nyerglerparoneayea: and then we'll expose those stats somehow?22:46
paroneayeanyergler: and then cc.engine will use pkg_resources to grab the file22:46
paroneayeaand parse it22:46
paroneayeato find out which ones are which percentag22:46
paroneayeae22:46
nyerglerseems like the parsing/etc is more appropriately located in cc.i18n22:47
nyergleror whatever its called22:47
nyerglerright?22:47
paroneayeasure, I can make a cc/i18n/util:get_statistics() function22:48
paroneayeawhich returns a dictionary22:48
nyerglergreat22:48
*** bassel has joined #cc23:11
mralexakozak: does IE, any version of, support the OpenSearch?23:19
mralex(i presume no)23:19
mralexoh, google tells me it does23:21
akozakmralex, yes23:24
akozak8+ IIRC23:24
mralexakozak: 7+ :) updating my css/js madlibs to account for IE.23:24
mralexhttp://labs.creativecommons.org/demos/search/search-i18n/index-i18n.php will say different things depending on what browser/version you have23:25
akozakah cool!23:25
akozakbtw the box shadow looks sort of bad in chrome23:26
akozakthe shadow doesn't have a radius23:26
mralexthey look fine in my version of chrome23:26
akozakhmmm23:26
akozakmralex, http://i.imgur.com/JGdOK.png23:27
akozak6.0.472.55 beta23:27
mralexlol weird23:27
mralexthat's supposed to be rounded cornered23:27
akozakthe bottom is23:28
mralexwhich is why the shadow is rounded23:28
akozakbutton23:28
akozaklooks fine if i remove -webkit-box-shadow23:28
mralexsure, but you seem to have a browser issue causing the corners of that box to not be rounded23:29
akozakhm weird23:30
JED3nkinkade: the column `payment_instrument_id` in civicrm_contribution indicates which processor handled the contribution right?23:32
*** oshani has joined #cc23:33
JED3does a NULL value for that column mean it was manually entered?23:33
*** oshani has quit IRC23:34
*** akozak_ has joined #cc23:35
nkinkadeJED3: I don't think so.23:38
nkinkadeI think payment instrument might be like "credit card", "check", "cash", etc.23:39
nkinkadepresumably there's a civicrm_payment_instrument table with the corresponding textual values.23:39
nkinkadeThere is no certain way to know the whether it was PayPal or GC.23:39
nkinkadeThere is an indirect way and that is be looking at the invoice_id in civicrm_contribution.23:40
nkinkadeThe invoice_ids for PayPal will be longer and the ones for GC will be 15 chars, all caps.23:40
*** akozak has quit IRC23:41
*** MarkDude has quit IRC23:44
JED3nkinkade: thx! i'm able to filter out the google one's based on the length23:50
nkinkadeI used that method just this afternoon to filter out some Google payments while cleaning up some data from a recent bug.  I think should be 100% reliable, though I don't like it in principle.23:57
nkinkadeOkay, I'm off to meet Jen for dinner.23:58
JED3nkinkade: have fun23:59
JED3nkinkade: thx23:59

Generated by irclog2html.py 2.6 by Marius Gedminas - find it at mg.pov.lt!