Thursday, 2010-08-05

*** mralex has quit IRC00:10
*** ssalevan has quit IRC00:50
*** nyergler_ has quit IRC01:30
*** papyromancer is now known as papy|zZzZz01:44
*** Odemia has quit IRC01:45
*** Odemia has joined #cc01:48
*** karlcow has quit IRC01:54
*** karlcow has joined #cc01:56
*** Odemia has quit IRC02:04
*** Odemia has joined #cc02:06
*** Odemia has quit IRC02:21
*** Odemia has joined #cc02:24
*** Kaetemi has quit IRC02:32
*** Odemia has quit IRC02:38
*** Odemia has joined #cc02:46
*** ArneBab has joined #cc03:03
*** Odemia has quit IRC03:26
*** ArneBab has quit IRC04:32
*** ArneBab has joined #cc04:37
*** Odemia has joined #cc04:58
*** bassel has quit IRC05:15
*** rcoup has quit IRC06:01
*** oshani has joined #cc06:07
*** oshani has left #cc06:08
*** Kaetemi has joined #cc06:16
*** erlehmann has joined #cc08:18
*** wormsxulla has quit IRC08:24
*** wormsxulla has joined #cc08:44
*** rcoup has joined #cc08:58
*** wormsxulla has quit IRC09:01
*** thisisparker has joined #cc09:04
*** ArneBab has quit IRC09:06
*** bassel has joined #cc09:13
*** wormsxulla has joined #cc09:56
*** wormsxulla has quit IRC10:52
*** wormsxulla has joined #cc10:53
*** MarkDude has quit IRC11:23
*** rcoup has left #cc11:25
*** MarkDude has joined #cc11:30
*** Odemia has quit IRC11:45
*** wormsxulla has quit IRC11:45
*** wormsxulla has joined #cc11:46
*** wormsxulla has quit IRC11:53
*** wormsxulla has joined #cc11:57
*** tvol has joined #cc12:06
*** sudiptamondal has joined #cc12:07
*** sudiptamondal has quit IRC12:26
*** wormsxulla has quit IRC12:42
*** wormsxulla has joined #cc12:44
*** bassel has quit IRC12:45
*** JoiIto has joined #cc14:04
*** wormsxulla has quit IRC14:35
*** ArneBab has joined #cc14:42
*** oshani has joined #cc14:48
*** wormsxulla has joined #cc14:58
*** Odemia has joined #cc15:09
*** papy|zZzZz is now known as papyromancer15:45
*** bassel has joined #cc15:55
*** nyergler has joined #cc15:57
*** MarkDude has quit IRC16:32
*** wormsxulla has quit IRC16:37
*** JED3 has quit IRC16:38
*** MarkDude has joined #cc16:48
*** MarkDude has quit IRC16:52
*** MarkDude has joined #cc17:00
*** wormsxulla has joined #cc17:13
*** oshani has quit IRC17:20
*** JED3 has joined #cc17:28
*** thisisparker has quit IRC17:37
*** karlushi has quit IRC17:37
JED3paroneayea: pung17:39
*** mralex has joined #cc17:45
*** MarkDude has quit IRC17:47
*** karlushi has joined #cc17:48
*** bassel has quit IRC17:55
*** bassel has joined #cc17:56
*** MarkDude has joined #cc18:01
paroneayeaJED3: pong18:09
JED3paroneayea: hey, you busy?18:09
paroneayeaJED3: nope18:09
JED3nm, trying to figure out the perf issues in the api18:09
JED3this jurisdictions isn't playing nice18:10
paroneayeamore context?18:11
JED3paroneayea: well first, i discovered this while trying to debug the issue
JED3so that license has a dc:replaceby in the rdf, i'm not i know exactly what that means18:17
paroneayeaoh dc:isReplacedBy18:19
JED3it doesn't behave like the other licenses using that18:19
paroneayeaI'm not really sure myself18:19
JED3 for example is valid18:19
JED3but not
paroneayeaah I bet I know why18:20
paroneayeaone sec let me test18:20
paroneayeaJED3: for some reason that license is missing its licenseClass18:25
paroneayeawhich means that when the cc.license code checks to see if any of the selectors have that license18:25
paroneayeait doesn't find it because no selector reports it as existing18:25
JED3paroneayea: ahh okay18:26
paroneayeaso we should fix that and add the licenseClass :)18:26
paroneayeaI wonder if any other licenses are missing their licenseClass...18:26
erlehmannwhich one of you is the API guy?18:29
JED3heh lord18:29
JED3erlehmann: it is i18:30
erlehmannJED3, would it hurt if under $api_url . '/locales' the locales came with some localized labels?18:30
JED3erlehmann: that's been requested before, i'll look into it again18:32
erlehmannit would certainly be helpful for my gsoc plugin. no sane user will be selecting zh_TW18:32
JED3erlehmann: but everyone else is perfectly capable of getting those labels themselves :)18:33
paroneayeaJED3: without looking, I am guessing: is the reason the api is having some speed issues due to librdf slow queries?18:33
JED3paroneayea: thats my only idea at this point18:33
JED3the jurisdictions querys namely18:33
erlehmannJED3, also, is /support/jurisdictions the right place to parse jurisdictions out of?18:33
paroneayeathat whole multiple-conditions-makes-your-query-as-slow-as-molasses18:33
JED3erlehmann: for a complete list of jurisdictions, yes18:34
erlehmannfine. i used it.18:34
JED3erlehmann: however if you're using it for a UI, look at /support/jurisdictions.js18:34
erlehmannJED3, i parse it myself and then build a UI.18:34
JED3paroneayea: hehe, i ended up not using the query remember?18:34
erlehmannand document.write is not nice anyway18:35
paroneayeaoh right18:35
JED3our version of librdf doesnt support negation using sparql18:35
paroneayeapython filtering :)18:35
JED3paroneayea: yup18:35
erlehmannalso, i don't have whole URLs for jurisdictions, but simple identifier. i take the stuff between the last two slashes AND HOPE IT DOESNT CHANGE18:35
JED3erlehmann: for the website url of a jurisdiction, you'll need to use the /license/standard/jurisdiction/XX API call18:37
JED3paroneayea: have you witnessed inconsistent responses times in RDF.Query?18:54
paroneayeaJED3: hm not sure18:54
paroneayeaI think I've mostly observed a long-query short-query divid18:55
*** erlehmann is now known as erschlafmann18:55
JED3paroneayea: but not inconsistent response times for the same query?18:55
paroneayeaI haven't noticed it... that's not to say it doesn't happen18:56
*** jed_ has joined #cc18:58
*** nyergler_ has joined #cc18:59
*** nyergler_ has joined #cc18:59
*** JED3 has quit IRC19:00
*** nyergler has quit IRC19:00
karlushi removed the use of Creative Commons Licenses19:01
karlushi>Who owns the concepts?19:01
karlushi>The Contributor(s) own the concept(s) generated on OpenIDEO which they non-exclusively license to the Challenge Host for possible publication.19:01
karlushithey added in their prose19:01
karlushi> At the end, a top concept is chosen. All concepts generated are shareable, remix-able, and reusable in a similar way to 'creative commons'.19:01
karlushi Terms of Use as yesterday do not mention Creative Commons19:02
karlushiSo instead of clarifying the use of creative commons, they decided to not use it. *sigh*19:03
paroneayeahey jed_19:29
*** jed_ is now known as JED319:29
paroneayeaor JED3 :)19:30
JED3jed_ is so professional sounding you can call me JED3 :)19:30
paroneayealooks like I might have found out why that happened with that license.19:30
JED3jed_ is my alter-ego for when the office connection has a case of the hiccups19:31
JED3paroneayea: o rly?19:31
paroneayeathe cc:licenseClass doesn't appear in index.rdf but it does appear in the file19:31
paroneayeacheck out:19:31
paroneayeathe reason this got goofed up was that the file is split into two pieces, a cc:License and a rdf:Description19:33
JED3paroneayea: yeah, i see the licenseclass19:33
paroneayeafor some reason no other files are split up like this19:33
paroneayeaanyway when it gets merged into index.rdf it only pulls out one of them, and so the data in cc:License is missed apparently (!)19:34
paroneayeaanyway I can fix this file and re-merge index.rdf19:35
JED3paroneayea: awesome19:36
JED3paroneayea: still can't figure out why the api is running like a turtle on dope when called for jurisdictions19:40
*** JED3 is now known as JED3-lunch19:40
paroneayeahaha turtles on dope19:42
paroneayeaI like how you just talked about avoiding unprofessional usernames then said "running like a turtle on dope"19:42
paroneayeathat's awesome19:43
paroneayeaI love #cc19:43
*** bassel has quit IRC20:09
JED3-lunchparoneayea: :D20:10
paroneayeaJED3-lunch: it should be fixed... also uploaded a new package20:17
*** JED3-lunch is now known as JED320:18
JED3paroneayea: thank you20:18
paroneayeaJED3: yup :)20:18
paroneayeaJED3: oh man20:20
paroneayeathere are more problems with this rdf file20:20
paroneayeaseveral assertions inside it that are by-nd-nc instead of by-nc-nd20:20
paroneayeanyergler_: yo20:20
nyergler_paroneayea, hola20:21
paroneayeaso if you read our previous conversation, there was an rdf file that was a bit misformed20:21
paroneayeafor cc by-nc-nd 2.0 jp20:21
*** bassel has joined #cc20:22
paroneayeahowever, I'm also noticing that it has several assertions that are ordered like by-nd-nc20:22
paroneayeaso I'm wondering if there's any intentional reason this is this way20:22
paroneayeaor if it's an error I should correct (as I assume)20:22
paroneayeacurrent result is that:20:22
nyergler_paroneayea, it just looks broken to me20:22
paroneayeaah ok20:22
paroneayeanyergler_: also, it strikes me that it wouldn't be hard to do something with license.rdf that's like what we're doing with cc.i18n20:24
paroneayeawhere there are automatically generated tarballs20:24
paroneayeamaybe that's overthinking though20:24
paroneayeardf doesn't change as much20:24
nyergler_paroneayea, true cc.i18n has the additional property that it needs to be post-processed before making the tarball (to make MO files)20:25
nyergler_i suppose we merge the RDF20:26
nyergler_hrm, don't know20:26
paroneayeaI wasn't thinking of auto-merging the index.rdf but that's a good idea also ;020:27
nyergler_paroneayea, i need to step away for a second, not sure if anything here needs a decision20:28
nyergler_(i'd say hold off on auto-building rdf ATM)20:28
*** bassel has quit IRC20:28
paroneayeaoh no need for a decision now, just pondering20:28
*** MarkDude has quit IRC20:33
*** MarkDude has joined #cc20:49
*** jgay has joined #cc20:54
*** jgay has quit IRC21:06
*** MarkDude has quit IRC21:08
JED3paroneayea: pyng21:10
paroneayeaJED3: pong21:11
JED3paroneayea: more weirdness21:12
JED3>>> cc.license.by_uri('')21:13
JED3<License object ''>21:13
JED3>>> cc.license.by_uri('').title('en')21:13
JED3>>> cc.license.by_uri('').title('en')21:13
JED3u'Attribution-NonCommercial 3.0 Spain'21:13
paroneayealet's see..21:14
JED3brazil's rdf must be funky21:14
paroneayeaJED3: no i18n title in brazil's RDF21:15
JED3paroneayea: are you sure?21:15
JED3my index.rdf has dc:title's for br21:15
paroneayeaJED3: I mean, no dc:title xml:lang="i18n"21:15
paroneayeathe "template" for translations21:16
paroneayeasee other files21:16
JED3ahh okay21:16
JED3paroneayea: which file exactly?21:17
paroneayeafor whatever reason, these licenses are missing the i18n template assertion21:17
JED3oh nm i see what you're saying21:17
paroneayeathey are all taiwan and br21:18
JED3hmm so whyd it fail?21:18
paroneayeaso these are the files that were at an earlier time in the license.rdf-git/licenses/ directory21:19
paroneayeathey were added after the new packaging efforts21:19
paroneayeaer after the retooling21:19
paroneayeathere were thus a number of issues with them, apparently this being one of them21:20
paroneayeatrying to figure out how I added the i18n lang assertion in the first place, one sec21:20
paroneayeaah I see... tools/gen_i18n_titles.py21:22
paroneayeaI'll generate and add those.21:22
JED3paroneayea: okay, thx21:25
nyergler_a little more important, do we have any idea why i18n isn't happening on deeds?21:26
paroneayeanyergler_: what??21:26
nyergler_see your email21:27
nyergler_paroneayea, see my reply, just sent with a little more information.21:29
paroneayeaso I am guessing that this is relevant to the move from cc.i18npkg to cc.i18n21:32
nyergler_seems like a safe assumption21:32
paroneayeaI could quick-rever to before that though there have also been a few important bugfixes the last few days.21:33
nyergler_I'd prefer to know what happened if we can look @ it21:34
paroneayeayeah I'll do that21:34
paroneayeaor, am doing it now ;)21:34
*** Kaetemi has quit IRC21:38
JED3nkinkade: ping21:41
*** erschlafmann has quit IRC21:42
*** MarkDude has joined #cc21:45
paroneayeahm.  This is strange.21:45
paroneayeaoh man.21:45
paroneayeaI think I know what happened.21:46
nkinkadeJED3: Hi.21:46
paroneayeanyergler_: okay I think I know what happened21:47
nyergler_paroneayea, is that a good or a bad "i think i know" :)21:47
paroneayeait's a "this isn't too hard to fix if I'm right"21:47
JED3nkinkade: hey, so the API slow downs, how long do they normally last (how long is the api unresponsive) and how often does it happen?21:47
paroneayeaso remember how I said the defaultLanguages in jurisdictions.rdf were out of date21:47
paroneayeaso I was going to import from licenses.xml21:47
paroneayeawell what changed here is:21:47
*** tvol has quit IRC21:47
paroneayea(Pdb) license.jurisdiction.default_language21:47
paroneayeaused to be es-mx21:47
paroneayeaI think what happened was I didn't convert them from es_MX to es-mx style.21:48
nkinkadeJED3: I'm not sure.  I'm not polling all day long.21:48
nyergler_ah, so when you imported from licenses.xml, we didn't convert _ to -?21:48
nyergler_paroneayea, sounds plausible21:48
paroneayeashould only take me a second to confirm...21:48
nyergler_paroneayea, why don't you edit mx by hand on staging and see if that fixes it there?21:49
nkinkadeI just know that Mike first wrote to webmaster@, or caught me here on #cc, to say that the international/ pages were showing PHP errors and not loading fully.21:49
nkinkadeJED3: ^^21:49
paroneayeanyergler_: ah good idea21:49
nkinkadeSince then I've noticed it too, and earlier today I could hardly get a single intl. page to load properly.21:49
JED3nkinkade: damn, okay thanks21:49
JED3nkinkade: so do the errors normally look like the one you took a screen capture of?
nkinkadeJED3: I've seen many like that, but actually earlier today sometimes Varnish would just timeout and give an error.21:52
JED3nkinkade: okay, then its not just a specific call failing or slowing down, its the entire api21:53
*** erschlafmann has joined #cc21:55
*** erlehmann has joined #cc21:58
nkinkadeJED3: For what it's worth, the international/xx pages seem to be returning quickly now.21:58
nkinkadeEarlier that wasn't the case.21:58
JED3nkinkade: yeah, which is why i'm asking these questions21:59
JED3"psh its working just fine, mike and kinkade are crazy"21:59
nkinkadeThat makes me wonder if it's a load issue.  But when this stuff happens I log into both a5 and a8 and see nothing out of the ordinary.21:59
*** erschlafmann has quit IRC21:59
nkinkadeAt least nothing that top displays.21:59
nkinkadeI've even looked at network traffic to see if it's higher than normal.22:00
JED3nkinkade: and nothing telling?22:00
nkinkadeJED3: Yeah, nothing.  This morning when I had the issue a8 was pushing out 1.2MB/sec approx. Now it's at 1.03MB/sec.22:01
paroneayeanyergler_: yup that's it:
nyergler_paroneayea, aweomse22:04
nyergler_paroneayea, can you do a pass over the RDF files and fix those?22:04
paroneayeaworking on fixing up the RDF and making another release now.22:04
nyergler_just saw the commit22:28
nyergler_looks like we're adding some lines with the underscore style; ie,22:28
nyergler_+    <dc:title xml:lang="en_CA">Romania</dc:title>22:28
nyergler_don't know if that's right or wrong for xml:lang, just wanted to double check22:28
paroneayeayeah that's not technically changed in this but22:28
paroneayeayou are right22:28
paroneayeathere is that problem22:29
*** erlehmann has quit IRC22:29
*** ichi has joined #cc22:29
*** ichi has left #cc22:29
nyergler_so yes, a problem, but a different one?22:29
paroneayeaI think maybe the same mistake may be being made by the translation machinery?22:29
paroneayeaof not converting it22:29
nyergler_ok, well let's fix the deed issue first :)22:30
*** erlehmann has joined #cc22:30
paroneayeacache wiped...22:31
nyergler_paroneayea, awesome, thanks!22:32
paroneayeasorry about that lapse22:33
*** erlehmann_ has joined #cc22:34
*** erlehmann has quit IRC22:36
paroneayeanyergler_: also I just checked and it turns out this lapse caused a failure in cc.license's tests22:36
paroneayeawe should get buildbot or something running soon :)22:36
nyergler_paroneayea, so in other words, CI would have saved our ass here22:36
paroneayeawe'll say yay to an alternate dimension where cc tech team already had CI working, and this problem never happened22:37
paroneayeamaybe in that alternate dimension they accidentally took down the server for 4 hours today though for completely other reasons22:37
paroneayeathose twerps.22:37
nyergler_all high and mighty with the CI server22:38
nyergler_but still fucking things up22:38
*** erlehmann__ has joined #cc22:44
JED3paroneayea: pinf22:45
paroneayeaJED3: pong22:46
*** erlehmann_ has quit IRC22:46
JED3paroneayea: sigh nevermind22:49
paroneayealol ok22:49
JED3the change in license.rdf for "Babel>1.0" stopped a buildout22:49
nyergler_nkinkade, you still around?22:49
JED3but its because i dont have the basket link in my cfg22:50
nkinkadenyergler_: Yeah.22:50
JED3can we go ahead and move the eggs from your a9 dir and onto the basket?22:50
JED3how much work would that be?22:50
paroneayeanot too much work.  just need to copy them over and make sure that all buildout.cfg's pointing at my dir are updated for basket22:51
paroneayeawhich most are I think22:51
paroneayeaI updated every one I thought needed to be updated just a couple of days ago22:51
JED3mine wasn't ;)22:51
paroneayeabut I'm not sure22:52
paroneayeathat's why I didn't wipe the directory :)22:52
paroneayeaalso I accidentally just uploaded a couple more cc.licenserdf's there22:52
JED3okay cool well i'll make sure the scraper and the api are up to date?22:52
paroneayeaso those need to be copied over22:52
JED3whoops premature ? mark22:52
paroneayeaJED3: sure.22:52
paroneayeaI gotta go actually for a bit22:52
*** MarkDude has quit IRC22:53
paroneayeathen I'm going to email akila about last steps for gsoc22:53
paroneayeathen I'm out till tomorrow22:53
*** erlehmann__ has quit IRC22:54
paroneayeaactually before I hang up here..22:55
paroneayeanyergler_: I did a buildout and that test is still failing but for a different, related reason22:55
nyergler_paroneayea, ok, can you email/open ticket/something so that it doesn't get lost?22:56
nyergler_i understand it's EOB for you22:56
*** MarkDude has joined #cc23:03
*** erlehmann__ has joined #cc23:08
*** JoiIto has quit IRC23:26
*** nyergler_ has quit IRC23:44

Generated by 2.6 by Marius Gedminas - find it at!