=== npmccallum [~npmccallu@69-162-252-7.ironoh.adelphia.net] has joined #launchpad
=== cprov_ [~cprov@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== cprov_ is now known as cprov
=== cprov_ [~cprov@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== cprov_ is now known as cprov
=== cprov_ [~cprov@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== elmo_ [~james@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== debonzi [~debonzi@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
debonzilifeless, hi.. is the PQM blocked again?01:44
debonzielmo, ping01:47
debonzielmo_, ping01:47
elmo_I'll check01:57
elmo_but you guys should get some sleep01:57
elmo_yeah, lalo broke it01:58
=== cprov [~cprov@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
debonzielmo_, first I need to get some cprov stuf :)02:05
debonzielmo_, and you guy, before sleep, please fix it :)02:05
elmo_I can't dude, my connection keeps dieing02:09
elmo_is yours working?02:09
debonzielmo_, yes.. seems to be02:10
debonzielmo_, Im using the cable on02:10
lifelessI'll fix it02:22
=== debonzi [~debonzi@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== cprov [~cprov@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== elmo_ [~james@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad
=== justdave [~dave@] has left #launchpad ["Poof"]
=== justdave [~dave@] has joined #launchpad
lifelessI don't have time to be futzing with launchpad at the moment.04:12
lifelessis there stuff from Marks edict that is needed for the canonical.arch subdir or canonical.soyuz subdir ?04:12
cprovlifeless: btw, RAW runs with some hacks on imports but I still missing doc/guides04:30
cprovlifeless: you're welcome04:31
stublifeless: canonica.soyuz probably. Don't know about canonical.arch.06:12
stubI'm not actually sure what problem he is trying to solve, so I'm not sure what exactly needs to be done.06:13
lifelessI've got the monthly ACS meeting tonight.06:14
lifelessoh bah. ignore that, I'll just ring SteveA tonight.06:15
=== mdz [~mdz@69-167-148-207.vnnyca.adelphia.net] has joined #launchpad
=== BradB|London [~bradb@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
=== cprov [~cprov@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
=== stub [~stub@dsl-] has joined #launchpad
=== stub [~stub@dsl-] has joined #launchpad
cprovstub: morning, I need your suggestion about how to integrate a full DB dump with real packages imported in 10:36
stubI don't think we can do that with the new system Mark setup for the sample data - it will have to all be rekeyed somehow or tossed (at least those tables being used in production - this may not be a problem as it is probably better sample data)10:38
cprovstub: I've already done something on this direction, have you seen the current.sql ?10:39
stubYes, but I don't know what changes you are talking about10:40
cprovstub: we have now the Ubuntu universe imported to soyuz context using a nice script called gina10:42
cprovstub: soyuz context = source/bin/person/etc10:43
stubYes - but did you suck that in from the production database dump?10:43
stubI'm not sure what your original question was about now.10:44
=== lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
=== kiko [~kiko@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
cprovstub: just to clarify, I removed all package info from the original DB, then insert everithing from the real universe, now I have a dump of this table (filled)10:46
kikowe have a 130 (or so) MB file10:47
stubIf you removed all package info from he original DB, then wouldn't you have destroyed the Malone sample data while you are at it?10:48
cprovstub: the question is : How to integrate it as our Production DB or whatever ?10:48
stubSo we want to install this on Emperor?10:48
cprovstub: nop, I leave you  mozilla(dummy) sourcepackage there :)10:48
kikostub, it still needs some validation, but maybe10:49
kikothe question is how to populate new databases -- do we use this data?10:49
stubTo install it on emperor, I would need to run a script that inserts records into the database but doesn't remove anything.10:49
kikobecause it's a *lot* of data10:49
stubOh... so this is for the dev database, not production?10:50
kikothat's doable, we can isolate copy froms there10:50
=== debonzi [~debonzi@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
kikostill for devs10:50
kikowhen it goes into production, we'll have gina running in a cronjob10:50
kikoit can incrementally process packages10:50
stubok. I think I'm on the correct page now :-)10:50
stubI wouldn't want that entire database installed by default, so running 'make' in database/schema should just import current.sql.10:51
stubRunning 'make full' could import the large dataset.10:51
kikobut I still want to make sure it's all good data10:51
kikoor maybe make soyuz-extra10:51
stubDoes that sound usable?10:51
kikogiven that "full" is a bit ungeneric10:52
cprovoopps ... we already have soyuz-extra :)10:52
kikoI don't know, should full bring more data?10:52
kikois our target having more blobs of rosetta/malone data?10:52
kikostub, I have 13,000 packages for you :)10:52
stubThe 'full' target can have dependancies on 'full-soyuz' etc.10:53
kikocome to think of it10:53
cprovyeah, now the current.sql is enough for all the other apps( it's suposed to be) the real imported data is like a alpha DB 10:53
stubso 'make full-soyuz' would just import the extra soyuz data and 'make full' extra soyuz data and perhaps some rosetta stuff.10:53
stubHmm... actually...10:54
stubIt would be nice if this wasn't stored in the launchpad--devel branch so people don't have to download it.10:55
cprovstub: look nice to me ... but we still having 130 Mb files (:-0)10:55
cprovstub: off course : |10:56
stubSo it is a seperate package you can enable in your configs/canonical.com/launchpad/development file that doen't involve modifying launchpad--devel at all ?10:57
cprovstub: or simple download it from some of canonical machines (KISS)10:58
cprovstub: anyway ... do you think it would be possible ? any suggestion ?10:58
stubI would create a new branch designed to be installed into the top level launchpad directory containing a simple makefile and the output of pg_dump11:00
stubAs long as the pg_dump format is the default (text), diffs should make updates efficient.11:01
cprovstub: yes 11:01
kikowe're discussing it with mark & steve11:01
SteveAhi stub11:13
SteveAare you available for a phone call with mark in a bit?11:13
stubHow bit a bit? (Me was thinking of dinner)11:13
SteveAtomorrow morning, we'll be moving interfaces, page templates, and database classes around11:14
SteveAand consolidating them into one place11:14
SteveAwant us to just move the malone ones?11:14
SteveAif so, can you get all your changes checked in by then?11:14
stubMorning UTC you mean?11:14
SteveAstarting 10amutc11:15
SteveAstub: how about you go get dinner, and call mark on his cellphone when you're done?11:15
stubI'll be done in 30 mins11:16
SteveAoh, not 10am utc11:16
SteveA10am BST11:16
SteveAthat's 9am UTC I think11:16
SteveAok, I'll tell mark 30 mins11:16
=== stub sods off
=== carlos [~carlos@69.Red-80-33-181.pooles.rima-tde.net] has joined #launchpad
debonzicarlos, morning11:36
carlosdebonzi: hey, yesterday I saw that we born the same day :-P11:37
carlosnot sure if the same year :-)11:37
debonzirealy.. ? Nice.. :)11:37
kikodebonzi, carlos: spiv and I were born on the same date as well11:37
carlos11st November11:37
carloshmm or it's 11th ?11:38
dilysNew bug 2057 for Launchpad/Rosetta: Hilight interpolations in c-format messages11:39
carlosdaf: ?11:39
carloshmm, that bug will be interesting11:40
dafcarlos: yeah, interesting :)11:53
carlosdaf: I'm still thinking about using the same "special chars" for the new line and space for the entry field11:54
dafcarlos: it's an interesting idea11:56
dafcarlos: would that work through JavaScript?11:56
carlosphone, sorry11:56
=== carlos is back
carlosdaf: Yes, I suppose it should be done with JavaScript11:58
dilysNew bug 2058 for Launchpad/Rosetta: Show template completeness statistics on translation page11:58
carlosshould I file a bug report for limi about it?11:58
=== cprov [~cprov@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
kikowho got this dilys thing?11:59
dafcarlos: file a bug on it, but don't assign it to anyone yet12:00
kikodaf! 12:00
kikodaf rox0r12:00
dilysNew bug 2059 for Launchpad/Database: Better constraints on 'name' columns12:00
kikoI need to file some bugs now12:01
kikoto get the title of top bugspammer12:01
kikonpmccallum, be sure to ping me when you're in 12:01
carloskiko, our personal irc flooder :-P12:04
dilysBug 2045 resolved: Debug skin & authentication don't mix12:05
dilysNew bug 2060 for Launchpad/Rosetta: Glyphs should be changed also in the edition fields12:21
=== lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
dilysBug 1968 resolved: SQLObject needs to use the DEFAULT value for a column as defined in the database12:23
=== sabdfl [~mark@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
sabdflstub: around?12:26
stubsabdfl: Sure12:26
=== elmo_ [~james@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
sabdflstub: will call shortly, higher bandwidth :-)12:27
kikostub, you da man12:29
stubClosing all those bugs except for yours ;)12:32
carlosdaf: the upload link is still there in the alpha server...12:32
dafkiko: http://www.joelonsoftware.com/articles/Unicode.html12:40
dafcarlos: is it linked from anywhere?12:40
carlosdaf: yes12:41
kikooh no12:42
kikoI'm going to have to *understand* unicode now :-(12:42
dafcarlos: at least it doesn't do anything :)12:43
carlosdaf: :-P12:43
kikostub, I'll reassign it12:43
carlosdaf: seems like gettext wraps the lines before we do it:12:48
carlosmsgid ""12:48
carlos-"Failed to locate a program for configuring the date and time. Perhaps none "12:48
carlos-"is installed?"12:48
carlos+"Failed to locate a program for configuring the date and time. Perhaps none is12:48
carlosmsgid ""12:49
carlos-"The use of this key was deprecated in GNOME 2.6 in favour of the 'format' "12:49
carlos-"key. The schema is retained for compatibility with older versions."12:49
carlos+"The use of this key was deprecated in GNOME 2.6 in favour of the 'format' key. "12:49
carlos+"The schema is retained for compatibility with older versions."12:49
dafperhaps the default is off by one?12:49
daf79 instead of 80, or whatever it is?12:49
carlosYes, it seems to be that12:50
carlosI'm going to change it12:50
carlosdaf: I think it's 7812:53
carlos78 + 2 for the char quotes12:53
carlos78 + 2 for the quotes12:53
kikostub, dude12:54
kikopsql:comments.sql:99: ERROR:  relation "scrapedproject" does not exist12:54
kikostub, are those droppings yours? :)12:54
stubspiv: Did you get my comments on the scrapedproject yesterday?12:55
stubkiko: Leave the dropping please12:55
kikookay but it's starting to smell12:56
cprovstub: aha, don't worry he is maniac12:56
kikowho? lalo?12:56
cprovkiko: hehe12:56
stublalo is starting to smell????12:56
kikoyour droppings12:56
kikostub, ping me when you get cprov's request merged, I want it so gina can breathe easier12:59
stubthe gpg one? doing that now.12:59
dafcarlos: so, what are you working on right now?01:00
carlosat this moment I'm reviewing the exported .po files from our alpha server01:01
carloswhen I finish, I will retake the bug #197001:02
kikothanks stub01:04
kikoit is such a pleasure working with you man01:04
kikoyou just get all of our crap moving01:05
kikouhm, I didn't mean it to come out that way01:05
dafcarlos: I think #2058 would be a pretty quick fix01:05
carlosI will look at it then before #197001:06
dafalso, could you comment on #2042?01:07
carloswe should priorize all our bugs for the beta release...01:07
carlosdaf: done, I will answer lalo's question as soon as I finish the po export review01:09
dafcarlos: thanks01:11
stubcprov: What is the value for a 'D' algorithm GPG key? I havn't got the constants in my copy of lp/dbschema.py01:13
cprovstub: I didn't commit yet, but for instance they are attached on bug 205501:14
cprovstub: btw, we need also lower() constraint in Person.name, could you arrange it too ?01:17
stubcprov: It is already there actually01:17
cprovstub: yep, I just saw, sorrry 01:18
carlosdaf: the devel server is down01:23
daf    NameError: name 'UTC_NOW' is not defined                                                                                                                              01:23
daf^^^ that's why01:24
carlosdaf: stub did some changes about it, perhaps you need latest version from rocketfuel01:24
daflatest RF crashes with that error01:24
carlosdaf: then, look at the import section01:25
daflib/canonical/database/constants.py", line 9:01:25
daf    nowUTC = UTC_NOW                                                                                                                                                      01:25
stubMe fix01:26
=== stub really should test 'simple' changes
=== carlos detected a new bug with the export code :-?
carlosdaf: could you do a check for me, please?01:32
dafcarlos: sure01:33
carlosdaf: get a evolution .po file from your rosetta server01:33
daffor which language?01:33
carlosand check that msgid_plural is not always the msgid value01:34
carlosdaf: es.po 01:34
carlosbut I suppose it should be the same always01:34
dafmsgid != msgid_plural01:34
stubkiko, daf: patches with PQM01:34
carlosthe alpha server does it correctly, but my local server has that bug01:34
kikostub, you da rockstar01:34
carlosdaf: I kwonot they are different 01:35
carlos /s/kwonot/know/01:35
dafwhat do you mean by "should be the same always"?01:35
carlosif you have the bug, they should be always the same01:35
carlosmsgid "%d contact"01:35
carlosmsgid_plural "%d contact"01:35
carlosmsgstr[0]  "%d contacto"01:35
carlosmsgstr[1]  "%d contactos"01:35
carloswell, that one is not a good example01:36
carlosbecause it's also broken in the alpha server01:37
carlosso perhaps it's aproblem with sample data01:37
carlosbut this one has the bug:01:37
carlosmsgid ""01:37
carlos"Opening %d contact will open %d new window as well.\n"01:37
carlos"Do you really want to display this contact?"01:37
carlosmsgid_plural ""01:37
carlos"Opening %d contacts will open %d new windows as well.\n"01:37
carlos"Do you really want to display all of these contacts?"01:37
carlosthat's the correct value01:37
carlosand I get:01:37
carlosmsgid ""01:37
carlos"Opening %d contact will open %d new window as well.\n"01:37
carlos"Do you really want to display this contact?"01:37
carlosmsgid_plural ""01:37
carlos"Opening %d contact will open %d new window as well.\n"01:37
carlos"Do you really want to display this contact?"01:37
dafbug! bug!01:38
carlosdaf: could you confirm it? 01:38
=== carlos wants to know if it's in rocketfuel or in his local archive
dafkiko: er, the pot calling the kettle black?01:38
carloskiko: :-D01:38
dafcarlos: should I star-merge from RF before I check?01:39
carlosdaf: no, that way we could try to detect where is the problem01:39
carloscheck it now01:39
carlosif it works correctly, star-merge and if it fails we could look only to the new patchsets you got from rocketfuel01:39
dafare any of our tests failing? :)01:43
carlosdaf: the functional tests should fail...01:43
carloslet me check01:43
carlosdaf: yes, it fails01:44
dilysNew bug 2061 for Launchpad/Launchpad: SQLObject better Error Handler engine 01:45
dafcarlos: ah, good01:48
dafcarlos: it seems to work here01:51
dafmsgid ""01:51
daf"Opening %d contact will open %d new window as well.\n"01:51
daf"Do you really want to display this contact?"01:51
dafmsgid_plural ""01:51
daf"Opening %d contacts will open %d new windows as well.\n"01:51
daf"Do you really want to display all of these contacts?"01:51
dafa regression!01:51
carlosdaf: is it ok for you to merge now with rocketfuel or is it a problem now?01:52
carlosfor you01:52
dafI can do that now01:54
carlosok, thanks01:55
carlosdaf: are you getting the msgsets ordered by sequence?: https://rosetta.shuttleworthfoundation.org/projects/gnome/evolution/evolution-2.0/translate02:00
=== limi [~limi@] has joined #launchpad
=== lalo [~lalo@] has joined #launchpad
dilysNew bug 2062 for Launchpad/Soyuz: dbschema doesn't tell you it wants multiple lines when it dies02:04
carloslalo: hey!, hello02:05
=== lalo :-)
laloalmost two days without internet! that's torture02:10
lalosomeone should pass a law that cutting a person's internet is against human rights02:10
limiit's actually a human right in some european country, can't remember which :)02:11
laloI must move, then.02:12
=== lalo packs up
dafcarlos: yeah02:15
dafcarlos: I'm probably sorting them the wrong way :)02:15
carlosdaf: I think you are not sorting them02:16
dafI'm check02:16
dafI'll check02:16
carlosdaf: Will we have a meeting now or could I go to have lunch?02:16
daflet's have a quick meeting02:17
laloyay meeting02:17
carlosdaf: then something is not working as it should :-P02:18
dafhmm, belay that02:19
dafI think you're right; it's not sorted02:20
dafI was looking at the wrong thing02:20
daf            return RosettaPOMessageSet.select(query)[key] 02:20
carlosdaf: could we start the meeting?02:27
dafok, what have we been working on today?02:28
carlosI was reviewing the export code02:29
carlosand detected two bugs, one fixed, another one is in being debugged (the one I asked you)02:30
dafand you found a bug?02:30
dafwhich was the fixed one?02:30
carlosthe line wrap02:30
dafah, right02:30
carlosI fixed also the functional test because It did not saw the wrap bug02:31
daflalo: what have you been up to? :)02:32
laloI progressed as far as I could without net connection... then I was forced to take 1.5 days off :-)02:33
laloyesterday night I committed what I had02:33
dafI saw your comment on the bug -- can you and Carlos liaise on working out the rosettaUpdates part?02:34
carlosMy next task in my TODO list is answer lalo's question02:35
laloI realized rosettaCount is actually two categories of msgsets02:36
lalo1. msgsets that we have translations for in Rosetta, and were not in the PO02:36
lalo2. msgsets that had no translation in the PO and have some in Rosetta02:36
lalo1 is what I'm counting now, but I don't know how to find 2 with sql02:36
carloslalo: origin02:37
carlosbut we could talk about it after the meeting02:37
=== carlos wants to go to have lunch :-P
=== lalo is having bf *during* meeting :-)
lalowhat more? are we done?02:38
lalodaf: do you want me to move stuff around after we're done with stats?02:38
dafI've been doing various stuff related to the alpha02:38
laloI suppose you answered my email, but I haven't got to it yet02:39
dafI just took on 205702:39
daflalo: I think the people here in London are thinking of doing all the moving, but I'm not sure02:40
dafSteveA: is that correct?02:40
daflalo: at any rate, can you confirm with Steve before you start?02:40
daflalo: I assume you've seen his mail to the list02:41
laloif it's in the last 12h, not yet02:41
SteveAwe will do all the moving here02:42
dafSteveA: great02:42
daflalo: do you have enough tasks to keep you occupied?02:42
SteveAbetween 0900 and 1200 UTC tomorrow02:42
lalodaf: yes02:42
carlosdaf: any chance to get a list of bugs sorted by priority?02:43
lalosince I'm at a later timezone, I can make myself responsible for fixing anything that happens to break after the moving02:43
laloalthough I don't expect anything to break :-)02:43
dafcarlos: why not?02:44
carlosdaf: so?02:45
dafcarlos: well, we can assign priorities to bugs, can't we? :)02:45
=== SteveA pings daf
carlosMy question is more: "Who should do that?" than "How could we do that?"02:46
dafI see02:48
dafwell, don't be afraid to assign priorities to bugs02:48
dafI'll try and use the priority field more myself02:48
daflimi: I know your priority is Malone at the moment, but have you had a chance to look at any of the Rosetta bugs assigned to you?02:50
limidaf: almost finished the autofocus02:51
daflimi: cool!02:51
carlosdaf: anything else?02:53
daflimi: I suspect #1950 would be easy to fix -- I could do it, but when I tried it looked really ugly02:53
dafcarlos: no, I think we're done02:53
dafoh, by the way:02:54
dafI'll be travelling back to Wales this afternoon, so I'll be offline for a few hours02:54
carlossee you later, lunch time02:54
=== lalo finally gets to read work mail
limidaf: looking...03:02
=== limi wants to skin bugzilla :P
dilysNew bug 2063 for Launchpad/Launchpad: prettify "logged in as" thingy03:03
limidaf: yes, should be easy - let me see if I can fit it in today or early monday03:04
dafspiv: so, apparently you can get information on bugs as XML from Bugzilla very easily, but not XML notifications03:04
=== limi is working on Lurker at the moment
daflimi: great!03:04
laloooh, I must be really good at python03:16
kikoeither that or you've had too much mescaline03:18
laloI seem to have been granted access to Guido's time machine03:19
lululimi: Mark is also able to do the upload for you, when you're ready Limi.03:22
lalodaf: do you want to get a say in my priorities, or can I just pick my own way trough it?03:22
daflalo: time machine?03:23
daflalo: I'd like to have a say03:23
laloyou don't know Guido's time machine? man, you're not properly educated in Python lore yet :-)03:23
limilulu: well, it's more involved than an upload03:23
laloit's an old in-joke in the Python community for when a bug is fixed in a commit a few days before it's reported03:24
limiwe need to link in the style sheet in Lurker template etc03:24
lalo(or feature request)03:24
laloGuido took that to a form of art - he sometimes does that with *releases*03:24
dafhaha :)03:26
lalomy tasks for today are: (a) #1975 (make script able to update statistics en masse); (b) move my transaction hack from poimport.py to sqlbase.py; (c) work on the rosettaCount query with Carlos03:26
=== limi is now known as limi|lunch
lalo(c) will be done, I suppose, whenever Carlos can :-)03:26
dafare there bugs open for each of those tasks?03:27
lalo(a) and (c) but not (b)03:30
laloalthough if I made a bug for (b) I could probably make a half-dozen bugs depend on it03:30
lalo[off: and access to the Time Machine is supposedly reserved for the greatest gurus, so when someone demonstrates having access for the first time, it's usually occasion for commemoration ;-)] 03:31
dafwell, if it has a short lifespan, it's probably not worth doing the dependency stuff03:31
daf[what was the bug in this case? :)] 03:31
lalo[#2042, but I didn't fix the bug itself, the time-machine usage was on a technical detail of how to fix it] 03:32
laloactually doing the transaction thing would *fix* a few bugs straight away, I believe - at least one03:33
dafwhich one?03:33
lalo#1986. In fact, I could "turn" #1986 into the bug for this task, with a simple rename.03:35
laloI suppose I'll do that - saying #1986 depends on having transactions is being too anal; I think #1986 *is*, pragmatically, the lack of transactions.03:35
laloI need more clothes, brb03:36
dafthat seems reasonable to me03:37
dafor, as you say, make a new bug and make #1986 depend on it03:38
laloI'll rename #198603:38
dilysNew bug 2064 for Launchpad/Soyuz: Import "chosen ones" as default users with chosen usernames03:40
daflalo: liberate those transactions! :)03:42
lalodude. apps which mess with the window stacking should be taken out and shot. :-/03:42
dafwhich app are you thinking of?03:42
laloI hate it when I'm typing something, then Firefox pops up because it finished loading an app, and my Enter keypress event ends up going to the button labeled "Delete all your important files, send your credit card number to a very visible public location, and mail a death thread to George W. Bush"03:44
lalonot that I have a credit card, but you get my point.03:47
carloswe have a "problem"03:51
carlosthere is no way to calculate the rosettaCount number03:51
carloswith our current database03:51
laloI suspected that :-P03:52
lalonot even, maybe, with a subtransaction?03:52
carlosbecause the sequence will be != 0 if the .po file has a null translation but has the msgid03:52
carloshmmm, perhaps we could play with dates03:52
lalo(09:46:32) lalo: I realized rosettaCount is actually two categories of msgsets03:52
lalo(09:46:47) lalo: 1. msgsets that we have translations for in Rosetta, and were not in the PO03:52
lalo(09:47:03) lalo: 2. msgsets that had no translation in the PO and have some in Rosetta03:52
lalo(09:47:16) lalo: 1 is what I'm counting now, but I don't know how to find 2 with sql03:52
carloslalo: I saw that03:54
laloI know03:54
laloI'm repeating for reference, as it's relevant03:54
carlosthe problem is that we don't have a way to know if that translation was already reimported from the database03:54
laloI'll go make mate, think on, I'll brb03:54
carlosperhaps a datefirstseen == datelastactive could do the work03:55
carlosuntil we split the msgid and the translations03:55
carlosif we add a translation from rosetta they will be the same03:55
carlosif we export a .po and import it again03:55
carlosthe dates will be different03:56
carlosyes, it should do the work03:56
dafperhaps this should be documented on the wiki under RosettaTranslationStates03:58
carloslet me check that I'm right and if it works, we add it :-)03:58
lalocarlos: what if we only imported the .po once?03:59
lalo(and added no translations trough the web)04:00
lalothen datefirstseen == datelastactive04:00
carloslalo: that's the origin field :-)04:00
carlosorigin = 1 --> comes from a .po file04:00
carlosorigin = 2 --> comes from Rosetta04:00
carlosthat field will not change04:00
carlosso datefirstseen == datelastactive && origin == 204:01
carloswell, and inlastrevision=TRUE04:01
lalodaf: when I was writing the code, I detected a conceptual problem04:01
laloit boils down to - when you import a file, you can be either updating from revision control, or submitting translations as an user04:02
lalodo we handle that?04:02
dafwhen we do an import, we currently assume it's from revision control04:02
lalobecause the only place where this makes a difference is statistics04:02
laloso for now we keep assuming that?  If we do (which is fine given alpha/beta constraints), we should say so in the upload page04:03
lalo(I don't know if we already do, I'm just dumping my brain)04:03
dafthe upload page doesn't do anything04:03
laloah, well. Then that's ok :-P04:04
laloback to the problem at hand04:04
dafwhen the upload page is implemented (there's a bug open and a partial implementation in a branch), that assumption will need to change04:04
lalohmm. that would require substantial changes to pofile_adapters.  Would you like me to do that change before I leave?04:05
carlosdaf: I think we decided to handle the web upload as SCM uploads after talking about it with Mark04:05
lalo(like, probably, next week)04:05
dafcarlos: for now, yes04:06
dafcarlos: but in future we should be able to do other uploads too04:06
carloslike xml-rpc ? :-P04:06
dafno, I mean uploads of translator owrk04:06
dafI mean not-from-SCM translations04:07
carlosdaf: if we develop an application that connect to rosetta using xml-rpc (like gtranslator or kbabel..), that's other origin04:07
dafoh, right, yeah04:08
lalook, set 2 is something like: WHERE PotSet.sequence > 0 AND PotSet.primeMsgID = POSet.primeMsgID AND POSet.fuzzy = FALSE AND POSet.sequence > 0 AND (SELECT COUNT from POTranslationSighting POSighting WHERE POSighting.POMessageSet = POSet.id AND POSighting.inLastRevision = TRUE) = 0 AND (SELECT COUNT from POTranslationSighting RosettaSighting WHERE POSighting.POMessageSet = POSet.id AND POSighting.active = TRUE) > 004:12
lalobut the syntax I used for subqueries is pseudolanguage, I don't think that would actually work04:13
=== lalo decides to just test it
carloslalo: the POSet.sequence > 0 should be removed04:14
carlosalso,  (SELECT COUNT from POTranslationSighting POSighting WHERE POSighting.POMessageSet = POSet.id AND POSighting.inLastRevision = TRUE) = 004:14
carlosthat's incorrect04:14
lalothat's what I just said :-)04:15
carlosI mean04:15
carlosit should be > 0 :-)04:15
laloah, ok04:15
laloright, I think I typed on inertia :-)04:15
laloit's not SELECT COUNT, tough - what is it?04:16
carlosSELECT COUNT(*)04:16
laloit's not "=0" - it's correct04:17
lalocheck again04:17
laloyour first correction, it's already the way you say04:17
carlos<carlos> lalo: the POSet.sequence > 0 should be removed ?04:18
laloDUDE, I'M AWESOME :-p04:18
laloWHERE PotSet.sequence > 0 AND PotSet.primeMsgID = POSet.primeMsgID AND POSet.fuzzy = FALSE AND POSet.sequence > 0 AND (SELECT COUNT from POTranslationSighting POSighting WHERE POSighting.POMessageSet = POSet.id AND POSighting.inLastRevision = TRUE) = 0 AND (SELECT COUNT from POTranslationSighting RosettaSighting WHERE POSighting.POMessageSet = POSet.id AND POSighting.active = TRUE) > 004:18
lalooops, pasted the old one again, sorry04:18
laloSELECT COUNT(*) from POMsgSet PotSet, POMsgSet POSet WHERE PotSet.sequence > 0 AND PotSet.primeMsgID = POSet.primeMsgID AND POSet.fuzzy = FALSE AND POSet.sequence > 0 AND (SELECT COUNT(*) from POTranslationSighting POSighting WHERE POSighting.POMsgSet = POSet.id AND POSighting.inLastRevision = TRUE) = 0 AND (SELECT COUNT(*) from POTranslationSighting RosettaSighting WHERE RosettaSighting.POMsgSet = POSet.id AND RosettaSighting.active = TRUE) > 0;04:18
lalothis one runs and prints the right result04:18
laloawesome :-P04:19
carlosPOSet.sequence > 0 <--- That's incorrect04:19
carlosbecause the that pomsgset could be out of the .po file last time we imported it04:19
carlosyou only need to check the potset04:20
laloyes, well04:20
laloI'm looking for only subset 2 here04:20
laloI suppose if I remove that clause I get the entire correct query for rosettaCount, but I haven't looked into that yet04:20
carlosplease, could you explain me this:04:21
carlos(SELECT COUNT(*) from POTranslationSighting POSighting WHERE POSighting.POMsgSet = POSet.id AND POSighting.inLastRevision = TRUE) = 004:21
carlosthe inLastRevision != than "it exists in the .po file"04:22
laloif this message set has any translations that come from SCM, we're not interested on it04:22
carlosyou have a conceptual error04:23
carlosthat's origin != 104:23
carlosor even better origin == 204:23
laloI assume inLastRevision means "it exists in the *latest* import file", which is more useful than "it exists in the .po file"04:23
carlosno, that's not true04:24
carlosthat flag is to know, when we have more than one translation04:24
carloswhich one is the newer one04:24
=== lalo changes the query
lalohmm, no, that's "active" and "dateLastActive"04:25
lalo    inLastRevision = Attribute("True if this sighting is currently in the upstream POFile, otherwise false.")04:26
lalo    origin = Attribute("Where the sighting originally came from.")04:26
carloslet me check...04:26
laloactually it would be wrong to query on origin, because a translation may have been first submitted into Rosetta, then later incorporated into SCM, and origin would still be Rosetta04:26
lalobut I should probably add "active = TRUE" to both subqueries04:27
carloslalo: that's why you need to check that it's not already in a po file we imported04:28
carloshmm 04:28
lalowhich is precisely what I'm doing :-)04:28
carlosI'm not completely sure about all this04:28
lalome neither04:29
lalomy sql level is far from wizard04:30
carlosok, that's what I know about all this (or I thought I know)04:30
carlosdaf: ping04:30
carlosdaf: we need you04:30
lalobut I'm quite sure about the meaning of the columns04:30
dafcarlos: what's the problem?04:30
lalodaf: I suppose you have to scroll back :-) up to "oops, pasted the old one again, sorry", I believe04:31
carlospotranslationsighting.active is equal TRUE when it's a useful translation and equal to FALSE if it's not valid (like a German translation that was imported by error into a Spanish translation)04:32
carlospotranslationsighting.inlasttranslation is equqal TRUE if it's that translation is the latest one revised so we should only export that one04:32
lalocarlos: yes - so we want to completely discard these, or we'll have bogus numbers on the query04:33
carlosand equal FALSE if we change the translation with a new one (from a po import or from rosetta)04:33
lalono, that's not correct, at least not according to the interface and to how we've been using it04:33
laloa translation from Rosetta is *not* inLastRevision04:34
carlosI mean04:34
lalobecause it's not in revision at all04:34
carloshow are you mapping it with our database?04:34
lalo"revision" means SCM in this context04:34
carloslalo: no, that's why we renamed from "inPoFile" to "inLastRevision"04:35
dafI think the documentation for "inLastRevision" is misleading04:35
carloslalo: it makes no sense to have a flag that only tells us if a translation is in SCM but not in rosetta, that's the point of origin04:36
laloorigin means where it's from *originally*04:36
carlosdaf: the one in interfaces.py?04:36
dafcarlos: yes04:36
carlosdaf: do we agree then04:37
laloI don't know why we want to know that, but that's what it is :-)04:37
carloslalo: Could you explain me why do you think the rename from inPoFile to inLastRevision is so natural to use the way you think?04:38
lalothe code states somewhere that origin never changes04:38
carloslalo: origin must not be changed04:38
carloswe agree on that04:38
lalothen I'm right04:38
carlosno :-P04:39
laloif you create a translation in Rosetta, then incorporate it in SCM, and import the PO, origin is still Rosetta04:39
carlosor I'm wrong, but I don't see it yet :-)04:39
carloslalo: right04:39
lalobut inLatestRevision is True04:39
carloswhen you insert a new translation from Rosetta04:39
carlosinLastestRevision should be changed to True04:39
carlosfor the new insert04:39
carlosand the old ones changed to False04:39
lalocan you rephrase your question about the column renaming? I couldn't parse it04:40
carloswhen I told you about the rename, you told me "<lalo> precisely"04:41
carlosso I suppose that with the rename you see it as a better name to use it that way04:41
carlosand I think we renamed it to prevent the rationale you are thinking :-)04:41
laloprecisely - inPOFile could mean "In any PO File", while inLatestRevision means "in the PO File present in the latest revision of whatever SCM upstream uses"04:41
laloyou remember wrong :-)04:42
carlosdaf: could we know your point of view?04:42
carloslalo: perhaps04:42
lalowe renamed to make it *clearer* that it's about SCM04:42
dafinLastRevision is not about SCM04:43
dafbecause it's also necessary for Rosetta translations04:43
lalohow so?04:43
laloif it's not about SCM, then it's misnamed, and I'm pretty sure that's not the case.04:43
dafwhen you create new transaltions through the web, you need to indicate that the translation sightings you're creating are current04:43
dafwhich is done by setting inLastRevision to True04:44
laloI believe that's an incorrect usage of the field04:44
carloslalo: take it as "in last revision of the translation"04:44
daf(I think)04:44
dafnow I'm confused also04:44
laloif neither of us is sure, we should ask Mark04:44
carlosdaf: that's what I thought04:44
lalo(in fact I *am* sure, but I'm minority :-P)04:45
carloslalo: yes, perhaps it's a good thing to do04:45
daflalo: if I'm wrong, so is the code04:45
dafthis should not be unclear04:45
lalosame for me :-004:45
carlosand fix our interface documentation04:45
laloif I'm wrong, there are a few places that must be fixed04:45
dafok, a question:04:45
dafif inLastRevision doesn't indicate that a translation sighting is the one that is current for a message set, what does?04:46
laloand if you have more than one active, you want the one with the latest lastActive timestamp04:47
carloslalo: then, how could we mark some translations that are completely wrong to be "deprecated"?04:47
daflalo: that's a field on message sets, isn't it?04:47
lalomaybe there is one in message set too, but I'm talking about translation sighting04:48
carlosactive and inlastrevision comes from potranslationsighting04:48
carlosthere is another inlastrevision for the pomsgsets04:49
=== lalo runs "make fullcheck" for great justice and tries not to think about how nightmarish it would be to make a functional test for statistics
carlosno, sorry04:49
carlosa typo in my sql query :-P04:49
dafcarlos: I think Lalo is correct04:50
carlosdaf: then, could you (or lalo) write it down in any place04:51
carlosso it's clear?04:51
carlosfor instance the interfaces.py04:51
carlosBTW, I think we should validate it with Mark before doing any change04:51
lalosabdfl: ping?04:54
lalo(doesn't hurt to try)04:54
dafhe's on the phone04:55
=== npmccallum [~npmccallu@69-162-252-7.ironoh.adelphia.net] has joined #launchpad
lalodaf: and what about the other two tasks? any preference?05:15
laloor maybe you have some other task for me I wasn't thinking about?05:16
dafno, I don't05:19
laloargh. Made 4 commits with the wrong tla id :-/05:19
dafI don't think I have a preference for the order05:19
daflalo: all-to-easily done05:19
=== daf twiddles his thumbs waiting for PQM
=== lalo ponders subscribing arch-users
=== limi|lunch is now known as limi
=== lalo decides he definitely should
carlosdaf: did you tested the po export I asked this morning?05:22
=== carlos is not sure
dafno, I didn't05:22
dafI've just merged from RF05:22
dafI'll do it now05:22
carlosok, thanks05:23
dilysBug 2057 resolved: Highlight interpolations in c-format messages05:26
carlosdaf: hmmm, good feature :-P, but please, change the color :-D05:27
dafI was feeling lazy :)05:28
dafif you like, feel free to make it better05:28
dafcarlos: export seems to still work for me05:28
carlosdaf: :-P05:29
lalohmm. LP doesn't run for me :-(05:29
carlosthe functional tests works?05:29
daflalo: UTC_NOW error?05:30
daflalo: fixed in RF05:30
lalowha? but I just merged in05:32
=== lalo merges again
sabdfllalo: pong05:51
lalowe're confused about the inLastRevision field of POTranslationSighting, again :-)05:51
laloI believe it means that translation is in SCM - is that correct?05:52
sabdflafaicr it means it was present in the last SCM version we parsed05:52
lalos/believe/am pretty sure/05:52
carloswell, it should mean the same in pomsgset, potranslationsighting and pomsgidsighting05:52
carlossabdfl: and the active field?05:52
sabdflthat's to indicate the one we would give as the *current best*05:53
sabdflso say we parse a PO file from SCM and find a new translation05:53
sabdflthat becomes true in both inlastrevision and active05:53
sabdflthen, someone gives us a new translation over the web05:53
sabdflthat becomes active=true, but inlastrevision=false05:54
sabdflwile the old one remains inlastrevision=true05:54
sabdflthen the maintainer downloads the po from us05:54
sabdflwe give him the translations that have active-true05:54
sabdflso he gets the new one05:54
sabdflhe commits it05:55
sabdflwe sync from scm05:55
sabdfland parse the po file05:55
sabdflwe see the new translation in there, so we mark it inlastrevision=true05:55
sabdflof course that means the old one is now inlastrevision=false05:55
carlosthen, we should set active = false for the old ones when we get a new translation05:55
carlosif that's the case, we don't have a way to "disable" bad translations (like german strings imported by error as Spanish ones)05:57
lalodidn't we use to have an obsolete field?05:57
lalowhat I'm doing now is to assume there are more than one with "active = TRUE" and fetch the one with the latest lastActive timestamp05:58
carloslalo: it was renamed to active :-)05:58
lalocarlos: ah.05:58
lalomy head hurts :-p05:58
sabdfllalo: no, there should only ever be one with active=true and one with inlastrevision=true05:59
carlossabdfl: then, what should we do with the case I'm talking about?06:00
laloso we still need a field for a translation that should be completely disregarded06:00
lalo"deprecated" or "deleted" or something06:01
sabdflcarlos, lalo, i have to be in #ubuntu-meeting for a meeting, catch me afterwards06:01
carlossabdfl: later06:01
carloslalo: could you write down all this so we don't get confused again? (I'm asking you because seems like you know it well)06:02
carlosalso, we should review all code and fix it so it works that way :-(06:03
dafcarlos: can you file a bug on that?06:03
carlosdaf: sure06:03
laloI'll expand on the docstrings that are currently in the interfaces. But these docstrings are correct06:03
lalo(the ones that exist)06:03
carloslalo: ok06:04
carlosthat's even better than a document06:04
dafcarlos: changing the colour of the hilighting is a one-line change to launchpad.css, by the way06:04
daflalo: great!06:04
=== carlos miss the comments in the DB schema file
dafcarlos: yeah, we should add some there06:04
carlosdaf: don't worry, the hard part is done the UI changes could come from limi :-P06:05
dafcarlos: true06:05
dafcarlos: I think limi is really busy at the moment, though06:05
dafcarlos: Malone and Lurker are higher priorities for him right now06:05
carlosdo we have already a timeline for the beta and the final release?06:05
carlosI know06:06
dafonce we decide on all the features we want in the beta, then we can estimate how long each one will take and how long the beta will take06:06
carlosI thought we had a timeline for the final release already 06:07
carlos(something like warty release date)06:07
carlosthat's why I'm asking06:08
dafI can't remember :)06:08
lalo     Untranslated messages: -32     06:08
carloslalo: yes, that one is funny06:09
carlosseems like somepeople is going to do the same we are doing with arch but with subversion06:12
carlosfor GNOME06:12
lalohmm, I can't set language preferences at all in my local lp06:13
carloslalo: it should not be a problem anymore, I did a patch so pt_BR should work06:14
carlosand it's in rocketfuel since yesterday or even earlier06:14
lalobut I can't choose *at all* right now, regardless of country code06:15
laloeg if I choose Arabian and Chinese it breaks06:15
carloscould you show me a traceback?06:15
=== lalo tries
laloI can't get to the debug skin either :-/06:16
dilysNew bug 2065 for Launchpad/Rosetta: Review all code so we are using the database as it should06:17
carloslalo: It works here with latest code from rocketfuel06:18
lalohttp://localhost:8085/++skin++Debug/rosetta/ -> 40406:20
lalois this the right url?06:20
dafmake debugging-on?06:20
=== daf will be away for a few hours
dafsee you all later06:20
carlosnice, the ++skin++Debug works with authentication now :-P06:20
carlosdaf: later06:20
laloit is disabled by default? I didn't know about that change06:20
lalolater daf, boa viagem06:21
carloslalo: It was disabled always06:22
carlosyou need to copy a file (that's what make debugging-on does)06:22
carlosto activate it06:22
lalobut it used to work for me06:22
carlosdid you checked out launchpad recently?06:22
laloah, duh06:23
carlosthat's the problem :-P06:23
laloI threw away my old tree06:23
carlosyou have lost the file :-)06:23
lalothat's when :-)06:23
lalocarlos: can you get to my IP maybe?
lalogreat :-)06:25
carlosI'm in06:25
carlosbut it's slow...06:26
carlosI see the error06:26
carloslalo: are you using your sqlobject branch?06:27
carlosbecause I think that's the problem, that you are not using it...06:27
carlosnot sure, but it's similar to a problem I had with the scripts until I moved to your sqlobject branch06:28
laloI believe I am06:31
lalolet me check06:31
laloah, well, of course I am, otherwise my script would not be running06:32
carlosand are you using latest version of the database?06:32
carlosthen I don't understand it06:33
lalowell, latest as of two or three days. Should I recreate it?06:33
carlosany pending change to commit into rocketfuel?06:33
carlosyes, please06:33
carlosI'm not sure if it could be an issue06:33
carlosbut I did somechanges that perhaps sqlobject is not able to handle without the indexes or unique keys in the database06:34
laloyes, but none in code that is at all touched by this prefs page - I have changes to pofile_adapters, poimport.py and sql.RosettaPOFile.updateStatistics()06:34
carlosok, let's try the db tehn06:34
=== lalo recreates the db
lalocarlos: that fixed it   O.o06:44
lalocarlos: I get an error if I try to add english :-P06:51
carlossame error?06:51
carloslet me check06:51
lalolooks like - I don't have the old one around anymore to compare06:51
carlosyes, same error06:53
carlosand it fails also here06:54
carlosI think I know where is the problem, but I don't know why It's a problem...06:55
carloscould you file a bug report?06:56
lalohmm, I don't seem to have plural form info in the db - is it being imported by the makefile?06:59
laloah, well, nm07:02
laloI do have that info07:02
laloonly I thought it did have info on some of the languages I chose, and it doesn't :-)07:03
=== carlos is back
dilysNew bug 2066 for Launchpad/Rosetta: Error if I add English to my languages07:15
dilysNew bug 2067 for Launchpad/Rosetta: Translate page: Batch widget on the bottom too07:18
lalook, now I have translations I can test statistics against.07:19
lalo#2042 now depends on #2065 :-/07:36
dilysBug 2058 resolved: Show template completeness statistics on translation page07:39
laloI'll take a break, I'm hurting all over07:40
carloslalo: yes, we should have it fixed for the beta07:42
carlossorry, I saw #2065 and I thought about #1965 :-P07:44
=== lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has left #launchpad []
laloit would be rather interesting for a bug to depend on 1965 :-P "I can't fix this unless the beta is released"07:55
spivlalo: I guess the final release would ;)07:57
=== cprov [~cprov@host217-37-231-28.in-addr.btopenworld.com] has joined #launchpad
sabdflok, i'm back08:10
carlossabdfl: we were talking about a missing field08:19
carloswe need a way to set a potranslationsighting as "wrong" so we don't care about it ever08:20
carlosif we see it again from a pofile we could forget about it because we already know it will be wrong08:20
carloswith the scenario we were talking we don't have such flag08:20
carlos(or I don't see it)08:20
=== stub [~stub@dsl-] has joined #launchpad
=== cprov [~cprov@81-86-111-202.dsl.pipex.com] has joined #launchpad
=== cprov [~cprov@224-21-161-212.DSL.ONCOLT.COM] has joined #launchpad

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!