IRC Archive for channel #xwiki on 23 May 2013

Last modified by Vincent Massol on 2013/05/23 23:53

<sdumitriu> has quit
00:25 <sdumitriu> has joined #xwiki
00:52 <polx> has quit
03:06 <tekzilla> has quit
04:30 <Denis1> has joined #xwiki
05:32 <tekzilla> has joined #xwiki
05:51 <Denis1> has quit
05:58 <cjd> has quit
06:39 <huhongbo> has joined #xwiki
07:18 <huhongbo> has quit
07:38 <cjd> has joined #xwiki
07:45 <polx> has joined #xwiki
08:24 <evalica> has joined #xwiki
08:36 <mflorea> has joined #xwiki
08:44 <Denis1> has joined #xwiki
08:44 <Denis1> is now known as <Denis>
08:49 <vmassol> has joined #xwiki
08:49 <vmassol> good morning
08:49 <vmassol> BFD day today! :)
08:49 <vmassol> we're only 20 behind
08:49 <vmassol> we can make it
09:10 <msmeria> has joined #xwiki
09:19 <rrodriguez> has quit
09:50 <xwikibot> has joined #xwiki
09:58 <sburjan> has joined #xwiki
10:24 <SvenDowideit> has quit
10:27 <SvenDowideit> has joined #xwiki
10:30 <sburjan> Hello devs. From your point of view, which name is more correct to refer to the page tabs (Comments, Attachments, History, Information). I am thinking about Page Tabs or Bottom Tabs. WDYT ?
10:31 <sburjan> I need it for the Test Reporting App, because now I have a space called "Comments Annotations Attachments History Information Tests" which is kind of too long
10:43 <tmortagne> has joined #xwiki
10:51 <evalica> the id is called Doc Extra. You can call them Doc Extra Tabs
10:51 <evalica> #docExtraTabs
11:08 <evalica> has quit
11:10 <evalica> has joined #xwiki
11:35 <tdelafosse> has joined #xwiki
11:37 <vmassol> msmeria: hi, could you tell me how to reproduce http://jira.xwiki.org/browse/XWIKI-8971# ?
11:39 <Slashman> has joined #xwiki
11:39 <msmeria> vmassol: set up a new wiki from a war file, click cancel for the DW and import the appropriate xar from the administration and then navigate to those pages
11:40 <msmeria> you'll see there's no parent set for them
11:40 <vmassol> ok trying to reproduce now
11:40 <vmassol> what XAR?
11:40 <vmassol> the ui-all one?
11:40 <msmeria> yes
11:42 <vmassol> msmeria: import as backup pack or not?
11:43 <vmassol> you're logged with superadmin when you do the import?
11:43 <msmeria> as backup
11:43 <msmeria> for 4.5.3 I think it worked with Admin..
11:43 <vmassol> ok, trying with superadmin
11:44 <vmassol> no orphaned page
11:44 <vmassol> the docs you mentioned have parents
11:44 <msmeria> on 4.5.3 ?
11:45 <vmassol> 5.1-SNAPSHOT
11:45 <vmassol> there's no XWikiServerClassSheet btw
11:45 <vmassol> nor XWikiApplicationClassSheet
11:45 <msmeria> ok then. I guess it was magically fixed :)
11:45 <vmassol> nor XWikiApplicationClass
11:45 <vmassol> no
11:45 <vmassol> the fact that you mention XWikiApplicationClass means
11:46 <vmassol> that your steps are not correct
11:46 <vmassol> this comes from the app manager
11:46 <vmassol> and wiki manager
11:46 <vmassol> with your steps they're not installed
11:46 <vmassol> so you cannot have them
11:46 <msmeria> I tested on a XEM
11:47 <vmassol> rrrrrr
11:47 <vmassol> grrrr
11:47 <msmeria> :)
11:47 <vmassol> how am I supposed to know that?
11:47 <msmeria> my bad. sorry
11:47 <vmassol> could you fix teh jira issue and mention everything?
11:47 <vmassol> :)
11:47 <vmassol> so that I can retry
11:47 <msmeria> should've mentioned in the jira
11:47 <tmortagne> vmassol, msmeria: don't forget my comment about the fact that 4.5 and 5.x are very differents on these documents
11:48 <vmassol> tmortagne: if it's fixed on 5.x I don't rthink we need to fix that on 4.5.x wdyt?
11:48 <vmassol> (seems pretty minor to me)
11:48 <tmortagne> vmassol: just saying that you are trying to reproduce something in very different conditions
11:49 <tmortagne> in case you wanted to reproduce it to make sure you have it
11:49 <vmassol> Denis: ping
11:49 <vmassol> forget that
11:49 <vmassol> :)
11:49 <tmortagne> and ys we don't care of missing parent on those document if it's only on 4.5 branch
11:50 <vmassol> tmortagne: ok , thanks
11:51 <vmassol> Denis: why is http://jira.xwiki.org/browse/XWIKI-9041 an improvement now?
11:51 <vmassol> either it's a bug that still exists or it doesn't exist anymore and can be closed
11:54 <KermitTheFragger> has joined #xwiki
12:21 <vmassol> msmeria: what XAR should I import?
12:22 <msmeria> ui-all
12:23 <vmassol> which one?
12:23 <vmassol> xem?
12:23 <vmassol> xe?
12:23 <msmeria> xem
12:23 <msmeria> you're testing on xem now, right ?
12:23 <vmassol> yes
12:23 <vmassol> http://maven.xwiki.org/snapshots/org/xwiki/manager/xwiki-manager-ui-all/5.1-SNAPSHOT/ ?
12:23 <msmeria> so xem
12:23 <msmeria> http://maven.xwiki.org/snapshots/org/xwiki/manager/xwiki-manager-ui-all/5.1-SNAPSHOT/xwiki-manager-ui-all-5.1-20130523.074816-69.xar
12:23 <msmeria> yeah
12:25 <vmassol> ok I have 2 orphans
12:25 <vmassol> fixiing
12:34 <polx> has quit
13:09 <mflorea> has quit
13:09 <Slashman_> has joined #xwiki
13:12 <Slashman> has quit
13:12 <vmassol> evalica: should this be marked as bfd http://jira.xwiki.org/browse/XWIKI-6508 ?
13:12 <vmassol> ah improvement
13:12 <vmassol> sorry
13:12 <vmassol> )
13:12 <vmassol> :)
13:12 <evalica> thanks vincent
13:12 <evalica> :)
13:12 <evalica> yes is an impr
13:35 <Slash__> has joined #xwiki
13:36 <vmassol> sdumitriu: hi
13:36 <vmassol> any idea regarding my comment at http://jira.xwiki.org/browse/XWIKI-9157
13:36 <vmassol> ?
13:37 <sdumitriu> What comment?
13:37 <vmassol> wrong url
13:37 <vmassol> :)
13:37 <vmassol> http://jira.xwiki.org/browse/XWIKI-8937
13:38 <Slashman_> has quit
13:39 <vmassol> I really don't know this code at all btw....
13:40 <vmassol> tmortagne: that's for you :)
13:40 <vmassol> (edy's commit)
13:41 <sburjan> Hello devs. Can someone tell me best practices to get the proper date type when writing a XWIki Custom Mapping in Hibernate ? Do I have to look at what Hibernate supports, right ? I am talking about the xml file
13:41 <sburjan> *data types
13:41 <sdumitriu> vmassol: The only clean way of tracking visits is to store a random, unique ID in a cookie
13:42 <vmassol> sdumitriu: ok so you're suggesting to be even more radical and drop the ip + ua algorithm
13:42 <sdumitriu> IP + User agent doesn't make a good identifier, since everybody in a large company will have the same Gateway IP and the same browser
13:42 <vmassol> indeed
13:42 <vmassol> not sure why we have this...
13:44 <sdumitriu> vmassol: Where's that code from? Which class?
13:44 <vmassol> sdumitriu: StatsUtil
13:45 <sdumitriu> findVisit, found it
13:45 <vmassol> found this too in the same class:
13:45 <vmassol>                 visitStats = findVisitByIPUA(ip + ua, context);
13:46 <vmassol> the code I've shown was in : createNewVisit
13:48 <vmassol> actually seems tmortagne coded this in XWIKI—2249
13:48 <vmassol> (or maybe moved it?)
13:48 <vmassol> XWIKI-2249
13:49 <vmassol> yes he moved it from XWikiStatsServiceImpl
13:51 <sdumitriu> vmassol: There's another bug that makes this whole code completely useless: XWIKI-6140
13:51 <vmassol> I see
13:51 <vmassol> depends how fast we want to fix this
13:51 <sdumitriu> Nobody complained about it in a long time...
13:51 <vmassol> oana tells me she's having the problem frequently
13:52 <vmassol> and still has it
13:52 <vmassol> for one instance she's working on
13:52 <vmassol> she has users who keep having this problem
13:52 <sdumitriu> Oh, you mean XWIKI-8937
13:52 <vmassol> yes
13:52 <vmassol> because that leads to an error
13:52 <vmassol> a failure to save to the db
13:52 <sdumitriu> Well, for me the safest change at the moment is to compute a hash of the UA
13:53 <sdumitriu> Trimming isn't very clean, although we do it in other places as well
13:53 <sdumitriu> IIRC, we already trim the UA when storing the visit
13:53 <vmassol> it's simpler but not clean, however if we plan to change the algorithm later....
13:53 <vmassol> we could go with some quick trimming
13:54 <sdumitriu> Yes, maybe that would be the quickest
13:54 <vmassol> my only issue is that I don't master this and I'm sure there are no tests
13:55 <vmassol> I'm worried that there might be severla places to fix
13:55 <vmassol> like the 2 I've found in that StatUtil class
13:56 <vmassol> also are there any consequences to changing the algorithm on existing instances?
13:56 <vmassol> ie in the DB the id will be different
13:56 <vmassol> so stats might be flawed
13:57 <vmassol> ?
13:57 <vmassol> well better flawed stats than not working at all )
13:57 <vmassol> :)
13:57 <vmassol> actually since it's not saved for large values, there won't be any flawed stats
13:57 <vmassol> (if we trim)
13:59 <sdumitriu> Exactly, trimming is also backwards compatible on DBs that do enforce the 255 limit
14:00 <mflorea> has joined #xwiki
14:28 <sdumitriu> vmassol: The PDF export doesn't save attachment on the disk anymore, right?
14:28 <vmassol> hmmm I don't fully remember, would need to check, finishing something first
14:32 <vmassol> hmm anyone has seen this https://gist.github.com/vmassol/5635722/raw/dc9ed00a61ca0fe93025934a7734055340048c41/gistfile1.txt ? It happens at startup on the first request (home page) on a XEM where I have manually import the XEM ui-all XAR
14:32 <vmassol> tmortagne: any idea?
14:34 <tmortagne> vmassol: it's a new setup ? I saw this on old databases like xwiki.org and myxwiki.org ans assumed it was some database inconsistency
14:34 <vmassol> new setup
14:34 <vmassol> easy to reproduce
14:34 <vmassol> haven't debugged yet
14:34 <vmassol> (was on somerthing else when I noticed it)
14:35 <tmortagne> note: saw errors like this but no idea if it was the same properties
14:35 <tmortagne> now localy I never saw this
14:35 <tmortagne> and I'm always testing with XEM
14:35 <vmassol> I built XEM HSQLDB from sources
14:35 <vmassol> then skip DW
14:35 <tmortagne> (XEM on tomcat/mysql)
14:35 <vmassol> import XEM ui-all
14:35 <vmassol> stop it
14:35 <vmassol> restart and call home page
14:36 <tmortagne> maybe it's only on hsqldb
14:38 <tmortagne> vmassol: it seems to be all about stringlist properties
14:38 <tmortagne> maybe it's related to the revert of sdumitriu commit
14:38 <vmassol> maybe
14:38 <vmassol> I built just manager
14:38 <tmortagne> since according to the error hibernate does no find the value in the table were it should be
14:38 <vmassol> so I got its dep from our remote repo
14:38 <tmortagne> so it sound  a lot like it does no look at the right table
14:39 <vmassol> I'll rebuild later on to check if I still have it
14:43 <vmassol> sdumitriu: is this good enough or do I need to compute 255 bytes and not chars?
14:43 <vmassol>     private static String computeUniqueID(String ip, String ua)
14:43 <vmassol>         return StringUtils.substring(ip + ua, 0, 255);
14:43 <sdumitriu> It's chars
14:43 <vmassol> so it's ok right?
14:43 <sdumitriu> Yes
14:43 <vmassol> cool
14:43 <vmassol> I'm going to commit this with the hope it works fine
14:43 <vmassol> I've tested manually
14:44 <vmassol> couldn't find any issue with my simple tests
15:27 <vmassol> 10 more bugs to go to get even!
16:22 <polx> has joined #xwiki
16:23 <sburjan> guys, I have a dilemma.
16:24 <sburjan> I am running these lines on XEM and they don't work
16:24 <sdumitriu1> has joined #xwiki
16:24 <evalica> vmassol: the Radeox Chart Macro and the 'new' Chart macro are 2 different things, right? If the issues don't specify what macro it used, but they are old issues so probably it was the Radeox Macro, should I close the issues as 'Won't fix' or 'Cannot reproduce'? Do we still want to make improvements on the Radeox Macros?
16:24 <sburjan> {{velocity}}
16:24 <sburjan> $targetDocRef
16:24 <sburjan> $targetDoc.save()
16:24 <sburjan> {{/velocity}}
16:24 <sdumitriu1> evalica: We're trying to remove the Radeox Chart macro
16:24 <sdumitriu1> vmassol even removed it once, but had to bring it back
16:25 <sdumitriu1> So close as won't fix
16:25 <sburjan> i get the output of $targetDocRef, but after I got the "$targetDoc.save()" string on my screen. The document is not created
16:25 <sburjan> am I missing something ?
16:25 <sdumitriu1> evalica: Or, try to see if it applies to the new chart macro as well, in which case you can comment saying that, and leave it open
16:26 <evalica> sdumitriu: because they were 'generic' issues stating Chart macro in general, I've close them as Cannot Reproduce, since the new macro works well. So they are 2 different things. Thanks :)
16:40 <msmeria> has quit
16:51 <evalica> has quit
16:59 <sdumitriu1> vmassol: Anything left for 4.5.4?
17:00 <sdumitriu1> I can start the release now
17:00 <vmassol> sdumitriu1: don't think so
17:00 <vmassol> cool
17:04 <sdumitriu1> vmassol: Any idea what this means? http://ci.xwiki.org/job/xwiki-enterprise-test-extension-4.5.x/117/console
17:04 <sdumitriu1> Agent slave jar needs updating?
17:05 <vmassol> but since it's a BFD I didn't spend time on it
17:05 <vmassol> FYI I've updated a2
17:07 <vmassol> we need to get jeknins back to stable tomorrow
17:07 <vmassol> we have 51M1 on Monay
17:07 <vmassol> so tommorrow needs to be a stabilization day
17:41 <rrodriguez> has joined #xwiki
17:52 <sburjan> we are at 21 bugs closed, congratulations !
17:53 <Slash__> has quit
17:54 <Slashman> has joined #xwiki
18:13 <jbwiv> vmassol: you around?\
18:14 <vmassol> busy but around :)
18:20 <sdumitriu1> Why isn't tmortagne here...
18:21 <tmortagne> what do you mean ?
18:21 <vmassol> :)
18:21 <sdumitriu1> Ah, didn't see you
18:21 <sdumitriu1> tmortagne: This looks broken
18:21 <sdumitriu1> http://l10n.xwiki.org/xwiki/bin/view/L10N/Search?application=Platform&name=&lang=&key=platform.extension.distributionWizard.extension.outdatedextensionsStepSummary&origtext=&transtext=
18:21 <sdumitriu1> This is a @deprecated key
18:21 <sdumitriu1> And it was automatically copied
18:21 <sdumitriu1> But it seems to have been copied from the wrong key
18:22 <KermitTheFragger> has quit
18:23 <tmortagne> it's not deprecated, you mean it's the new name of a deprecated key ?
18:24 <sdumitriu1> tmortagne: Yes, new name for a deprecated key
18:24 <tmortagne> http://l10n.xwiki.org/xwiki/bin/view/L10N/Search?application=Platform&name=&lang=&key=platform.extension.distributionWizard.extensionsStepSummary&origtext=&transtext= seems to contains the same thing at least for some properties
18:24 <sdumitriu1> But the translations weren't copied from the old key, but from a different one
18:25 <sdumitriu1> Yes, some are OK
18:25 <tmortagne> as far as I can see some copy were correct and some not
18:25 <sdumitriu1> But others mean "Annotations"
18:25 <tmortagne> not sure what could have happen
18:25 <sdumitriu1> The romanian one means "Cancel"
18:26 <sdumitriu1> nl -> [Edit]
18:26 <tmortagne> the way it's working I don't see how it could have copied the wrong key (several keys in this case)
18:26 <sdumitriu1> pt -> Update
18:26 <sdumitriu1> So it's a mess...
18:27 <sdumitriu1> I'm removing the wrong ones, so that at least we don't commit wrong translations
18:27 <tmortagne> the one that were here are right for what I can see, the wrong one are the one that were not translated
18:27 <tmortagne> trying to understand what happended
18:28 <tmortagne> the wrong translation is definitely associated to "Copied from deprecated key" comment at least
18:28 <sdumitriu1> Maybe it gets the previous one that was translated
18:28 <tmortagne> it would be the same for all translations
18:29 <tmortagne> *all keys
18:29 <tmortagne> looking at the code
18:29 <tmortagne> (http://l10n.xwiki.org/xwiki/bin/view/L10NCode/L10NGroovy)
18:33 <rrodriguez> has quit
18:37 <polx> has quit
18:42 <sburjan> any idea if we have UTF8 encoding in the Tomcat from xwiki.org ?
18:44 <sdumitriu1> sburjan: xwiki.org is in latin1
18:45 <sburjan> sdumitriu1: thanks
19:01 <sdumitriu1> tmortagne: Another issue, several "vi" translations disappeared, for example http://l10n.xwiki.org/xwiki/bin/Platform/PlatformXWikiCoreResources_-1913250921_platform-index_vi
19:01 <sdumitriu1> They're not even in the trash
19:01 <tmortagne> from where did you get the link http://l10n.xwiki.org/xwiki/bin/Platform/PlatformXWikiCoreResources_-1913250921_platform-index_vi ?
19:02 <sdumitriu1> From _fr, replaced the language
19:02 <tmortagne> http://l10n.xwiki.org/xwiki/bin/XE/XEXWikiCoreResources_-1913250921_platform-index_vi
19:02 <tmortagne> looks like it has been deleted before I move it to platform app
19:03 <sdumitriu1> http://pastebin.com/Wc5WxdHR
19:03 <sdumitriu1> Hm, why was it deleted?
19:03 <tmortagne> no idea
19:04 <tmortagne> maybe a bug in rename
19:04 <tmortagne> which would have deleted the document even if it failed to copy it
19:04 <sdumitriu1> The others were deleted one day later: http://l10n.xwiki.org/xwiki/bin/XE/XEXWikiCoreResources_-1913250921_platform-index_fr
19:04 <tmortagne> actually I think I have an idea
19:05 <tmortagne> the first move has been broken because there was no place left on the server
19:05 <tmortagne> I guess the copy failed but renamed deleted the document anyway
19:05 <sdumitriu1> Right
19:05 <tmortagne> s/renamed/rename/
19:05 <sdumitriu1> How do we get back those translations?
19:05 <tmortagne> depend how many we have in this situation
19:05 <tmortagne> one is easy
19:06 <tmortagne> just need to move the document to it's right place and change name field
19:06 <tmortagne> fixing this one
19:07 <sdumitriu1> 51 total
19:08 <tmortagne> the best would probably be a script searching for all deleted document in XE space and starting with XEXWikiCoreResources and moving then to platform if they don't already exist
19:08 <tmortagne> but that would take ages
19:09 <tmortagne> that's the only language were you have the issue ?
19:10 <sdumitriu1> I think so
19:10 <sdumitriu1> Didn't finish reviewing changes yet
19:11 <sdumitriu1> http://l10n.xwiki.org/xwiki/bin/view/Platform/PlatformWorkspaceApplication_1157455186_workspacemanager-error-notinvirtualmode_fr was also deleted from the local file, I don't know if it was correctly moved someplace else or if there was a one-time failure in the export
19:11 <sdumitriu1> Ah, just finished reviewing
19:11 <sdumitriu1> So only those translations are missing
19:13 <sdumitriu1> Ah, untracked files
19:13 <tmortagne> there is no reason you don't have http://l10n.xwiki.org/xwiki/bin/view/Platform/PlatformWorkspaceApplication_1157455186_workspacemanager-error-notinvirtualmode_fr since there is nothing special about thi stranslation as far as I can se on the document
19:14 <tmortagne> it should end up in the eport
19:14 <tmortagne> export
19:14 <sdumitriu1> I agree
19:14 <sdumitriu1> It was the only one that was weirdly missing
19:14 <sdumitriu1> I skipped it
19:17 <tmortagne> actually looks like http://l10n.xwiki.org/xwiki/bin/view/Platform/PlatformWorkspaceApplication_1157455186_workspacemanager-error-notinvirtualmode_en has been deleted
19:17 <tmortagne> that's probably why you don't have it exported
19:17 <tmortagne> you have it in other languages ?
19:17 <sdumitriu1> Not sure, the _fr is the only one actually translated
19:21 <tmortagne> this key is not part of the default language on github it seems
19:21 <Slashman> has quit
19:21 <sdumitriu1> OK
19:21 <sdumitriu1> So it's normal
19:21 <tmortagne> the importer never deleted antything in this case it set it as dead
19:21 <sdumitriu1> I left it for 4.5 since it might be still used there
19:21 <tmortagne> so it's not exactly normal either
19:22 <tmortagne> but in this case it does not seems to have been deleted by the importer anyway
19:22 <tmortagne> no idea how Caty ended up deleting that document
19:23 <tmortagne> restoring it
19:24 <tmortagne> I have to go, will check tomorrow for the vi keys
19:25 <sdumitriu1> Thanks tmortagne
19:25 <sdumitriu1> Bye
19:25 <tmortagne> as for the wrong deprecated key copy I will have to debug since I really don't understand how this can happen by reading the code
19:25 <tmortagne> has quit
19:34 <sdumitriu1> vmassol:
19:34 <sdumitriu1> [WARNING] 'build.plugins.plugin.(groupId:artifactId)' must be unique but found duplicate declaration of plugin org.apache.maven.plugins:maven-enforcer-plugin @ org.xwiki.commons:xwiki-commons-core:4.5.4-SNAPSHOT, /home/hudsonagent/releases/xwiki-trunks/xwiki-commons/xwiki-commons-core/pom.xml, line 149, column 15
19:35 <vmassol> I can look tomorrow, I won't have the time now, have to go in a few minute/seconds
19:38 <tdelafosse> has quit
19:43 <abusenius> has joined #xwiki
19:47 <mflorea> has quit
19:47 <sdumitriu1> vmassol: Still there?
19:48 <sburjan> has quit
20:08 <cjd> has quit
20:16 <cjd> has joined #xwiki
20:52 <polx> has joined #xwiki
20:59 <polx> has quit
21:07 <jbwiv> is now known as <timtimtim>
21:08 <timtimtim> is now known as <timtimc>
21:14 <polx> has joined #xwiki
21:27 <mflorea> has joined #xwiki
21:32 <polx> has quit
22:03 <evalica> has joined #xwiki
22:04 <evalica> has quit
22:33 <vmassol> has quit
23:49 <rrodriguez> has joined #xwiki
23:53 <mflorea> has quit

Get Connected