IRC Archive for channel #xwiki on 16 December 2011

00:22 <rrodriguez> has joined #xwiki
00:33 <jvdrean> has quit
00:35 <rrodriguez> has quit
00:44 <rrodriguez> has joined #xwiki
02:20 <sburjan`> has quit
03:24 <tekzilla> has quit
03:26 <tekzilla> has joined #xwiki
05:13 <DrLou> has quit
05:51 <CIA-115> Sergiu Dumitriu master * reb6ff18 / (11 files in 4 dirs): [release] Updated translations - http://git.io/YLqZ5Q
05:51 <CIA-115> Sergiu Dumitriu stable-3.3.x * r6bd81df / (11 files in 4 dirs): [release] Updated translations - http://git.io/_Z72bA
06:00 <Denis> has joined #xwiki
06:03 <Denis1> has quit
07:01 <CIA-115> Sergiu Dumitriu master * rc79d05f / (2 files): [release] Updating translations - http://git.io/Tw0ZDQ
07:01 <CIA-115> Sergiu Dumitriu stable-3.3.x * ra0ff861 / (2 files): [release] Updating translations ...
07:28 <CIA-115> Sergiu Dumitriu xwiki-commons-3.3 * r6444b5a / pom.xml : [release] Preparing release xwiki-commons-3.3 - http://git.io/4uCCsA
07:30 <CIA-115> Sergiu Dumitriu xwiki-commons-3.3 * r09af15d / (31 files in 31 dirs): [maven-release-plugin] prepare release xwiki-commons-3.3 - http://git.io/Lm9KMw
07:34 <CIA-115> Sergiu Dumitriu xwiki-rendering-3.3 * r6975b32 / pom.xml : [release] Preparing release xwiki-rendering-3.3 - http://git.io/OOhztw
07:34 <CIA-115> Sergiu Dumitriu xwiki-rendering-3.3 * rdc56f07 / (26 files in 26 dirs): [maven-release-plugin] prepare release xwiki-rendering-3.3 - http://git.io/qnZfiA
07:46 <vmassol> has joined #xwiki
07:47 <rrodriguez> has quit
08:22 <mflorea> has joined #xwiki
08:27 <sburjan`> has joined #xwiki
08:29 <Denis2> has joined #xwiki
08:30 <@sdumitriu> vmassol: Any idea what this means?
08:30 <@sdumitriu> mmap failed for CEN and END part of zip file
08:30 <@sdumitriu> I get this when trying to release xwiki-platform
08:30 <@sdumitriu> It occurs at:
08:30 <@sdumitriu> maven-bundle-plugin:2.3.5:manifest (bundle-manifest) @ xwiki-platform-oldcore
08:31 <@sdumitriu> For: /home/release/.m2/repository/com/novell/ldap/jldap/4.3/jldap-4.3.jar
08:31 <@sdumitriu> But the jar is fine, I can extract it both with unzip, jar, and java -jar
08:31 <Denis> has quit
08:36 <tmortagne> has joined #xwiki
08:36 <vmassol> hmm never seen this maven error
08:38 <vmassol> I see there's a 2.3.6 version of the maven bundle plugin but I don't know if it helps
08:38 <vmassol> http://svn.apache.org/repos/asf/felix/releases/maven-bundle-plugin-2.3.6/doc/site/jira-report.html
08:43 <@sdumitriu> I could try
08:46 <@sdumitriu> Huh? "No plugin found for prefix 'maven-bundle-plugin'"
08:47 <@sdumitriu> Ah, my bad
08:48 <@sdumitriu> But when I run it manually, it works well
08:49 <Denis2> is now known as <Denis>
08:58 <@sdumitriu> vmassol: But do we need to generate an OSGi manifest for oldcore?
08:58 <vmassol> we need it for oldocre as much as for the other jars
08:59 <vmassol> in theory I doubt anyone is using our jars in osgi but who knows
08:59 <vmassol> it's good to have them
08:59 <@sdumitriu> Yeah, and I have a feeling it's not related to the bundle plugin anyway
08:59 <vmassol> (be a good osgi citizens)
09:00 <@sdumitriu> Could be related to the fact that it's a i386 machine, unlike the other agent machines
09:00 <@sdumitriu> I started the release on agent1 now to see if it works there
09:07 <+tmortagne> sdumitriu: if you are talking about agent-3 it has some issue it seems since it often fail to build test and even platform (generally the WYSIWYG war)
09:07 <vmassol> note: we have 2 func tests that started failing on 3.3: http://ci.xwiki.org/job/xwiki-enterprise-test-selenium-3.3/
09:08 <vmassol> s/2 func tests/2 test runs/
09:08 <vmassol> testCreateAndDeleteGroup
09:08 <vmassol> we might have a regression there
09:08 <@sdumitriu> I see that there's only one left now
09:12 <@sdumitriu> That there's no build manager this week
09:15 <jvdrean> has joined #xwiki
09:27 <@sdumitriu> Going to sleep, my eyes are very itchy
09:28 <sdumitriu> has quit
09:28 <vmassol> ok nn, sdumitriu when …..
09:28 <vmassol> too late....
09:34 <sburjan> has joined #xwiki
09:42 <+sburjan> good morning
09:42 <+sburjan> vmassol: I don't have the Garrlery Macro issue anymore. Did you do something to it ?
09:43 <vmassol> sburjan: gm, nope
09:43 <+sburjan> I tried with Chrome 16 and FF 3.6 and it looks ok now
09:43 <vmassol> it's working but the L&F is really bad
09:43 <vmassol> way too much back
09:43 <vmassol> *black
09:45 <vmassol> btw 7 UI tests failing on postgresql 9.1, compared to only 3 for HSQLDB (these 3 are always failing on my mac) - This means 4 test failing on postgres
09:46 <vmassol> out of 136 tests
09:46 <vmassol> going to run the full test suite on oracle 10g now
09:46 <+sburjan> you're running them locally, right
09:46 <vmassol> yes
09:47 <+tmortagne> vmassol: how do you run the test or postgres and oracle ? you setup a running XWiki by hand first ?
09:48 <vmassol> yes
09:48 <+tmortagne> ok
09:48 <vmassol> I call the tests from my IDE actually but you should be able to do something similar from the command line by changing a bit the pom.xml (we could define a profile for that too)
09:48 <+tmortagne> probably
09:50 <jvdrean> has quit
09:52 <vmassol> tmortagne: oh thought it was sorin asking me actually ;)
09:52 <+tmortagne> :)
09:52 <vmassol> which is why I mentioned the command line...
09:53 <+tmortagne> I was wondering if you already did some work on the pom.xml to be able to execute test for postgres and oracle ;)
09:53 <vmassol> not yet
09:53 <vmassol> but I'd really like
09:53 <vmassol> that
09:53 <vmassol> so that we can setup agents
09:54 <vmassol> actually it's supposed to work
09:54 <+tmortagne> yep would be nice
09:54 <vmassol> except that we don't clean the DB
09:54 <vmassol> in the database module
09:54 <vmassol> we jsut import the xar
09:54 <+tmortagne> well it's not really database module job I think
09:54 <+tmortagne> it should be done at test level
09:54 <vmassol> well we can refactor
09:55 <vmassol> but this is what the database modules were doing when I coded them
09:55 <vmassol> ie importing data in the DB
09:55 <vmassol> basically after the database module is executed
09:55 <vmassol> you have a filled DB
09:55 <vmassol> for some DB we package them as zip, etc
09:56 <+tmortagne> yes but in the case of test we don't execute database module on jenkins
09:56 <vmassol> and for others they're not packaged since it's hard to do
09:56 <vmassol> we do I think
09:56 <vmassol> since we run the full XE build
09:56 <vmassol> it's just
09:56 <vmassol> that we don't run the right profile
09:56 <vmassol> we only run the hsqldb one ATM
09:56 <+tmortagne> IMO database module should stay only for packaged databases and we should use some import/cleaner maven plugin in the test init
09:57 <+tmortagne> no we don't
09:57 <+tmortagne> there is a build for XE and then one build by test project
09:57 <+tmortagne> alone
09:57 <vmassol> the test projet depend on the XE one
09:57 <+tmortagne> so we never build the database module with the tests
09:57 <vmassol> they're only triggered
09:57 <vmassol> when XE rebuilds
09:57 <+tmortagne> there  is no reason for them to build on the same agent than XE
09:58 <+tmortagne> so no it's not going to work
09:58 <vmassol> indeed that's one issue
09:58 <+tmortagne> and again IMO  it's test project job to setup its test environment
09:58 <+tmortagne> you should be able to build it alone
09:58 <vmassol> I don't agree
09:59 <vmassol> at least for hsqldb
09:59 <vmassol> the reason I did it like this
09:59 <vmassol> was that we're testing exactly what is shipped
09:59 <vmassol> (again for hsqldfb and other embeddable DBs only)
09:59 <vmassol> for other dbs I never did anything really
09:59 <+tmortagne> I don't understand, we build the hsqldb test projects alone all the time
10:00 <+tmortagne> if you do a mvn clean install the database is reseted
10:00 <vmassol> it's depending on the hsqldb database module
10:00 <vmassol> there's no other import
10:00 <+tmortagne> lkike it depends on platform
10:00 <+tmortagne> like
10:00 <+tmortagne> because it's easier like that
10:00 <+tmortagne> we can't pakage database for others
10:01 <+tmortagne> expect javadb
10:01 <+tmortagne> so we have to import
10:01 <vmassol> yep this is what I said above
10:01 <+tmortagne> but the idea is the same: the database is reseted by test project
10:01 <vmassol> the other option is to create VMs
10:01 <vmassol> as part of the build
10:02 <vmassol> it's quite interesting too
10:02 <+tmortagne> that's probably way more complex
10:02 <vmassol> since it would allow us to package a full environment
10:02 <vmassol> for end users
10:02 <vmassol> I remember I found a maven plugin for doing that a while back
10:02 <vmassol> yes it's more complex
10:03 <vmassol> and probaly more time consuming too
10:03 <vmassol> it's also way cleaner and has some other advantages
10:03 <vmassol> it's the saint graal for me :)
10:06 <vmassol> http://vagrantup.com/ for ex (that's not the one I had found earlier on)
10:16 <+sburjan> isn't there any maven plugin that resets db before starting the test suite ?
10:16 <vmassol> there is
10:16 <vmassol> http://www.dbunit.org/
10:16 <vmassol> (for ex)
10:17 <vmassol> (I've used it a long time ago)
10:17 <silviar> has joined #xwiki
10:17 <+sburjan> so this would clean the db before _each_ test or before _each_ test suite ?
10:17 <vmassol> there might be more alternatives now
10:17 <vmassol> test suite
10:17 <vmassol> it's too long otherwise
10:18 <+sburjan> yeah ...
10:20 <CIA-115> tmortagne master * r8f5826d / xwiki-platform-core/xwiki-platform-extension/xwiki-platform-extension-repositories/xwiki-platform-extension-repository-xwiki/xwiki-platform-extension-repository-xwiki-server-api/src/main/java/org/xwiki/extension/repository/xwiki/internal/resources/ExtensionVersionFileRESTResource.java : Update comment - http://git.io/wa_Wqg
10:26 <jvdrean> has joined #xwiki
10:27 <+sburjan> http://www.cubrid.org/
10:34 <vmassol> nice presentation: http://www.slideshare.net/jedi4ever/manage-your-devtestprod-like-a-cloud
10:52 <lucaa> has joined #xwiki
11:32 <+jvdrean> hi, if I use {{include document="Included.Document" context="new"/}} do I have a way to know that I'm included from another document when in a Included.Document script ?
11:38 <+jvdrean> $context.tdoc does the trick
11:42 <+lucaa> odd, it shouldn't :)
11:42 <+lucaa> tdoc is only supposed to be the translated $context.doc :)
11:48 <+jvdrean> do you know any proper way to do it ?
11:50 <vmassol> jvdrean: you could look for METADATA info in the XDOM
11:50 <vmassol> (not sure if it's the best solution though but it would work)
11:51 <vmassol> looking for the Metadata.SOURCE above your block
11:56 <vmassol> sburjan: 6 tests failijg on oracle, so that's 3 in real…. so HSQLDB = 0, postgresql = 4, oralce = 3 failing
12:05 <evalica> has joined #xwiki
12:11 <vmassol> lunch time, bbl
12:15 <mflorea> has quit
12:25 <+sburjan> vmassol: I wonder where oracle fails
12:32 <+jvelociter> Hello
12:33 <vmassol> hello
12:35 <vmassol> deleting a page for ex
12:35 <vmassol> (sburjan)
12:36 <vmassol> scheduler test too
12:37 <vmassol> and whatlist
12:37 <+sburjan> I am going to lunch, I will check deleting a page when I get back
12:37 <vmassol> you won't see much
12:38 <vmassol> except that it doesn't show the recycle bin
12:38 <+sburjan> so no console messages ?
12:38 <vmassol> no
12:38 <+sburjan> so we don't fully support oracle in 3.3 ?
12:38 <vmassol> we don't fully support any DB in 3.3
12:38 <vmassol> I can find you issues with all DBs
12:39 <+sburjan> ok, then we need that thing you proposed on devlist
12:39 <vmassol> what's important is to tell users what's know to not work I think
12:40 <+sburjan> assuring support on browsers is hard enough already, adding also the db's .. it's gonna make my life a hell. big +1 to find a way to run agents on diferent db's
12:40 <vmassol> sburjan: you haven't done much to make your life simple so far
12:41 <+sburjan> wdym ?
12:41 <vmassol> if I were you i'd have automated so many more things from the onset
12:41 <vmassol> ie running the test suite on various browsers
12:41 <vmassol> and various DBs
12:41 <vmassol> you seem to be waiting for someone to do it for you :)
12:42 <+sburjan> well I was referring to manual testing
12:42 <vmassol> yes exactly
12:42 <vmassol> why do that?
12:42 <vmassol> it's like washing the dishes manually vs using a wahsing machine ;)
12:42 <+sburjan> back be in 30
12:42 <vmassol> ok
12:43 <vmassol> (note: there are still stuff that cannot be autimated but all stuff that can be automated should be automated)
12:53 <vmassol> sburjan: the scheduler job issue is the delete page issue since the test deletes a job and restores it
12:55 <Enygma`> has joined #xwiki
12:59 <silviar> has quit
13:00 <vmassol> hmm I don't understand the testWatchThisPageAndWholeSpace test
13:00 <vmassol> it sets watch on a page and a space
13:00 <vmassol> and then trigger the job
13:00 <vmassol> but I don't see where it makes modifications to the watched elements
13:01 <vmassol> thus I don't see how we can receive the email
13:01 <vmassol> tmortagne: seems you coded it, maybe you can help me?
13:02 <vmassol> (actually not coded but moved it to sel2)
13:06 <CIA-115> Vincent Massol stable-3.3.x * r49d1e81 / xwiki-enterprise-test/xwiki-enterprise-test-ui/src/test/it/org/xwiki/test/ui/watchlist/WatchThisPageAndWholeSpaceTest.java : Improved comments a bit - http://git.io/GKrFvA
13:07 <CIA-115> Vincent Massol master * rbe854c2 / xwiki-enterprise-test/xwiki-enterprise-test-ui/src/test/it/org/xwiki/test/ui/watchlist/WatchThisPageAndWholeSpaceTest.java : Improved comments a bit - http://git.io/20GD5g
13:08 <vmassol> guys if you can answer on the "user profile" split on the devs list, I could start doing it so that silvia can start documenting the user profile feature… :)
13:18 <mflorea> has joined #xwiki
13:30 <+jvelociter> mflorea: Hi !
13:30 <+mflorea> Hi Jerome
13:30 <+jvelociter> question : is it possible with the sheet system to have my sheet rendered in the context of the document it is applied to ?
13:30 <+jvelociter> (as opposed to the context of the sheet itself as it is the case it seems)
13:31 <+jvelociter> here's my problem: at some point I use $doc.content in the sheet, so that the content of the document is rendered. But relative images/attachemnts URI are wrong
13:32 <+jvelociter> (and I don't want to fix them since I want my pages to be editable via normal WYSIWYG)
13:33 <+jvelociter> do you see my issue ?
13:34 <+mflorea> I'm not sure I fully understand. When a sheet is applied, it is rendered in the context of the document it is applied to. i.e. $doc is the document the sheet is applied to, not the sheet.
13:34 <+mflorea> so the context document is the document the sheet is applied to
13:35 <+mflorea> I don't understand why you get wrong URLs
13:36 <+mflorea> let me try something locally
13:41 <+mflorea> jvelociter: I can reproduce, let me debug to understand the reason.
13:44 <+tmortagne> vmassol: what the mail send is the following modifications I think:
13:44 <+tmortagne>         // Watch Test.TestWatchThisPage
13:44 <+tmortagne>         ViewPage page = getUtil().createPage("Test", "TestWatchThisPage", "TestWatchThisPage ui", null);
13:44 <+tmortagne>         page.watchDocument();
13:44 <+tmortagne>         // Watch TestWatchWholeSpace
13:44 <+tmortagne>         page = getUtil().createPage("TestWatchWholeSpace", "Test1", "TestWatchWholeSpace ui", null);
13:44 <+tmortagne>         page.watchSpace();
13:44 <+tmortagne> I guess watchlist does not start watching when you enable it but actually send you the last 24h when the daily watchlist scheduler job run for example
13:46 <+jvelociter> mflorea: yes, actually it looks like a bug
13:46 <+mflorea> yes
13:46 <+jvelociter> let me know, I can provide you with a XAR if you want
13:46 <+tmortagne> it actually behave like {{include document="sheet"}}
13:47 <+tmortagne> because of the source is used to resolve the references
13:47 <+tmortagne> what vmassol was talking about
13:47 <+mflorea> then I need to set the source to the doc the sheet is applied to
13:48 <+tmortagne> or simply remove the SOURCE metadata from the sheet XDOM
13:48 <+tmortagne> so that it does not override
13:48 <+mflorea> ok
13:49 <+tmortagne> depend what you do exactly with the sheet XDOM actually I guess
13:49 <+tmortagne> if it's inserted in another one of executed alone
14:03 <+mflorea> tmortagne: there's no removeMetaData so I guess the only option is to use addMetaData to overwrite the "source" key
14:05 <+tmortagne> or add removeMetaData
14:05 <+mflorea> :)
14:05 <+tmortagne> vmassol ?
14:05 <vmassol> otp
14:06 <+mflorea> I'm wondering what's best: to keep the source meta data specified or not.
14:07 <+mflorea> I think it's better to set the source meta data to the document the sheet is applied to
14:07 <+tmortagne> well if we don't want to override the best should be to remove it from this POV
14:07 <+tmortagne> since you already setup everything needed to execute the XDOM in the context of the proper document
14:08 <+tmortagne> mflorea: how is this XDOM produced ?
14:08 <+tmortagne> since the rendering parsers don't set the source
14:08 <+mflorea> yeah, but still, someone could use the document displayed independently, obtain the XDOM and include it else where. In this case it's better to have the source meta data
14:09 <+mflorea> the XDOM is produced by rendering the sheet document (in the context of the document the sheet is applied to)
14:09 <silviar> has joined #xwiki
14:10 <+mflorea> i.e. the document the sheet is applied to is put on the context and then the sheet is displayed with the default content displayer
14:10 <+tmortagne> I mean how exactly, "rendering the sheet document" does not tell me what code actually create the DOM
14:10 <+mflorea> document content displayer
14:11 <+mflorea> documentDisplayer.display(sheet, displayParameters)
14:11 <+tmortagne> what is sheet ?
14:11 <+tmortagne> looking
14:12 <+mflorea> sheet is a document
14:12 <+mflorea> the sheet is displayed as any other document. The difference is that it is not rendered in its own context. The context document != sheet document
14:13 <silviar> has quit
14:13 <+mflorea> look in SheetDocumetDisplayer#display(DocumentModelBridge document, DocumentModelBridge sheet, DocumentDisplayerParameters parameters)
14:14 <+mflorea> and DocumentContentDisplayer#display(DocumentModelBridge document, DocumentDisplayerParameters parameters)
14:14 <+mflorea> (the first is in oldcore, the second is in display module)
14:15 <+tmortagne> ok so the SOURCE metadat is set by XWikiDocument#getXDOM
14:17 <+mflorea> yes
14:18 <+tmortagne> problem is that you don't have the XDOM in SheetDocumentDisplayer
14:18 <+tmortagne> so you can't modify it here
14:18 <+mflorea> I have it
14:18 <+mflorea> I return it in display
14:18 <+tmortagne> I see you call
14:18 <+tmortagne> return documentDisplayer.display(sheet, sheetDisplayParameters);
14:18 <+mflorea> yes
14:18 <+tmortagne> display executed the transformations already
14:19 <+mflorea> ah, so you say that its too late
14:19 <+tmortagne> not for references I think but the SOURCE is not only used for references
14:19 <+mflorea> hmm, why is it too late to set the source there?
14:20 <+mflorea> indeed
14:20 <+tmortagne> any macro could look at it and behave like references
14:21 <+tmortagne> now the thing is that the source is right and should probably stay like this
14:21 <silviar> has joined #xwiki
14:21 <+tmortagne> what needed is to implements the separation between base reference and source
14:22 <+mflorea> indeed, because right now source is used as base ref
14:22 <+tmortagne> vmassol proposed it in a mail I think already but it's not done yet
14:22 <+tmortagne> yep
14:22 <+tmortagne> was a mail about include macro refactoring for the same issues basically
14:22 <vmassol> is that the mail that Denis didn't like? :)
14:22 <vmassol> (the one about the display macro)
14:22 <DrLou> has joined #xwiki
14:23 <vmassol> I wanted to implement it two weeks ago but I re-read the thread and found that we didn't all agree about it
14:23 <+tmortagne> what Denis didn't like the difference between $doc.getURL and [reference]
14:23 <vmassol> so I dropped it from my todo queue
14:23 <+tmortagne> there was lot of thing in that mail
14:23 <vmassol> yes
14:23 <+tmortagne> the separation between source and base reference is still valid
14:24 <+tmortagne> the proposal should probably be cut in peaces
14:25 <vmassol> "I guess watchlist does not start watching when you enable it but actually send you the last 24h when the daily watchlist scheduler job run for example" hmmm this means the test can fail depending of when you run it....
14:26 <+tmortagne> vmassol: why ?
14:26 <+tmortagne> we could implement just this part for now so that mflorea can use it in displayers
14:27 <+tmortagne> and make the part about making include macro configurable  on that since Denis was only talking about the default behavior
14:27 <vmassol> imagine that the 2 test docs are created at 23:59:59 and that the trigger of the  daily notifier happens at 24:00:01
14:27 <+tmortagne> would still be in the last 24h
14:28 <+tmortagne> not the same day ok but I doubt it's based on the same day
14:28 <vmassol> hmm ok
14:28 <+tmortagne> but I don't really know much about it
14:28 <+tmortagne> just guessing based on the test
14:28 <vmassol> in any case that's not the reason why it fails for me
14:28 <+mflorea> jvelociter: the conclusion is that it's not easy to fix on my side. I need some changes in the rendering.
14:29 <+jvelociter> mflorea:
14:29 <+jvelociter> OK
14:29 <+jvelociter> but I think it should have high priority
14:29 <+mflorea> yes, I'll raise a jira isue
14:29 <+tmortagne> yep
14:29 <+jvelociter> it kind of breaks the whole purpose of the sheet system right now
14:29 <+jvelociter> cool
14:31 <+Denis> tmortagne: what I have also said, is that if we need a flexible solution, this should go in parameter of the link, not the whole page
14:31 <+tmortagne> vmassol: make me rememeber that the hack for the livetable in that test can now be removed
14:31 <vmassol> would be nice indeed
14:31 <+tmortagne> Denis: we can have both
14:31 <vmassol> I'm trying to debug why the watchlig mail isn't sent rightnow
14:32 <+Denis> tmortagne: sure, but I would not agree on a default that is contradictory between getURL et []
14:32 <+tmortagne> vmassol: no mail at all, and the things are properly watched ?
14:32 <vmassol> yes
14:32 <vmassol> I got this on oracle
14:32 <+tmortagne> Denis: that's the current default ;)
14:32 <vmassol> was ok on hsqldb
14:32 <vmassol> but I have run it only once
14:33 <+Denis> tmortagne: this is bad
14:33 <+tmortagne> maybe it's the request which try to find all the modifications to send
14:33 <+tmortagne> (vmassol)
14:33 <vmassol> maybe
14:33 <vmassol> there are no errors in the logs
14:35 <+tmortagne> Denis: well the idea of the mail was to go back to not having reference resolved based on the source by default and provide a configuration to allow it from what I remember
14:36 <+tmortagne> even said in my last mail
14:36 <+tmortagne> "
14:36 <+tmortagne>  So you propose to get back to previous behavior by default, right ? Given the fact that it has been changed very recently maybe we should have the property false right away and say it was a mistake ?
14:36 <+tmortagne> "
14:36 <+tmortagne> so for me it was ok for everyone
14:38 <vmassol> ok it's a fluke
14:38 <silviar> has quit
14:38 <vmassol> I got the test to work on oracle too
14:38 <vmassol> so it doesn't depend on the db actually
14:38 <+tmortagne> are you able to make it flicker with hsqldb ?
14:38 <vmassol> sburjan: so the good news is that there's only 1 test not working with oracle: the page deletion
14:38 <+tmortagne> weird that this happen on mail sending
14:39 <vmassol> it corectly deletes the page
14:39 <vmassol> but there's no recycle bin
14:39 <+sburjan> vmassol: that's good news
14:39 <vmassol> it means 3 tests are failing on postgreSQL at the moment
14:40 <lucaa> has quit
14:42 <+mflorea> jvelociter: XWIKI-7301
14:43 <silviar> has joined #xwiki
14:46 <lucaa> has joined #xwiki
14:49 <+mflorea> tmortagne: should I commit the temporary/partial fix for now (set the source meta data in SheetDocumentDisplayer)?
14:52 <+tmortagne> mflorea: unless you want to implement the separation between source and base reference (which is not very hard) ;)
14:55 <+Denis> tmortagne: Yes, I am +1 to revert this mistake
14:57 <lucaa> has quit
14:57 <+Denis> tmortagne: I have been really clear in my last mail of that thread, and I think I have cut it in pieces :)
14:58 <+tmortagne> Denis: was only talking about about changing which property to look for base reference, not changing any behavior fo include for now
14:58 <+tmortagne> but would be nice to apply vmassol proposal since from what I can see nobody was really against it, you said that you don't find the option very useful ok but that does not prevent much you don't have to use it ;)
14:59 <+Denis> tmortagne: I was not favorable since it increase complexity, but I have not said ?1, just almost ?1 ;)
14:59 <+Denis> tmortagne: if we do it, we should do all of it, including the link syntax
15:00 <+tmortagne> it's not critical, but if we change default include behavior it's important to do it as soon as possible
15:00 <+Denis> I fully agree
15:01 <+Denis> no one has never been able to show me the discussion about this change that should have never happen IMO
15:01 <+tmortagne> would have been nice to do right right away in 3.3 since the behavior change as been introduce in 3.2
15:02 <+Denis> I agree, I though vmassol would have pursue this, I was busy on other matter and I admit that I have forgot about it
15:03 <+Denis> is it too late ?
15:05 <+tmortagne> I think sdumitriu already started the 3.3 release
15:05 <vmassol> yes
15:05 <vmassol> and 3.3 is already too late
15:07 <+tmortagne> vmassol: can do what you proposed for 3.3/3.4 for 3.4/3.5
15:08 <vmassol> I proposed it for 3.3/3.4? (don't rememer)
15:08 <vmassol> but yes
15:09 <vmassol> evalica: I like the colors on http://incubator.myxwiki.org/xwiki/bin/view/Improvements/34Proposal ;)
15:10 <+tmortagne> vmassol: was a long time a ago now ;)
15:10 <+evalica> :) vmassol: suppose to be a surprise - still need to finish my proposal mail - but I'm glad you like it
15:11 <vmassol> tmortagne: actually I was thinking that I'd like to implement the technical space proposal for 3.4 if I get some time
15:11 <vmassol> (since it's becoming more important now that we have the EM UI)
15:12 <vmassol> (and I still have the UI extension stuff to do too)
15:12 <vmassol> (well too many things for sure….)
15:12 <+tmortagne> yea as usual too many idea and not enough peoples
15:14 <+tmortagne> mflorea: so either introduce just the new property in Metadata and modify include/html renderer to use it since you don't need more in your case or only do the partial fix with a big TODO
15:16 <vmassol> tmortagne: testWatchThisPageAndWholeSpace succeeds (I'm on hsqldb) but there are tons of error in the console
15:16 <+tmortagne> vmassol: what errors do you get ?
15:17 <vmassol> this is just one execution: http://pastebin.com/tWMNsXU7
15:17 <+mflorea> tmortagne: I'll do the partial fix for now
15:18 <+tmortagne> vmassol: "Could not connect to SMTP host: 127.0.0.1, port: 3025" so it really try to send the mails but I guess there is an issue in the way we use greenmail
15:19 <+tmortagne> maybe we don't stop it properly and when you start several times the same test it fail to init greenmail the next times
15:19 <vmassol> tmortagne: the test passes
15:19 <vmassol> I had stopped XE
15:20 <vmassol> before running it
15:20 <vmassol> I ran it only once here
15:20 <+tmortagne> ok that a lot of error then for one mail
15:20 <vmassol> I'm just surprised it passes with so many errors
15:20 <+tmortagne> it should not try to send so many of them
15:21 <+tmortagne> or maybe it's the same mail but it try to send it several times and finally succeed
15:21 <vmassol> maybe
15:21 <vmassol> haven't debugged yet
15:21 <+tmortagne> but it's a lot of tries
15:22 <vmassol> btw I need to leave early this afternoon (going to the north of  France for the weekend - first christmas this weekend ;)) …
15:22 <+tmortagne> it does not seems to wait a bit beetween the erreors
15:23 <lucaa> has joined #xwiki
15:27 <lucaa> has quit
15:42 <lucaa> has joined #xwiki
16:09 <vmassol1> has joined #xwiki
16:11 <vmassol> has quit
16:21 <vmassol1> has quit
17:29 <question4U> has joined #xwiki
17:33 <CIA-115> Sergiu Dumitriu xwiki-platform-3.3 * r5af4826 / (188 files in 188 dirs): [maven-release-plugin] prepare release xwiki-platform-3.3 - http://git.io/_MI2jQ
17:33 <CIA-115> Sergiu Dumitriu xwiki-platform-3.3 * re486862 / pom.xml : [release] Preparing release xwiki-platform-3.3 - http://git.io/oFzGoQ
17:48 <pgmjsd> has joined #xwiki
17:50 <CIA-115> Sergiu Dumitriu xwiki-enterprise-3.3 * rc75b21e / (40 files in 40 dirs): [maven-release-plugin] prepare release xwiki-enterprise-3.3 - http://git.io/9GXEgw
17:50 <CIA-115> Sergiu Dumitriu xwiki-enterprise-3.3 * r45a9b16 / pom.xml : [release] Preparing release xwiki-enterprise-3.3 - http://git.io/YyvgZw
18:05 <sburjan> has quit
18:11 <CIA-115> Sergiu Dumitriu xwiki-manager-3.3 * r2683672 / pom.xml : [release] Preparing release xwiki-manager-3.3 - http://git.io/2Qf9nw
18:11 <CIA-115> Sergiu Dumitriu xwiki-manager-3.3 * ra8991c2 / (16 files in 16 dirs): [maven-release-plugin] prepare release xwiki-manager-3.3 - http://git.io/CX_r2w
18:22 <tmortagne> has quit
18:30 <mflorea> has quit
18:38 <sdumitriu> has joined #xwiki
19:15 <jvdrean> has quit
20:28 <Denis1> has joined #xwiki
20:28 <Denis> has quit
20:35 <evalica> has quit
20:47 <sburjan`> has quit
20:47 <sburjan`> has joined #xwiki
20:49 <Enygma`> has quit
21:09 <lucaa> has quit
22:15 <Enygma`> has joined #xwiki
22:27 <Enygma`> has quit
22:36 <jvelociter> has quit
22:48 <silviar> has quit
23:12 <rrodriguez> has joined #xwiki