IRC Archive for channel #xwiki on 17 January 2012
Last modified by Vincent Massol on 2012/10/18 19:14
00:14 <sburjan`> has joined #xwiki
00:17 <jvdrean> has quit
01:17 <CIA-121> Sergiu Dumitriu master * rc3986e2 / xwiki-platform-core/xwiki-platform-colibri/src/main/resources/colibri/colibri.css : XWIKI-6824: Standard disabled button does not look disabled ...
02:23 <tekzilla> has quit
02:25 <tekzilla> has joined #xwiki
05:54 <DrLou_> has quit
05:59 <Denis> has quit
06:00 <Denis> has joined #xwiki
07:25 <mflorea> has joined #xwiki
07:32 <rrodriguez> has quit
07:48 <vmassol> has joined #xwiki
07:54 <sburjan`> has quit
08:08 <sburjan`> has joined #xwiki
08:41 <CIA-121> Sergiu Dumitriu xwiki-commons-3.3.1 * rb1f7a47 / pom.xml : [release] Preparing release xwiki-commons-3.3.1 - http://git.io/phgiCg
08:41 <CIA-121> Sergiu Dumitriu xwiki-commons-3.3.1 * r0eccc05 / (31 files in 31 dirs): [maven-release-plugin] prepare release xwiki-commons-3.3.1 - http://git.io/hLh3hA
08:48 <tmortagne> has joined #xwiki
08:49 <CIA-121> Sergiu Dumitriu xwiki-rendering-3.3.1 * ra77472f / pom.xml : [release] Preparing release xwiki-rendering-3.3.1 - http://git.io/-crA6g
08:49 <CIA-121> Sergiu Dumitriu xwiki-rendering-3.3.1 * r6cf510b / (26 files in 26 dirs): [maven-release-plugin] prepare release xwiki-rendering-3.3.1 - http://git.io/zapF9Q
08:50 <+tmortagne> sdumitriu: you sure it's ok to release 3.3.1 ? I can see there is some tests failing on 3.3 branch on jenkins
08:52 <@sdumitriu> Seemed OK to me
08:52 <+tmortagne> ok, did not looked in detail
08:53 <vmassol> good morning guys
08:53 <vmassol> cool that you're releasing 3.3.1 sdumitriu!
08:53 <vmassol> guys we need to find a solution for this false positive: http://ci.xwiki.org/job/xwiki-enterprise-test-webstandards/org.xwiki.enterprise$xwiki-enterprise-test-webstandards/956/testReport/junit/org.xwiki.test.webstandards/CustomDutchWebGuidelinesValidationTest/Validating_Dutch_Web_Guidelines_validity_for____xwiki__bin__import__XWiki__Import_editor_globaladmin_section_Import_executed_with_credentials_Admin_admin/
08:53 <vmassol> we need to identify why exactly there's a broken connection
08:54 <vmassol> and then wait for request/response to finish
08:54 <vmassol> I think we're already reading everything from the response so it has to be something else
08:54 <vmassol> is there an ajax request on the import page somewhere?
08:54 <vmassol> s/broken connection/connection aborted/
08:54 <@sdumitriu> The logged URL should have been different
08:54 <CIA-121> Denis Gervalle feature-security-authorization * rd161663 / (1505 files in 182 dirs): Merge branch 'master' into feature-security-authorization (+20 more commits...) - http://git.io/xq5RCg
08:56 <vmassol> I can fix the error logged, need to understand first when it happens
08:56 <vmassol> found the place in XWikiAction
08:58 <vmassol> first I'm not sure it should be an error
08:59 <vmassol> could be a warning but I'm not sure
08:59 <@sdumitriu> I'm not even sure a warning is needed
08:59 <vmassol> since it's currently logged as an error I'll just print the stack trace for now
08:59 <@sdumitriu> I added that log a very long time ago
08:59 <vmassol> to know more about when it happens
08:59 <vmassol> wdyt?
09:00 <@sdumitriu> I don't think it's needed
09:00 <vmassol> I've checked the code that throws it
09:00 <@sdumitriu> We won't get any more information from there
09:00 <vmassol> and it's a real error
09:00 <@sdumitriu> It doesn't tell what the client was doing
09:00 <vmassol> in most places
09:00 <vmassol> sure but knowing from where it's sent might give us a clue
09:01 <vmassol> about what caused it
09:01 <@sdumitriu> Nope
09:01 <@sdumitriu> It's thrown when trying to write the response
09:01 <@sdumitriu> At some point response.getOutputStream().write will fail
09:01 <@sdumitriu> Not because of something the code itself does
09:02 <@sdumitriu> But because the client got tired of waiting for data and stopped the page
09:02 <vmassol> eah
09:02 <vmassol> yeah
09:02 <vmassol> checked more and indeed it seems to always been sent when writing the response
09:03 <@sdumitriu> Historically, I introduced that log line because it used to be a real stacktrace displayed automatically by the container
09:03 <@sdumitriu> http://jira.xwiki.org/browse/XWIKI-1922
09:04 <vmassol> ok so you're ok to make this a debug log?
09:04 <@sdumitriu> info
09:04 <vmassol> hmmm
09:04 <vmassol> I don't think info is correct
09:05 <vmassol> info is informational for me, like printing the version number
09:05 <vmassol> if it's a problem it can be: error, warning or debug if not important and only used to debug stuff
09:05 <vmassol> I don't think printing this will give a good info
09:05 <vmassol> good and useful
09:05 <vmassol> from what you tell me we'd only need this in debug mode
09:05 <vmassol> and even then I'm not even sure
09:05 <vmassol> I proposed debug to be on the safe side
09:07 <vmassol> ok sdumitriu?
09:10 <@sdumitriu> OK
09:18 <vmassol> hmm we should probably still look for a way to tell httpclient to wait for all threads to finish
09:18 <vmassol> (sdumitriu)
09:19 <vmassol> I'm slightly worried that my fix isn't the good one...
09:19 <vmassol> and that this log was symptomatic of something to fix
09:20 <vmassol> I'm still wondering if a warn is not better
09:35 <CIA-121> Marius Dumitru Florea master * rc6da441 / (3 files): XE-1078: Upgrade to Selenium 2.16.1 and Selenium Maven Plugin 2.2 ...
09:45 <CIA-121> Marius Dumitru Florea master * r03409e6 / xwiki-enterprise-test/xwiki-enterprise-test-wysiwyg/src/test/it/org/xwiki/test/wysiwyg/CacheTest.java : XE-1078: Upgrade to Selenium 2.16.1 and Selenium Maven Plugin 2.2 ...
09:49 <CIA-121> Sergiu Dumitriu xwiki-platform-3.3.1 * r5ec4d06 / pom.xml : [release] Preparing release xwiki-platform-3.3.1 - http://git.io/rYIJ4A
09:49 <CIA-121> Sergiu Dumitriu xwiki-platform-3.3.1 * r5fb867a / (188 files in 188 dirs): [maven-release-plugin] prepare release xwiki-platform-3.3.1 - http://git.io/mOhCTg
09:53 <sburjan> has joined #xwiki
09:53 <+sburjan> good morning everyone
09:54 <@sdumitriu> Good morning Sorin
10:04 <CIA-121> Sergiu Dumitriu xwiki-enterprise-3.3.1 * rf392444 / pom.xml : [release] Preparing release xwiki-enterprise-3.3.1 - http://git.io/VyKx1Q
10:04 <CIA-121> Sergiu Dumitriu xwiki-enterprise-3.3.1 * r264ad39 / (40 files in 40 dirs): [maven-release-plugin] prepare release xwiki-enterprise-3.3.1 - http://git.io/y3o8lg
10:07 <gdelhumeau> has joined #xwiki
10:11 <CIA-121> Sergiu Dumitriu xwiki-manager-3.3.1 * r96d47b0 / pom.xml : [release] Preparing release xwiki-manager-3.3.1 - http://git.io/M7bXyQ
10:11 <CIA-121> Sergiu Dumitriu xwiki-manager-3.3.1 * r17a8d84 / (16 files in 16 dirs): [maven-release-plugin] prepare release xwiki-manager-3.3.1 - http://git.io/6gbK8A
10:12 <rrodriguez> has joined #xwiki
10:19 <gdelhumeau> has quit
10:19 <gdelhumeau> has joined #xwiki
10:20 <CIA-121> Vincent Massol master * rbe47e46 / xwiki-platform-core/xwiki-platform-oldcore/src/main/java/com/xpn/xwiki/web/XWikiAction.java : XWIKI-7408: Don't display "connection aborted" error logs - http://git.io/BnrpDQ
10:25 <gdelhumeau> has quit
10:27 <rrodriguez> has quit
10:30 <jvdrean> has joined #xwiki
10:32 <evalica> has joined #xwiki
10:33 <+sburjan> sdumitriu: are you releasing 3.3.1 or 3.2.1 ?
10:33 <vmassol> 3.3.1
10:33 <@sdumitriu> 3.3.1
10:34 <@sdumitriu> I don't know if there will be an 3.2.1
10:34 <+sburjan> I see users asking for 3.2.1 .. maybe we could recomment 3.3.1 instead of 3.2.1
10:34 <Enygma`> has joined #xwiki
10:35 <rrodriguez> has joined #xwiki
10:37 <@sdumitriu> sburjan: It would be nice if you could test a bit the release
10:37 <+sburjan> sdumitriu: is it already released ? 3.3.1
10:37 <@sdumitriu> It's built
10:38 <@sdumitriu> Didn't announce it yet, though
10:38 <+sburjan> ok, downloading and starting to test
10:38 <@sdumitriu> http://maven.xwiki.org/releases/org/xwiki/enterprise/xwiki-enterprise-jetty-hsqldb/3.3.1/
10:38 <+sburjan> don't announce it yet :P
10:38 <+sburjan> give me a few hours
10:38 <@sdumitriu> Yep, I'll go get some sleep
10:38 <@sdumitriu> Thanks, bye
10:38 <+sburjan> okay, ttyl
10:38 <sdumitriu> has quit
11:21 <rrodriguez> has quit
11:38 <vmassol> so guys what's left to do for the 3.4RC1 release?
11:39 <vmassol> just the failing tests?
11:39 <vmassol> snif :( sergiu still hasn't fixed the extension tests
11:46 <vmassol> mflorea: could you give us a status?
11:47 <vmassol> seems http://ci.xwiki.org/job/xwiki-manager-test-webstandards/ is failing to start xwiki
11:47 <+mflorea> tmortagne should look at the manager tests
11:47 <vmassol> "sh: Can't open start_xwiki.sh"
11:47 <+mflorea> I'm fixing the wysiwyg tests atm, and then I'm going to look at the ui-tests
11:48 <+mflorea> I already fixed selenium (1) tests
11:48 <vmassol> ui tests are hanging?
11:48 <vmassol> let us know if you need help somewhere
11:49 <vmassol> same error with xmlrpc tests: http://ci.xwiki.org/job/xwiki-manager-test-xmlrpc/org.xwiki.manager$xwiki-manager-test-xmlrpc/539/testReport/junit/org.xwiki.test.integration/XWikiTestSetup/junit_framework_TestSuite_dcb52ae/
11:49 <vmassol> "sh: start_xwiki.sh: No such file or directory"
11:49 <+mflorea> yes, could be related to tmortagne's latest commit on manager
11:49 <vmassol> let's see what cause the failure
11:49 <vmassol> that should be easy to see with jenkins
11:49 <vmassol> checking
11:50 <vmassol> tmortagne: caused by http://ci.xwiki.org/job/xwiki-manager/4082/changes#detail0
11:50 <+mflorea> re ui-tests, indeed, they are hanging, the second time, in LoginTest, not the same test though, but I can't reproduce locally
11:51 <vmassol> mflorea: you asked in email, but it's very easy to see what causes an issue with jenkis
11:51 <+mflorea> will look into it later
11:51 <+mflorea> how?
11:51 <vmassol> for ex take this: http://ci.xwiki.org/job/xwiki-manager-test-xmlrpc/539/
11:51 <vmassol> this is what is failing
11:51 <+mflorea> the changes are not always in the failing module
11:51 <vmassol> you can see there is: "Started by upstream project http://ci.xwiki.org/job/xwiki-manager-test-pom/ build number http://ci.xwiki.org/job/xwiki-manager-test-pom/539/"
11:51 <vmassol> so you click on 539
11:51 <vmassol> then "Started by upstream project http://ci.xwiki.org/job/xwiki-manager/ build number http://ci.xwiki.org/job/xwiki-manager/4088/"
11:52 <vmassol> you click on 4088
11:52 <vmassol> "Started by upstream project http://ci.xwiki.org/job/xwiki-enterprise/ build number http://ci.xwiki.org/job/xwiki-enterprise/8473/"
11:52 <vmassol> you click on 8473
11:52 <+mflorea> ok, so you go recursively
11:52 <vmassol> etc
11:52 <vmassol> and then you reach what caused the problem
11:52 <+mflorea> I wouldn't call this easy :)
11:52 <vmassol> it takes like 10 seconds
11:52 <vmassol> I call this easy
11:52 <+mflorea> but indeed, it's posible
11:53 <vmassol> actually the ex I gave wasn't good
11:53 <+mflorea> well, if you don't act quickly you can loose the cause because other build overwrite it
11:53 <vmassol> you need to take the first failure after a succesful build
11:53 <+mflorea> like when there are other 10 build in the mean time
11:53 <vmassol> yes we need to have a successful build to see what causes the problem
11:54 <+mflorea> yep
11:54 <+mflorea> re help, maybe you can look at why ui-tests are hanging on LoginTest
11:55 <+evalica> vmassol: question about IE6 jiras - since we are not supporting IE6 anymore what do we do with the jira issues that affect versions smaller that XE 3.2 ? close them? or consider the relevant for the specific version (for example 2.2 when IE6 was supported)? I guess close them since nobody will add patches to fix those problems. The closing reason should be 'won't fix', right?
11:55 <+mflorea> actually, the first time was in RegisterTest
11:55 <vmassol> ok I'll preapre my workspace for that
11:55 <vmassol> evalica: mark them as won't fix
11:55 <+evalica> thanks
11:55 <vmassol> with a link to our browser straegy
11:55 <+mflorea> thanks
11:55 <vmassol> page
12:00 <+tmortagne> right forgot to modify test dependencies
12:02 <+tmortagne> hmm actually the id did not changed
12:02 <+tmortagne> s/id/ids/
12:04 <+tmortagne> so nothing changed from tests POV
12:05 <vmassol> tmortagne: well somehow the commit I mentioned broke webstandards and xmlrpc tests
12:05 <vmassol> they were working just before your change and they have failed continuously just after
12:06 <vmassol> I haven't looked at your commit precisely so I don't know how this was caused
12:08 <+tmortagne> the commit just moved some maven module under the same folder to separate them from the ui-all but no id has been changed or the way to build them, just moved some dependencies
12:10 <+tmortagne> checking the produced packages
12:10 <+tmortagne> looks like start_xwiki.sg is missing
12:12 <+tmortagne> found it
12:12 <+tmortagne> parent issue
12:16 <CIA-121> tmortagne master * r787843e / (5 files in 5 dirs): Fix parents - http://git.io/ZImpUg
12:17 <+sburjan> tmortagne: something is broken with EM/XR
12:18 <+sburjan> I am testing 3.3.1 and also running yesterday's 3.4-SNAPSHOT and I am not able to install extensions
12:18 <+sburjan> on 3.4-SNAPSHOT I get: The following new extensions will be installed:
12:18 <+sburjan> $extension.id.id ($extension.id.version)
12:18 <+evalica> vmassol: do you know do we have a page that shows what are the supproted skins?
12:19 <+evalica> I would like to make one similar to the ne we have for browsers
12:19 <+tmortagne> sburjan: you get that in 3.4 ? there used to be a bug in the UI but I refactored it yesterday evening so you should get a new version probably if yours is from yesterday
12:19 <+tmortagne> 3.3.1 is supposed to be the same thing as 3.3
12:20 <+tmortagne> except some small bugfixes
12:21 <+tmortagne> downloading a 3.3.1 version
12:21 <+tmortagne> how do you get the issue ?
12:21 <+sburjan> just a sec because I am still trying to reproduce
12:21 <+sburjan> failed once, worked twice
12:22 <+tmortagne> on 3.3.1 ?
12:22 <+sburjan> yes
12:22 <+sburjan> so don't download
12:22 <+tmortagne> looks like the thing I fixed yesterday on 3.4, I tough this issue was only in 3.4
12:23 <+sburjan> but it's flickering ..
12:23 <+tmortagne> maybe xwiki.org is not responding very well
12:23 <+tmortagne> what do you do ?
12:23 <+tmortagne> search or advanced install ?
12:24 <+sburjan> on 3.3.1 I got: Caused by: java.io.IOException: Invalid answer (500) fo the server when requesting
12:24 <+sburjan> simple search + install
12:24 <+sburjan> also the "fo" seems a typo
12:26 <+tmortagne> this is a server error so probably an issue on xwiki.org side
12:27 <+tmortagne> an looks like in that case the install preparation UI has a bug which makes it display incorectly
12:27 <+tmortagne> but in that case you should still have the apply button, don't you ?
12:28 <+tmortagne> and generally the real install is working well
12:28 <+sburjan> I get the stacktrace after I click the Apply
12:28 <+tmortagne> ok
12:28 <+tmortagne> so maybe xwiki.org is really in a bad state then
12:29 <+sburjan> having other issues too .. I get a velocity error when uninstalling .. but I can't see the stacktrace
12:29 <+sburjan> let me investigate
12:30 <+sburjan> does xwiki create some local repo ? or uninstalling an extension removes the xar from disk ?
12:30 <+tmortagne> uninstall does not remove the extension from the local repository
12:30 <vmassol> evalica: well we support the skins we bundle: http://enterprise.xwiki.org/xwiki/bin/view/Main/Features
12:31 <vmassol> we should update http://platform.xwiki.org/xwiki/bin/view/Features/Skins to list only the supported skins and just add a link to all skins on extensions
12:31 <+tmortagne> sburjan: I have to go, back in about 2h
12:32 <+sburjan> ok
12:32 <vmassol> evalica: actually in the first link I've only shown apps not skins
12:32 <vmassol> so we just need to modify http://platform.xwiki.org/xwiki/bin/view/Features/Skins
12:33 <vmassol> let me update it
12:34 <+evalica> I will also want to add the end of support for certain skins: like albatross -and the no-innovation in toucan - also the supported resolution would be inetersting - I need to do some research on this datas and I will add them too
12:35 <vmassol> this info should be put in the skin itself
12:35 <vmassol> (IMO)
12:35 <vmassol> supported resolution?
12:35 <+evalica> yes .. like until what version did we supported 800x600? what is now the supported resolution?
12:35 <+evalica> colibri and toucan have fixed widths .. etc
12:36 <vmassol> it used to be 1024x768 AFAIR a long time ago but I don't know now
12:37 <+evalica> right now it works also for 970px - even if this is not a resolution - but could be interesting also for mobile support
12:37 <+evalica> anyway - thanks
12:39 <+sburjan> vmassol: I see a strange thing happening. I install Presentation Application, and when I uninstall it I get a Velocity error which I can't expand. If I stop the wiki and delete the local repo, restarting the wiki will not list Presentation Application as Installed anymore. Does Presentation Application something more special than the other apps ?
12:42 <+sburjan> vmassol: also restarting the wiki still lists the extension, but this time uninstall button works
12:43 <vmassol> evalica: I'm done on http://platform.xwiki.org/xwiki/bin/view/Features/Skins
12:43 <vmassol> evalica: would be great if you could update the colibri screenshot...
12:43 <vmassol> and update http://extensions.xwiki.org/xwiki/bin/view/Extension/Colibri+Skin
12:45 <+evalica> k - let me first parse my selected 30 jiras :)
12:46 <vmassol> cool you're cleaning up jiras
12:49 <+sburjan> bbl, lunch
12:56 <vmassol> mflorea: which test of LoginTest was hanging?
12:56 <vmassol> (I have my setup ready, starting to test)
12:56 <+mflorea> http://ci.xwiki.org/job/xwiki-enterprise-test-ui/899/console
12:57 <+mflorea> I seeo.x.t.u.LoginTest - testRedirectBackAfterLogin started
12:57 <vmassol> k
12:57 <+mflorea> but this morning it was another one, from RegisterTest
12:57 <+mflorea> testRegisterJohnSmith precisely
12:58 <+mflorea> both pass locally so I don't know what's the problem
13:17 <vmassol> mflorea: I ran 100 times testRedirectBackAfterLogin locally and got no error...
13:18 <+mflorea> then it must be something wrong with the agent
13:18 <vmassol> I ran it by itself
13:18 <vmassol> it could also be caused by other tests that run before it
13:18 <+mflorea> indeed
13:18 <+mflorea> re agent, both time it was agent-2 (FF 9..0.1)
13:19 <+mflorea> we could run on agent-1 (FF3.6) to see if it hangs there too
13:19 <vmassol> I looked at the code and couldn't find a place where it could hang
13:20 <vmassol> actually there's one
13:20 <vmassol> WebElement tokenInput = getDriver().findElement(By.xpath("//input[@name='form_token']"));
13:20 <vmassol> in recacheSecretToken()
13:21 <vmassol> but it shouldn't
13:21 <vmassol> I've only seen hangs in the past for findElements <— 's'
13:22 <vmassol> findElement should time out fine normally
13:24 <vmassol> btw tmortagne when executing LoginTest.testRedirectBackAfterLogin() we have this error all the time in the console:
13:24 <vmassol> 2012-01-17 13:13:43,897 [http://localhost:8080/xwiki/bin/cancel/XWiki/XWikiPreferences?ajax=1] ERROR o.i.ObservationContextListener - Can't find any begin event corresponding to [class org.xwiki.bridge.event.ActionExecutedEvent (cancel)]
13:25 <vmassol> I've seen this one several times, we need to look at it
13:25 <vmassol> especially since it's an ERROR
13:25 <vmassol> (we shouldn't have anything in the console anyway)
13:25 <vmassol> there's still this one too that you made happen and which has been going on for very long now: "2012-01-17 13:13:46,655 [http://localhost:8080/xwiki/bin/loginsubmit/XWiki/XWikiLogin] WARN o.x.v.i.DefaultVelocityEngine - Deprecated usage of method [com.xpn.xwiki.api.XWiki.parseMessage] in /templates/login.vm@29,33"
13:26 <vmassol> I think we should make some effort to no have spurious logs in the console
13:37 <vmassol> evalica: I'm fighting all the time with the horizontal menu on xwiki.org
13:37 <vmassol> I think we'll need to revert if we cannot fix it
13:37 <vmassol> it's so much a pain
13:37 <vmassol> (if it happens to me it must happen to lots of people too)
13:40 <+evalica> vmassol: I don't know how to fix it but I've observed that if you click on the menu item than it works (the sub menu hovering)
13:40 <vmassol> I tried that but it doesn't work for me
13:41 <vmassol> when I hover the menu entry disappears so I cannot click on it, I can click on the menu itself but it doesn't hlep
13:41 <vmassol> in any case we cannot accept this as a solution
14:16 <+sburjan> there is definitely something fishy with EM/XR. I want to install an extension and I get 2 failures, and the 3rd attempt works like a charm
14:21 <+sburjan> tmortagne: ping me when y ou're back
14:36 <sburjan> has quit
14:37 <+tmortagne> vmassol: "Can't find any begin event corresponding to [class org.xwiki.bridge.event.ActionExecutedEvent (cancel)]" is definitely an error and should be fixed. Thing is it's not a general issue but, a case by case things so when you find one and know how to reproduce it please create a jira issue. It generally mean some action is doing something weird or the Execution context is not properly restored after being cloned
14:37 <vmassol> mflorea: apart from adding debug logs in various places in testRedirectBackAfterLogin I'm out of ideas…
14:39 <sburjan> has joined #xwiki
14:39 <CIA-121> Marius Dumitru Florea master * rf22e8c0 / (2 files): XE-1078: Upgrade to Selenium 2.16.1 and Selenium Maven Plugin 2.2 ...
14:39 <CIA-121> Marius Dumitru Florea master * r81d7973 / xwiki-platform-core/xwiki-platform-gwt/xwiki-platform-gwt-user/src/main/java/org/xwiki/gwt/user/client/ui/rta/SelectionPreserver.java : XWIKI-7409: Caret is lost when applying heading style on empty body ...
14:41 <vmassol> tmortagne: ok, just reported XWIKI-7410
14:41 <sburjan> has quit
14:42 <+mflorea> vmassol: I (hopefully) fixed the WYSIWYG tests. I'm moving to UI tests (AppWithinMinutes failing tests ). I don't have ideas re hanging tests..
14:42 <sburjan> has joined #xwiki
14:42 <vmassol> going to google a bit and search in the webdriver group
14:42 <+tmortagne> vmassol: ok thanks
14:43 <+sburjan> tmortagne: it seems that if I install/uninstall same extension several times I get a java out of heap memory on a jetty+hsqldb xwiki. Several times meaning ~ 5 cycles of install/uninstall
14:44 <+tmortagne> sburjan: what kind of extension ?
14:44 <+sburjan> I have played with Presentation Application
14:45 <+sburjan> rebooted my pc, let me see it again
14:46 <+tmortagne> btw about " If I stop the wiki and delete the local repo, restarting the wiki will not list Presentation Application as Installed anymore.", EM store in the local repository what's installed and what's not if you removed it does not have any idea there is something
14:46 <vmassol> btw mflorea the hanging issue with findElements I mentioned earlier is here: https://groups.google.com/forum/#!topicsearchin/webdriver/hang/webdriver/BnGY75mFTD0 (was fixed in Selenium 2.0rc3)
14:46 <+tmortagne> still 3.3.1 ?
14:47 <+tmortagne> refactored a lots of stuff in 3.4 to make it stronger and did not backported much on 3.3 branch
14:47 <+sburjan> 3.4-SNAPSHOT. Just reproduced: Uninstall -> Install -> Uninstall (velocity error)
14:47 <+sburjan> not even 2 full cycles
14:48 <+sburjan> but I am unable to get the stacktrace because the velocity error does not expand ...
14:48 <+tmortagne> seems weird even if the application was kept fully in memory you should not get outofmemory
14:48 <+tmortagne> I don't think EM UI has expendable errors
14:48 <+sburjan> tmortagne: I have another stack :)
14:49 <+tmortagne> either you get all or just the message generally
14:49 <+sburjan> tmortagne: try this: Install Presentation Application. Search it again, and press Install. You should be warned that it is already installed. But I get this message in a stacktrace
14:49 <+tmortagne> that's normal for now
14:50 <vmassol> mflorea: maybe the issue is something like https://groups.google.com/forum/#!topicsearchin/webdriver/hang/webdriver/OYRBNhWZFr4
14:50 <+sburjan> in 3.3.1 I didn't got any stack, I got a nice message t hat is already installed
14:50 <+tmortagne> the UI should not even allow you to click on install on the first place
14:50 <+tmortagne> you get an error because trying to install an already installed extension is an error
14:50 <+sburjan> but it does
14:50 <+tmortagne> I thin kthere is a jira issue already about install button
14:51 <+tmortagne> yes that's normal too
14:51 <+tmortagne> 3.3.1 prepare the install in (not very good) velocity
14:51 <+tmortagne> 3.4 use the new API to generate an install plan and get detailed error
14:52 <+tmortagne> you have simple message in 3.3.1 mostly because it's it does not have access to more
14:52 <+tmortagne> we need a general UI oriented error message filter
14:52 <+sburjan> tmortagne: could you try to reproduce the install/uninstall cycle issue ? Try installing/uninstalling Presentation Application several times. I want to know if you have this issue too
14:52 <+tmortagne> vmassol already create an issue about that
14:53 <+tmortagne> sburjan: yep i'm preparing my setup right now
14:53 <+sburjan> I restarted my container.. and got an error when Uninstall -> Install -> Uninstall
14:53 <+sburjan> using 3.4-SNAPSHOT 168
14:55 <+sburjan> I also get this in 3.3.1
14:56 <+sburjan> but as you said, they use diferent API's, but there seems to be a common problem
14:56 <+tmortagne> you are talking about outofmemory or something else ?
14:56 <+sburjan> yes
14:57 <+sburjan> got the trace.. let me pastebinit
14:57 <+sburjan> http://pastebin.com/tqxGTvFp
14:58 <+tmortagne> (installing presentation application)
14:58 <+sburjan> so install-> uninstall and so on .. you don't have to leave the admin app
14:59 <+tmortagne> I just got a java.lang.OutOfMemoryError: PermGen right away
14:59 <+tmortagne> so not the same as you
14:59 <+sburjan> but still out of mem :)
14:59 <+tmortagne> since you get a Java heap space
14:59 <+tmortagne> it's very different things
15:00 <+tmortagne> you don't fill both with the same kind of things
15:00 <lucaa> has joined #xwiki
15:00 <+sburjan> I have no custom memory settings for my JVM or xwiki
15:00 <+tmortagne> now I don't see how a xar could increase the permgen
15:01 <+lucaa> hi guys (rendering guys -- vmassol, tmortagne ). I have a short question. If I have a linkblock who's resource type is ResourceType.INTERWIKI, can I be sure that it's resource reference is absolute? (and not relative to current wiki / current space) ?
15:01 <+sburjan> not expert on these things ... but good that you got an error too .. too bad not the same one :(
15:02 <+tmortagne> lucaa: really you have a interwiki, how you do that ? xwiki parser or renderers don't support it yet
15:02 <+sburjan> because this is the 4th time I am able to reproduce, so there is definitely something. My machine has 8gb of RAM, so I'm pretty sure i'm not out of memory
15:03 <+tmortagne> sburjan: yea, trying to find what is consuming the permgen but it's not going to be easy I think
15:03 <+tmortagne> do you know this application well ?
15:03 <+lucaa> tmortagne: no I don't, I need to transform all the links in my document in links with absolute references and I was wondering which ones I need to handle (doc i know, attach I know, and I was wondering about the others)
15:03 <+sburjan> tmortagne: not at all.
15:03 <+tmortagne> sburjan: will check first what is this application
15:04 <+sburjan> tmortagne: I'll try reproducing with other apps
15:04 <+tmortagne> lucaa: since it's supposed to target some other wiki (not even a xwiki) I guess you could see it as absolute
15:04 <+tmortagne> sburjan: ok
15:05 <+lucaa> tmortagne: aham, ok. Actually my questions would be, if I copy paste that wiki syntax to another doc and I render it, will the link point to the same location? I guess the answer is yes
15:05 <+tmortagne> lucaa: said another way it's supposed to be an extrenal link by definition
15:05 <+lucaa> tmortagne: ok, even better :)
15:06 <+tmortagne> when it will be supported...
15:13 <vmassol> ouch another broken page on xwiki.org: http://www.xwiki.org/xwiki/bin/view/References/WebHome
15:13 <vmassol> it has a blog xobject
15:13 <vmassol> fixing
15:15 <+sburjan> tmortagne: tried with ircbot app, and I am unable to reproduce. trying another app. any suggestion of a larger app ?
15:15 <vmassol> nice xwiki wiki: http://www.afdcc.com/
15:17 <+tmortagne> sburjan: presentation application is pretty small as far as I can see
15:18 <+tmortagne> don't know any big application
15:18 <+tmortagne> admin app maybe
15:18 <vmassol> presentation app is quite large
15:18 <+tmortagne> it's not that big but bigger than IRC for sure
15:18 <vmassol> it has one page doing several MBs
15:18 <+tmortagne> hmm
15:18 <vmassol> with tons of image attachments
15:18 <+tmortagne> so maybe thats the issue then
15:18 <+sburjan> hmmmm... tried ircbot, FAQ, and now with Monitoring Tools
15:19 <+tmortagne> ok so the issue is not really the size of the application but the size of one of the application pages I think
15:19 <+sburjan> unable to reproduce with these
15:19 <+tmortagne> I debugiing it anyway
15:20 <+tmortagne> sburjan: yea I think you will not be able to reproduce it unless you find or produce an application with a big pages like presentation application has
15:20 <+tmortagne> lets see what is loosing memory then
15:20 <+tmortagne> sburjan: would you have a stack trace of when you get an error uninstalling ?
15:21 <+tmortagne> trying to import the xar with normal import UI first
15:22 <+tmortagne> I sould have seen that the xar file was more than 9M...
15:22 <+tmortagne> sburjan: was it the first time you were trying to install this application ?
15:22 <+tmortagne> or did it used to work ?
15:22 <+sburjan> tmortagne: yes. first time .. I don't remember installing it before
15:22 <+tmortagne> ok lets fins what's wrong
15:23 <+tmortagne> it's a good test case for the EM internal importer
15:23 <+sburjan> managed to get the stack trace by going to /xwiki/bin/view/XWiki/InstalledExtensions and Uninstalling it from there. The error from here expands so I can see the stack
15:25 <CIA-121> Marius Dumitru Florea master * r09534a7 / xwiki-enterprise-test/xwiki-enterprise-test-extension/src/test/it/org/xwiki/test/ui/extension/ExtensionManagerAdminTest.java : XE-1062: Fix failing Extension Manager tests ...
15:27 <+sburjan> tmortagne: I see you reported it on JIRA :)
15:27 <+tmortagne> sburjan: you are talking about http://jira.xwiki.org/browse/XWIKI-7411 ?
15:27 <+sburjan> yes
15:27 <+tmortagne> just hit this issue when trying to import the xar
15:28 <+tmortagne> but does not have anything to do with the issue we have with EM
15:28 <+tmortagne> just a UI bug
15:29 <+tmortagne> hmm Presentation.XWikiPresentation is veeeery big, it's been loading for almost 10 minutes now...
15:30 <+tmortagne> the outofmemory issue is maybe just caused by this page being stored in the cache...
15:30 <vmassol> cjdelisle: any progress on http://jira.xwiki.org/browse/XWIKI-6686 ? do you have a plan to fix this? :)
15:30 <+tmortagne> I'm going to get a timeout
15:30 <+tmortagne> pergen
15:31 <+tmortagne> sburjan: just reproduce the pergen issue without using EM at all
15:31 <+tmortagne> just importing the xar with the admin and trying to access Presentation.XWikiPresentation
15:31 <+tmortagne> so I doubt there is much we can do at EM level
15:31 <@cjdelisle> hmm IIRC we found an issue where pages were being loaded twice in the tests, I wonder if that had an effect..
15:31 <+tmortagne> the issue is comming from the fact that the page is taking way too much memory in the cache I think
15:32 <vmassol> cjdelisle: maybe the test could be enabled again
15:32 <@cjdelisle> if not, I'm not sure this one will be quick to solve, it's something in XWiki.java#rollback()
15:32 <+sburjan> tmortagne: I understand. Well, goot it's not an EM issue
15:32 <+tmortagne> and even worst if it's stored several times indeed
15:33 <vmassol> mflorea: what about:
15:33 <vmassol> // @Ignored because it fails erroniously when run on the CI server.
15:33 <vmassol> // see: http://jira.xwiki.org/jira/browse/XE-949
15:33 <vmassol> /*
15:33 <vmassol> public void testHoldDeleteKeyBeforeMacro()
15:33 <vmassol> any idea how to fix this now?
15:33 <vmassol> you have a few others you've excluded:
15:33 <vmassol> // @Ignored because this test causes a history dropdown menu to open and it breaks.
15:33 <vmassol> // see: http://jira.xwiki.org/jira/browse/XE-947
15:33 <vmassol> /*
15:33 <vmassol> public void testShortcutsForSaveAndView() throws MalformedURLException
15:33 <vmassol> // @Ignored because this test causes a history dropdown menu to open and it breaks.
15:34 <vmassol> // see: http://jira.xwiki.org/jira/browse/XE-947
15:34 <vmassol> /*
15:34 <vmassol> public void testShortcutsForSaveAndContinue()
15:34 <+mflorea> I don't remember excluding them, I think it was cjdelisle
15:34 <vmassol> another one:
15:34 <vmassol> @Ignore
15:34 <vmassol> public void failingTestDeleteMergeOk2() throws Exception
15:34 <+mflorea> that's not mine
15:35 <vmassol> should this sync module be removed?
15:35 <vmassol> and moved to contrib
15:35 <vmassol> s/module/plugin/
15:35 <@cjdelisle> I did because they were flickering tests
15:35 <@cjdelisle> flickering/failing erroniously
15:36 <+sburjan> tmortagne: is it normal that not all the core extensions have license information ?
15:36 <+mflorea> vmassol: we could move it
15:36 <vmassol> mflorea: is it working?
15:36 <+sburjan> tmortagne: example: XWiki Platform - XML - Script Service, XWiki Platform - Formula - Macro, etc
15:36 <+tmortagne> sburjan: yes, the core extension scanner get all the information it can find in the pom.xml file located in the jar file but does not resolve the parents
15:36 <CIA-121> Ecaterina Moraru (Valica) master * r60316ed / xwiki-platform-core/xwiki-platform-colibri/src/main/resources/colibri/colibri.css : XWIKI-6339: Entering a category with a very long name completely breaks the layout ...
15:36 <+tmortagne> and license information are often in the parent pom
15:37 <+tmortagne> problem is that if I make the scanner look at parent pom its probably going to take ages to init XWiki the first time
15:37 <+mflorea> only if you compile the wysiwyg editor by yourself, I send a vote a long time ago to exclude it from the list of wysiwyg editor plugin that get compiled
15:37 <+sburjan> I understand
15:38 <vmassol> jvdrean is back! yeah ;)
15:38 <+tmortagne> my plan is to generate a file at build time with informations like that so that the scanner can find some missing stuff in it
15:38 <vmassol> mflorea: and the vote was positive to exclude it?
15:38 <+mflorea> and ofc, I haven't tested it, so even if you build the wysiwyg editor to include, I can't say it would work
15:38 <+mflorea> yes
15:38 <vmassol> (oh sorry misread)
15:38 <vmassol> (to exclude it from the compiled list, not to move it out)
15:38 <+mflorea> yep
15:39 <+mflorea> i.e. right now it is in the source code, but it doesn't get compiler to JavaScript
15:39 <+mflorea> so it's not available
15:40 <vmassol> mflorea: do you want to send a vote to move it out to contrib?
15:40 <vmassol> it could be published as an extension on extension.xwiki.org
15:40 <vmassol> (in the "other" category)
15:41 <vmassol> (till you find a way to be able to add plugins dynalically! :))
15:41 <+mflorea> do we have a retired repo in the contrib? because there's no one to maintain it
15:41 <vmassol> yes we do
15:41 <+tmortagne> vmassol: there is a plugin category
15:41 <vmassol> tmortagne: not xwiki plugin
15:41 <+tmortagne> s/category/type/
15:41 <vmassol> wysiwyg plugin
15:41 <+tmortagne> ha ok
15:41 <+tmortagne> indeed
15:41 <+mflorea> yes, but it would be misleading, because you can't install it as an extension
15:42 <vmassol> extensions are of all kinds
15:42 <+tmortagne> mflorea: what kind of packaging is it ? jar ? just some files somewhere ?
15:42 <vmassol> the othe category is good for that
15:42 <vmassol> other
15:42 <+tmortagne> mflorea: there is a lot of stuff you can't install on extensions.xwiki.org
15:42 <vmassol> there's no packaging tmortagne
15:42 <vmassol> you need to compile from sources
15:42 <+tmortagne> like a lot of documentation for example
15:42 <vmassol> it's a "source extension" if you prefer ;)
15:43 <vmassol> (almost like a snippet but not quite)
15:43 <+tmortagne> ok so it's only source
15:43 <+tmortagne> yea kind of a cleaner snippet
15:43 <+mflorea> tmortagne: source code :) . let's say jar, but in order to use it, you have to modify the wysiwyg-client module to add it as a dependency + add a ref to some class, then build wysiwyg client and war
15:44 <+mflorea> vmassol: I can send a vote and move it but it's not a priority now, maybe after the 3.4RC1 release
15:45 <vmassol> sure
15:45 <+mflorea> strange failure on ui-tests, http://ci.xwiki.org/job/xwiki-enterprise-test-ui/900/console , those jars are accessible to me
15:46 <+tmortagne> first time I see a code 502 :)
15:47 <+mflorea> http://www.checkupdown.com/status/E502.html
15:53 <+tmortagne> probably a temporary thing
15:53 <+mflorea> we'll see
16:05 <+tmortagne> I have to go, back around 18h I think
16:06 <+mflorea> tmortagne: I see that manager selenium tests use their own selenium configuration
16:06 <+mflorea> I'll see if I can use the config from enterprise
16:06 <+mflorea> so that we use the same selenium and selenium maven plugin version
16:07 <+tmortagne> mflorea: theses test are very old, probably the project itsself has probably been refactored several time without changing the tests
16:07 <+tmortagne> I doubt it need any special configuration
16:07 <+mflorea> ok, thanks for the info
16:07 <+tmortagne> it's pretty simple standard selenium stuff
16:11 <vmassol> has quit
16:13 <vmassol> has joined #xwiki
16:17 <+sburjan> I see this on GreepPepper website: The macro include now available with the GreenPepper XWiki plugin
16:19 <DrLou> has joined #xwiki
16:25 <+mflorea> vmassol: I'd like to update xwiki-manager to use the same version/config of selenium webdriver / maven plugin. The easiest is to move the corresponding dependencyManagement and pluginManagement sections from xwiki-enterprise-test to xwiki-enterprise pom, so that they are inherited by xwiki-manager and then by xwiki-manager-test pom. WDYT? Any other better idea?
16:25 <vmassol> I don't like it at all
16:25 <vmassol> test stuff should remain in test I think
16:25 <+mflorea> ok, but how do we about duplication?
16:26 <vmassol> well you need to explain the problem first :)
16:26 <+mflorea> maven has the "import" dependency scope, but this will work only for dependencyManagement section, not for pluginManagement
16:26 <vmassol> ok I understand the problem I think
16:26 <vmassol> the solution we're using so far
16:26 <+mflorea> the problem is that the selenium config/version is duplicated between enterprise and manager
16:27 <vmassol> is to put in platform whatever is common
16:27 <vmassol> since if it's used by more than 1 it measn there's a common need
16:27 <vmassol> but this is for java
16:27 <vmassol> for maven I'd need to think more about it
16:27 <vmassol> ah yes
16:28 <vmassol> it could be done as a maven plugin I guess
16:28 <+mflorea> right now manager uses selenium 2.5.0 and selenium maven plugin 2.0, compared with 2.16.1 and 2.2 on enterprise
16:28 <vmassol> mflorea: IMO duplication is ok
16:28 <+mflorea> and the config is duplicated
16:28 <vmassol> when there are no good solution
16:28 <vmassol> till maven introduces mixings
16:28 <vmassol> mixins
16:29 <vmassol> the only solution I can think without looking at details right now is to have our maven plugin for running func tests
16:29 <vmassol> same as what I've done for packaging basically
16:29 <+mflorea> I was looking for a quick solution :)
16:29 <vmassol> quick solution
16:29 <vmassol> is to keep dups
16:30 <vmassol> that's the quickst
16:30 <vmassol> by far
16:30 <vmassol> :)
16:30 <+mflorea> ok, then I'll have to update the selenium version on manager
16:30 <vmassol> depends how much you want to share mflorea
16:31 <vmassol> if it's just the selenium version it's ok
16:31 <+mflorea> at least the selenium version and the selenium maven plugin version
16:31 <vmassol> but plugin mgnt is too much
16:31 <+mflorea> so that I make the upgrade in one place
16:31 <vmassol> and would cause issues
16:31 <vmassol> s/would/could/
16:32 <vmassol> I think this is ok
16:32 <vmassol> actually if you look at our top level pom
16:32 <vmassol> hmm no
16:32 <vmassol> need to check something first
16:32 <vmassol> before saying something wrong
16:32 <+mflorea> :)
16:33 <vmassol> yes it's correct
16:33 <vmassol> so
16:33 <vmassol> in top level pom we have dep mgmt for unit test libraries
16:33 <vmassol> so provided we agree that our main tool for func testing is sel2 (which we do agree ATM I believe)
16:34 <vmassol> then I'm ok to have that info in top level XE pom
16:34 <vmassol> (could even maybe be in a higher level pom)
16:34 <vmassol> in the near future, manager shouldn't depend at all on XE
16:34 <+mflorea> I see
16:35 <vmassol> maybe top level pom is a bit too much
16:35 <vmassol> but maybe platform pom would be ok
16:35 <vmassol> especially since we're now doing func testing in platform
16:35 <vmassol> (with sel2)
16:36 <vmassol> hmm but xe doesn't depend on platform pom
16:36 <+mflorea> but how would they "end up" in manager
16:36 <+mflorea> yep
16:36 <+mflorea> exactly
16:36 <vmassol> so it would have to be in top level pom ...
16:36 <vmassol> which is ok
16:37 <vmassol> s/ok/acceptable
16:37 <+mflorea> :)
16:37 <vmassol> (acceptable till maven introduces mixins)
16:38 <vmassol> wondering how I get the sel2 version in linkchecker tests
16:38 <+sburjan> vmassol: testing 3.3.1 with Oracle. I have a small error when importing the XAR, having to do with Lucene Index Updater and Blog.Catgegories icon.png. Should I report this ? Do you remember having it ?
16:39 <+mflorea> we would move the selenium.version prop from enteprise test to top level pom
16:39 <+mflorea> but why do you need the version?
16:40 <vmassol> i'm trying to find out how I did it...
16:40 <vmassol> ie how do I run sel2 tests on linkchecker modue
16:41 <+mflorea> ah
16:42 <vmassol> ah ok found
16:42 <vmassol> you forgot to update it
16:42 <vmassol> :)
16:42 <vmassol> <!-- Dependencies for the UI test framework -->
16:42 <vmassol> <dependency>
16:42 <vmassol> <groupId>org.seleniumhq.selenium</groupId>
16:42 <vmassol> <artifactId>selenium-java</artifactId>
16:42 <vmassol> <version>2.14.0</version>
16:42 <vmassol> </dependency>
16:42 <vmassol> in platform-test module
16:42 <vmassol> so a top level prop sounds good
16:42 <+mflorea> yep, so it's better to have this in a single place
16:42 <+mflorea> good
16:43 <vmassol> Enygma`: I got the same problem yesterday: http://jira.xwiki.org/browse/XWIKI-7413
16:44 <vmassol> I failed to create a wiki, delete it and recreate it
16:44 <+mflorea> I'll make the move and I'll update manager and linkchecker
16:44 <vmassol> mflorea: ok thanks
16:44 <vmassol> Enygma`: thanks for reporting it! :)
16:44 <vmassol> (I wanted to investigate but you did it, this is cool)
16:45 <+mflorea> btw, ui-tests are hanging at the same place, I think, http://ci.xwiki.org/job/xwiki-enterprise-test-ui/901/org.xwiki.enterprise$xwiki-enterprise-test-ui/console :(
16:45 <+mflorea> LoginTest - testRedirectBackAfterLogin started
16:45 <vmassol> hmm
16:46 <vmassol> we'll have to put lots of SOP or equivalent
16:46 <vmassol> to narrow the issue
16:46 <+Enygma`> vmassol: welcome :)
16:46 <+Enygma`> I think it`s present since at least may 2011
17:12 <evalica> has quit
17:12 <evalica> has joined #xwiki
17:17 <evalica> has quit
17:18 <CIA-121> Vincent Massol master * rd691ed5 / xwiki-enterprise-ui/src/main/resources/XWiki/XWikiSyntax.xml : XE-1089: Add example of email link with a subject - http://git.io/nS1j1A
17:20 <Enygma`> has quit
17:20 <+sburjan> vmassol: didn't even know about this being possible . re the subject to e-mail :)
17:20 <vmassol> it's standard mailto url
17:20 <vmassol> nothing to do with xwiki actually :)
17:22 <+sburjan> yes, I see we have it already documented on syntaxes page.. which is good
17:23 <vmassol> I'm applying the strategy I'd like everyone to follow
17:23 <vmassol> ie when a question is ask
17:23 <vmassol> answer with a link
17:23 <vmassol> *asled
17:23 <vmassol> *asked
17:23 <+sburjan> just helped the first person on the user lists :)
17:24 <vmassol> nice start sburjan
17:25 <vmassol> so guys wdyt about "index" instead of "navigation"… my last but one mail on the list
17:38 <vmassol> tmortagne: any opinion on "index" vs "navigation"? I'd like to progress and one opinion would help ;) (this is a pre req for me to finish the linkchecker ui test)
17:39 <vmassol> btw do you think I should commit this after RC1 or wait for 4.0M1 ?hmm I should probably wait for the branch creation and commit on trunk for 4.0M1
17:39 <+tmortagne> vmassol: I think I prefer "navigation" but I'm ok with "index"
17:39 <vmassol> mflorea: maybe you could create the branch now or soon?
17:39 <vmassol> argh
17:39 <vmassol> I was pretty sure index was better
17:39 <+tmortagne> vmassol: I would wait for the branch
17:40 <+mflorea> I'd like to wait for the build to stabilize
17:40 <+tmortagne> index sounds a bit technical for me but not an english expert either...
17:40 <vmassol> technical?
17:40 <vmassol> it's an index
17:40 <vmassol> as in books
17:40 <vmassol> it's called index in English
17:40 <vmassol> it's not technical
17:40 <+tmortagne> well index in book is very specific
17:40 <vmassol> yep
17:40 <+tmortagne> I think I perefer an application about navugation in general
17:40 <vmassol> this is about listing a collection of something
17:41 <vmassol> to get to it quickly by scanning the index
17:41 <vmassol> the reason I prefer index is that navigation is a lot more generic
17:41 <vmassol> and I fear we're going to put lots of stuff in it
17:41 <+tmortagne> well yes that's why I prefer it :)
17:41 <vmassol> (everyone hainvg its own definition of what navigation is)
17:41 <+tmortagne> it's not that much things either
17:42 <vmassol> right now we have 10 pages
17:42 <vmassol> for all docs + spaceindex + spaces
17:42 <vmassol> it's quite a nice bundle
17:42 <vmassol> nice and tight
17:43 <+tmortagne> what is spaces ?
17:44 <vmassol> list of spaces
17:44 <vmassol> I was hesitating for that one
17:44 <vmassol> but it's still an index
17:44 <vmassol> (index of spaces)
17:46 <+tmortagne> as I said I'm ok with index too, if you are sure it's the best choice I will not fight with my poor english level ;)
17:46 <vmassol> I'm not sure
17:46 <vmassol> :)
17:46 <vmassol> that would be too easy
17:47 <vmassol> I wonder what else we would put in the nav app that wouldn't be in the index app
17:47 <vmassol> spaces is actually the space macro
17:47 <+tmortagne> what would be about navigation but would not fit in index bundle ? do we have anything already ?
17:47 <vmassol> so that one could in a macro app
17:47 <vmassol> hmm we said we didn't want that
17:47 <vmassol> we want business apps
17:47 <vmassol> nothing ATM
17:48 <+tmortagne> yep -1 for macro app
17:48 <+tmortagne> panels app is already crappy enough
17:48 <vmassol> yeah it'll get removed in due time
17:48 <vmassol> well not removed
17:48 <vmassol> but slimmed down
17:48 <+tmortagne> maybe the thing the directly go to a page by name
17:48 <+tmortagne> you know with a shortcut with G
17:48 <vmassol> ah
17:48 <vmassol> imagine we had a Naviagtion Panel
17:48 <+tmortagne> but not sure what is it
17:48 <+tmortagne> maybe a template
17:48 <vmassol> would that go in the index app
17:49 <vmassol> I think it could
17:49 <+tmortagne> what would be a navigation panel ?
17:49 <+tmortagne> a tree ?
17:49 <+tmortagne> like we use to have
17:49 <vmassol> the current one
17:49 <vmassol> yes
17:49 <vmassol> current = the retired one
17:49 <+tmortagne> hope it would not be the same one ;)
17:51 <sburjan> has quit
17:51 <CIA-121> Marius Dumitru Florea master * rcac4d62 / pom.xml : XCOMMONS-78: Add Selenium dependency management to the top level pom - http://git.io/JWNVpg
17:51 <+tmortagne> if the menu was in a page for example it would be in navigation application and not in index probably
17:51 <CIA-121> Marius Dumitru Florea master * r41733c9 / xwiki-platform-core/xwiki-platform-test/pom.xml : XCOMMONS-78: Add Selenium dependency management to the top level pom ...
17:51 <CIA-121> Marius Dumitru Florea master * r83d42d4 / xwiki-enterprise-test/pom.xml : XCOMMONS-78: Add Selenium dependency management to the top level pom ...
17:51 <CIA-121> Marius Dumitru Florea master * r1e4484b / xwiki-manager-test/pom.xml : XCOMMONS-78: Add Selenium dependency management to the top level pom ...
17:53 <vmassol> not sure what you mean tmortagne
17:53 <+tmortagne> the menu
17:53 <+tmortagne> on the top
17:53 <vmassol> I think the meaning I have for navigation app is actually an index app
17:53 <vmassol> which is why I prfer index probably
17:53 <vmassol> ah yes the menu would be in a menu app probably
17:54 <vmassol> or anothyer name
17:54 <vmassol> but nav would be too generic for it I think
17:54 <+tmortagne> if it's a menu app it will not contains much unless the menu is configurable and there is a whole menu configuration UI with it
17:54 <vmassol> yep
17:54 <vmassol> we need it configurable
17:54 <+tmortagne> we need a lot of things :)
17:55 <vmassol> yes :)
17:55 <+tmortagne> xwiki.org seems dead
17:55 <vmassol> we know we'll have this one
17:55 <vmassol> (the configurable menu)
17:55 <vmassol> xwiki.org is ok for me
17:55 <vmassol> (it's even fast here)
17:55 <+tmortagne> right it's www.xwki.org thaat does not work :)
17:56 <+tmortagne> yea it's fast
17:57 <vmassol> mflorea: I'd put some comment to explain that it's for Functional UI tests
17:57 <vmassol> (it's a detail though)
17:57 <+mflorea> in top level pom?
17:57 <vmassol> yes
17:57 <+mflorea> ok
18:02 <vmassol> ok I'll keep navigation for now locally we'll see what others say
18:02 <vmassol> mflorea: when can you create a branch for 3.x?
18:02 <+mflorea> tomorrow
18:03 <vmassol> ok I'll wait and stash till then
18:03 <vmassol> thanks
18:03 <+mflorea> np
18:04 <+mflorea> I don't understand why platform build fails. it complains the version for selenium dep is missing, but it should take it from dep management on commons pom.
18:04 <vmassol> tmortagne: btw all translations keys are currently with xe.index.* for alldocs, spaceindex and co
18:04 <vmassol> (so someone else liked the name "index" at some point ;))
18:04 <+mflorea> it fails on CI, I have no problems locally after building commons poms
18:05 <vmassol> mflorea: you committed in the right order?
18:05 <vmassol> ie top levle pom commit first
18:05 <+mflorea> yes, and I forced a second commons build but it did not help
18:05 <+tmortagne> vmassol: I sais navigation was more generic that that theses page were not about index ;)
18:05 <vmassol> ah actually since you committed in several commit it's normal
18:06 <vmassol> tmortagne: yeah
18:06 <vmassol> mflorea: first commit it top level fine
18:06 <vmassol> so this starts a job on it
18:06 <vmassol> the queue is created only after the top level pom is built
18:06 <+mflorea> ah
18:06 <vmassol> and only for the next descendants
18:06 <vmassol> not the full transitive descendants
18:06 <vmassol> I think
18:07 <vmassol> then the next commit comes
18:07 <vmassol> and there are agents available
18:07 <vmassol> so they try to build it
18:07 <vmassol> if that's the case jenkins' maven integration should be smarter
18:08 <sdumitriu> has joined #xwiki
18:08 <+mflorea> it's a pain because we're building so much, and sometimes is unnecessary. For instance if I commit a fix for a wysiwyg test, this triggers enterprise build and then all tests are run again, while I need only wysiwyg tests to run again
18:08 <vmassol> yep, we should investigate this a bit to see if there's something planned or that we can do
18:16 <+mflorea> vmassol: I believe you change to syntax help breaks some Dutch WebGuidelines http://ci.xwiki.org/job/xwiki-enterprise-test-webstandards/966/
18:16 <vmassol> sh****t
18:16 <vmassol> I took a measure risk in not checking
18:16 <+mflorea> :)
18:17 <abusenius> has joined #xwiki
18:17 <vmassol> moore's law....
18:18 <+mflorea> looks like you are forced to show the email address in the label
18:18 <vmassol> yeah I need to understand that a bit more
18:18 <vmassol> and why it wasn't breaking before
18:18 <vmassol> I just added a second example with ?subject=...
18:19 <+mflorea> I guess it considers ?subject part of the email address and it complains that this part is not visible in the label..
18:20 <vmassol> ah indeed I forgot that I had the label matching the mailto
18:20 <vmassol> probably for the same reason
18:20 <vmassol> hmm might be a bug of the checker
18:24 <vmassol> /**
18:24 <vmassol> * Links to e-mail addresses: the e-mail address to which the message is addressed must be visible in the link text.
18:24 <vmassol> */
18:24 <vmassol> yeah I think it's a bug of the checker
18:24 <+mflorea> indeed
18:25 <vmassol> fixing
18:25 <vmassol> hmm
18:25 <vmassol> I see a second check below which we might hit
18:25 <vmassol> /**
18:25 <vmassol> * Links to e-mail addresses: the URL in the href attribute of a link to an e-mail address may only contain the
18:25 <vmassol> * mailto protocol and an e-mail address.
18:25 <vmassol> */
18:26 <+mflorea> yes, this is listed in the test failure message
18:26 <vmassol> so this ones seems to say that I've done shouldn't be done at all
18:26 <vmassol> *that what
18:27 <vmassol> hehe such a simple commit...
18:27 <vmassol> apparently
18:27 <+mflorea> yes, this means you can't pass a query string to the email address
18:27 <vmassol> so I have 2 choices
18:28 <vmassol> - revert the commit without understanding more and instead add a FAQ entry explaining that it's possible but not WCAG compliant
18:28 <vmassol> - try to understand why it's forbidden
18:28 <vmassol> I need to understand it if I want to write the FAQ entry anyway
18:28 <vmassol> ;)
18:28 <vmassol> sh*****it
18:28 <vmassol> fun....
18:29 <vmassol> jvdrean: you wouldn't remember why we have this by any chance?
18:29 <vmassol> (I mean except that it was in the dutch algo we got)
18:30 <vmassol> ah I'll try to run the dutch checker online
18:30 <vmassol> see if it has more info
18:32 <vmassol> tmortagne: you're configuring jenkins re dep building strategy?
18:32 <+tmortagne> vmassol: yep, I'm not done yet (been grabbed on skype)
18:33 <vmassol> found doc: http://versie1.webrichtlijnen.nl/english/manual/development/production/link-navigation/email-addresses/#r-pd-8-17
18:33 <vmassol> reading
18:33 <vmassol> ok so the only reason is:
18:33 <vmassol> "Additional parameters behind the email address – for instance, for indicating the subject line or the text of the email message – are not supported by all email programmes."
18:34 <vmassol> ok so they're recommending to use a form instead of a link
18:34 <vmassol> mflorea: are you also ok that I revert my commit?
18:35 <+mflorea> yes
18:35 <vmassol> ok doing it
18:38 <+tmortagne> vmassol: jenkins iss all set
18:39 <+tmortagne> s/iss/is/
18:42 <+jvdrean> just caught up
18:45 <vmassol> jvdrean: it's ok
18:45 <vmassol> problem solved
18:46 <CIA-121> Vincent Massol master * rcdce468 / xwiki-enterprise-ui/src/main/resources/XWiki/XWikiSyntax.xml : XE-1089: Add example of email link with a subject ...
18:47 <+jvdrean> yep, I've seen it was. So the first test was wrong right ?
18:47 <+jvdrean> did I just had the same test for both rules ?
18:47 <vmassol> yeah kind of but it doesn't matter
18:48 <vmassol> since it kind of validates the second rule too
18:48 <+jvdrean> yes
18:48 <vmassol> it would be better to fix it I guess
18:48 <vmassol> if you have time
18:48 <vmassol> ...
18:54 <vmassol> hmm where is the share page feature located in the code base?
18:55 <vmassol> (also looking for where it's documented on xwiki.org)
18:59 <vmassol> sburjan`: any idea?
18:59 <vmassol> found it
18:59 <vmassol> http://platform.xwiki.org/xwiki/bin/view/Features/DocumentLifecycle#HSharePagebyMail
19:00 <tmortagne> has left #xwiki
19:01 <+jvdrean> vmassol: fixed
19:01 <CIA-121> Jean-Vincent Drean master * r03f5089 / (2 files in 2 dirs): [misc] Improved Dutch Web Guidelines test on mailto links - http://git.io/47tGZQ
19:01 <vmassol> oool thanks :)
19:13 <evalica> has joined #xwiki
19:17 <+mflorea> cool, I can run tests exactly as out CI agents with xvfb-run
19:18 <evalica> has quit
19:39 <vmassol> that's nice
19:39 <vmassol> can you reproduce the issue mflorea?
19:39 <vmassol> (the hanging)
19:40 <+mflorea> I haven't tried, but I can reproduce the failing wysiwyg tests which pass otherwise
19:40 <vmassol> that's great
19:40 <+mflorea> without xvfb
19:40 <+mflorea> I'll try later the ui-tests
19:41 <+mflorea> the good part is that I can leave them run in background and work on something else in the mean time
19:45 <jvdrean> has quit
19:47 <vmassol> has quit
19:52 <lucaa> has quit
21:05 <pgmjsd> has quit
21:09 <jvdrean> has joined #xwiki
21:22 <vmassol> has joined #xwiki
21:26 <pgmjsd> has joined #xwiki
22:22 <rrodriguez> has joined #xwiki
22:52 <jvdrean> has quit
23:06 <abusenius> has quit
23:08 <mflorea> has quit
23:58 <jvdrean> has joined #xwiki
