IRC Archive for channel #xwiki

00:01 <lucaa1> has joined #xwiki
00:01 <lucaa> has quit
00:25 <lucaa1> has quit
00:48 <evalica> has quit
01:24 <roxxo> Would anyone happen to have a suggestion on where to start debugging a traceback received on a page export->RTF? I experienced it with 2.4.3, then upgraded to 2.7.0 today, and still get the same traceback
01:25 <cjdelisle> Can you pastebin the stack trace?
01:25 <roxxo> cjdelisle: sure, one sec
01:27 <roxxo> http://pastebin.com/4rhT4LmW
01:30 <cjdelisle> !jira jodconverter
01:31 <cjdelisle> it's a popular exception
01:31 <cjdelisle> https://encrypted.google.com/search?q=at+org.artofsolving.jodconverter.office.PooledOfficeManager.execute(PooledOfficeManager.java%3A97)
01:32 <roxxo> oh wow
01:33 <roxxo> I'll investigate that then, thanks.
01:33 <cjdelisle> I think the thing to do is come back in about 10-12 hours and the people who know all about the office import/export will be here.
01:33 <cjdelisle> Most everyone is asleep right now.
01:33 <roxxo> ah, will do. are they on GMT?
01:34 <cjdelisle> Paris and Romania mostly.
01:34 <cjdelisle> s/measurments/zones/
01:42 <roxxo> ah well, I shall return tomorrow - peace out
01:44 <cjdelisle> good night/morning/whatever it is :)
02:02 <Pon> oo,
02:02 <Pon> its good morning, its 10am here in Tokyo.
02:03 <cjdelisle> did you get an answer to your wuestion about link rewriting?
02:03 <abusenius> has quit
02:03 <Pon> dont think so.
02:03 <Pon> oh, ShortURLs?
02:03 <Pon> I tried that
02:04 <cjdelisle> I know it can be done since xwiki.org uses it but I have not had to deal with that code specifically so I am feeling around a little bit here.
02:04 <Pon> I'm wondering why full urls are included in stuff at all, surely everything should be relative to some point.
02:05 <Pon> eg, instead of   the xredirect=http://whatever   it should be without the host qualifier.
02:05 <Pon> same for the wysiwyg
02:05 <cjdelisle> Hmm. we do use relitive urls. Perhaps zeus is rewriting in the wrong direction?
02:06 <Pon> you dont though,
02:06 <Pon> even bypassing zeus (only works on my 10. net) I see a full url showing up in  the login bit
02:06 <cjdelisle> I see, logout is not reletive.
02:06 <cjdelisle> (xredirect)
02:07 <Pon> y
02:07 <Pon> and the editor one is hidden inside javascript (an awesome 2mb page!!!)
02:07 <Pon> gah, >100 emails since last night.
02:07 <cjdelisle> Yes that is indeed awesome, it is generated by google gwt.
02:08 <cjdelisle> I'm sorry you ever had to look at that, it's about equivilent to machine code.
02:09 <Pon> its quite impressively bloated.
02:09 <Pon> I think the javascript needs to go on a total starvation diet.
02:10 <Pon> well, in the interests of staying current, I will try and get M2 and put that on.
02:10 <Pon> if it works and stuff, I can probably work on making a Japanese translation ^^;
02:11 <cjdelisle> I like js, I don't like the way it is most often used but I like the idea of moving most presentation work over to the user side.
02:11 <cjdelisle> Cool re Japanese translation.
02:11 <Pon> y, get the far end to do the heavy lifting.
02:11 <cjdelisle> I'm restarting my server now to see if I can reproduce the issue.
02:12 <Pon> shouldnt have to restart it ^^;
02:12 <Pon> it seems pretty consistent
02:13 <cjdelisle> Well I'm trying to make it use localhost instead of 127.0.0.1 (since I don't have a proxy handy)
02:14 <Pon> either should work ^^;
02:14 <Pon> basically the full name of the request is going in there ^^;
02:14 <cjdelisle> hmm I'm getting a bunch of:
02:14 <cjdelisle> Illegal character in path at index 48: /skins/colibri/$xwiki.getSkinFile('colibri.css',
02:15 <cjdelisle> in the log
02:15 <Pon> wao
02:18 <Pon> did you turn some extra logging on?
02:18 <sburjan`> has quit
02:20 <cjdelisle> No, that's an exception so it should show up.
02:24 <Pon> ok, I'll fire up M2,
02:24 <Pon> I'm certain I was seeing exactly the same issues in M1 as 2.7 so I wonder if anything has changed.
02:25 <cjdelisle> Nothing has changed in that code AFAIK.
02:25 <Pon> well, we'll soon see. shouldnt take more than about 15 minutes ^^;
02:30 <Pon> I see this wonderful error in the server log
02:33 <Pon> and on access, stacks of INFO complaining about velocity
02:33 <Pon> [#|2011-02-10T10:31:45.710+0900|INFO|glassfish3.0.1|javax.enterprise.system.std.com.sun.enterprise.v3.services.impl|_ThreadID=154;_ThreadName=Thread-1;|2011-02-10 10:31:45,710 [http://localhost:8080/xwiki/bin/view/Main/] WARN  internal.DefaultVelocityEngine  - Deprecated usage of method [com.xpn.xwiki.api.XWiki.getXMLEncoded] in xwiki:Main.WebHome@2,42
02:33 <cjdelisle> Yes, core moves faster than templates.
02:34 <Pon> ok the "class="tme" still references localhost.
02:34 <cjdelisle> We need to go through with a find -exec sed command and auto fix those.
02:34 <Pon> maybe you need a bug report so that people know its there ^^;
02:35 <Pon> cor, clicking on "blog" produced a few thousand lines of error!
02:36 <cjdelisle> I think that is in the release notes, the old xar does not work with the new .war.
02:36 <cjdelisle> Velocity changes the way they handeled references and it broke lots of code.
02:37 <cjdelisle> Updating to the newest version of the wiki content should get rid of _most_ of that (there are still a few deprecation warnings)
02:37 <Pon> ok
02:38 <Pon> I like the loop error there, that was 30000 lines of log ^^;
02:38 <cjdelisle> com.xpn.xwiki.XWiki.formatDate <-- that?
02:39 <Pon> y
02:40 <cjdelisle> Is zeus on the same box as xwiki?
02:40 <Pon> y
02:40 <Pon> zeus runs a whole pile of sites,
02:40 <Pon> it has a gateway for /xwiki    -->  localhost:8080   (glassfish v3 instance)
02:41 <Pon> I wish it would tell me what URL it was being asked for
02:41 <Pon> ok, you can play if you want,
02:41 <Pon> http://www.anime.org/xwiki/
02:42 <cjdelisle> that domain must have been in high demand.
02:42 <Pon> cor, the admin interface looks a lot nicer.
02:42 <Pon> I've had it for oh, 15 years now.
02:42 <Pon> it has a few users,
02:42 <Pon> a cosplayer, an anime director ...
02:45 <cjdelisle> ok
02:45 <Pon> should I open a bug report with these? would that help?
02:45 <Pon> go to LOG-IN
02:45 <cjdelisle> Do you know if you can get zeus to rewrite this?
02:45 <cjdelisle> Host: 127.0.0.1:8081
02:45 <Pon> if you look in the source for that page,
02:45 <Pon> <input name="xredirect" type="hidden" value="http://localhost:8080/xwiki/bin/WebHome"/>
02:45 <cjdelisle> I think the problem is the Host: header.
02:46 <Pon> I would love to know what it is actually sending though.
02:46 <Pon> through
02:46 <cjdelisle> tcpdump -i lo -A -s 5000 | grep 'Host:'
02:46 <cjdelisle> That should be able to confirm it.
02:47 <Pon> on ZWS, all you can specify for the gateway is   path -> host/port
02:47 <Pon> and rewrite stuff.
02:47 <cjdelisle> requrit headers?
02:47 <Pon> the rewrite engine is elsewhere.
02:48 <Pon> I think it gets called after.
02:48 <Pon> anwyay, I'll try that tcpdump first
02:48 <cjdelisle> I think you'll see a lot of this:
02:48 <cjdelisle> Host: 127.0.0.1:8081
02:49 <Pon> Host: localhost:8080
02:50 <cjdelisle> Although I don't know for sure, I think the people who are running xwiki behind proxies are using proxies which do not rewrite the header so the external header makes it in to the xwiki instance.
02:51 <Pon> probably
02:51 <Pon> makes it a little difficult to use with ZWS in the way ^^;
02:51 <Pon> So, we know where it gets its URL base from,
02:53 <Pon> the rewrite engine comes too early to help. if I rewrite the Host: line there, it will be over written.
02:53 <Pon> still, progress ^^;
02:53 <cjdelisle> http://knowledgehub.zeus.com/articles/2010/04/05/augmenting_the_web_with_zeus_traffic_man
02:54 <cjdelisle> http.setHeader( "Host", $host );
02:54 <Pon> traffic manager costs serious money.
02:54 <cjdelisle> Ahh, I see.
02:55 <cjdelisle> hehe ngrep!
02:55 <Pon> I have had this zeus license since it was beta (I still have the official beta tester backpack, and boxers!)
02:55 <Pon> "Zeus Web Server - Godlike Performance" (tacky ^^!)
02:56 <cjdelisle> hmm
02:58 <cjdelisle> http://tardate.blogspot.com/2007/03/request-header-rewrites-with-java.html
02:58 <cjdelisle> A servlet filter, this looks doable
02:59 <Pon> I would like just an Xwiki.fakehost=<whatever> ^^;
03:00 <cjdelisle> xwiki.home should be that but I suspect it might have been broken when the url factory was rewritten.
03:01 <cjdelisle> No tests broke and it was a little used function so it was forgotten.
03:01 <cjdelisle> (this is just what I suspect, I have nothing to go on.)
03:01 <Pon> I set xwiki.home=http://www.anime.org/
03:01 <Pon> no luck there at all
03:04 <Pon> ok fail on zws rewrite engine.
03:05 <Pon> I should report this as a zeus bug (feature!) too
03:05 <cjdelisle> I'm looking at the code right now, it looks like we have a bug.
03:06 <cjdelisle> But I should be able to make a modified class file you can drop in and get it running.
03:06 <Pon> must be nice to be able to read java.
03:06 <Pon> wao
03:12 <cjdelisle> ok, I see what the problem is.
03:13 <cjdelisle> That feature doesn't work unless used with a multiwiki "farm" which also explains why it was not noticed when it stopped working.
03:14 <cjdelisle> Most users who are using a proxy are also running multi wiki.
03:14 <Pon> so maybe if I turn on multiwiki and do nothing else, it will just work?
03:16 <cjdelisle> It only seems to work for subwikis.
03:16 <cjdelisle> Often the main wiki will be made unavailable for security reasons.
03:19 <cjdelisle> !jira xwiki.home parameter
03:22 <roxxo> has quit
03:27 <Pon> if you want a bug report, I can file one for you ^^;
03:31 <cjdelisle> I have a fix but it breaks 2 tests and I want to make sure it is safe.
03:32 <Pon> amazing ^^;
03:32 <Pon> UTF8?????????
03:32 <Pon> that will give the irc bot something to worry about ^^;
03:35 <cjdelisle> tcpdump -i lo -A -s 5000 | grep 'x-forwarded-host'
03:35 <cjdelisle> try that one
03:48 <Pon> nothing
03:49 <Pon> nope, not a sausage
03:49 <cjdelisle> It seems that xwiki is expecting x-forwarded-host header to exist and zeus is not sending it.
03:50 <Pon> what should be in that field?
03:50 <Pon> I can maybe fake that.
03:50 <cjdelisle> it should be whatever host the user went to (anim.org)
03:50 <Pon> like that?
03:51 <Pon> oo
03:51 <cjdelisle> I think so.
03:51 <Pon> that seems to fix it
03:51 <Pon> have a poke at the site if you want.
03:51 <Pon> http://www.anime.org/xwiki/
03:51 <cjdelisle> I neeeed sleeo
03:51 <cjdelisle> *sleep
03:51 <Pon> oh right, it must be close to midnight for you
03:52 <cjdelisle> only 10 but I go to bed and get up early for my time zone.
03:52 <roxxo> has joined #xwiki
03:53 <cjdelisle> Go ahead and open a bug on that since, if nothing else, the description of xwiki.home should be made more explainitory.
03:53 <cjdelisle> good night.
03:53 <cjdelisle>  /morning
03:54 <Pon> ok.
03:54 <Pon> thank you for all the help ^^!
03:54 <cjdelisle> you're welcome
03:56 <Pon> I guess that would be xwiki core right?
05:24 <venkatesh> has joined #xwiki
05:25 <venkatesh> has quit
05:25 <venkatesh> has joined #xwiki
05:33 <Lockie> has joined #xwiki
06:01 <Denis> has quit
06:09 <Denis> has joined #xwiki
07:00 <Lockie> has quit
07:20 <sburjan`> has joined #xwiki
07:50 <rstavro> has joined #xwiki
07:53 <vmassol> has joined #xwiki
07:59 <vmassol> bug fixing day!
07:59 <Pon> all the velocity "deprecated" errors ...?
08:00 <mflorea> has joined #xwiki
08:10 <jstoldt> has joined #xwiki
08:11 <jstoldt> good morning
08:14 <Pon> well, its almost evening here ^^!
08:14 <Pon> oh, the other day, today I got help from djcelisle and it seems to work ^^!
08:43 <AndreasJonsson> has joined #xwiki
08:47 <Pon> 18879 root     1678M 1498M sleep   59    0   0:46:20 0.1% java/352
08:47 <Pon> go go java!
08:54 <sburjan> has joined #xwiki
08:54 <+sburjan> good morning
08:56 <Pon> evening!
08:56 <Pon> almost 5pm!
09:03 <jstoldt> Pon: what did you do to get it working?
09:03 <jstoldt> just out of interest
09:03 <rstavro> has quit
09:03 <Pon> apparently, its expecting  x-forwarded-host   or so to contain its "outside" URL.
09:04 <jvdrean> has joined #xwiki
09:04 <Pon> so persuading zws to put that in made it work.
09:04 <Pon> cjdelisle was a big help!
09:04 <jstoldt> so it was a problem with the gateway after all?
09:09 <Denis> has quit
09:10 <Pon> yes and no
09:10 <Pon> xwiki.home doesn't work ^^;
09:29 <arkub> has joined #xwiki
09:52 <jvdrean> has quit
09:53 <tmortagne> has joined #xwiki
09:55 <rstavro> has joined #xwiki
10:01 <jstoldt> mflorea:
10:01 <+mflorea> Hi
10:01 <jstoldt> found a bug in the office importer app
10:01 <jstoldt> hi
10:01 <jstoldt> when you choose a non-ascii name, the name will be created with a wrong/broken title
10:02 <jstoldt> trying to figure out if i already reported that
10:02 <jstoldt> in jira
10:02 <+mflorea> what name are you referring to?
10:02 <+mflorea> target page name?
10:02 <jstoldt> exactly
10:02 <jstoldt> also, i noticed that the title of the target page is empty when using the importer app
10:03 <jstoldt> i.e. create page from office doc
10:03 <+mflorea> re non-ascii chars, look at the office import app in jira http://jira.xwiki.org/jira/browse/XAOFFICE
10:04 <jstoldt> ^^
10:04 <jstoldt> also note that i closed XWIKI-5706
10:05 <+mflorea> re page title, are you sure it happens for all docs, I think it looks for the first heading in the doc content and sets it as title
10:05 <jstoldt> okay, that might be it... because i tried presentations
10:05 <jstoldt> still, room for improvement there
10:06 <+mflorea> yes, but where should I take the title from in case of a presentation
10:06 <+mflorea> I have only slide images
10:06 <jstoldt> well, when creating a page via a "blind link" it sets the title to what the page name is, automatically
10:06 <+mflorea> I see
10:06 <jstoldt> which is something i'd also expect for the importer app
10:07 <+mflorea> you can open a jira issue then, re page title for office presentation import
10:07 <jstoldt> will do
10:08 <+mflorea> thanks
10:12 <jstoldt> XAOFFICE-58
10:16 <jstoldt> if i mark issues as "duplicates", should i close them?
10:17 <+tmortagne> jstoldt: yep
10:17 <jstoldt> okay
10:17 <+tmortagne> and properly link it to the duplicated issue
10:17 <+mflorea> yes, link as duplicate and close as duplicate
10:18 <jstoldt> hmmm can't close XAOFFICE-45
10:18 <jstoldt> Neither XAOFFICE-49
10:18 <+mflorea> I can do it
10:19 <jstoldt> ok
10:19 <jstoldt> i linked them already
10:20 <+mflorea> done
10:20 <jstoldt> grea :)
10:20 <jstoldt> *great
10:21 <jstoldt> going to create an issue for the title char thing i reported
10:21 <jstoldt> XAOFFICE-33 relates to it but it's not the exact same thing
10:24 <jstoldt> XAOFFICE-59
10:33 <jvdrean> has joined #xwiki
10:40 <lucaa> has joined #xwiki
10:41 <jvelociter> has joined #xwiki
10:45 <jstoldt> btw, i think sdumitriu or somebody else dropped a line about a bugfix version being released on thursday the other day
10:45 <jstoldt> does that mean there will be a 2.7.1 release soon?
10:46 <+mflorea> yes
10:46 <Enygma`> has joined #xwiki
10:46 <+mflorea> check the devs mailing list
10:49 <jstoldt> nice
10:50 <+sburjan> I found something odd. I have everything updated on my ui (ran svn up and maven stat) .. nothing new, everything seems updated. But my eclipse has some errors in EditObjectTest.java .. on method clickEditObject. I also checked the code from SVN, and it's the same. Odd thing is that that method doesn't exist anymore because vmassol refactored all the clickEdit methods (removed edit) from method names.
10:50 <evalica> has joined #xwiki
10:51 <+sburjan> My question is .. how the hell hudson didn't failed the test
10:52 <+tmortagne> sburjan: well hudson fail to build
10:53 <+tmortagne> it did not even started the tests
10:53 <+tmortagne> which is normal if it has  the same issue you have
10:53 <+sburjan> I am looking at Build #365 (Feb 9, 2011 1:11:48 AM)
10:53 <+tmortagne> yes build #365 is red as far as i can see
10:54 <+tmortagne> did not looked at the details but it definitely failed
10:54 <+sburjan> oh so red spheres means failed build ? and test not run ?
10:54 <+tmortagne> yes
10:54 <+sburjan> tmortagne: thanks ;) didn't know
10:54 <+sburjan> now it makes sense :)
10:54 <+tmortagne> red: failed build
10:54 <+tmortagne> orange: failed tests
10:54 <+tmortagne> green: ok
10:54 <+sburjan> I was waiting to see a stack trace or something
10:55 <+tmortagne> i mean blue
10:55 <+tmortagne> not green
10:55 <+tmortagne> i agree that there is not much blue on XE tests :)
10:55 <+sburjan> :)
10:55 <+sburjan> so in this case, should I create a patch to fix the build ?
10:55 <+sburjan> I only have to change 2-3 lines
10:56 <+sburjan> need a committer :P
10:57 <+tmortagne> in which project is the issue ?
10:57 <+tmortagne> ui-tests ?
10:57 <+sburjan> yes
10:57 <+sburjan> class EditObjectsTest.java
10:57 <+tmortagne> ok openning it
10:58 <+tmortagne> you already know what to change for sure ?
10:58 <+sburjan> lines: ObjectEditPage oep = vp.clickEditObjects(); and  oep = vp.clickEditObjects();
10:58 <+tmortagne> i can do it in that case (i don't know this project very well)
10:58 <+sburjan> clickEditObjects() was refactored to EditObjects()
10:59 <+sburjan> looking for the source of the issue
10:59 <+sburjan> I think vmassol didn't notice those
10:59 <+sburjan> just a sec
11:00 <+sburjan> 34486 08.02.2011 15:15:20, by vmassol  this is the commit that changed the methods
11:00 <+sburjan> I see vmassol refactored that class to
11:01 <+sburjan> so I guess someone committed over it with the older values, can't figure it out
11:02 <+tmortagne> so what is the patch exactly ?
11:02 <+sburjan> 34568 08.02.2011 23:35:34, by cjdelisle
11:02 <+sburjan> this commit broke it :)
11:03 <+sburjan> yey, found it
11:03 <+sburjan> so basically you have to replace :
11:03 <+sburjan> 1. ObjectEditPage oep = vp.clickEditObjects(); -> ObjectEditPage oep = vp.editObjects();
11:04 <+sburjan> 2. oep = vp.clickEditObjects(); -> oep = vp.editObjects();
11:04 <+sburjan> these 2 lines from EditObjectsTest.java class
11:06 <+sburjan> IMO this is too small for an issue ... cjdelisle didn't have an issue also
11:08 <+tmortagne> ok fixing it
11:09 <+tmortagne> not sure this method is really finished
11:09 <+evalica> cjdelisle: are you around? can you please change the status of the channel adding "Bug Fixing Day" ? :)
11:09 <+tmortagne> there is an unused variable
11:09 <+tmortagne> objectForm
11:10 <+tmortagne> anyway fixing the build for now
11:11 <+tmortagne> sburjan: should be ok now
11:13 <+tmortagne> mflorea: did you know http://webodf.org/ ?
11:13 <+tmortagne> thinking about office viewer
11:14 <+tmortagne> seems nicer than convert it to html if it's working well
11:14 <+mflorea> tmortagne: sounds interesting
11:19 <vmassol> guys don't forget to use the bugfixingday tag in jira
11:25 <+evalica> vmassol: we are gonna use that just for closed issues right?
11:26 <vmassol> yes
11:26 <+sburjan> also we could look for old issues to see if they are still available
11:52 <abusenius> has joined #xwiki
11:58 <+evalica> we should fix thing today right? in order to see an improvement - not report everything I see :P
11:59 <vmassol> yes that's worse :)
11:59 <vmassol> because that makes the bug count grow instead of reducing
12:09 <jstoldt> iirc sergiu also mentioned reporting bugs would be good ;)
12:09 <jstoldt> maybe i can come  up with some more stuff :-P
12:10 <vmassol> well the general idea is to catch up on our fixed bugs vs newly created bugs
12:10 <jstoldt> i know ;)
12:10 <vmassol> http://jira.xwiki.org/jira/secure/Dashboard.jspa?selectPageId=10000#Created-vs-Resolved-Chart/10470
12:11 <vmassol> we need to fix 80 to be even
12:11 <vmassol> for the 1 year period
12:11 <+tmortagne> now if there is bug to  report not sure it's good to not do it...
12:12 <jstoldt> well, i usually ask you if i should report it after i mention it here and you keep telling me to create issues, so don't shoot the messenger. ;-P
12:14 <+sburjan> Ubuntu 11.04 will ship with LibreOffice instead of OpenOffice, so we'll have to support both :(
12:14 <jstoldt> sburjan: not a real bother
12:14 <jstoldt> i am using LibO for a number of weeks already
12:14 <+sburjan> not now, because they still are similar, but they co in opposite direction .. talking about implementation
12:14 <jstoldt> and the only bother was that i still needed some old config files from OOo
12:14 <vmassol> tmortagne: right but if someone wants to participate in tis bugfixingday it's aobut fixing
12:15 <+sburjan> so LibO is better than OO ?
12:15 <+tmortagne> sburjan: like most of other linux distrib and anyone who want to use the right one ;)
12:15 <vmassol> can we can do a bugreportingday another day ;)
12:15 <jstoldt> imo yes, sburjan
12:15 <vmassol> s/can/we can/
12:15 <jstoldt> i've actually been using go-oo for some time already
12:15 <+tmortagne> vmassol: sure it's not a day to search for bugs but we shouldd not say to people to not report any bug if they have one
12:15 <vmassol> jstoldt: if you want to help for this day you should fix some small stuff, could be documentation improvements, or other small stuff
12:15 <jstoldt> and i think it was really an improvement
12:16 <vmassol> tmortagne: of course we're not going to refuse people fillng in bugs
12:16 <vmassol> but they won't be participating to the bugfixingday initiative that's all
12:16 <jstoldt> vmassol: just point me to something... reviewing 850 tickets is already a days work for me...
12:16 <+sburjan> tmortagne, jstoldt : yes, but I am saying that LibO will come bundled with 11.04 instead of OO. IMO this is a big step .. Cannonical switched from OO to LibO
12:16 <+sburjan> ofc you can install whi you want in the end
12:16 <jstoldt> sburjan: they said so months back
12:16 <+sburjan> I was talking about the bunde :P
12:17 <+sburjan> *bundle
12:17 <jstoldt> so it's really no suprise here
12:17 <+tmortagne> sburjan:cannonical is part of Libo contributors so it was more or less to come ;)
12:17 <jstoldt> besides, i usually opt for the more open source kind of software
12:17 <+sburjan> didn't knew that. All i know Cannonical has Goo,  a custom OO
12:17 <+sburjan> a modified OO more exactly
12:18 <AndreasJonsson> I looked a bit at the old XWIKI-228, I recommend closing it without fix.
12:18 <jstoldt> go-oo was a novel distribution
12:18 <+tmortagne> they custom pretty much everything anyway
12:18 <+tmortagne> to add reporting tools etc.
12:19 <+sburjan> I wonder how they keep keeping up with so many custom-applications
12:19 <jstoldt> hmmm i really see that more bugs are being reported since i started working with xwiki... funny the coincidence :-D
12:19 <+sburjan> the Cannonical guys .. cloning appz and customizing for their own distro
12:19 <+sburjan> I never understood how this works
12:19 <jstoldt> prolly lots of automization...
12:19 <+tmortagne> yep
12:20 <+tmortagne> for most of application i think they patch them during the build
12:20 <+sburjan> they would need an army on how many appz are in main universe
12:20 <+tmortagne> and usually they contribute directly to theses application for there needs anyway
12:20 <+tmortagne> they just don't wait for the patch to be applied
12:20 <jstoldt> or use own repositories to which the pull changes from the official reps and only review them if problems occur
12:23 <+evalica> AndreasJonsson: about XWIKI-228 can't test it - but if you are saying it works for you - you should add a comment to the issues saying that you couldn't reproduce the problem
12:25 <+sburjan> tmortagne: seems ui-tests are building now :P
12:25 <+tmortagne> sburjan: I hope so :)
12:25 <+sburjan> we have failed tests, but not failed build :)
12:26 <SvenDowideit> has quit
12:26 <AndreasJonsson> evalica: I have not tested it, only inspected the code.  Lots of changes have been made since the issue was reported. Doesn't seem likely that the error is still there.
12:28 <SvenDowideit> has joined #xwiki
12:28 <+evalica> AndreasJonsson:  if you comment on it - the update will appear on [email protected] and if a commiter knows about the problem and agrees - will close the issue
12:29 <SvenDowideit> has quit
12:30 <SvenDowideit> has joined #xwiki
12:38 <esdaniel> has joined #xwiki
12:44 <esdaniel> jvelociter: nice script: XWiki-plus-groovy-is-love-the-10-lines-RSS-aggregator-macro - does it require any dependencies? Having trouble tyring to invoke on a clean-install
12:45 <vmassol> hmm can we ban his ip or sometrhing?
12:47 <abusenius> has quit
12:48 <+jvelociter> esdaniel: haven't tried it for a while
12:49 <+jvelociter> but there should not be much changes
12:49 <+jvelociter> there are no other dependencies than the ones already in XE
12:49 <+jvelociter> I can try it on a fresh install later on
12:55 <esdaniel> jvelociter: thanks for clarifying, i suspect it's a case of 'user' error and i need to straighten that out, do let me know your test works though, please.
12:57 <+jvelociter> ok, no pb
13:05 <jstoldt> I'll be out for today
13:05 <jstoldt> bye
13:12 <jstoldt> has quit
13:14 <mflorea> has quit
13:29 <abusenius> has joined #xwiki
13:31 <cjdelisle> good morning/afternoon
13:32 <sdumitriu> has joined #xwiki
13:32 <cjdelisle> 05:16 <+evalica> cjdelisle: are you around? can you please change the status of the channel adding "Bug Fixing Day" ?  :)
13:33 <+evalica> yeah - you were admin at that hour :P
13:33 <+evalica> and it's "Bug Fixing Day" yey
13:33 <cjdelisle> s/admin/asleep ;)
13:33 <+evalica> so now ... what do you think? ;)
13:34 <+evalica> *operator
13:34 <cjdelisle> http://www.xwiki.org/ | Bug Fixing Day! | Most developers are online usually between 08-21 GMT | This channel is logged http://dev.xwiki.org/xwiki/bin/IRC/
13:34 <cjdelisle> ?
13:34 <cjdelisle> gotta look up how long a topic can be, I think it's pushing it
13:34 <+evalica> or just put it first :P
13:36 <+sburjan> cjdelisle:  it fits
13:36 <cjdelisle> lol chanserv
13:36 <cjdelisle> was going to let you guts do it, the bot says otherwise
13:36 <cjdelisle> s/u/y/
13:36 <cjdelisle> argh
13:37 <cjdelisle> every morning the keys on my keyboard are moved around!
13:37 <+sburjan> cjdelisle:  /topic http://www.xwiki.org/ | Bug Fixing Day! | Most developers are online usually between 08-21 GMT | This channel is logged http://dev.xwiki.org/xwiki/bin/IRC/
13:38 <+evalica> cjdelisle: Thank you :)
13:38 <cjdelisle> Might be a cool idea to change the topic when releases come out aka #postgresql
13:39 <+evalica> yeap
13:40 <cjdelisle> XWIKI-5981
13:40 <cjdelisle> I have a patch but I don't know if it's safe.
13:41 <cjdelisle> I need to know what the difference is between XWikiContext.getOriginalDatabase() anf XWikiContext.getMainWiki()
13:46 <vmassol> lucaa: not sure it's good to use the com.xpn package in the dashboard macro. For ex it's not been used in the annotation-io lodule
13:46 <vmassol> s/lodule/module/
13:47 <cjdelisle> haha so it's not just my keyboard
13:48 <vmassol> :)
13:48 <vmassol> restarting adium, it's really slow here
13:48 <vmassol> has quit
13:50 <vmassol> has joined #xwiki
13:50 <cjdelisle> http://pastebin.com/ULLHFTsr <-- I made this change locally, it should solve XWIKI-5981 and passes all tests but I'm not confident in it at all.
13:51 <cjdelisle> Maybe tmortagne or mflorea or whoever knows a lot about urls could take a look.
13:52 <cjdelisle> The problem is that for a single wiki, changing xwiki.home has no effect.
13:52 <cjdelisle> (maybe that is intentional in which case all we need is documentation)
13:53 <cjdelisle> bbiab, coffee
14:16 <DrLou> has joined #xwiki
14:30 <venkatesh> has quit
14:37 <vmassol> #bugfixingday is currently loosing: Week 8, 2010, 15 open, 9 closed…. come on guys, let's make an effort !
14:37 <vmassol> (see http://jira.xwiki.org/jira/secure/Dashboard.jspa?selectPageId=10000#Created-vs-Resolved-Chart/10470)
14:37 <vmassol> oops wrong date
14:38 <@sdumitriu> vmassol: I disagree a bit with XWIKI-5982
14:38 <@sdumitriu> The standard is to always put the page name in the title
14:38 <vmassol> Week 7, 2011, 23 created, 28 closed
14:38 <@sdumitriu> -core.rename.title = Renaming {0}
14:38 <@sdumitriu> +core.rename.title = Rename Page
14:38 <vmassol> sdumitriu: I don't know any standard but I do know what we currently have on existing pages
14:38 <vmassol> Create Page
14:38 <vmassol> Copy Page
14:38 <vmassol> Create Space
14:38 <vmassol> I aligned it on them
14:39 <vmassol> it feels wrong to have non static titles too
14:39 <@sdumitriu> I'm talking about history, comments, attachments (as standalone pages, not in tabs)
14:39 <@sdumitriu> So there are two standards
14:39 <vmassol> hmm didn't check those but in any case it couldn't work
14:39 <vmassol> see my last mail for a different idea
14:40 <@sdumitriu> Create is special, since it doesn't affect a document
14:40 <@sdumitriu> Why couldn't it work?
14:40 <vmassol> because we need a visual way to display a reference in this case
14:40 <vmassol> you're not going to wrtite:
14:40 <@sdumitriu> Renaming $doc.name
14:40 <@sdumitriu> You are on a page
14:41 <vmassol> Renaming page mypage for space space1, space2, space and wiki mywiki
14:41 <@sdumitriu> You have the wiki-space-doc in the menu
14:41 <@sdumitriu> And the breadcrumbs
14:41 <@sdumitriu> You shouldn't display the full reference in the title
14:41 <vmassol> ok yes we could do that right now it was the full name
14:41 <mflorea> has joined #xwiki
14:41 <vmassol> I aligned on copy actually
14:41 <@sdumitriu> Actually no, not doc.name
14:41 <vmassol> maybe that's the error
14:41 <@sdumitriu> Renamin $doc.plainTitle
14:41 <vmassol> hmmm
14:42 <vmassol> not sure about title
14:42 <vmassol> in that case if it's title then maybe we also need the source to be displayed
14:42 <@sdumitriu> Yes
14:42 <@sdumitriu> I agree with the rest of the changes
14:42 <vmassol> ok
14:42 <vmassol> I can change that
14:42 <@sdumitriu> (adding space and name in the content)
14:42 <@sdumitriu> My problem was with the title
14:43 <vmassol> the good part is that it's not going ot break people who don't upgrade ;)
14:43 <@sdumitriu> Emilie forgot her password again?
14:43 <vmassol> seems so
14:43 <vmassol> ok making the change sdumitriu
14:43 <@sdumitriu> Thanks
14:43 <vmassol> we need to change copy accoridnlgy too
14:43 <@sdumitriu> Yes, I'll do that
14:43 <@sdumitriu> Or do you want to do it?
14:44 <vmassol> ok I can do it
14:45 <+sburjan> sdumitriu: if you need the test machine please tell me so I can turn it on
14:45 <@sdumitriu> sburjan: Thanks
14:45 <vmassol> sdumitriu: Rename {0} instead of Renaming {0}, ok?
14:45 <+sburjan> do you need it ?
14:45 <@sdumitriu> sburjan: Not right now
14:45 <+sburjan> ok
14:45 <@sdumitriu> vmassol: Fine with me either way
14:46 <vmassol> since we say "Create page" and not "Creating page"
14:49 <vmassol> sdumitriu: do you like that we have links in the title?
14:50 <vmassol> ex: Attachments for http://localhost:8080/xwiki/bin/view/Main/Pagewith.in.name.
14:50 <@sdumitriu> It's nice, it's a quick way to get back to the document view
14:50 <vmassol> ok I'll do it for consistency but I'm not a big fan
14:50 <@sdumitriu> Before we had the new menu with link on the page name there was no way to get back
14:51 <florinciu> has joined #xwiki
14:51 <@sdumitriu> In the meantime, two other ways to get back have been introduced
14:51 <@sdumitriu> So now it is a bit redundant, but I like it
14:52 <@sdumitriu> mflorea: You should add bugfixingday keyword to the issues you close, even if they're Duplicate
14:52 <@sdumitriu> XAOFFICE-49
14:53 <@sdumitriu> XAOFFICE-45
14:53 <+mflorea> ok, but this is not bug fixing :)
14:54 <@sdumitriu> Bug triage is also useful
14:54 <florinciu> has quit
14:54 <+mflorea> ok, done
14:54 <@sdumitriu> We're closing the gap on created vs closed
14:57 <+jvelociter> hi rendering wizzards, tmortagne vmassol and all
14:57 <+jvelociter> I'm trying to write a table row entry on several lines, so that it looks nice in the code
14:57 <+jvelociter> is there a way I can escape new lines ?
14:58 <+jvelociter> like you would do in config files
14:58 <+jvelociter> something=blablabla,\
14:58 <+jvelociter>                      somethinfelse
14:59 <@sdumitriu> jvelociter: Why not |(((line
14:59 <@sdumitriu> line
14:59 <@sdumitriu> line)|next table cell
14:59 <+jvelociter> (I've tried the escape char at the end of the line but it does not work)
14:59 <vmassol> jvelociter: yes you need to use (((…)))
15:00 <+jvelociter> but that will create a group inside my cell
15:00 <+jvelociter> no ?
15:00 <vmassol> yes that's what you want: group several paragrpahs together
15:00 <+jvelociter> no that's not what I want
15:00 <vmassol> let me check something
15:01 <+jvelociter> I just want to have a long long line written on several lines in the code
15:01 <vmassol> jvelociter: this owrks fine:
15:01 <vmassol> |=h1|=h2
15:01 <vmassol> |line1
15:01 <vmassol> line2|line3
15:01 <vmassol> now in xwiki synatx 2.0 you need a group if you want to include an empty line
15:01 <+jvelociter> vmassol: ok but line1 and line2 are rendered on different lines, right ?
15:01 <vmassol> yes
15:02 <+jvelociter> yes but that's not what I want
15:02 <+jvelociter> sorry if I misexpressed myself
15:02 <vmassol> what you want odesn't exist
15:02 <+jvelociter> ok
15:02 <vmassol> you cannot break a single line in several lines
15:02 <+jvelociter> that's what I needed to know
15:03 <vmassol> not sure why you'd need that actually
15:03 <+jvelociter> for readability purpose
15:03 <vmassol> well several lines is more readable usually
15:03 <vmassol> :)
15:03 <+jvelociter> you don't understand
15:03 <vmassol> ah yes
15:03 <+jvelociter> readability of the code
15:03 <vmassol> well
15:03 <+jvelociter> not the content
15:03 <vmassol> you can use the velocity macro ;)
15:03 <+jvelociter> yes I have another subterfuge
15:04 <+jvelociter> I will use #define
15:04 <vmassol> and use the correct filter
15:04 <+jvelociter> which filter do I want ?
15:04 <vmassol> let me check
15:04 <vmassol> the html one I think
15:04 <+jvelociter> {{velocity filter="html" }}
15:04 <+jvelociter> ?
15:04 <vmassol> http://extensions.xwiki.org/xwiki/bin/view/Extension/Velocity+Macro+Filter
15:04 <vmassol> yes
15:04 <+jvelociter> great
15:04 <vmassol> http://extensions.xwiki.org/xwiki/bin/view/Extension/Velocity+Macro+Filter#Hhtml
15:06 <vmassol> I guess we could make that filter notion more generic and not only for velocity….
15:06 <vmassol> but it's a bit hard probbaly
15:06 <vmassol> anyway not for now
15:06 <vmassol> :)
15:07 <vmassol> evalica: is this an idea: http://jira.xwiki.org/jira/browse/XABLOG-61 ?
15:07 <vmassol> looks more like an improvement
15:07 <vmassol> to me
15:07 <vmassol> or a trivial bug fix
15:07 <+evalica> you're right thanks
15:07 <+evalica> I'll fix it
15:08 <cjdelisle> I'd like to take a stab at XWIKI-5981, can someone tell me if xwiki.home is supposed to have any effect for single wiki installations?
15:08 <vmassol> no idea cjdelisle
15:11 <cjdelisle> Does anyone know what the difference is between XWikiContext.getOriginalDatabase and XWikiContext.getMainWiki?
15:12 <cjdelisle> undocumented code blargh
15:13 <cjdelisle> Suppose I could point one to the other and see who complains ;)
15:16 <+tmortagne> cjdelisle: XWikiContext.getOriginalDatabase is not the main wiki
15:16 <+tmortagne> AFAIK it's supposed to be modified each time you switch the wiki
15:17 <cjdelisle> Ok. Is there something I can write into a comment in XWikiContext?
15:17 <+tmortagne> or it's the request wiki maybe
15:17 <+tmortagne> but for sure it has nothing to do with main wiki
15:18 <cjdelisle> getOriginalDatabase == getDatabase at time of request?
15:18 <+tmortagne> as I said i'm not sure, all i know is that it's not the same thing as main wiki
15:18 <+sburjan> sdumitriu: "Verify that we can skip the default delete result page and instead redirect to any page we want." Is this URL valid according to the specification ? http://localhost:8080/xwiki/bin/delete/Main/DeleteAttachment&xredirect=http://localhost:8080/xwiki/bin/view/Main/
15:19 <+sburjan> because it doesn't redirect if I use it like this
15:19 <cjdelisle> XWikiServletURLFactory.getServerURL() does not take into account the setting for "xwiki.home" unless db is not equal to getOriginalDatabase.
15:20 <cjdelisle> So in a non-farm screnario, xwiki.home has no effect.
15:21 <cjdelisle> if (xwikidb == null || xwikidb.equals(context.getOriginalDatabase())) { return this.serverURL; }
15:21 <cjdelisle> I can't understand what that line is intended to do.
15:26 <vmassol> cool one more color theme :)
15:30 <cjdelisle> hmm, the changes were merged in from the 1.5 branch which is now gone.
15:38 <+sburjan> vmassol: any idea how can I tset if an action I want to make has to _fail_ in order to pass the test, but without using try catch ?
15:39 <vmassol> well the UI never fails so you cannot use try/catch for sure....
15:39 <vmassol> (in unit test that's what you should do though)
15:39 <cjdelisle> @Test(expected=NullPointerException.class)
15:39 <vmassol> in ui tests you tests the output condition
15:40 <+sburjan> I am not logged, and want to delete a page, the ui doesn't have ui. so callding delete() should return in error .. "Element not found"
15:40 <cjdelisle> (only for a unit test though as said)
15:40 <+sburjan> the ui doesn't have Delete Page option. This is what I need to make sure
15:40 <vmassol> that's easy
15:41 <vmassol> but to do it properly you should add a PO method IMO
15:41 <vmassol> so that you don't have to  know about the xpath to the element
15:41 <vmassol> in the test
15:41 <cjdelisle> canDelete() ?
15:42 <vmassol> something like isDeleteAvailable()
15:42 <vmassol> canDelete if you want
15:42 <+sburjan> so i should extend ViewPage PO and add a new method which returns bool _if_ the xpath is not visible or something like that ?
15:42 <vmassol> yes
15:42 <+sburjan> okay. but I'll stil lhave tu use try catch throw in PO
15:42 <vmassol> return getDriver().findElements(By.xpath("…")).size() > 0;
15:43 <vmassol> no
15:43 <+sburjan> thanks :)
15:47 <+sburjan> what do you say about allowedToDelete() ?
15:47 <vmassol> I prefer can or is prefix
15:47 <vmassol> it's more standard
15:48 <vmassol> btw it's not about being allowed
15:48 <vmassol> it's about being able to click a delete button
15:48 <+sburjan> ok
15:48 <+sburjan> canDelete() then :)
15:51 <+sburjan> vmassol: the visibility of a WebElement can be checked with isEnabled() ?
15:52 <+sburjan> because I already have that button as a WebElement, but can't cast size() on it, and I wouldn't like to have another piece of code looking for that
15:52 <+sburjan> like findElements(....) as you suggested above
15:52 <+sburjan> but it seems I don't have any other way
15:54 <vmassol> isVisible could work indeed
15:54 <vmassol> you need to cast the WebElement
15:54 <vmassol> you cast to RenderedWebElement
15:54 <+sburjan> okay.. thanks
15:54 <vmassol> isDisplayed()
15:54 <vmassol> see http://stackoverflow.com/questions/2646195/how-to-check-if-an-element-is-visible-with-webdriver
15:55 <vmassol> only issue is that this might not work with the htmlunit driver but I havn"'t checked
15:55 <abusenius> has quit
15:56 <+sburjan> I'll test it
15:56 <vmassol> answer: it works apparently: http://webdriver.googlecode.com/svn/javadoc/org/openqa/selenium/RenderedWebElement.html
15:56 <vmassol> http://webdriver.googlecode.com/svn/javadoc/org/openqa/selenium/htmlunit/RenderedHtmlUnitDriverWebElement.html extends it
15:57 <vmassol> safari driver doesn't seem to support it though: http://webdriver.googlecode.com/svn/javadoc/org/openqa/selenium/safari/SafariWebElement.html
15:57 <vmassol> I don't know if their idea is that all driver should support it or not
15:57 <vmassol> probably I'd say
15:58 <vmassol> otherwise you just do as I mentioned and it'll always work
15:58 <+sburjan> already done it to RenderedWebElem
15:58 <+sburjan> now testing
15:58 <+sburjan> wish me luck :)
15:58 <vmassol> again casting to RenderedWebElement is not portable so IMO we should avoid it as much as possible
15:59 <vmassol> especially since we have no reason to do that here
16:02 <+sburjan> okay
16:03 <+sburjan> but it would be cool to test on IE's .. if we are writing portable tests
16:03 <+sburjan> maybe set up a IE test machine (we'll talk about that very soon)
16:43 <abusenius> has joined #xwiki
17:08 <sburjan> has quit
17:32 <Enygma`> has quit
17:35 <StefanO> has joined #xwiki
17:35 <roxxo> hey guys, I'm back to bug the jodconverter xwiki exports about a popular traceback when exporting to rtf
17:35 <roxxo> s/exports/experts/
17:36 <StefanO> Hello guys! Anca and I found a bug in Prototype 1.7+Scriptaculous1.9 regarding Sortables
17:36 <StefanO> after dragging the elements stack on top of each other instead of rearranging as they should
17:37 <roxxo> http://pastebin.com/4rhT4LmW <-- this is the result of exporting to RTF, although PDF and HTML exporting create files. Could it be an openoffice problem?
17:38 <StefanO> this doesn't occur with older versions
17:38 <StefanO> i tested in both xwiki 3.0 and in a local project
17:39 <+lucaa> well, in xwiki 3.0 is actually proto 1.7 + scriptaculous 1.8.3
17:39 <+lucaa> and that;s one issue, that scriptaculous supports proto 1.7 only from 1.9
17:39 <+lucaa> so, along with http://jira.xwiki.org/jira/browse/XWIKI-5753 we should also upgrade scriptaculous
17:40 <@sdumitriu> Back then scriptaculous wasn't released yet
17:40 <+lucaa> ah, ok that's probably it
17:40 <@sdumitriu> So, does it work with 1.9?
17:40 <@sdumitriu> Did you test?
17:40 <+lucaa> sdumitriu: can you help with this as part of bugfixing day? I need to finish something else now, unfortunately
17:40 <+lucaa> StefanO says not, I'll leave you with him
17:40 <+lucaa> (he tested)
17:41 <@sdumitriu> Hah, XWIKI-5753 closed on 22Dec, Scriptaculous released on 23Dec
17:41 <+lucaa> :)
17:41 <StefanO> actually Scriptaculous 1.9 says REQUIRED_PROTOTYPE: '1.6.0.3',
17:41 <StefanO> as a minimum requirement
17:42 <StefanO> but it comes bundled with 1.7
17:43 <+lucaa> sdumitriu: are we using scriptaculous drag & drop for the panels wizard, by any chance?
17:43 <StefanO> and Proto1.7 doen't work with both Scriptaculous 1.9 and 1.8.3
17:43 <@sdumitriu> No
17:44 <@sdumitriu> The panel wizard uses hand made JS written in 2006
17:44 <+lucaa> mh...
17:45 <StefanO> i reverted to an older bundle P 1.6.0.1 + S 1.8.1 and it works
17:45 <evalica> has quit
17:45 <+lucaa> I need this drag & drop to work for the dashboard editing
17:46 <@sdumitriu> Maybe we're doing something wrong
17:46 <+lucaa> while I found a 'hack' to fix it, I don't know if it's the best choice
17:46 <@sdumitriu> Drag and drop should work, I doubt that Scriptaculous really shipped a broken version
17:46 <@sdumitriu> I mean completely broken
17:47 <@sdumitriu> Maybe something in our CSS or HTML structure triggers a bug
17:47 <+lucaa> StefanO: can you test your local out of xwiki version for this combination?
17:47 <+lucaa> p1.7 + s 1.9?
17:47 <StefanO> i did
17:47 <StefanO> same issue
17:47 <+lucaa> and it works?
17:47 <+lucaa> aah, interesting
17:48 <+lucaa> can you try to recheck the params that we're passing to the Sortable,create?
17:48 <StefanO> and i doubt you dashboard.js script could be the problem
17:48 <@sdumitriu> Let's be clear on something: we're not downgrading prototype
17:48 <+lucaa> no we're not
17:49 <+lucaa> StefanO: well, just look again through it and check that they're fine. Then if it still fails, try to minimize that example as much as possible, with the minimal things required for a Sortable to fail
17:49 <+lucaa> before that though, look for the scriptaculous bugtracker and see if there's anything reported on the topic
17:49 <+lucaa> if not, report it
17:49 <+lucaa> please
17:50 <+lucaa> in the mean time for the dashboard, we'll hack it
17:50 <StefanO> ok, i'll try all possible combinations and make a report on that
17:50 <+lucaa> no, no
17:50 <+lucaa> not all possible combinations
17:50 <+lucaa> just one which corresponds to the documentation and contains the simplest Sortable that fails
17:50 <+lucaa> the minimal number of options, etc
17:50 <StefanO> well we have to see if it's prototype or scriptaculous the one that fails
17:51 <+lucaa> it's scriptaculous on prototype, so I guess it's the scriptaculous guys that handle i
17:51 <+lucaa> it
17:51 <+lucaa> moving on private
17:51 <StefanO> ok
18:03 <StefanO> has quit
18:05 <@sdumitriu> Interesting, in two years only one line of code was changed in Scriptaculous
18:05 <@sdumitriu> No, two
18:05 <npm_> is now known as <npm>
18:08 <@sdumitriu> XWIKI-5987
18:08 <vmassol> sburjan`: can you handle the mail entitled "Documentation - Administration" on the users list?
18:09 <vmassol> (ie fix the doc)
18:09 <vmassol> (and reply to the mai)
18:09 <vmassol> mail
18:15 <+sburjan`> vmassol: ok.. let me reag it first
18:16 <+sburjan`> read
18:25 <cjdelisle> Does RTF export work?
18:26 <mflorea> has quit
18:30 <cjdelisle> roxxo: I just tried rtf and it seems to work for me, is it one specific page that fails?
18:37 <evalica> has joined #xwiki
18:40 <sdumitriu> has quit
18:53 <jvelociter_> has joined #xwiki
18:55 <jvelociter_> has quit
18:59 <tmortagne> has quit
19:09 <roxxo> cjdelisle: what page did you try? I've tried several different pages, including the default 2.7 Sandbox.WebHome
19:09 <cjdelisle> do you have openoffice configured to do import/export?
19:23 <roxxo> cjdelisle: hmm... good question. I started openoffice manually, and RTF isn't displayed in the File->Export.. menu, but it is displayed in the Save As.
19:25 <cjdelisle> Ok so you are using openoffice integrated with xwiki.
19:25 <jvdrean> has quit
19:25 <cjdelisle> I really know nothing about the office integration.
19:28 <roxxo> cjdelisle: right, it's connected to the Internally managed OO Server
19:30 <rstavro> has left #xwiki
19:30 <cjdelisle> You probably should have been here about 7 hours ago, most everyone has left already esp. the openoffice experts.
19:30 <roxxo> aw man
19:31 <roxxo> do they usually come back at a time other than 5AM? :)
19:31 <sdumitriu> has joined #xwiki
19:33 <cjdelisle> You could try sending mail to the user list.
19:43 <roxxo> cjdelisle: the suits want a demo by tomorrow :(
19:43 <roxxo> maybe pdf will suffice, what version of xwiki + openoffice + OS are you running when rtf export works fine?
19:46 <roxxo> oh man, jira XWIKI-5866 is a known bug
19:46 <cjdelisle> I am using 3.0M2 (which shouldn't matter) I do not have openoffice installed.
19:47 <cjdelisle> was /away last couple minutes seeing what firefox does when passed:
19:47 <cjdelisle> stream.print('<html><head></head><body>'); for(;;) { stream.print('<div>'); }
19:47 <cjdelisle> it takes it surprisingly well
19:48 <roxxo> wow
19:48 <roxxo> are you being mean to firefox ...?
19:48 <cjdelisle> testing...
19:49 <cjdelisle> strong xml parser
19:50 <cjdelisle> now stream.print('<html><head></head><body><a href="'); for(;;) { stream.print('z'); }
19:51 <cjdelisle> Are you using openoffice for import functionality?
19:55 <cjdelisle> hmm. java has a push speed of just about 7.5Mbps
19:56 <roxxo> cjdelisle: I just uninstalled openoffice, and disconnected the OpenOffice Server, and RTFs are created.. sweet
19:56 <roxxo> not so sweet, however: OO crashes when loading the resulting RTF
19:57 <cjdelisle> probably XWIKI-5251
19:57 <roxxo> yup, that's the one
19:59 <cjdelisle> I understand that people have gotten it to work with the openoffice server.
20:00 <cjdelisle> jstoldt was setting that up but he's left 6 hours ago.
20:02 <roxxo> aw man
20:03 <cjdelisle> what's your operating system? windows?
20:06 <roxxo> I'm running this on Ubuntu Lucid
20:07 <cjdelisle> what is the package name of the openoffice server?
20:13 <evalica> has quit
20:15 <cjdelisle> It looks like I have it installed, if you can tell me how to set it up, I might be able to help you with your problem.
20:15 <cjdelisle> (however I am lazy and do not want to RTFM)
20:24 <arkub> has quit
20:26 <roxxo> cjdelisle: sorry, had to put out a fire.. I just ran `aptitude install openoffice.org`, then in xwiki->Administer Wiki->OpenOffice Server set the Server Path to /usr/lib/openoffice, and Server profile to /home/xwiki/.openoffice.org/3 (though these were already filled in)
20:28 <cjdelisle> ok, one minute.
20:30 <cjdelisle> what version are you using?
20:31 <cjdelisle> (xwiki)
20:31 <roxxo> 2.7.0, though the same problem occurred with 2.4.3
20:34 <lucaa> has quit
20:41 <cjdelisle> heh you're going to love this...
20:42 <cjdelisle> it works for me
20:42 <cjdelisle> you must be crazy :P
20:44 <cjdelisle> here's what I did:
20:44 <cjdelisle> start xwiki
20:44 <cjdelisle> (it was running)
20:44 <cjdelisle> goto Administor->OpenOffice Server
20:45 <cjdelisle> Notice there are no fields to fill in, What is this?!
20:45 <cjdelisle> Select start and click update
20:46 <cjdelisle> look at error message because the openoffice dir is incorrect.
20:46 <cjdelisle> can't change dir because the I can't fill in the fields, must be something wrong with my installation (yes, I abuse xwiki too.)
20:46 <cjdelisle> easiest thing to do --> open a terminal
20:47 <cjdelisle> su
20:47 <cjdelisle> ln -s /usr/lib/openoffice/ /opt/openoffice.org3
20:47 <vmassol> sdumitriu: I'm finishing copy.vm changes but I have a question
20:48 <cjdelisle> back to xwiki
20:48 <@sdumitriu> vmassol: Shoot
20:48 <vmassol>  ${services.csrf.isTokenValid("$!{request.getParameter('form_token')}")} returns true if there's no form_token parameter, is that normal?
20:48 <vmassol>  ${services.csrf.isTokenValid("")} returns true
20:48 <cjdelisle> notice it needs the profile directory to exist as well
20:49 <cjdelisle> mkdir /usr/share/tomcat5.5/.openoffice.org
20:49 <cjdelisle> mkdir /usr/share/tomcat5.5/.openoffice.org/3
20:49 <vmassol> sdumitriu: I was expecting false, am I understanding something wrongly?
20:50 <cjdelisle> vmassol: it always returns true because otherwise there would be failures everywhere.
20:50 <cjdelisle> (peper tiger)
20:50 <cjdelisle>   a
20:50 <vmassol> cjdelisle: what's the point of this test then? (clearly there's something I don't get)
20:50 <@sdumitriu> vmassol: If you didn't enable csrf, then it's normal
20:50 <vmassol> oh
20:50 <vmassol> it's false by default?
20:50 <@sdumitriu> It's disabled by default
20:50 <vmassol> ah
20:50 <vmassol> ok I get it now
20:51 <vmassol> thanks
20:51 <cjdelisle> roxxo: still here?
20:54 <roxxo> cjdelisle: yep, I'm half paying attention to this stupid meeting, and repeating your above steps
20:55 <roxxo> ITIL needs to be banned
20:56 <cjdelisle> ITIL?
20:57 <roxxo> haha
20:57 <roxxo> it's just management fluff certification about "processes" and "services"
20:58 <cjdelisle> so anyway it complained about lack of content in the profile so I did:
20:58 <cjdelisle> cp /home/user/.openoffice.org2/* /usr/share/tomcat5.5/.openoffice.org/3 -R
20:58 <cjdelisle> and
20:58 <cjdelisle> chown tomcat55 /usr/share/tomcat5.5/.openoffice.org/3 -R
20:59 <cjdelisle> it seems to work despite being openoffice2
21:00 <roxxo> I have oo3 installed, and /usr/lib/openoffice exists and contains program,share,basis-*, is that what you get?
21:00 <roxxo> I wasn't able to edit the paths through the xwiki admin interface either
21:01 <cjdelisle> They can be edited in xwiki.properties but it was easier for me to symlink than restart tomcat
21:02 <AndreasJonsson> Regarding csrf, I've enabled csrf and occasionally run into failures.  Is there any plans to have it enabled by default in any release in the near future?
21:02 <cjdelisle> I'm using debian lenny and it seemed that I had oo2 by default.
21:03 <cjdelisle> How do you define "near"? ;)
21:03 <AndreasJonsson> 3.x
21:03 <cjdelisle> IMO good chance of it before 4.0
21:04 <AndreasJonsson> Great!
21:05 <cjdelisle> hey, you're a security guy
21:05 <cjdelisle> was playing with this earlier:
21:05 <cjdelisle> stream.print('<html><head></head><body>'); for(;;) { stream.print('<div>'); }
21:05 <cjdelisle> thought you might get a kick out of it
21:05 <@sdumitriu> cjdelisle: Why is jbrBridge operator?
21:06 <cjdelisle> because it auths as me
21:06 <@sdumitriu> Aha
21:06 <+abusenius> re csrf, I'd be +1 for enabling it at least during 3.1 and at least start fixing failures
21:06 <cjdelisle> all runs through the same irssi instance.
21:06 <+abusenius> it's just that I personally have no time :(
21:07 <@sdumitriu> AndreasJonsson: I'll apply CSRF patches, can you point me to some?
21:07 <cjdelisle> Maybe we could turn it on for milestones but off for releases so we get testing done.
21:07 <+abusenius> *at least during 3.1 ... 3.1M2
21:07 <@sdumitriu> Like XAADMINISTRATION-218
21:08 <AndreasJonsson> I'm tagging my reports in jira with 'csrf'
21:08 <+abusenius> cjdelisle: yes, that's what I mean
21:08 <AndreasJonsson> I havent actually provided patches for all of them, though.
21:09 <@sdumitriu> AndreasJonsson: K, I'll look at them
21:10 <+abusenius> btw. AndreasJonsson thanks for testing, it's great that someone actually uses csrf protection :)
21:16 <AndreasJonsson> No problem!
21:17 <AndreasJonsson> cjdelisle: Will that cause a stack overflow in the parser?
21:17 <@sdumitriu> I wonder if it's something we can do right now...
21:17 <@sdumitriu> AndreasJonsson: Did you encounter any serious issues other than those that you reported?
21:17 <AndreasJonsson> No
21:18 <jvdrean> has joined #xwiki
21:25 <roxxo> cjdelisle: maybe it has to do with xwiki's tomcat being installed under $HOME, though that should take care of everything permission-related, what files do you have in your openoffice.homepath and openoffice.profilePath?
21:26 <cjdelisle> home path/
21:26 <cjdelisle> ?
21:26 <cjdelisle> you mean server path?
21:31 <+sburjan`> it's the same
21:31 <cjdelisle> server path is ls /usr/lib/openoffice/
21:31 <cjdelisle> help  presets  program  share
21:32 <cjdelisle> and for profilepath I made a dir and copied what was in /home/user/.openoffice.org2/
21:33 <cjdelisle> whatever came with debian
21:35 <+sburjan`> cjdelisle: you're using debian ?
21:35 <cjdelisle> yes
21:37 <+abusenius> sdumitriu: the problem with enabling csrf by default is not really the bugs, but failing tests imo
21:37 <+abusenius> there are many false positives
21:40 <AndreasJonsson> We should consider introducing a max nesting level for block elements in all parsers.
21:45 <roxxo> cjdelisle: hmm, I'll try that
22:06 <@sdumitriu> jvelociter: XAADMINISTRATION-161
22:06 <@sdumitriu> What do you mean by that?
22:06 <@sdumitriu> What exactly is the problem?
22:10 <vmassol> anyone has a multiwiki instance ready to test my latest commit? (copy feature between wikis)
22:11 <@sdumitriu> sburjan`: I guess now is too late to ask for the IE test machine, right?
22:12 <mflorea> has joined #xwiki
22:21 <cjdelisle> sdumitriu: do you know anything about the xwiki.home param and whether it's supposed to work in a non-farm instance?
22:21 <@sdumitriu> Nope
22:22 <cjdelisle> I'm thinking that if there is only one wiki and it is set then all links should point to it.
22:23 <cjdelisle> zeus proxy does not set x-forwarded-host and it rewrites the host: parameter, xwiki.home makes sense in a case like that.
22:23 <cjdelisle> But it currently doesn't work that way.
22:54 <+jvelociter> sdumitriu: I think I was refering to the inconstent look between regular admin sections and the ones introduced by configurable class
22:54 <+jvelociter> but maybe it's less/not relevant anymore
23:14 <vmassol> nice: 31 issues closed: http://jira.xwiki.org/jira/secure/IssueNavigator.jspa?reset=true&jqlQuery=resolved+%3E%3D+-5d+AND+keywords+~+bugfixingday+ORDER+BY+assignee+ASC
23:14 <vmassol> 17 real ones
23:16 <vmassol> only 5 committers to have closed issues though…. out of 15....
23:17 <vmassol> (sergiu, caty, marius, thomasM and me). jvelociter I think you could have tagged your issues as bugfixingday too
23:18 <vmassol> cjdelisle: you didn't close any issue yet today? ;)
23:18 <vmassol> (or you forgot to tag them?)
23:19 <cjdelisle> I'm looking at XWIKI-5981 but it's a touchy issue and I want to get it right.
23:19 <cjdelisle> (also I am slacking)
23:19 <vmassol> right, you didn't pick an easy one
23:20 <vmassol> jvdrean: come on you still have time to pick one that takes 5 minutes to fix! ;)
23:20 <cjdelisle> But we get lots of free explainiation comments out of the deal.
23:20 <vmassol> :)
23:20 <vmassol> anca's been slacking too….
23:22 <vmassol> we're 18 committers in the hall of fame acdtually
23:22 <vmassol> (including guilaume)
23:22 <vmassol> cjdelisle: is the description I put for you on the hall of fame page ok: http://dev.xwiki.org/xwiki/bin/view/Community/HallOfFame ?
23:23 <vmassol> (you can change it if you want)
23:23 <cjdelisle> "the grinch"?
23:23 <vmassol> :)
23:23 <vmassol> talking about affiliation
23:24 <cjdelisle> It doesn't matter to me.
23:24 <vmassol> (not filiation ;))
23:24 <vmassol> cjdelisle: sure just wanted to make sure you're comfortable with what I put
23:26 <cjdelisle> I was also thinking about tackleing that junit memory leak issue, it's so enchanting to imagine some arraylist somewhere that is not being emptied but I keep thinking the bug must exist because of some limitation.
23:28 <vmassol> cjdelisle: you like complex stuff I can see…. ;)
23:29 <cjdelisle> Need challenge or else I get bored.
23:29 <vmassol> #bugfixingday is about fixing the max of easy stuff in one day IMO
23:29 <vmassol> to make a dent in the # of total open bugs
23:30 <vmassol> anyway I'm going to leave you guys, I'm not as night owl as you are…. going to be now
23:30 <vmassol> s/be/bed
23:31 <vmassol> nn, take care and enjoy
23:31 <vmassol> has quit