IRC Archive for channel #xwiki
Last modified by Vincent Massol on 2012/10/18 18:44
headache - (00:08): got to go
headache - (00:08): goodnight
sdumitriu - (00:08): Goodnight
headache - (00:08): :)
headache left at 00:08 (Quit: Sto andando via
CalebJamesDeLisl left at 00:14 (Ping timeout: 240 seconds
CalebJamesDeLisl joined #xwiki at 00:15
nickless left at 00:27 (Ping timeout: 264 seconds
nickless joined #xwiki at 00:27
nickless is now known as nickless_ ([email protected]
npm left at 00:51 (Quit: Leaving.
npm joined #xwiki at 00:59
jvelociter left at 02:42 (Quit: jvelociter
jvelociter joined #xwiki at 02:48
jvelociter left at 02:51 (Client Quit
nickless_ left at 04:22 (Ping timeout: 246 seconds
Denis left at 06:02 (Read error: Connection reset by peer
Denis joined #xwiki at 06:10
DV_ joined #xwiki at 06:24
nuvolari - (06:35): rawr! :D g'morning
sdumitriu - (06:36): Goodnight
DV_ left at 06:38 (Ping timeout: 256 seconds
DV_ joined #xwiki at 06:44
mflorea joined #xwiki at 06:54
bblfish left at 07:38 (Ping timeout: 264 seconds
vmassol joined #xwiki at 07:39
bblfish joined #xwiki at 07:43
kibahop joined #xwiki at 08:01
kibahop left at 08:22 (Quit: Leaving.
kibahop joined #xwiki at 08:26
kibahop left #xwiki at 08:26
anamarias joined #xwiki at 08:41
silviar joined #xwiki at 09:13
DV_ left at 09:13 (Read error: Connection reset by peer
DV_ joined #xwiki at 09:13
headache joined #xwiki at 09:17
headache - (09:17): hello
vmassol - (09:17): headache: hi
CalebJamesDeLisl - (09:31): Good morning.
sdumitriu left at 09:32 (Ping timeout: 260 seconds
CalebJamesDeLisl - (09:33): re the spam accounts: I'm wondering what type of people we are trying to prevent by forcing first/last name, it seems to me that only the most unsophisticated spammer would be unable to type john smith into the form.
vmassol - (09:33): right
vmassol - (09:34): but add the captcha to it and it increases our chances
CalebJamesDeLisl - (09:34): It seems that we need to get notice about new users so if their names are mrViagra or something offensive we can delete them.
vmassol - (09:34): also even with firstname/lastname there are several per day
vmassol - (09:34): we don't often have these kind of names
vmassol - (09:35): right now most spam users are auto generated names that don't mean anthing
CalebJamesDeLisl - (09:35): Also there was an advantage to not requiring the first/last name because they wouldn't enter it then could be detected, deleted.
CalebJamesDeLisl - (09:35): If it's mandatory then they will enter something.
vmassol - (09:35): they're always entering them!
vmassol - (09:35): this is automated caleb
vmassol - (09:35): we cannot do anything for non automated spam
CalebJamesDeLisl - (09:35): Ok.
vmassol - (09:36): well we can, but only with an acceptation workflow which is a bit of a pain
CalebJamesDeLisl - (09:36): Some of the names just look like grieftors and bored teens.
CalebJamesDeLisl - (09:37): if it's automated then captcha is a big plus.
vmassol - (09:37): right now they don't seem to be doing anything with the majority of the users created there
CalebJamesDeLisl - (09:37): Hmm
vmassol - (09:37): but better not to leave them here just in case they're preparing some massive attack in the future ;)
CalebJamesDeLisl - (09:38): Wouldn't they wget a site for baby names or something rather than the obvious names they seem to choose.
lucaa left at 09:38 (Quit: Leaving.
Enygma` joined #xwiki at 09:39
vmassol - (09:39): yeah
vmassol - (09:39): they'll improve over time I guess
CalebJamesDeLisl - (09:39): I wonder who these people are? people in third world contries getting paid for signed up usernames?
CalebJamesDeLisl - (09:40): s/contries/countries
CalebJamesDeLisl - (09:40): (security nut says know your enemy :) )
vmassol - (09:41): no idea
CalebJamesDeLisl - (09:42): I'm not big on forcing first/last name because some (very cool) people will feel it violates their privacy.
CalebJamesDeLisl - (09:42): And I don't see the benefit
CalebJamesDeLisl - (09:42): :)
CalebJamesDeLisl - (09:43): wouldn't it be awesome: !xwikibot delete user someSpammer
vmassol - (09:44): CalebJamesDeLisl: personally I think you should be accountable for what you write
DV_ left at 09:45 (Read error: Connection reset by peer
CalebJamesDeLisl - (09:45): which is your argument for first/last names?
vmassol - (09:45): so forcing first/last name shows we're not some "hippy" warez stuff
vmassol - (09:45): like we have irc users named without their real name
vmassol - (09:45): I don't like this
vmassol - (09:45): if I help people, I'd like to know who I'm talking to
CalebJamesDeLisl - (09:46): I just look at the ip address :)
vmassol - (09:46): we're not some underground warez stuff ;)
CalebJamesDeLisl - (09:47): The problem as I see it is there is no way to be sure, Eve could enter anything while good people will try to be honest and may be put off.
vmassol - (09:47): sure
vmassol - (09:47): you can enter John Smith
vmassol - (09:47): what I don't understand
vmassol - (09:47): and what I'd love to understand
vmassol - (09:48): is why people create accounts on xwiki.org if they don't want to participate
vmassol - (09:48): maybe just to post comments
vmassol - (09:48): in which case we can now try to open them with captcha
CalebJamesDeLisl - (09:48): Given that the usernames get signed up and untouched I'm willing to work with the assumption that there is a market for usernames somewhere.
vmassol - (09:48): but even if you remove the need to log in for comments
vmassol - (09:48): you'll see that 90% of users created don't do anything with the wiki
vmassol - (09:48): so that beats me
vmassol - (09:49): especially since there's a big disclaimer on the reg page
vmassol - (09:49): explaining that you shouldn't create accounts unless you want to participate in the contnet
CalebJamesDeLisl - (09:50): Before I started participating here, I used to use reddit which didn't ask for any info, just a captcha, I always thought that was very cool of them.
CalebJamesDeLisl - (09:51): I wrote a patch to change the way votes were valued to prevent spam but that's a different story.
CalebJamesDeLisl - (09:51): Suppose we locked accounts for non participation.
CalebJamesDeLisl - (09:51): ?
CalebJamesDeLisl - (09:51): Maybe a captcha was needed to reopen them.
CalebJamesDeLisl - (09:52): Gentle on users, tough on bots.
vmassol - (09:52): the only advantage of gettings all those users
headache is now known as gvallarelli ([email protected]
gvallarelli - (09:52): :D
vmassol - (09:52): is that it tests the scalability of our user handling code
CalebJamesDeLisl - (09:53): I havn't looked at it lately but I think it should be fine.
gvallarelli is now known as giuseppeV ([email protected]
CalebJamesDeLisl - (09:54): I ran an instance with 200,000 documents and 500,000 links on a pentium4
vmassol - (09:54): it is
vmassol - (09:54): but it wasn't in the past
vmassol - (09:54): we fixed that with the livetable
vmassol - (09:54): (and the paging in it)
CalebJamesDeLisl - (09:54): I see.
giuseppeV is now known as headache ([email protected]
CalebJamesDeLisl - (09:55): Maybe a username which is inactive for x amount of time is locked, then after y amount of time it is deleted.
headache - (09:55): i don't agree with deletion
vmassol - (09:55): sure,all this sounds good, it just depends how much work we want to do right now
headache - (09:56): maybe disabled
CalebJamesDeLisl - (09:56): headache: They can be undeleted.
headache - (09:56): or deleted after a lot time a year or more
DV_ joined #xwiki at 09:56
vmassol - (09:56): btw Im' not sure the undelete works right now
vmassol - (09:56): (for users)
vmassol - (09:57): since they're removed from their groups when deleted
mpavel joined #xwiki at 09:57
CalebJamesDeLisl - (09:57): Hmm, well a PR script could fix that.
vmassol - (09:57): script?
vmassol - (09:58): we just need to handle it in events
vmassol - (09:58): but the pb is that we loose information
mpavel left #xwiki at 09:58
CalebJamesDeLisl - (09:58): Ahh yes.
CalebJamesDeLisl - (10:00): Anyway I think that forcing first/last name is not very effective against spammers and may turn off some cool people, it doesn't IMO make the site seem more legitimate because it isn't verified, however I'm happy to make the change if you like.
vmassol - (10:00): fine we can start with captcha and see how it goes if you want
vmassol - (10:00): could you set that up?
CalebJamesDeLisl - (10:01): thanks :)
CalebJamesDeLisl - (10:01): can do. You can give me access to the registration page or I can upload the changed version.
vmassol - (10:01): ah you don't have access, let me check that
CalebJamesDeLisl - (10:02): I see the registration page already shows a captcha on xwiki.org
vmassol - (10:02): oh it's done already
vmassol - (10:02): which is why I don't have any spammer to delete
vmassol - (10:02): (I was wondering why this morning :))
vmassol - (10:02): http://www.xwiki.org/xwiki/bin/register/XWiki/Register
CalebJamesDeLisl - (10:02): Yes I have noticed no funny names lately.
vmassol - (10:03): the message is wrong btw caleb: "Welcome to the registration form. This will allow you to edit pages, once the admin gives you appropriate rights."
vmassol - (10:03): this messgae is misleading
CalebJamesDeLisl - (10:03): It is set in applicationResoureces.properties I can change it in the registration page and commit the change to the core.
vmassol - (10:03): ouch http://www.xwiki.org/xwiki/bin/view/Main/News is broken
vmassol - (10:04): ok would be good
CalebJamesDeLisl - (10:04): What do you have in mind?
vmassol - (10:04): when not logged in http://www.xwiki.org/xwiki/bin/view/Main/News fails
vmassol - (10:04): same when logged in actually
CalebJamesDeLisl - (10:05): "Sign up here to get access to modify pages"?
CalebJamesDeLisl - (10:07): "Welcome to the registration form. Registration will allow you to participate in this wiki."?
vmassol - (10:08): the latter sounds good
vmassol - (10:08): maybe:
CalebJamesDeLisl - (10:08): Ok, I'll change it.
vmassol - (10:08): s/Registration/Signing up/
vmassol - (10:08): ?
CalebJamesDeLisl - (10:08): Much agreed.
evalica joined #xwiki at 10:08
CalebJamesDeLisl - (10:09): technically. Registration == turning over property (a ship) for safe keeping (by the harbormaster)
headache - (10:09): hi evalica i've read your comment about a mockup i've realized for xwiki.org homepage
headache - (10:10): however i have no balsamiq license so i cannot some errors, the mockup is only a screenshot
headache - (10:10): *can't fix
CalebJamesDeLisl - (10:11): "Welcome to the sign up form. Once you join you will be able to participate in this wiki." ?
vmassol - (10:14): sounds good
CalebJamesDeLisl - (10:14): Hmm, no access to XWiki.Registration for committers group.
tmortagne joined #xwiki at 10:15
jvelociter joined #xwiki at 10:15
evalica - (10:16): headache: the comment was a suggestion to keep in mind
headache - (10:17): evalica, ok :)
jvelociter left at 10:18 (Client Quit
xenon75 joined #xwiki at 10:19
xenon75 left at 10:19 (Client Quit
xenon75 joined #xwiki at 10:19
florinciu joined #xwiki at 10:21
lucaa joined #xwiki at 10:26
anamarias - (10:35): tmortagne: on the large xar import snippet, in the comment at the bottom you say the content was changed to 2.0 syntax
anamarias - (10:35): http://code.xwiki.org/xwiki/bin/view/Snippets/LargeXARImportScriptSnippet
anamarias - (10:35): Thomas Mortagne, 2009/12/17 18:26
anamarias - (10:35): Removed attachement and converted the content to xwiki/2.0, no issue anymore :)
anamarias - (10:35): but the groovy code is still for 1.0 syntax
tmortagne - (10:36): anamarias: that's a very old comment
tmortagne - (10:36): there is no xar anymore
tmortagne - (10:37): (and i did convert to 2.0 ;))
anamarias - (10:37): hmm
anamarias - (10:37): so what's the actual recent snippet then?
tmortagne - (10:38): anamarias: i don't understand tehre is only one snipper on http://code.xwiki.org/xwiki/bin/view/Snippets/LargeXARImportScriptSnippet
anamarias - (10:38): but the code is for xwiki 1.0 syntax, right?
tmortagne - (10:38): byw my last comment was "Removed attachement and converted the content to xwiki/2.0, no issue anymore :)"
anamarias - (10:39): if i convert the code to work in 2.0 syntax
tmortagne - (10:39): my comment did not had anything to do with the snippert itself
tmortagne - (10:39): but with the snipper page
anamarias - (10:39): I should replace all <% with {{groovy}}
anamarias - (10:39): ah, now I get it about the comment
vmassol - (10:39): anyone has an idea why http://www.xwiki.org/xwiki/bin/view/Main/News is broken?
vmassol - (10:39): seems like the macro isn't there anymore
tmortagne - (10:39): anamarias: you don't have to convert to 2.0, would be nice but i's working well in 1.0
vmassol - (10:39): #getDocumentsModifs
tmortagne - (10:40): anamarias: and i prefer to keep it in 1.0 for older xwiki installs
nickless_ joined #xwiki at 10:40
vmassol - (10:40): I don't see where it's imported from actually
tmortagne - (10:40): maybe sometyhing removed from macro.cm
tmortagne - (10:40): macro.vm
vmassol - (10:41): ah ok didn't know it was in macros.vm
tmortagne - (10:41): me neither
tmortagne - (10:41): i said maybe :)
vmassol - (10:41): ok :)
vmassol - (10:41): AFAIR it was in a page
anamarias - (10:42): tmortagne: and how would it look like in 2.0, just that i'll know ?
anamarias - (10:42): smth like:
tmortagne - (10:42): well put it in a groovy macro and it should be all i think
tmortagne - (10:42): let me see
anamarias - (10:42): {{groovy}} if (smth) { {{/groovy}} {{html}} some html {{/html}} {{groovy}} // continue code ...
anamarias - (10:42): ?
anamarias - (10:43): there is some html code inside it
tmortagne - (10:43): yes html too
tmortagne - (10:43): i mean i don't think there is any wiki syntax in it
anamarias - (10:43): i'm interested how the macros are nested
tmortagne - (10:43): hmm maybe list syntax actually
tmortagne - (10:44): yes tehre is some list syntax but it's the same in 1.0 and 2.0 so it's ok
CalebJamesDeLisl - (10:44): If you are using groovy try: {{groovy}} if (smth) {println("{{html}} some html {{/html}}"); {{/groovy}}
CalebJamesDeLisl - (10:44): I think that works.
tmortagne - (10:44): anamarias: {{groovy}}{{html}}{{/html}}{{/groovy}}
tmortagne - (10:44): because you produce html with groovy
tmortagne - (10:44): the top macro is executed first
tmortagne - (10:45): actually no
tmortagne - (10:45): it's groovy so you haver to explicitely print the html macro
tmortagne - (10:46): something like
tmortagne - (10:46): {{groovy}}
tmortagne - (10:46): print "{{html}}"
tmortagne - (10:46): print "{{/html}}"
tmortagne - (10:46): {{/groovy}}
tmortagne - (10:47): basically from rendering POV {{groovy}}{{/groovy}} only contains groovy, then it parse the result of the groovy macro
lucaa - (10:47): devs do we have a plan for the 2.3M1 rel?
lucaa - (10:47): monday?
vmassol - (10:49): lucaa: we're waiting for you actually so you need to tell us when you can make it,
vmassol - (10:49): ?
lucaa - (10:50): well, now that I have my votes, I only need to move the code to the platform and enterprise
lucaa - (10:50): I can do it over the weekend
vmassol - (10:50): (it was supposed to be released on 8th March so we're currently eating our M2 time)
tmortagne - (10:50): speaking of release, i'm finishing ti read my morining mails and i will start to prepare the 2.2.3 release, say it if there is anything still blocker
vmassol - (10:51): lucaa: ok for a release on Monday then for me
vmassol - (10:51): tmortagne: great
lucaa - (10:51): vmassol: at least I hope I have all the votes I need... otherwise I'll apologize on monday for all unvoted commits
CalebJamesDeLisl - (10:52): vmassol: edit access on XWiki.Registration?
vmassol - (10:53): you still need it?
vmassol - (10:53): oh for the message
vmassol - (10:53): doing it
vmassol - (10:54): CalebJamesDeLisl: hmm ot sure how it works
vmassol - (10:54): i see http://www.xwiki.org/xwiki/bin/view/XWiki/Registration
vmassol - (10:54): but it's not the page used for registration apparently
CalebJamesDeLisl - (10:54): Yes it is, there is an includeForm.
CalebJamesDeLisl - (10:54): At some point I'd like to create a set of new sheets for applications, snippets, etc in 2.0 format so edit on code:XWiki.* would be cool.
vmassol - (10:55): in the past it was like this:
glerouge joined #xwiki at 10:55
vmassol - (10:55): if a XWiki.Registration page existed, the register action would use it otherwise default to the templates
vmassol - (10:55): how does it work now?
CalebJamesDeLisl - (10:55): That's still how it works. It just uses an include instead of calling the renderer now.
CalebJamesDeLisl - (10:56): (an include inside of an if)
vmassol - (10:56): ok so you have a different behavior depending on the action passed?
vmassol - (10:57): (whether it's "register" or "view")
CalebJamesDeLisl - (10:57): no because you go to /register/Register or to /view/Registration
vmassol - (10:58): hmm there's no permission on XWiki.Registration
jvdrean joined #xwiki at 10:58
vmassol - (10:58): checking the xwiki space
CalebJamesDeLisl - (10:58): But yes the registration page works fine through the view action, a few small changes to registerinline.vm and RegisterAction could be removed.
CalebJamesDeLisl - (10:59): RegisterAction duplicates public APIs.
vmassol - (10:59): indeed only admin group can edit in the XWiki space
jvdrean left at 11:00 (Read error: Connection reset by peer
vmassol - (11:00): CalebJamesDeLisl: I've added you to the admin group
vmassol - (11:00): (so be careful ;))
CalebJamesDeLisl - (11:01): Which username?
CalebJamesDeLisl - (11:01): cjdelisle?
vmassol - (11:01): no
vmassol - (11:01): the other one
vmassol - (11:01): XWiki.CalebJamesDeLisle
CalebJamesDeLisl - (11:01): Ok, I'll create another one, I try to sandbox myself most of the time.
CalebJamesDeLisl - (11:03): re code.xwiki.org, what do you think of adding XWiki.ApplicationClassSheet20 (syntax2.0) and making it the default?
vmassol - (11:03): good to me
CalebJamesDeLisl - (11:03): right now adding applications or snippets is broken because everything defaults to syntax2.0
vmassol - (11:03): go ahead CalebJamesDeLisl if you want to do it
vmassol - (11:04): it's a good idea
CalebJamesDeLisl - (11:04): Next time I do a bunch of documentation I will.
tmortagne - (11:04): CalebJamesDeLisl: +1 and we should do the same for snippet, plugins etc...
CalebJamesDeLisl - (11:07): If I can edit a non-existant page (not saving) that means I have create access?
CalebJamesDeLisl - (11:10): Just previewed $hasCreatePage, looks like I can create those (at some point :) )
KermitTheFragger joined #xwiki at 11:11
nickless_ left at 11:16 (Ping timeout: 265 seconds
vmassol - (11:16): tmortagne: question for you: is our xwiki syntax parsing faster now after your upgrade of the wikimodel jar?
vmassol - (11:16): (ie does it include the new javacc modifications done by … - can't recall his name)
tmortagne - (11:16): vmassol: should be but it's difficult to tell exactly
vmassol - (11:17): tmortagne: we can look at test execution times
lucaa left at 11:17 (Quit: Leaving.
tmortagne - (11:17): vmassol: yes it's the patch from Andreas
lucaa joined #xwiki at 11:17
tmortagne - (11:17): that i applied
tmortagne - (11:17): he also fixed ))) bug
vmassol - (11:17): ok
tmortagne - (11:17): but it was mainly about applying he's great work
vmassol - (11:18): 16 seconds on http://hudson.xwiki.org/job/xwiki-platform-core/org.xwiki.platform$xwiki-core-rendering-tests/5845/testReport/
vmassol - (11:19): but same here: http://hudson.xwiki.org/job/xwiki-platform-core/org.xwiki.platform$xwiki-core-rendering-tests/5836/testReport/
CalebJamesDeLisl - (11:19): Better change the heading while I'm at it... "Sign up" ? "Join XWiki.org" ? "Join the community" ?
tmortagne - (11:19): vmassol: probleme is that most of the time is taken by component initi and things like that i guess
vmassol - (11:19): sure but there's a non negligible part spent on the parsing
florinciu left at 11:19 (Read error: Connection reset by peer
vmassol - (11:20): we don't see subseconds
vmassol - (11:20): but it doesn't seem it's improving parsing times dramatically at meast
vmassol - (11:20): *least
tmortagne - (11:20): we would need a real rendering parsing perf test, sdumitriu wrote one to compare to xwiki/1.0 i think
vmassol - (11:21): CalebJamesDeLisl: you could explain that an account is only needed for participating, ie working on improving the doc and that it's not required for reading, or something like that. I don't remember the message I had put before (it's been removed)
tmortagne - (11:21): vmassol: well the parsing is a lot cleaner but not sure LOOKHEAD is that wrong for perf
mflorea left at 11:21 (Quit: Leaving.
florinciu joined #xwiki at 11:21
tmortagne - (11:21): s/is/was/
vmassol - (11:21): tmortagne: right
CalebJamesDeLisl - (11:22): it's been removed? where was it?
vmassol - (11:22): on that page
vmassol - (11:22): could be found in history maybe
CalebJamesDeLisl - (11:22): that page? XWiki.Registration?
vmassol - (11:23): yes
vmassol - (11:23): (depends if the history was lost when the page was upgraded)
CalebJamesDeLisl - (11:23): I see it's from the last install. Sure it wasn't in registerinline.vm?
vmassol - (11:24): it was in the wiki for sure
CalebJamesDeLisl - (11:24): I'll look.
vmassol - (11:24): I had added XWiki.Registration for that purpose
CalebJamesDeLisl - (11:24): "Join the community" good heading?
vmassol - (11:25): I'd prefer to show that it's a join as in : help improve the content
vmassol - (11:25): (showing that you don't need to join if all you want is read the wiki)
CalebJamesDeLisl - (11:25): I was thinking Join the community as the heading then help improve the content in the welcome message (smaller)
vmassol - (11:26): maybe we should have a bullet list explaining what benefits it gives to register
vmassol - (11:26): not sure
vmassol - (11:26): join the community is fine otherwise caleb
CalebJamesDeLisl - (11:27): Ok, I'll change that and look at the revision history. (Strange to get my own messages offering me to admin the page)
vmassol - (11:27): :)
CalebJamesDeLisl - (11:29): Your change is still there. Keep the warning box or put it in text in the welcome message?
vmassol - (11:30): as you want
vmassol - (11:30): you can reword, change too
CalebJamesDeLisl - (11:31): I think yellow warnings are too scary for a registration page.
jvdrean joined #xwiki at 11:43
jvelociter joined #xwiki at 11:59
vmassol left at 12:02 (Quit: Leaving.
vmassol joined #xwiki at 12:04
mflorea joined #xwiki at 12:05
tmortagne - (12:07): did something changed something on http://l10n.xwiki.org ? flags are not shown anymore
vmassol - (12:14): l10n is quite slow btw
tmortagne - (12:14): that's normal
tmortagne - (12:14): i just imported some langauge
tmortagne - (12:14): language
vmassol - (12:14): ok
tmortagne - (12:14): just wait everithing to be in caches
vmassol - (12:14): I'm checking http://l10n.xwiki.org/xwiki/bin/view/Main/AllDocs to see if there were changes
tmortagne - (12:14): there http://l10n.xwiki.org/xwiki/bin/edit/L10N/Flags?editor=object which is supposed to be loaded always but it does not seems so
tmortagne - (12:14): looks like jsx that does not work for some reason
vmassol - (12:16): if you filter on L10N in the space column you can see changes
tmortagne - (12:16): looks like a bug, not sure it's because of someone change
vmassol - (12:16): but the wiki wasn't upgraded was it?
tmortagne - (12:17): i don't think so
tmortagne - (12:17): but maybe restared
DV_ left at 12:17 (Ping timeout: 258 seconds
DV_ joined #xwiki at 12:17
lucaa left at 12:21 (Quit: Leaving.
evalica1 joined #xwiki at 12:21
lucaa joined #xwiki at 12:21
sdumitriu joined #xwiki at 12:23
vmassol - (12:27): CalebJamesDeLisl: seems we haven't had any new user since 1st of MArch on xwiki.org
CalebJamesDeLisl - (12:27): I have to sign up a sandbox user so I'll make sure it's working.
vmassol - (12:28): actually it's not true
vmassol - (12:28): it's just the news page that is not reporting correctly
vmassol - (12:28): this user was created on the 19th for ex: http://www.xwiki.org/xwiki/bin/view/XWiki/kladko?viewer=history&showminor=true
evalica left at 12:29 (*.net *.split
plunden left at 12:29 (*.net *.split
vmassol - (12:29): we're cheating to find users on the News page but the algo isn't good
vmassol - (12:29): #set ($sql = "where 1=1 and doc.comment = 'Created user' order by doc.date desc")
vmassol - (12:30): so if the user creates a watchlist for ex
vmassol - (12:30): it's moved to modifications and no longer appears in new users
vmassol - (12:30): I wonder why we don't simply search for XWikiUsers objects hmmm
bblfish - (12:44): There is an error on http://webid.myxwiki.org/xwiki/bin/view/Main/WebHome
bblfish - (12:44): "Error number 4001 in 4: Error while parsing velocity page /templates/contentview.vm Wrapped Exception: Failed to evaluate content with id"
bblfish - (12:45): any idea how that happened?
lucaa - (12:45): hi bblfish it's probably a script failure
bblfish - (12:46): I suppose this one http://webid.myxwiki.org/xwiki//templates/contentview.vm
lucaa - (12:46): I mean it's surely a script failure
lucaa - (12:46): lemme check
bblfish - (12:46): but how did it come to fail
vmassol - (12:46): hmmm
bblfish - (12:46): ?
lucaa - (12:46): it's either in a doc or in the template
vmassol - (12:46): seems to be a bug in XE 2.3M1 somehwere
lucaa - (12:46): it can be a script in a xwiki document
bblfish - (12:46): did you just upgrade xwiki?
vmassol - (12:46): Caused by: java.lang.NullPointerException
vmassol - (12:46): at com.xpn.xwiki.doc.XWikiDocument.getXDOM(XWikiDocument.java:6914)
vmassol - (12:46): at com.xpn.xwiki.doc.XWikiDocument.getDefaultEditMode(XWikiDocument.java:5531)
vmassol - (12:46): at com.xpn.xwiki.api.Document.getDefaultEditMode(Document.java:1673)
vmassol - (12:46): at sun.reflect.GeneratedMethodAccessor2106.invoke(Unknown Source)
vmassol - (12:46): at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
vmassol - (12:46): at java.lang.reflect.Method.invoke(Method.java:597)
vmassol - (12:46): at org.apache.velocity.util.introspection.UberspectImpl$VelMethodImpl.doInvoke(UberspectImpl.java:389)
vmassol - (12:46): yes we did
vmassol - (12:46): yesterday afternoon
bblfish - (12:47): ah ok
vmassol - (12:47): btw we have a CSS issue too when clicking on the error to open the stack trace
vmassol - (12:47): any CSS lover here who could help fix that one? ;)
bblfish - (12:49): I am a bit surprised because I have the feeling that it was working not that long ago. Someone seems to have reported that they got a WebId there
vmassol - (12:49): tmortagne: any idea?
lucaa - (12:50): vmassol: if you refer the fact that the line goes 'out of the screen' I think it's the <pre> around the stacktrace that does it, which is why it doesn't do normal line wrap...
lucaa - (12:50): don't have a solution right now, I would need to try stuf
CalebJamesDeLisl - (12:51): vmassol: Did you get that error out of the logs? I'm getting a different error ("Failed to parse input source" @ XWikiDocument.java:6992)
lucaa - (12:51): CalebJamesDeLisl: vmassol, yes me too, I don't have the NPE
tmortagne - (12:51): vmassol: according to the stack trace you copy pasted you should an error log before that
lucaa - (12:51): in the page I mean
tmortagne - (12:52): if (this.xdom == null) {
tmortagne - (12:52): try {
tmortagne - (12:52): this.xdom = parseContent(getContent());
tmortagne - (12:52): } catch (XWikiException e) {
tmortagne - (12:52): LOG.error("Failed to parse document content to XDOM", e);
tmortagne - (12:52): }
tmortagne - (12:52): }
tmortagne - (12:52): return this.xdom.clone();
TaurusLee left at 12:52 (Ping timeout: 256 seconds
tmortagne - (12:52): nullpointer here means that parse failed
tmortagne - (12:52): parseContent
vmassol - (12:53): tmortagne: you can see the error here: http://webid.myxwiki.org/xwiki/bin/view/Main/WebHome
MichelleShi joined #xwiki at 12:53
vmassol - (12:54): CalebJamesDeLisl, lucaa: check the bottom of the page after you clicked on the error
lucaa - (12:54): vmassol: I checked all
vmassol - (12:54): and you don't see what I pasted at the bottom ?
lucaa - (12:54): nop
lucaa - (12:54): maybe we're guest, and you're logged in
lucaa - (12:55): it seems to reproduce
lucaa - (12:55): I get the same on all refreshed
lucaa - (12:55): s
lucaa - (12:55): which is this: http://pastebin.com/54YC7p5W
vmassol - (12:56): that's weird
bblfish - (12:56): I think I get the same
lucaa - (12:56): aaaah
lucaa - (12:56): I logged in
lucaa - (12:56): and now I see your css issue
lucaa - (12:56): and the NPE
lucaa - (12:56): :)
vmassol - (12:56): ok
bblfish left at 12:57 (Quit: Leaving.
kibahop joined #xwiki at 12:58
vmassol - (13:00): tmortagne: apparently parseContent could return null
vmassol - (13:00): (if I a read the code correcdtly)
vmassol - (13:01): hmm this would mean that parser.parse() would return null
vmassol - (13:01): could it be some pb caused by the wikimodel upgrade of yesterday?
vmassol - (13:02): (the NPE is on this.xdom.clone())
florinciu left at 13:02 (Quit: Leaving.
CalebJamesDeLisl - (13:04): Looks like every single script on webid is broken.
lucaa - (13:04): vmassol: the CSS issue is related to the NPE, when i was guest I got nicely formatted errors
bblfish joined #xwiki at 13:05
bblfish - (13:06): sorry just installed latest os update, anything new?
vmassol - (13:07): we need tmortagne, I think it's related to the upgrade of yesterday
vmassol - (13:07): it looks as if, under some condition, the updated parser can return null
vmassol - (13:07): I think he went out for lunch
bblfish - (13:08): ok, thanks for the update
kibahop left #xwiki at 13:08
CalebJamesDeLisl - (13:10): Hmm, only Main.WebHome, when logged in and trying to edit (wiki mode), all scripts on the page break.
CalebJamesDeLisl - (13:13): This works, (2 versions old) http://webid.myxwiki.org/xwiki/bin/viewrev/Main/WebHome?rev=34.1
CalebJamesDeLisl - (13:16): Here's the change that broke it: http://webid.myxwiki.org/xwiki/bin/view/Main/?viewer=changes&rev1=34.1&rev2=35.1
lucaa - (13:17): CalebJamesDeLisl: I think that's just a side effect, there is indeed something wrong in the templates. for example I get an exception in the blue action bar when I view the changes
lucaa - (13:17): Error number 4001 in 4: Error while parsing velocity page Wrapped Exception: Failed to evaluate content with id []
lucaa - (13:17): instead of the menus
lucaa - (13:17): and the same NPE in the stack
CalebJamesDeLisl - (13:18): In the templates or in the parser. I was trying to figure out how to reproduce the trouble.
lucaa - (13:19): bblfish: can I rollback?
lucaa - (13:19): to 35.1, I'd like to see it 'in context'
bblfish - (13:19): hi
CalebJamesDeLisl - (13:20): Would preview work for what you need?
lucaa - (13:20): to 34.1 to be precise
bblfish - (13:20): the front page?
bblfish - (13:20): not sure what you want to roll back
lucaa - (13:20): well, CalebJamesDeLisl that's what I would like to check, if it's only the preview that makes 34.1 work or it would work anyway
lucaa - (13:20): Main.WebHome of webid
bblfish - (13:20): and I am not sure what 34.1 was
lucaa - (13:21): to version 34.1
lucaa - (13:21): http://webid.myxwiki.org/xwiki/bin/view/Main/WebHome#History
bblfish - (13:21): yes certainly
lucaa - (13:21): and these would be the changes:
lucaa - (13:21): http://webid.myxwiki.org/xwiki/bin/view/Main/?viewer=changes
bblfish - (13:21): ah ok, that's a small change
bblfish - (13:22): looks like the change was correct too
lucaa - (13:22): very interesting, it works http://webid.myxwiki.org/xwiki/bin/view/Main/
bblfish - (13:22): weird
bblfish - (13:23): is this wrong? [[RDFa>>http://www.w3.org/TR/xhtml-rdfa-primer/]]
CalebJamesDeLisl - (13:23): bblfish: this is the change that broke it: http://webid.myxwiki.org/xwiki/bin/view/Main/?viewer=changes&rev1=34.1&rev2=35.1
lucaa - (13:23): nop
lucaa - (13:24): yes, that's it, I can't re-rollback to 35 now nor 36 since there;s that error...
lucaa - (13:24): I can manually re-apply changes
lucaa - (13:26): CalebJamesDeLisl: aUser is you?
CalebJamesDeLisl - (13:27): Yes it's me
lucaa - (13:27): I
CalebJamesDeLisl - (13:27): I can't login for more than one page load without getting kicked out though.
lucaa - (13:27): can I force Main.WebHome?
lucaa - (13:27): are you using chrome/
lucaa - (13:27): ?
CalebJamesDeLisl - (13:28): No FF. Can't live without noscript :)
lucaa - (13:28): interesting, i knew that issue on chrome
lucaa - (13:28): can I force edit Main.WebHome?
CalebJamesDeLisl - (13:28): Oh right yes.
CalebJamesDeLisl - (13:29): The lock.. forgot about that.
bblfish - (13:29): CalebJamesDeLisl: it may be worth trying adding a first and last name to aUser to see if that solves your problem
CalebJamesDeLisl - (13:29): bblfish: I'll give it a shot but I think it's a cookie domain issue.
nickless_ joined #xwiki at 13:36
lucaa - (13:38): now that's interesting
lucaa - (13:39): Main.WebHome is working now
lucaa - (13:40): and when I try to apply these changes: http://webid.myxwiki.org/xwiki/bin/view/Main/?viewer=changes&rev1=36.1&rev2=41.1 to get back to the exact 36.1 version, I can't save, I get an exception
lucaa - (13:40): a NPE
bblfish - (13:40): mysterios
CalebJamesDeLisl - (13:41): It does some rendering when you do save and view, did you try save and continue.
CalebJamesDeLisl - (13:41): ?
lucaa - (13:42): I can, but I mean I get the exception from some spaces added at the end of ===
bblfish - (13:43): !
lucaa - (13:45): I could save & continue and now I can also render
CalebJamesDeLisl - (13:45): The kicked out problem comes from being logged in at www.myxwiki.org and at webid.myxwiki.org. cookie domain issue.
lucaa - (13:45): I mean view it, and there are no diffs
CalebJamesDeLisl - (13:46): I'm editing WebHome just to play with preview, feel free to force the lock.
vmassol - (13:46): we need to ping tmortagne when he's back from lunch since this looks like a regression in the parser IMO
vmassol - (13:47): (I'm just back from lunch, trying to catch up)
lucaa - (13:49): omg, is this random? I can view current page but when I go to history and I want to view revision 45.1 (current) I get an exception
CalebJamesDeLisl - (13:50): Playing with preview, it seems that the === Code === line is the only one which can't have spaces after ===
lucaa - (13:51): I managed to reproduce for Todo too...
lucaa - (13:51): I give up
lucaa - (13:51): and we need to do smth about the diffs, I don't understand this one, for example:
lucaa - (13:51): http://webid.myxwiki.org/xwiki/bin/view/Main/?viewer=changes&rev1=41.1&rev2=42.1
lucaa - (13:52): except for the case where it's an encoding difference
arkub joined #xwiki at 13:57
CalebJamesDeLisl - (13:57): If the section has a * (bullet) in it then a space after the == works, otherwise it breaks.
bblfish - (13:58): could it be that the space is some weird blank encoding
bblfish - (13:59): there are some spaces that are not ascii spaces
lucaa left at 13:59 (Quit: Leaving.
lucaa joined #xwiki at 13:59
CalebJamesDeLisl - (14:00): I think my space bar is the normal ascii space. This is definitely a parser problem.
MichelleShi - (14:02): hi, how to change the skin color of my wiki in myxwiki.org? I have to select the color in Administrator Console, but it doesn't work.
lucaa - (14:03): I can also reproduce locally: I got the content from http://webid.myxwiki.org/xwiki/bin/viewrev/Main/WebHome?viewer=code&rev=36.1 (36.1 is the initial failing version, before my rollback) and put it in a page and I can
lucaa - (14:03): can't save
lucaa - (14:03): because I get a NPE
lucaa - (14:03): you should know that save & continue says "saved" though, although it doesn't save anythin
lucaa - (14:03): g
lucaa - (14:04): ay ay, indeed save and continue POST response (that I looked at with firebug) is the exception, although it's a misleading 200 OK
lucaa - (14:05): (we also need to fix this one)
lucaa - (14:06): the exception on save is when saving links (backlinks) which needs to parse the XDOM, so it _will_ occur, regardless of save & continue (without rendering) or save & view with rendering
anamarias - (14:07): lucaa: how do I get the name of the document in which a panel is located in Velocity? (my code is in the panel)
anamarias - (14:07): in JS I know you look for <meta name="document" content="NewVocabulary.WebHome"/>
vmassol - (14:07): anamarias: $doc
anamarias - (14:07): hmm
anamarias - (14:07): $doc gives me the name of the panel document
vmassol - (14:08): it should give the name of the doc where it's included
lucaa - (14:08): anamarias: are you testing your panel in the Panel view page or in an actual document?
anamarias - (14:08): I'm doing ${doc.fullName} in the content atrribute of the PanelClass
anamarias - (14:08): in the actual document
lucaa - (14:09): maybe it could be something like $cdoc $idoc ...
lucaa - (14:09): but I wouldn't know to tell you which one
CalebJamesDeLisl - (14:10): bblfish: re CA certificates: I did some thinking, we shouldn't need to authenticate the server because the client does not send the public key in plaintext so it is impossible for a middleman to modify what the client sent (put their own key on it) and the MITM attack does not work. Am I correct?
CalebJamesDeLisl - (14:11): (to be more correct the MITM attack will require middlemanning the connection between the authenticating server and the key server)
bblfish - (14:11): the client does send public key in plain text, but it's not a problem
CalebJamesDeLisl - (14:12): So I think MITM fails even if the server key is compromised which is awesome.
bblfish - (14:12): If you can find a way for a browser to connect to a server that does not have a CA signed key without showing an error message that would be great
anamarias - (14:12): yup, $doc is correct in the content, but I see here's a <script language="JavaScript"> and inside the script $doc no longer points to the embedding document
CalebJamesDeLisl - (14:12): We create our own CA :D
anamarias - (14:13): what happens with the scripts ? from the content ?
lucaa - (14:13): anamarias: nothing
lucaa - (14:13): what panel is that syntax 1.0 ?
lucaa - (14:13): or 2.0?
anamarias - (14:13): well, inside the <script> tag, $doc is the panel document name, not the embedding doc's
CalebJamesDeLisl - (14:13): bblfish: Yesterday I installed the CAsert certificate in firefox and it was actually easier than viewing an unsigned website.
lucaa - (14:13): there should be no <script language at any point, one should use jsx and then do an include
bblfish - (14:13): "We Create out own CA" then you are back to the problem of how you get that CA's public key in the browsers, and how you have solved the problem of distribution, as you now once again have a bottleneck
bblfish - (14:14): also your above solution is insecure
anamarias - (14:14): I know lucaa , just wondering what happens in this case :p
lucaa - (14:14): anamarias: what syntax is that?
anamarias - (14:14): the panel doc is 1.0
bblfish - (14:15): if it is going to apply to all sites, as it could allow anyone to pretend to be xwiki.com
CalebJamesDeLisl - (14:15): Well the user joins a website they create their own key, that's a hassle, why not add a little more hassle and ask them to install a key.
bblfish - (14:15): and put links to malware in xwiki
bblfish - (14:16): they would have to install that key for every web site they visit
lucaa - (14:16): anamarias: well that's very strange since syntax 1.0 interprets groovy, then velo, then wikisyntax and in the end the resulted html is printed. So there is no awareness of <script> tags at velo rendering level
bblfish - (14:16): this problem will be much easier to solve CalebJamesDeLisl with DNSsec
arkub left at 14:16 (Read error: Operation timed out
CalebJamesDeLisl - (14:16): I know if we create a key and publish the private key we break all authentication.
bblfish - (14:16): where you will be able to place your public key securely in the dns attributes
lucaa - (14:17): except for the case when something very special happens for the <script> tag...
CalebJamesDeLisl - (14:17): anyone can pretend to be a bank etc.
CalebJamesDeLisl - (14:17): Perhaps we could create a key which said "this key cannot authenticate this site" in the address bar.
anamarias - (14:18): lucaa: thanks, I'll just move all the js to a jsx obj and get the doc name from the meta tag
anamarias - (14:18): that's how it was suppose to be anyways
CalebJamesDeLisl - (14:19): bblfish: ping me when you're back.
plunden joined #xwiki at 14:19
lucaa - (14:19): anamarias: better. The pb with <script in content is that it's hard to follow the executed js, and also, in 1.0 docs, there can be pbs with syntax which has meaning as wikisyntax too (think of array access with [], which is link in wikisyntax1)
nickless_ is now known as nickless ([email protected]
nickless is now known as nickless_ ([email protected]
nickless_ - (14:21): <_<
nickless_ is now known as nickless ([email protected]
DV_ left at 14:28 (Read error: Connection reset by peer
tmortagne - (14:31): vmassol: what content can a test to reproduce the issue ? also as i said i doubt parse return null but i think it throws an exception
vmassol - (14:32): tmortagne: you'll need to read the irc chat backlog, lucaa and CalebJamesDeLisl have isolated the pb above I think
lucaa - (14:32): tmortagne: copy paste of this: http://webid.myxwiki.org/xwiki/bin/viewrev/Main/WebHome?viewer=code&rev=36.1
lucaa - (14:32): I reproduced it on local too with that content
tmortagne - (14:32): lucaa: ok thanks, i'm testing
lucaa - (14:33): but yes, read the log, we kindof almost isolated something (I didn't understand much and gave up at some point)
tmortagne - (14:33): hmm this code view is very bad, i will have to remove all the lines numbers
lucaa - (14:33): there's a button for that
tmortagne - (14:33): indeed did not seen the link
lucaa - (14:33): (over the textarea)
CalebJamesDeLisl - (14:33): tmortagne: === some section === <- a space after the last = causes the break.
CalebJamesDeLisl - (14:34): But it will not break if there is a * (bullet) in the section.
tmortagne - (14:34): CalebJamesDeLisl: ok thanks
tmortagne - (14:34): checking on wikimodel uni tests
tmortagne - (14:35): hmm it "=== some section === " seems to work on wikimodel at least
lucaa - (14:36): CalebJamesDeLisl: I just tried that (the space thing) and I can't reproduce it
tmortagne - (14:36): ok
lucaa - (14:36): with content in the para after, with space without space, everything works, on the very same instance that the content pasted from webid fails
CalebJamesDeLisl - (14:36): Ok it must have something to do with something else on the page.
tmortagne - (14:36): checking with the big content
jvdrean1 joined #xwiki at 14:37
DV joined #xwiki at 14:38
lucaa - (14:38): tmortagne: my snapshot on which I could reproduce is 2.3-20100317.192952-233
tmortagne - (14:38): ok i can reproduce, also there is too issue here, whatever happen in saveLinks it should not break the whole save
tmortagne - (14:39): two
jvdrean left at 14:39 (Ping timeout: 268 seconds
tmortagne - (14:39): trying to find the exact issue
tmortagne - (14:42): ok so it's not returning null, the parser produce an exception, i will try to reduce to the exact issue
CalebJamesDeLisl - (14:43): http://pastebin.com/626pRCGn
CalebJamesDeLisl - (14:44): note the space after === Test ===
CalebJamesDeLisl - (14:45): The last line "word" is not needed, but pastebin eats trailing empty lines.
vmassol - (14:46): [maintenance] will restart hudson since it's having problems
tmortagne - (14:46): CalebJamesDeLisl: no need for a white space after = actually
tmortagne - (14:47): or maybe yes
tmortagne - (14:47): that's weird
tmortagne - (14:47): sometime it fail sometime it's not, i will test in wikimodel to not be distracted
lucaa - (14:47): CalebJamesDeLisl: the space between first === and Test is the problem
lucaa - (14:47): I removed it and retypesd it and stoped failing
lucaa - (14:47): (or am I going mad)
lucaa - (14:47): ?
lucaa - (14:48): slightly
savi joined #xwiki at 14:48
CalebJamesDeLisl - (14:48): Gremlins.
savi left #xwiki at 14:48
lucaa - (14:48): (nop, I'm going mad)
CalebJamesDeLisl - (14:49): Must be a thread timing error, sometimes fails sometimes not.
MichelleShi left at 14:49 (Quit: ??
tmortagne - (14:51): can't find smaller than: "= H = \nP\n\n"
tmortagne - (14:52): (at wikimodel level)
evalica1 left at 14:53 (Ping timeout: 265 seconds
lucaa left at 14:53 (Quit: Leaving.
CalebJamesDeLisl - (14:54): Yup that's what I am now testing with.
CalebJamesDeLisl - (14:55): "= = \nP\n\n"
tmortagne - (14:56): actually Adreas already did find and fixed the issue but google code doe snot send mail for closed issue it seems, http://code.google.com/p/wikimodel/issues/detail?id=169
tmortagne - (14:56): i'm fixing it
Enygma` left at 15:02 (Quit: Leaving.
CalebJamesDeLisl - (15:02): the problem was not present in 2.3-27617
vmassol - (15:04): sdumitriu: I'm cleaning myxwiki.org of all "spam" pages. How do I remove the "Blog" space completely, including hidden docs? http://myxwiki.org/xwiki/bin/view/Main/Dashboard
vmassol - (15:05): btw it's strange for users, since if you list all pages in the blog spaces you don't see any so you need to know there are hidden docs to understand why it's listed
plunden left #xwiki at 15:06
florinciu joined #xwiki at 15:17
nuvolari - (15:19): howcome do I get the general idea that glassfish is magnitudes more difficult to work with? :/
sdumitriu - (15:22): vmassol: Hidden pages are bad, we need to have real support for hidden pages soon
tmortagne - (15:23): yes the current implementation of hidden page broke many things
sdumitriu - (15:24): vmassol: You can Ctrl+G and type Blog. to see remaining blog pages
vmassol - (15:24): ok trying that
vmassol - (15:25): wow quite a lot to manually delete....
jvdrean1 - (15:28): restarting hudson
vmassol - (15:28): wait
vmassol - (15:28): you shouldn't have to do that
vmassol - (15:28): it's restating by itself
vmassol - (15:28): jvdrean1:
jvdrean1 - (15:33): if it supposed to do it by itself it did not this time
vmassol - (15:33): it worked in the past, strange
vmassol - (15:33): are you sure you waited enough?
vmassol - (15:33): it stays unavail for a few minutes normally
vmassol - (15:33): and then start again
jvdrean1 - (15:33): at least they weren't any job building
vmassol - (15:34): ???
vmassol - (15:34): when I checked the page it was down
vmassol - (15:34): how could there be builds?
jvdrean1 - (15:34): when you checked the page I already had stopped it I suppose
vmassol - (15:35): I checked about 2 secons after you posted!
vmassol - (15:35): anyway....
jvdrean1 - (15:35): anyway I don't think there's a big deal here, I just wanted to have it back quickly to help Florin with its windows agent
vmassol - (15:35): yes but if there's a pb with reatsrt we should fix it
vmassol - (15:36): we'll see next time
jvdrean1 - (15:36): TBH I don't remember this auto-restart feature, what I remember is a link in the admin to restart from the web
vmassol - (15:36): I've used it several times already
TaurusLee joined #xwiki at 15:38
TaurusLee left #xwiki at 15:39
vmassol - (15:46): tmortagne: found a regression in the parsing I think (maybe it's the same pb): http://myxwiki.org/xwiki/bin/view/IRC/IRCBot
vmassol - (15:46): ah no
vmassol - (15:46): forget that
tmortagne - (15:46): cool
vmassol - (15:46): it was some leftover ))) that I had not noticed on the page
vmassol - (15:46): since they were hidden before I didn't see them
tmortagne - (15:46): so it's a fix :)
vmassol - (15:46): yes
vmassol - (15:46): :)
tmortagne - (15:47): [release] starting Core release
vmassol - (15:48): btw fyi I had installed the IRC bot app on myxwiki.org to follow what's going on (it sends notifs on the myxwiki irc channel)
vmassol - (15:48): but since it's not restarted automaitcally for now and since myxwiki.org restarts often it's often down
tmortagne - (15:49): " myxwiki irc channel" ?
vmassol - (15:49): #myxwiki
tmortagne - (15:49): never hear about it
vmassol - (15:50): yes that's normal
vmassol - (15:50): I was doing some internal test about it
vmassol - (15:50): before talking about it
tmortagne - (15:50): ok
headache left at 15:52 (Remote host closed the connection
vmassol - (15:52): jvelociter: can I remove this space: http://myxwiki.org/xwiki/bin/view/Main/SpaceIndex?space=Macros ?
jvelociter - (15:53): vmassol: yes. please don't remove the macros in space XWIki/ yet
tmortagne - (15:56): as usual after an hudson uprade: xvnc lost
florinciu left at 15:56 (Read error: Connection reset by peer
jvdrean1 - (15:57): tmortagne: yes, the usual, want me to restart it ?
tmortagne - (15:58): was just a comment, i'm on hudson serveur anyway i will do it
vmassol - (15:58): fyi, removing IRC bot app from myxwiki.org, it probably wasn't such a good idea
DV left at 16:02 (Read error: Connection reset by peer
evalica joined #xwiki at 16:17
vmassol - (16:20): hmm I can't logout on myxwiki.org
mflorea - (16:23): sdumitriu: I'm debugging what happens with the response object when I visit a RSS link ( http://localhost:8080/xwiki/bin/view/Main/WebRss?xpage=plain&outputSyntax=plain ) and the response content type is set to text/html. Related to this, when I download an attachment ( http://localhost:8080/xwiki/bin/download/Sandbox/FF/xwiki-logo.png ) the content type is set three time, image/png in between of two text/html. This looks wrong to me. Is the
sdumitriu - (16:24): It should not be set to text/html
sdumitriu - (16:24): Who does the last call?
mflorea - (16:24): let me see
tmortagne - (16:30): [release] Note that i can't finish the release today because ow2 is down, it will still finish the maven part
mflorea - (16:33): sdumitriu: actually, for the image is fine, the last text/html is due to a redirect to the exception template (I'm throwing the exception so it's my fault). What about the RSS? Although the content type is set to text/html, FF detects it as application/xhtml+xml (in page info)
sdumitriu - (16:34): RSS should be rss
DV joined #xwiki at 16:34
mflorea - (16:36): you mean application/rss+xml ?
florinciu joined #xwiki at 16:36
sdumitriu - (16:36): $response.setContentType('application/rdf+xml')
mflorea - (16:38): some people (google search) say that the specifica mime type is not handled by some browsers (including IE). For these, they say text/xml is recommended. I haven't tested.
bblfish left at 16:39 (Read error: Connection reset by peer
florinciu - (16:41): hi
florinciu - (16:42): I'm adding the .net build plugins to hudson. It will need another restart. Is everybody ok with it?
tmortagne - (16:42): no
tmortagne - (16:42): i need xwiki-product-enterprise
DV left at 16:42 (Read error: Connection reset by peer
tmortagne - (16:43): after that you can do whatever you want with hudson
DV joined #xwiki at 16:44
jvelociter - (16:46): vmassol: tmortagne Q about the component manager
jvelociter - (16:46): can I have a class that implement two ComponentRole ?
jvelociter - (16:46): what's the result ? two components ?
tmortagne - (16:51): yes
tmortagne - (16:51): (two components)
jvelociter - (16:51): ok
tmortagne - (16:51): same fore role hints
jvelociter - (16:52): so if a component makes sense as a script service I can add ScriptService impl
jvelociter - (16:53): tmortagne hum, but is there a way I can have a hint for each impl ?
vmassol - (16:54): jvelociter: it's not recommended
tmortagne - (16:54): jvelociter: i don't understand that's what i just said: 1 component by role and by role hint
vmassol - (16:54): since you'll have 2 instances in memory
jvelociter - (16:56): tmortagne: yes but I have 2 roles. I want roleHint1 for role1 and roleHint2 for role2
jvelociter - (16:56): does it make sense ?
vmassol - (16:56): having a component with 2 roles is usually a bad practice
vmassol - (16:57): it means several concers are being mixed
vmassol - (16:57): *concerns
tmortagne - (16:57): jvelociter: you can't do that because the loader has no idea what hint id for what role
jvelociter - (16:57): vmassol: I see. On the other hand the alternative is to have a ScriptService impl that has the exact same interface as the interface it provides access too
jvelociter - (16:57): tmortagne: yes that was my question. ok
jvelociter - (16:57): vmassol: but that's ok I guess
vmassol - (16:57): you can reuse the same role
vmassol - (16:58): what's not good is if your component impl has more than 1 role
jvelociter - (16:58): vmassol: how can I reuse the same role ?
tmortagne - (16:58): a role is an interface
jvelociter - (16:58): I need to have the ScriptService role. so that's 1
vmassol - (16:58): have 2 clases implementing it
tmortagne - (16:58): you implemt the interface in two different classes
jvelociter - (16:58): that does not work
jvelociter - (16:59): since I'll need one that implements ScriptService which will make to 2 roles
jvelociter - (17:00): do you see what I mean ?
tmortagne - (17:00): wdym ? ScriptService is one role
tmortagne - (17:00): no i don't
jvelociter - (17:02): imagine you have a MyComponent interface, and a MyComponentImpl implementation that you also want to provide acccess to scripts.
vmassol - (17:02): bb in 20mn
jvelociter - (17:02): you can have "MyComponentImpl implements MyComponent, ScriptService"
jvelociter - (17:02): but then that's implementing 2 roles so that's not good
jvelociter - (17:03): or you can have a separate "MyComponentService implements ScriptService" that wrap calls to MyComponent, but that's a shame since it actually implements MyComponent
jvelociter - (17:03): see what I mean ?
jvelociter - (17:07): (tmortagne)
tmortagne - (17:07): i see that you have two different components
jvelociter - (17:07): yes ok. but both acts as a duck
LadySerena left at 17:08 (Quit: Time to sharpen my claws!
jvelociter - (17:08): so I could have just one implementation
tmortagne - (17:08): ScriptService is supposed to be a filter to manipulate a component from script
jvelociter - (17:08): but it's ok
jvelociter - (17:09): it was more a rethorical question because in the end I need to treat exceptions in the script service
jvelociter - (17:09): did not thought about that
tmortagne - (17:09): if you are sure anyone can access everything in MyComponent from script all i can see is a getMyComponent in your MyComponentService
tmortagne - (17:10): but usually that's not the goal of ScriptService
jvelociter - (17:10): what do you mean getMyComponent ?
jvelociter - (17:10): I will have my component as a requirement
tmortagne - (17:11): you get the entire MyComponent in a get of your service instead of filtering it
jvelociter - (17:11): ah I see
tmortagne - (17:11): but this is dangerous IMO
jvelociter - (17:11): yes I don't like it too much either
jvelociter - (17:11): because in the future it might not be the case anymore
jvelociter - (17:11): (having everything accessible)
tmortagne - (17:12): [release] maven release finished, will do the ow2 part on monday (ow2 is down)
tmortagne - (17:12): yep
tmortagne - (17:13): jvelociter: if the goal is to have both service and component to have the same api you can do the following
tmortagne - (17:13): * a simple interface with the api that is not a component role
tmortagne - (17:14): * extends this interface in both MyComponentService component and MyComponent role
jvelociter - (17:14): yes good idea. other components could ask for the script service if they need access to that API
tmortagne - (17:14): i mean implements in MyComponentService and extends in MyComponent
tmortagne - (17:14): no
tmortagne - (17:14): that's not what i said
jvelociter - (17:14): well
tmortagne - (17:15): i said: you have 3 different interfaces
tmortagne - (17:15): not 3 interface
tmortagne - (17:15): but 3 entity:
tmortagne - (17:16): * common api
tmortagne - (17:16): * role
tmortagne - (17:16): * script service
tmortagne - (17:16): component you not use a scriptservice that's wrong
tmortagne - (17:17): a interface role can extends another interface that is not a role
jvelociter - (17:17): ah I see the trick
jvelociter - (17:17): ok
tmortagne - (17:17): to force it to implements the right methods
mflorea left at 17:17 (Quit: Leaving.
jvelociter - (17:17): but then
tmortagne - (17:17): but again at the end i think you always have difference between component and the script service
jvelociter - (17:17): nothing implements the role
tmortagne - (17:18): let me say it another way:
tmortagne - (17:19): * CommonInterface
tmortagne - (17:19): * MyScriptService implements CommonInterface
tmortagne - (17:19): * MyRole extends CommonInterface
tmortagne - (17:19): * MyComponent implements MyRole
tmortagne - (17:19): with that you have MyComponent and MyScriptService that have the same api
tmortagne - (17:19): which is CommonInterface
tmortagne - (17:19): the component and the service
jvelociter - (17:20): (yes the question is rethorical as I said they are already different since scripts service should not throw all exceptions as there is not way to catch them in velocity)
tmortagne - (17:20): not sure about this no exception thing, it depends
jvelociter - (17:20): ok I see
jvelociter - (17:20): yes I said not all
jvelociter - (17:20): it depends
jvelociter - (17:22): but you don't want to throw an exception that you know can happen and that has sense for what you asked. For example throws NotPossibleToCreateSomethignBecauseItAlreadyExists. You don't want your velocity script to fail on that, you want to display a nice message
jvelociter - (17:23): stacktraces are useless when the real problem makes sense to users
jvelociter - (17:23): = if you can explain it simply, probably you don't want to show a stacktrace
tmortagne - (17:23): yes, what i say is that iut should not be a catch(Exception)
jvelociter - (17:23): yes of course
jvelociter - (17:24): by exceptions I mainly meant those thrown by the component API
tmortagne - (17:24): and maybe we should have several kind of services
tmortagne - (17:24): since for groovy it would be better to have excpetions
jvelociter - (17:24): yes that's the eternal question templating service vs. velocity
tmortagne - (17:24): same for python, php, ruby probably
jvelociter - (17:24): but honestly who uses groovy to write an XWiki app ?
jvelociter - (17:25): ;)
jvelociter - (17:25): ok me sometimes for wiki macros
vmassol - (17:25): back
tmortagne - (17:25): now groovy can use directly component manager if it does not like script service
jvelociter - (17:25): yes, true
tmortagne - (17:25): but it's less handy
tmortagne - (17:26): it's not even possible without the crappy static Utils.getComponent yet
Denis left at 17:26 (Quit: Leaving.
jvelociter - (17:26): 1 more line
tmortagne - (17:26): so it's not possible
jvelociter - (17:26): well the getComponent exists :)
jvelociter - (17:27): so it's possible
tmortagne - (17:27): Utils.getComponent is not normal it's used for old xwiki-core code
tmortagne - (17:27): starting to use it in groovy is not teh best way to get rid of it
jvelociter - (17:27): yes in groovy we do a deviant usage
silviar left at 17:27 (Quit: Leaving.
jvelociter - (17:27): but we don't have alternatives so it's nornal people use it
tmortagne - (17:28): we could easily provide alternative
jvelociter - (17:28): couldn't we reuse the annotations ?
tmortagne - (17:28): WDYM ?
jvelociter - (17:30): @Requirement
vmassol - (17:30): that works, you just need to register a component
vmassol - (17:30): so not very handy
tmortagne - (17:30): i still don't see the relationship with scripting
vmassol - (17:30): usually scriptlets are not components
vmassol - (17:30): I do this in the irc bot app but it's not really conveinent at all
jvelociter - (17:30): ok
jvelociter - (17:30): you register the component from code ?
vmassol - (17:30): yes
jvelociter - (17:31): ok
jvelociter - (17:31): yes not very light
jvelociter - (17:31): so what's the easy alternative tmortagne ?
vmassol - (17:32): I think thomas was referrring to registering the CM into the context
vmassol - (17:32): or making it avail through scritpt service
tmortagne - (17:32): there is several:
tmortagne - (17:32): * inject a "componentManager" in ScriptContext in the script macro
tmortagne - (17:32): * $services.getComponentManager that test for programming right
tmortagne - (17:32): yep
tmortagne - (17:33): * ComponentClass xobject that require a component
tmortagne - (17:34): less easy but easy enough ;)
tmortagne - (17:34): (that could inject it in the ScriptContext as a binding)
tmortagne - (17:35): s/could/would/
jvelociter - (17:36): ComponentClass that would be cool
jvelociter - (17:36): what do you bind the requirement too ?
tmortagne - (17:36): i have some ideas to have wiki component (like we have wiki macros) but never put it on design page
jvelociter - (17:36): it should rather be RequirementClass actually
jvelociter - (17:37): and we could have ComponentClass that declares component
tmortagne - (17:37): the object just declare the requirement
tmortagne - (17:37): you still need to provide it
tmortagne - (17:37): and the way to provide a variabvle to a script is ScriptContext
tmortagne - (17:38): sure RequirementClass anyway whatever the name: an object
jvelociter - (17:38): yes but what variable do you bind the requirement to ?
jvelociter - (17:38): or you select the name in the object
jvelociter - (17:39): ah it was s/too/to/
jvelociter - (17:39): sorry
tmortagne - (17:40): let do a quick design page, i love design pages that are never closed :)
vmassol - (17:43): yes we talked about it I remember
vmassol - (17:44): ie ability to create components in wiki page using objects
vmassol - (17:46): good name
vmassol - (17:46): :)
tmortagne - (17:47): he you are
jvelociter - (17:49): thanks tmortagne ;)
jvelociter - (17:50): ComponentRoleClass: indicate which role the page is implementing
jvelociter - (17:50): don't you mean rather : the page declares a ComponentRole ?
jvelociter - (17:51): (and let the role implemented by components be a field int he ComponentClass)
tmortagne - (17:52): jvelociter: no
tmortagne - (17:52): i followed the annotations
tmortagne - (17:52): actually yes
tmortagne - (17:52): i did a mistake
jvelociter - (17:53): ah ok:)
tmortagne - (17:53): i wanted to declare the implemnted role in ComponentRoleClass but better in ComponentClass
jvelociter - (17:54): yes I agree
tmortagne - (17:54): you can't define a component in wiki
tmortagne - (17:54): no easy to use :)
jvelociter - (17:54): you mean you can't technically ?
tmortagne - (17:55): yes
jvelociter - (17:55): ok
tmortagne - (17:55): well not from java
tmortagne - (17:56): not even sure from scripting
tmortagne - (17:56): (s/not/not use/)
tmortagne - (17:59): added the important part
florinciu left at 18:01 (Read error: Connection reset by peer
florinciu joined #xwiki at 18:06
tmortagne - (18:07): hmm, this made me thing of an important improvement for wiki macros
tmortagne - (18:07): think
DV left at 18:08 (Read error: Connection reset by peer
DV joined #xwiki at 18:09
xenon75 left at 18:09 (Quit: Leaving.
plunden joined #xwiki at 18:15
tmortagne - (18:19): [myxwiki] starting upgrade
vmassol - (18:20): cool
tmortagne - (18:26): hmm tomcat does not seems to want to stop
vmassol - (18:26): yes I had this pb too
vmassol - (18:26): have you stopped the contrab first?
tmortagne - (18:27): first time i have this issue, i worked well last time
vmassol - (18:27): I remembered I had to kill tomcat at one point
tmortagne - (18:27): yes i commented crontab
tmortagne - (18:28): let's kill it, a shame
tmortagne - (18:32): [myxwiki] upgrade done
plunden left #xwiki at 18:35
tmortagne left at 18:40 (Quit: Leaving.
sdumitriu left at 18:59 (Ping timeout: 260 seconds
KermitTheFragger left at 19:02 (Quit: Leaving
CalebJamesDeLisl - (19:18): "Five minutes later I was in the possession of a legitimate certificate issued to mozilla.com - no questions asked" https://blog.startcom.org/?p=145
DV left at 19:18 (Read error: Connection reset by peer
CalebJamesDeLisl - (19:18): Just thought I'd share.
evalica left at 19:20 (Read error: Connection reset by peer
glerouge - (19:20): CalebJamesDeLisl: :)
CalebJamesDeLisl - (19:20): +1 for the foaf+ssl, it distributes the security.
DV joined #xwiki at 19:21
florinciu left at 19:23 (Read error: Connection reset by peer
glerouge left at 19:36 (Quit: Leaving.
DV left at 19:37 (Read error: Connection reset by peer
DV joined #xwiki at 19:38
xenon75 joined #xwiki at 19:54
xenon75 left at 19:56 (Client Quit
DV left at 20:33 (Read error: Connection reset by peer
jvdrean1 left at 21:10 (Quit: Leaving.
nuvolari left at 21:11 (*.net *.split
jfx left at 21:11 (*.net *.split
xipe left at 21:11 (*.net *.split
jfx joined #xwiki at 21:14
xipe joined #xwiki at 21:14
nuvolari joined #xwiki at 21:15
jvelociter left at 21:15 (Read error: Connection reset by peer
xenon75 joined #xwiki at 21:32
DV joined #xwiki at 21:45
xenon751 joined #xwiki at 21:57
xenon75 left at 21:59 (Ping timeout: 256 seconds
anamarias left at 22:18 (Read error: Connection reset by peer
DV__ joined #xwiki at 22:44
DV left at 22:46 (Read error: Connection reset by peer
vmassol - (23:07): xipe: are you porchounet ?
vmassol - (23:07): :)
vmassol - (23:08): (thanks, that made me laugh on a friday evening!)
xipe - (23:08): no, i am not
xipe - (23:08): xwiki bug ?
xenon751 left at 23:17 (Quit: Leaving.
headache joined #xwiki at 23:21
headache - (23:22): hello
vmassol left at 23:24 (Quit: Leaving.
headache left at 23:39 (Quit: Sto andando via
