IRC Archive for channel #xwiki

Last modified by Vincent Massol on 2012/10/18 19:11

nickless left at 00:41 (Quit: Leaving
sdumitriu left at 03:18 (Ping timeout: 276 seconds
vmassol joined #xwiki at 07:41
gmpff left at 07:59 (Ping timeout: 240 seconds
vmassol left at 08:04 (Quit: Leaving.
kibahop joined #xwiki at 08:07
vmassol joined #xwiki at 08:10
plunden left #xwiki at 08:22
mflorea joined #xwiki at 09:05
plunden joined #xwiki at 09:12
Denis joined #xwiki at 09:20
CalebJamesDeLisl - (09:29): Good morning all.  Just found out that Guice is (somehow) generics aware.
vmassol - (09:29): hi CalebJamesDeLisl
vmassol - (09:30): yes it is using a wrapper
vmassol - (09:30): LiteralType if I recall corredctly
CalebJamesDeLisl - (09:31): Interesting.  Sorry to bring up things that you probably already went over.
CalebJamesDeLisl - (09:31): (Guice v. 299)
vmassol - (09:31): not a pb, I prefer other people to look at it too
vmassol - (09:31): the question for me right now
vmassol - (09:31): is not guice vs 299
vmassol - (09:32): but more guice vs our existing model and so far guice looses but there's one possibility that it could catch up
silviar joined #xwiki at 09:32
vmassol - (09:32): that possibility is peaberry since that would allow us not to have to do the osgi integration
vmassol - (09:32): but I'm not sure it's good enough
vmassol - (09:32): still researching that part
vmassol - (09:32): if you want some good presenation on this (but missing the speaker):
vmassol - (09:33):
vmassol - (09:33): and more specifically:
CalebJamesDeLisl - (09:33): Guice is far different from the current model? (I know less than I should)
vmassol - (09:33):
vmassol - (09:33): the only major difference is that guice has no annotations for the binding part
Enygma` joined #xwiki at 09:33
CalebJamesDeLisl - (09:34): I see, the notation.
evalica joined #xwiki at 09:34
vmassol - (09:34): yes guice has java apis only for the binding part
vmassol - (09:35): this one too:
CalebJamesDeLisl - (09:35): Where ever we go, I am excited about the security implications, it makes principle of least authority very easy.
vmassol - (09:35): I'm trying to understand it right now
vmassol - (09:35): (the dynamic guice app slides)
lucaa joined #xwiki at 09:36
vmassol - (09:36): I've started the osgi poc this week end too
vmassol - (09:36): (ie osgi beneath our programming model)
vmassol - (09:36): CalebJamesDeLisl: give me 3 mn, I need to disconnect and reconnect in my office...
CalebJamesDeLisl - (09:37): ok
vmassol left at 09:37 (Quit: Leaving.
vmassol joined #xwiki at 09:38
CalebJamesDeLisl - (09:39): I have been watching videos (lazy) so I am learning more about Guice since it's well discussed on youtube.
CalebJamesDeLisl - (09:40): Another feature which looks intriguing is the scope, request scope could replace the context -> less trusted code.
arkub joined #xwiki at 09:40
vmassol - (09:42): you're talking about jsr299 now?
vmassol - (09:42): because yes the scopes there are interesting for us too
CalebJamesDeLisl - (09:42): They both support it AFAIK.
vmassol - (09:42): personally I don't see the future for guice right now
CalebJamesDeLisl - (09:43): I would agree except that Google will not want to port their code.
vmassol - (09:43): it's not as if they'll have the choice
vmassol - (09:43): and it's not really google, it's more Bob
CalebJamesDeLisl - (09:44): I understand that most of the google projects are using it right now.
vmassol - (09:44): so it all depends how many fwks/libraries/servlet engines adopt 299
CalebJamesDeLisl - (09:45): If history is an indicator, the worst option will be standardized ;)
vmassol - (09:45): sure, that would be guice 3
vmassol - (09:45): ;)
vmassol - (09:45): (not answering your last point in my reply)
CalebJamesDeLisl - (09:46): Are there other compelling options other than Guice and 299?
vmassol - (09:46): osgi
vmassol - (09:47): more specifically
vmassol - (09:47): osgi DS, osgi dependency manager or iPojo
vmassol - (09:47): see
vmassol - (09:47): but the only compelling solution to me is 299 right now
vmassol - (09:47): I don't see why we would move from a non standard api (ours) to another non standard api
CalebJamesDeLisl - (09:48): We have a few options in how we decide which road to take. We can either look for a potential standard, or look for what provides for our needs best.
vmassol - (09:48): (the only reason I could  see would be to gain osgi support built in, as with guice + peaberry, if it proves well done)
kibahop left #xwiki at 09:48
vmassol - (09:49): CalebJamesDeLisl: you don't agree with what I proposed?
CalebJamesDeLisl - (09:49): Which part?
vmassol - (09:49): the roadmap
vmassol - (09:49): step1, 2, 3, etc
vmassol - (09:49): in any case
CalebJamesDeLisl - (09:50): I certainly wouldn't -1 it but I think we should do cost/benifits on each option.
vmassol - (09:50): for the short term, it all depends how successful or not is the poc I'm doing
vmassol - (09:50): CalebJamesDeLisl: that"'s exactly what I did
CalebJamesDeLisl - (09:51): I suspected as much, and I'm sorry to bring it up again, but I didn't really see that presented.
vmassol - (09:51): (but I didn't show it all for sure, was lazy to write that mail ;) Actually I wanted to blog about it but didn't take the time yet)
vmassol - (09:51): so that's why I said, ask questions in the meantile
vmassol - (09:51): meantimez
vmassol - (09:51): meantime
vmassol - (09:51): in any case you're 100% free to suggest other ideas
vmassol - (09:52): and we can discuss them
vmassol - (09:52): but talking without any evidence will not bring us too far
CalebJamesDeLisl - (09:52): Well, if we are to look purely from an implementation standpoint, Guice is integrated with GWT, and 299 will likely have some synergy with Hibernate and jbosscache.
vmassol - (09:53): I don't see why any of these would matter
vmassol - (09:53): they're really small stuff compared to the larger picture IMO
vmassol - (09:54): as an example our current comp model doesn't prevent us from using any of these
CalebJamesDeLisl - (09:54): The larger picture being that one of these standards is going to be the betacord?
vmassol - (09:54): first as I mentioend in the mail we need to focus on our needs
KermitTheFragger joined #xwiki at 09:54
vmassol - (09:54): the need is the extension manager
vmassol - (09:55): that's our #1 need
vmassol - (09:55): the comp model is really far far behind
vmassol - (09:55): since we already have a working one
CalebJamesDeLisl - (09:55): I think either of the options will be satisfactory for that need, am I wrong?
vmassol - (09:56): jsr299 not really
CalebJamesDeLisl - (09:56): Ahh yes because of the hot loaded components.
vmassol - (09:56): weld a bit more but not quite
vmassol - (09:57): all that remains for me to understand is what hack they use in guice for peaberry
vmassol - (09:57): since guice doesn't support dynamic comp either
vmassol - (09:57): and in the meantime I'm evaluating how much it costs to simply plugin osgi in our existing model
vmassol - (09:58): (learning osgi at the same time since that's needed anyway)
CalebJamesDeLisl - (09:58): Dynamic loading not supported in guice?
oanat joined #xwiki at 09:59
vmassol - (09:59): how do you do it?
CalebJamesDeLisl - (09:59): I thought I heard it was, but now I can't remember where.
vmassol - (09:59): it's definitely not built for doing that
vmassol - (10:00): now it's certainly since peaberry has found some way for doign it
vmassol - (10:00): s/since/possible since/
vmassol - (10:01): but it's not meant to be simple or even planned from the onset
vmassol - (10:01): otherwise it would be documented, in the base feature set, etc
vmassol - (10:01): (you wouldn't need another fwk to do it)
CalebJamesDeLisl - (10:01): I guess you would have to get the root injector and call inject because there would be no dependencies on the new class.
vmassol - (10:02): you can read:
CalebJamesDeLisl - (10:02): Lots of tabs in the browser...
CalebJamesDeLisl - (10:03): Here's a funny one:  little ajax code which pulls a twitter page and analyzes the intelligence of the followers.
CalebJamesDeLisl - (10:03): stupid += rCount(msg.toLowerCase(),/(lol|omg|wtf|ftw|\&lt\;3|luv|xo|yr|wit\W|\Wur\W|btw|rofl|lmao|stfu)/g) * 5;
npm left at 10:05 (Quit: Leaving.
CalebJamesDeLisl - (10:08): I see the problem, the bindings are static so you can't hot swap implementations.
CalebJamesDeLisl - (10:09): The obvious hack would be using providers which give us whatever implementation they want.
vmassol - (10:09): btw peaberry has its own guice jar
vmassol - (10:09): with several patches applied to it
vmassol - (10:10): for ex:
CalebJamesDeLisl - (10:10): As far as what standard will 'win' maybe the answer is none, maybe they will all me adjusted and melded together.
vmassol - (10:11): standards always win
vmassol - (10:12): it lowers the cost for everyone
vmassol - (10:12): at all levels
CalebJamesDeLisl - (10:13): Well one standard's name always ends up being used, but the benefits of guice may influence changes to 299 or vice versa.
vmassol - (10:13): definitely
vmassol - (10:14): it alreayd has
vmassol - (10:14): since 299 was done based on guice
vmassol - (10:14): :)
vmassol - (10:14): which is why I hate it
vmassol - (10:14): since it didn't take into account dynamic bean loading!
vmassol - (10:14): registration I mean
vmassol - (10:14): :)
vmassol - (10:14): since they looked at guice but not at Osgi for ex
vmassol - (10:14): so 299 is really a standardization of guice + seam
vmassol - (10:15): I hope a revised version of 299 will appear in the future to take into accoutn dynamic registration use cases
vmassol - (10:15): in any case I have no doubt that Weld and others will support it
vmassol - (10:15): whether it's in the spec or not
CalebJamesDeLisl - (10:16): What exactly do we need to load dynamically? new interfaces or are we hot switching implementation?
vmassol - (10:16): here's a thread:
vmassol - (10:16):
vmassol - (10:16): you'll find some use cases in there
vmassol - (10:16): that should be familiar to you
vmassol - (10:16): :)
CalebJamesDeLisl - (10:17): Yup, that was linked in the email.
SvenDowideit_ joined #xwiki at 10:18
CalebJamesDeLisl - (10:18): Guice has @Singleton and @LazySingleton, this makes me thing the world does not need to be built at startup.
SvenDowideit left at 10:18 (Ping timeout: 265 seconds
SvenDowideit_ is now known as SvenDowideit ([email protected]
vmassol - (10:18): yes that's Instance<?> in 299
vmassol - (10:18): (the lazy part)
CalebJamesDeLisl - (10:19): So (in guice) why not just get the static Guice.inject and use it to start the class in the middle.
CalebJamesDeLisl - (10:19): ?
CalebJamesDeLisl - (10:20): That is as long as we don't need to swap implementations of interface xyz.
vmassol - (10:20): the solution they use in peaberry AFAIU is to use a mediator
vmassol - (10:20): but I don't fully understand it yet
vmassol - (10:20): so I won't be able to answer your questions
CalebJamesDeLisl - (10:21): I think part of the problem is that the advocates of different standards like to throw rocks so little problems turn into major blockers.
SvenDowideit_ joined #xwiki at 10:23
CalebJamesDeLisl - (10:24): I see two possible use cases which have to be separated. One is to load a package which gets the benefits of injection but no internal code depends on it.
CalebJamesDeLisl - (10:25): The other would be to hot swap an internal class for an added class. If the class is singleton than this is impossible.
SvenDowideit left at 10:25 (Ping timeout: 276 seconds
SvenDowideit_ is now known as SvenDowideit ([email protected]
CalebJamesDeLisl - (10:26): Example of #2 would be loading an implementation of the XWikiCaptchaService which replaces the internal one.
CalebJamesDeLisl - (10:27): Of course captcha service is singleton so the puzzle answers which are in the cache go missing in the swap.
arkub left at 10:33 (Ping timeout: 258 seconds
gvallarelli joined #xwiki at 10:34
gvallarelli - (10:34): hello
florinciu joined #xwiki at 10:35
jvelo joined #xwiki at 10:37
arkub joined #xwiki at 10:48
Enygma` left at 10:57 (Quit: Leaving.
sdumitriu joined #xwiki at 11:01
SvenDowideit_ joined #xwiki at 11:11
SvenDowideit left at 11:13 (Ping timeout: 245 seconds
SvenDowideit joined #xwiki at 11:14
SvenDowideit_ left at 11:17 (Ping timeout: 265 seconds
SvenDowideit left at 11:18 (Ping timeout: 246 seconds
SvenDowideit_ joined #xwiki at 11:18
SvenDowideit_ is now known as SvenDowideit ([email protected]
SvenDowideit left at 11:25 (Ping timeout: 246 seconds
SvenDowideit joined #xwiki at 11:25
SvenDowideit left at 11:34 (Ping timeout: 246 seconds
SvenDowideit joined #xwiki at 11:35
CalebJamesDeLisl - (11:35): Don't know if you saw this:
CalebJamesDeLisl - (11:36): Looks like what I imagined, you have to call the root loader to load each module.  Unloading might be trickier.
mflorea left at 11:38 (Quit: Leaving.
jvdrean joined #xwiki at 11:41
jvelo - (12:02): vmassol, re "I'm not sure the import UI was heavily designed/reviewed / I have various nitpicks with it ;)"
jvelo - (12:02): please let me know about it
vmassol - (12:02): jvelo: hehe took you some time to react ;)
jvelo - (12:02): s/it/them/
vmassol - (12:02): that message was for you indeed :)
vmassol - (12:02): we talked about them already
jvelo - (12:02): was probably busy or not in the room
vmassol - (12:02): but yes I could make a list
vmassol - (12:02): and send it to the dev lists
jvelo - (12:02): if we did then I don
jvelo - (12:03): 't recall
vmassol - (12:03): we talked about ti when I upgraded
jvelo - (12:03): There is one issue I have in mind that JV & I discussed
vmassol - (12:03): with jv and you
jvelo - (12:03): it's the fact you can't expand all spaces
vmassol - (12:03): yes that's one
jvelo - (12:03): (to be able to search for names)
vmassol - (12:03): ok I'll try to find some time to list them
jvelo - (12:03): that's relatively easy to add, we need to find the proper place for the links in the UI
jvelo - (12:03): ok
vmassol - (12:03): (no guaratnee though, I reall yneed to start preparing my training session....)
glerouge joined #xwiki at 12:04
Ciprian joined #xwiki at 12:04
jvdrean left at 12:13 (Quit: Leaving.
florinciu left at 12:34 (Read error: Connection reset by peer
florinciu joined #xwiki at 12:34
mflorea joined #xwiki at 12:40
Enygma` joined #xwiki at 12:42
jvdrean joined #xwiki at 12:48
jvelo left at 12:52 (Ping timeout: 276 seconds
jvelo joined #xwiki at 12:52
florinciu left at 13:22 (Quit: Leaving.
florinciu joined #xwiki at 13:39
jvelo left at 14:10 (Ping timeout: 265 seconds
jvelo joined #xwiki at 14:25
florinciu left at 14:31 (Quit: Leaving.
jvelo left at 14:48 (Ping timeout: 246 seconds
florinciu joined #xwiki at 15:12
kibahop joined #xwiki at 15:16
jvelo joined #xwiki at 15:34
kibahop left #xwiki at 15:35
npm joined #xwiki at 15:49
nickless joined #xwiki at 15:55
temporalfox - (15:55): hi everyone
kibahop joined #xwiki at 15:56
lucaa - (15:56): hi temporalfox
temporalfox - (15:56): about me : I'm Julien Viet working for the eXo Platform company
temporalfox - (15:56): hacking a bit xwiki model and rendering engine to write a converter for docbook
temporalfox - (15:57): temporalfox is an acronym of eXo Platform
vmassol - (15:57): (for those who haven't noticed temporalfox is an anagram for exoplatform ;) julien: it's a cool name)
vmassol - (15:57): my name is: iwikx
vmassol - (15:57): hmm not as nice
vmassol - (15:57): ;)
lucaa - (15:58): but you vmassol, knew it, before right? your permutations capability is not that instant, is it?
vmassol - (15:58): hey I'm a scrabble champion!
Enygma` left at 15:58 (Quit: Leaving.
temporalfox - (15:58): I used this website to find the anagram :
vmassol - (15:58): was ranked 400th worldwide in the past....
lucaa - (15:58): ok, ok. Please pardon my underestimation :)
sdumitriu - (15:58): vmassol: You can say "Hi, I'm Xecs"
vmassol - (15:58): hehe
lucaa - (15:59): rofl
vmassol - (15:59): or Weenier Skirt Pix
vmassol - (15:59): ;)
vmassol - (15:59): (for xwiki enterprise :))
vmassol - (16:00):
vmassol - (16:00): cool site temporalfox
vmassol - (16:00): Rewrite Sex I Pink
temporalfox - (16:01): "Enter Pre Kiwi Six"
temporalfox - (16:01): sounds like take the red pill
temporalfox - (16:01): anyway :-)
temporalfox - (16:01): I would like to hear about the 2.1 model version and know if there is already documentation about the changes from 2.0
vmassol - (16:02): you're talking about the xwiki syntax right?
vmassol - (16:02): if so, then xwiki syntax 2.1 has not started yet
vmassol - (16:02): (it doesn't exist yet)
temporalfox - (16:03): yes the syntax and the new capabilities
vmassol - (16:03): we do have a list of issues raised in jira that are good condidates for the 2.1 version
temporalfox - (16:03): what kind of stuff do you have in mind for 2.1 syntax ?
vmassol is now known as Mac ([email protected]
Mac left at 16:04 (Remote host closed the connection
vmassol joined #xwiki at 16:04
vmassol is now known as ManicsSolvent ([email protected]
ManicsSolvent - (16:05): actually we should tag them in jira
ManicsSolvent - (16:05): temporalfox: what are you looking for exactly?
temporalfox - (16:05): nothing in particular
ManicsSolvent is now known as CentavosLimns ([email protected]
temporalfox - (16:06): just curiosity, the current 2.0 syntax is sufficient enough at the moment
CentavosLimns - (16:06): ok
CentavosLimns - (16:06): we need to start the effort for 2.1
temporalfox - (16:06): I haven't hit any real showstopper
CentavosLimns - (16:06): we just haven't started yet
CentavosLimns - (16:07): temporalfox: it's a good idea though. We should start to flag issues with syntax2.1 in jira
temporalfox - (16:08): actually there is one thing that would make sense to me, it's the concept of "title"
temporalfox - (16:08): there is no possiblity today to define a title for different kind of blocks
CentavosLimns - (16:08): here's one for ex:
CentavosLimns - (16:09): header?
temporalfox - (16:09): in docbook many elements have a <caption> tag, such as table, programlisting
temporalfox - (16:09): yes a bit
temporalfox - (16:09): but for other kind of blocks than section
temporalfox - (16:09): it's different
kibahop left at 16:09 (Ping timeout: 240 seconds
temporalfox - (16:09): you create a table and you give it a title
temporalfox - (16:10): or for a list
temporalfox - (16:11): for instance :
temporalfox - (16:11): at the very bottom of the page
temporalfox - (16:11): the table has a title
temporalfox - (16:11): today there is no first class concept of that in wiki syntax
temporalfox - (16:12): it makes sense too for images
CentavosLimns - (16:17): sorry otp
golya joined #xwiki at 16:17
jvelo - (16:20): CentavosLimns, what's with your username ?
CentavosLimns - (16:21): jvelo: you don't like it? :)
golya - (16:23): Hi! Can XWiki control permissions for specific pages?
glerouge - (16:23): yes
glerouge - (16:23): edit -> page rights
glerouge - (16:24): you need to be an admin and an advanced user to see that though
golya - (16:25): ah, thanks, found
plunden left #xwiki at 16:26
jvelo - (16:26): I prefered ManicsSolvent :)
jvelo - (16:26): ( CentavosLimns)
jvelo - (16:27): lol ok I've read history
golya - (16:32): And is there support for namespace hierarchy? So can I have Development/projects/Project1/Documentation/specification?
glerouge - (16:33): not yet
glerouge - (16:33): right now you can only have Space/Page
glerouge - (16:34): it's planned but for the future
golya - (16:35): thanks for the info
glerouge - (16:38): you can use tags in the meanwhile
CentavosLimns - (16:40): temporalfox: back
CentavosLimns - (16:40): not sure we want to add too much complexity
CentavosLimns - (16:41): so basically it's a caption
CentavosLimns - (16:41): need to think a bit about it
CentavosLimns - (16:41): could be done with custom parameters too
CentavosLimns - (16:42): (% caption="my caption" %)
CentavosLimns - (16:42): |= table header1 |= table header2
CentavosLimns - (16:42): ...
CentavosLimns - (16:43): since it could be considered metadata for the table
CentavosLimns - (16:47): nexus 1.6 is out:
CentavosLimns is now known as MacInsolvents ([email protected]
jvelo left at 16:50 (Quit: Ex-Chat
MacInsolvents is now known as InsolventScam ([email protected]
jvelo joined #xwiki at 16:50
InsolventScam - (16:51): nice tweet: :)
evalica left at 17:01 (Read error: Connection reset by peer
gvallarelli - (17:02): lol
florinciu left at 17:12 (Read error: Connection reset by peer
florinciu joined #xwiki at 17:14
oanat left at 17:34 (Quit: Leaving.
golya left at 17:35 (Quit: Page closed
tmortagne joined #xwiki at 17:47
InsolventScam - (18:18): hey guys, in case you don't know about it, you can have your avatar displayed when you post on the list and when it's viewed in nabble:
InsolventScam - (18:18): a few of us already have avatars
Ciprian left #xwiki at 18:20
nickless left at 18:26 (Ping timeout: 248 seconds
temporalfox - (18:28): vmassol: yes that makes sense
temporalfox - (18:29): vmassol: I'll find out if it makes sense for images too
InsolventScam - (18:29): temporalfox: I'm not 100% sure, would need to be discussed with others
bougie - (18:29): tmortagne: hello
InsolventScam - (18:29): in any case this would mean that renderers would need to be aware of that
tmortagne - (18:29): bougie: hi
InsolventScam - (18:29): tmortagne: you're on holiday today?
tmortagne - (18:30): yep
jvelo - (18:30): lol. you are a scam, Vincent
jvelo - (18:30): ;)
bougie - (18:30): sdumitriu tells me your an ldap expert :)
InsolventScam - (18:30): jvelo: yep as my name indicates ;)
tmortagne - (18:30): lets say a LDAP expert at XWiki level ;)
InsolventScam - (18:31): tmortagne: for when you're back then, we should discuss temporalfox's fox idea of a caption for syntax 2.1. I proposed something but I'd like to discuss it with you too.
bougie - (18:31):
bougie - (18:31): this link is up for you ?
tmortagne - (18:31): InsolventScam: i would need to read history (i dont have it)
tmortagne - (18:32): bougie: does not seems so
tmortagne - (18:32): (i don't have error but i will soon get a timeout i guess)
tmortagne - (18:33): bougie: actually i'm not even sure the link is converted
InsolventScam - (18:34): tmortagne: sure, no hurry, it's all recorded normally and it's not something important, just something we should discuss at some point
tmortagne - (18:35): sure
InsolventScam - (18:35): (btw InsolventScam is me, VincentMassol ;) just in case you're wondering)
tmortagne - (18:35): yes i was wondering at first but i translated ;)
nuvolari - (18:36): wawut? she's not here? :/
nuvolari - (18:36): lo everyone :)
InsolventScam - (18:36): tmortagne: and temporalfox is exoplatform....
tmortagne - (18:36): ok
InsolventScam - (18:36): (julien from exo actually)
InsolventScam - (18:37): (julien viet)
tmortagne - (18:37): ok :)
gvallarelli - (18:37): I'm editing profile but Xwiki doesn't save update information
InsolventScam - (18:37): yeah we're ahving fun today
tmortagne - (18:41): bougie: could you paste the the real url ? looks like does not work for me
tmortagne - (18:43): i mean use another service like pastebin or tinypaste
tmortagne - (18:44): ok actually ii lost internet...
jvelo - (18:44): it's still there don't worry
jvelo - (18:44): (the Internet)
tmortagne - (18:44): or maybe not
jvelo - (18:44): lol
tmortagne - (18:44): i was really stating to wondering, no activity on any chat plus impossible to access some public service
tmortagne - (18:44): :)
tmortagne - (18:48): InsolventScam: would be cool to order the other way (have more recent first)
InsolventScam - (18:48): yep
InsolventScam - (18:48): need to find that livetable option for that
InsolventScam - (18:48): do you know it?
InsolventScam - (18:48): let me look up the doc
tmortagne - (18:48): no idea but jvelo probably know
glerouge - (18:52): there's a parameter
glerouge - (18:52): something like sortColumn
glerouge - (18:52): and then ascending or descending for that column
glerouge - (18:53): it's in the doc
jvdrean left #xwiki at 18:54
jvdrean joined #xwiki at 18:54
InsolventScam - (19:03): got to go, if someone can do it great, otherwise I'll check later if I don't forget
bougie - (19:05): tmortagne:
glerouge - (19:05): done
tmortagne - (19:07): bougie: ok looking, in the meantime what is your problem exactly ?
tmortagne - (19:08): did not seen you already written it on the paste :)
jvdrean left at 19:09 (Quit: Leaving.
KermitTheFragger left at 19:10 (Quit: gone for food!
tmortagne - (19:11): bougie: what you did does not work ?
silviar left at 19:11 (Quit: Leaving.
tmortagne - (19:12): i don't understand what is the problem with "the memberUid attribut is in the OU roles and not in OU people"
tmortagne - (19:12): memberUid is the field name that the LDAP authenticator is supposed to look at in the group object
tmortagne - (19:12): it's used only for that, it has nothing to do with users objects
florinciu left at 19:14 (Read error: Connection reset by peer
tmortagne - (19:15): bougie: don't forget also the xwiki.authentication.ldap.group_classes option in case the LDAP group class used is not listed in default LDAP classes (the one in the configuration comment)
tmortagne - (19:15): it's probably your problem
gvallarelli - (19:15): got to go, goodbye
bougie - (19:15): so, i have to add  xwiki.authentication.ldap.group_classes=ou=roles,dc=appartland,dc=eu in my xwiki.cfg ?
arkub left at 19:16 (Ping timeout: 258 seconds
tmortagne - (19:16): bougie: no
tmortagne - (19:16): the LDAP class
tmortagne - (19:16): you have to use a LDAP client
tmortagne - (19:16): and looks at the group object
tmortagne - (19:16): and then find the proper class name
tmortagne - (19:17): it should be something like the ones you have as eexample in the xwiki.cfg
gvallarelli left #xwiki at 19:17
tmortagne - (19:20): bougie: looks at the field "objectClass"
tmortagne - (19:20): should should have several
tmortagne - (19:21): if you list them to me i should be able to tel you which one looks like a LDAP group class
tmortagne - (19:22): (problem is that there is no user/group standard in LDAP so i listed some well known classes but the list is not complete which is why there is this option)
bougie - (19:27): i do not have the default value for xwiki.authentication.ldap.group_classes in my xwiki.cfg :/
tmortagne - (19:27): standard xwiki.cfg look like this:
tmortagne - (19:28):  #-# [Since 1.5M1, XWikiLDAPAuthServiceImpl] #-# The potential LDAP groups classes. Separated by commas. # xwiki.authentication.ldap.group_classes=group,groupOfNames,groupOfUniqueNames,dynamicGroup,dynamicGroupAux,groupWiseDistributionList
tmortagne - (19:28):
tmortagne - (19:28): #-# [Since 1.5M1, XWikiLDAPAuthServiceImpl]
tmortagne - (19:28): #-# The potential LDAP groups classes. Separated by commas.
tmortagne - (19:28): # xwiki.authentication.ldap.group_classes=group,groupOfNames,groupOfUniqueNames,dynamicGroup,dynamicGroupAux,groupWiseDistributionList
mflorea left at 19:29 (Quit: Leaving.
tmortagne - (19:29): so if you have some LDAP UI use it to access your group LDAP object and you should find all you need
tmortagne - (19:29): is good but not very small
bougie - (19:30): I already use it :)
tmortagne - (19:30): ok
tmortagne - (19:33): bougie: i will probably have to go very soon, if you success would be great if you good document the LDAP server you are using in
tmortagne - (19:34): also if the issue is the group class you could create an issue on so that I add your object class in the list of default group classes
InsolventScam is now known as vmassol ([email protected]
tmortagne1 joined #xwiki at 19:52
tmortagne left at 19:52 (Read error: Connection reset by peer
bougie - (19:53): is xwiki has a ldap cache ?
bougie - (19:53): i think yes
tmortagne1 - (19:53): bougie: yes
tmortagne1 - (19:53): for the groups
tmortagne1 - (19:53): only for the groups
tmortagne1 - (19:53): otherwise it could be very costly
bougie - (19:53): how do i turn off ?
bougie - (19:54): (during tests)
tmortagne1 - (19:54): you can configure it xwiki.authentication.ldap.groupcache_expiration
tmortagne1 - (19:54): you can't turn it off i think but yyou can put something like 1 second
tmortagne1 - (19:55): i have to go now, bon courage ;)
bougie - (19:55): thx
tmortagne1 left at 19:55 (Client Quit
bougie - (19:57): yeah, it woks
bougie - (19:57): \o/
bougie - (19:57): works*
vmassol left at 20:01 (Ping timeout: 252 seconds
glerouge left at 20:02 (Quit: Leaving.
jvelo left at 20:43 (Quit: Ex-Chat
lucaa left #xwiki at 20:45
vmassol joined #xwiki at 20:59
vmassol - (21:44): sdumitriu: has just rebooted
sdumitriu - (21:45): I never said that _all_ the restarts were caused by /tmp, just "some"
sdumitriu - (21:46): The rest are a bit more often, and they are OOM, Heap space
vmassol - (21:46): I know, just didn't know if you're receiving these mails, so wanted to let you know
sdumitriu - (21:46): Yes, I do too
sdumitriu - (21:46): Thanks
sdumitriu - (21:47): Yep, a bit before the end:
sdumitriu - (21:47): Caused by: java.lang.OutOfMemoryError: Java heap space
vmassol - (21:47): right
vmassol - (21:47): good catch
vmassol - (21:48): so either we're leaking or we don't have enough memory set
vmassol - (21:48): ah it's mysql
vmassol - (21:48): so we can assume it's not leaking
vmassol - (21:48): (I hope)
vmassol - (21:48): hmm memory is not disk space, is it?
mflorea joined #xwiki at 21:52
plunden joined #xwiki at 22:21
lucaa joined #xwiki at 22:35
jvelociter joined #xwiki at 22:42
vmassol - (23:35): could be quite useful for us:
vmassol - (23:35): trying to find the license
vmassol - (23:35): BSD cool
vmassol - (23:35):
mflorea left at 23:45 (Quit: Leaving.
sdumitriu left at 23:49 (Ping timeout: 264 seconds

Get Connected