[00:01:41] <ALR> aslak: I fixed it. :)
[00:01:55] <ALR> Just added "comment" flag to Node
[00:01:59] <ALR> So in exporting:
[00:02:12] <ALR> targetChild = owned.createComment(source.text());
[00:02:16] <ALR> Instead of createElement
[00:02:20] <ALR> Bingo fixo
[00:02:41] <aslak> ALR, node.isComment  ?
[00:03:00] <ALR> aslak: yep.
[00:03:09] <aslak> "" node name ?
[00:03:13] <ALR> #comment
[00:03:18] <aslak> ok
[00:03:22] <ALR> I kept whatever was provided by the backend.
[00:03:39] <aslak> is comment is set, or checks name ?
[00:03:41] <ALR> But it's never exported out through the backend again, saving the error.
[00:03:59] <ALR> If comment is set, it doesn't check any name
[00:03:59] <aslak> is/the
[00:04:17] <ALR> But it'll throw IllegalArgumentException if you try to set a Node with children as a comment.
[00:04:25] <aslak> i mnea, you isCOmment(true) during import based on NodeTyp e?
[00:04:38] <aslak> +set
[00:04:42] <aslak> damn fingers.. :)
[00:04:52] <ALR> Yes
[00:04:59] <ALR>                   if (child.getNodeType() == org.w3c.dom.Node.COMMENT_NODE)
[00:04:59] <ALR>                   {
[00:04:59] <ALR>                      newTarget.setComment(true);
[00:04:59] <ALR>                   }
[00:05:24] <ALR> newTarget is our Node view.
[00:05:38] <aslak> yea
[00:06:43] <aslak> breaking the api tho.. :)
[00:06:58] <ALR> Huh?
[00:07:02] <aslak> wether you like it or not.. :)
[00:07:07] <ALR> What API breakage?
[00:07:15] <aslak> not.comment(true) if(node.comment())
[00:07:25] <aslak> damn it
[00:07:26] <aslak> not/node
[00:07:35] <ALR> Huh?
[00:08:08] <ALR> So where did I break the API exactly?
[00:08:28] <aslak> Node node.comment(true) && Boolean node.comment() would be following the current API of Node
[00:08:45] <aslak> setComment
[00:08:49] <ALR> Oh, you mean inconsistent.
[00:08:53] <aslak> yea
[00:08:56] <ALR> Yeah, umm...I changed those.
[00:08:58] <aslak> sorry,
[00:09:16] <aslak> aa ok
[00:09:23] <ALR> Like there's now:
[00:09:44] <ALR> attribute() < Returns Node, method chaining
[00:10:02] <ALR> getAttribute() < Returns a String, *not* method chaining
[00:11:00] <aslak> not pushed?
[00:11:09] *** jose_freitas has quit IRC
[00:11:59] <aslak> 58 branch i assume
[00:13:45] <ALR> aslak: Not pushed.  Doing more test writing and API cleanup
[00:13:55] <ALR> Then I need to squash
[00:14:05] <ALR> Few minutes
[00:14:11] <ALR> I broke Node.childrem
[00:15:03] <aslak> you broken the kids ?
[00:16:52] <ALR> Ha, no.
[00:16:59] <ALR> I don't like Node.create()
[00:17:11] <ALR> It returns a Node, but not *this* Node, the new one.
[00:17:26] <ALR> Also there's a create(Query) ... weird.
[00:18:11] <aslak> why would it create this node.. you just created a node
[00:18:32] <aslak> Node.create("a").get("a") ?
[00:19:14] <aslak> my mind and fingers are not synced to day at all
[00:19:21] <aslak> why would it return this node.. you just created a node
[00:21:03] <ALR> aslak: Hehe, it's not what I'd expect
[00:21:07] <ALR> It's just the naming of it.
[00:21:10] <aslak> and Query is there so you can create based on a QueryExpression. Node.create("child/child2/attr=value&some_text") and it creates the while sub tree
[00:21:22] <ALR> Instead of something like addChild()
[00:21:42] <ALR> For instance attribute(String ) returns "this"
[00:21:42] <aslak> seriously? calling method create(), your not expecting to get returned what was created?
[00:21:58] <ALR> No, I see the difference.
[00:22:02] <ALR> "attribute()" is a noun.
[00:22:06] <ALR> Create is still a verb.
[00:22:12] <ALR> You're right.
[00:22:53] <aslak> with addChild i would expect, addChild(Node)
[00:23:34] <ALR> Yeah
[00:23:41] <ALR> addChild(Node)
[00:23:47] <ALR> That what I was looking for
[00:23:48] <ALR> It's not there
[00:23:58] <ALR> Also I made getChildren() immutable.
[00:24:49] <ALR> having rebase problems
[00:26:52] <aslak> ROFL
[00:27:46] <ALR> aslak: https://github.com/shrinkwrap/descriptors/commit/4e5045be05cf980f5d1a3f6ba8802b759335d789
[00:27:47] <jbossbot> git [descriptors] 4e5045b.. Andrew Lee Rubinger [SHRINKDESC-31] Fix the failing test case by adding comment structure to Node
[00:27:48] <jbossbot> jira [SHRINKDESC-31] Descriptors clobbers file markup and formatting [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/SHRINKDESC-31
[00:28:10] <aslak> My Arq SPI talk at JavaOne, some way managed to go in as a BOF..  send a mail to JavaOne Speakers Service to see if we could change it to a session. nope, not possible
[00:28:17] <aslak> guess we're having a BOF then
[00:28:21] <ALR> aslak: NICE
[00:28:23] <ALR> Hahaha
[00:28:34] <ALR> I'm inviting myself to be a bird.
[00:28:38] <ALR> Of your feather.
[00:28:45] <aslak> hehe your welcome
[00:28:46] <aslak> :)
[00:29:03] *** mhuniewicz has joined #jbosstesting
[00:29:07] <ALR> aslak: So yeah, that's my commit above.
[00:29:08] <ALR> Simple.
[00:30:01] <mhuniewicz> aslak, Eclipselink gives me this: Exception Description: An exception was thrown while searching for persistence archives with ClassLoader: org.jboss.shrinkwrap.api.classloader.ShrinkWrapClassLoader@4ebac9b9
[00:30:14] <mhuniewicz> It's the same kinda problem we had with Hibernate, I guess?
[00:31:35] <ALR> aslak: Maybe.  Full trace?
[00:31:51] <ALR> Pastebin or Gist this time, maybe? :P
[00:32:09] <ALR> I mean mhuniewicz
[00:32:31] <mhuniewicz> I figured. ;)
[00:32:57] <mhuniewicz> http://pastebin.com/SJ5SG4g9
[00:33:38] <jbossbot> git [descriptors] push SHRINKDESC-58 e79b4a8.. Aslak Knutsen SHRINKDESC-31 Add TestCase to show problem with import/export of XML data with Comments
[00:33:39] <jbossbot> jira [SHRINKDESC-31] Descriptors clobbers file markup and formatting [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/SHRINKDESC-31
[00:33:39] <jbossbot> git [descriptors] push SHRINKDESC-58 4e5045b.. Andrew Lee Rubinger [SHRINKDESC-31] Fix the failing test case by adding comment structure to Node
[00:33:39] <jbossbot> git [descriptors] push SHRINKDESC-58 URL: http://github.com/shrinkwrap/descriptors/compare/22dfbf9...4e5045b
[00:33:39] <jbossbot> git [descriptors] push SHRINKDESC-58 6b12649.. Andrew Lee Rubinger [SHRINKDESC-31] Fix the failing test case by adding comment structure to Node
[00:33:40] <jbossbot> git [descriptors] push SHRINKDESC-58 URL: http://github.com/shrinkwrap/descriptors/compare/4e5045b...6b12649
[00:33:40] <ALR> Wow, NPE
[00:33:47] <mhuniewicz> Yeah.
[00:33:49] <ALR> I guess jbossbot is back
[00:33:55] <aslak> :)
[00:33:57] <mhuniewicz> And the annoying conversation thing is there as well.
[00:34:12] <ALR> mhuniewicz: Conversation thing?
[00:34:22] <mhuniewicz> ALR, it's 2 exceptions actually,
[00:34:33] <mhuniewicz> the second one comes from Arquillian and aslak knows more.
[00:34:34] <ALR> mhuniewicz: No, it's a nested one.
[00:34:42] <ALR> mhuniewicz: Line 85
[00:34:47] <ALR> NPE from ShrinkWrap?!  No!
[00:35:07] <ALR> Not in my project you don't.
[00:35:11] <ALR> That's getting fixed now.
[00:35:12] <mhuniewicz> No no.
[00:35:27] <mhuniewicz> Arquillian throws this up: java.lang.IllegalStateException: A request must be associated with the context in order to load the known conversations
[00:35:30] <ALR> mhuniewicz: No?  Tell me how line 85 isn't an NPE in ShrinkWrap ?
[00:35:54] <ALR> Oh I see, you posted 2 exceptions.  Why?
[00:36:07] <mhuniewicz> That's what the output is, my friend.
[00:36:10] <mhuniewicz> From a single test.
[00:36:18] <ALR> Probably one leads to the other.
[00:36:28] <ALR> Let's deal with this NPE first.
[00:36:31] <mhuniewicz> The first one causes the other but they are not directly related.
[00:36:51] <aslak> ALR, hmmmmm....
[00:37:07] <mhuniewicz> ALR, you're right, it seems to be coming from ShrinkWrap.
[00:37:17] <ALR> Not for long it doesn't.
[00:37:27] <aslak> didn't we fix this before.. i assume the request is coming in as / or similar, that's why it can't find the path
[00:37:34] <ALR> Now probably it's just a missing assertion on our part
[00:37:45] <aslak> i can swear i was woring with lincoln on this.. maybe his patch never was applied
[00:37:53] <ALR> It's likely I'll trade this NPE for an IllegalArgumentException because someone's giving us bad input
[00:37:58] <ALR> But let me see first
[00:40:52] <ALR> This is what we're dealing with:
[00:40:52] <ALR> https://github.com/shrinkwrap/shrinkwrap/blob/master/api/src/main/java/org/jboss/shrinkwrap/api/classloader/ShrinkWrapClassLoader.java#L134
[00:41:11] <ALR> Like a path that doesn't exist...
[00:41:42] <ALR> Basically when you pass a URL to an archive and it's not there...
[00:41:45] <mhuniewicz> Yep that's the one.
[00:44:10] <aslak> bed time.. night guys
[00:44:14] <ALR> aslak: Night
[00:44:23] <ALR> mhuniewicz: trying to figure how to duplicate this error
[00:44:25] <mhuniewicz> Good night
[00:44:33] <ALR> As we're the ones giving it a URL
[00:44:38] <mhuniewicz> Well I just pushed to github.
[00:45:05] *** aslak has quit IRC
[00:47:29] <ALR> mhuniewicz: Where is Eclipse Persistence source?
[00:47:49] <mhuniewicz> ALR, In my project you mean?
[00:48:03] *** mbg has quit IRC
[00:48:06] <ALR> No.
[00:48:18] <ALR> org.eclipse.persistence.internal.jpa.deployment.ArchiveFactoryImpl < Where does this class live?
[00:48:35] <mhuniewicz> eclipselink-2.3.0jar
[00:48:41] <ALR> The source for it.
[00:48:50] <ALR> GitHub?  SVN somewhere?  Closed source?
[00:49:48] <mhuniewicz> Oh sorry.
[00:49:50] <mhuniewicz> http://www.eclipse.org/downloads/download.php?file=/rt/eclipselink/releases/2.3.0/eclipselink-src-2.3.0.v20110604-r9504.zip
[00:50:06] <ALR> I was browsing http://dev.eclipse.org/viewcvs/viewvc.cgi/trunk/?root=RT_PERSISTENCE
[00:50:55] <ALR> Found it
[00:50:56] <ALR> http://dev.eclipse.org/svnroot/rt/org.eclipse.persistence/trunk/
[00:51:32] <ALR> Hmm, maybe not
[00:54:25] *** lightguard_jp has quit IRC
[00:54:31] <mhuniewicz> ALR, http://wiki.eclipse.org/EclipseLink/Source
[00:54:42] <ALR> Found it in your source JAR
[00:54:59] <ALR> mhuniewicz: https://gist.github.com/1098437
[00:55:40] <ALR> All they're trying to do is openStream
[00:55:44] <ALR> on a URL we gave them
[00:56:05] <ALR> mhuniewicz: Can you set a breakpoint in here?
[00:56:46] <ALR> Find out what the URL is they're trying to open, and also a listing of archive.toString(true) would be helpful
[00:56:55] <ALR> Somehow we're feeding bad data?
[00:57:06] <ALR> I'm not sure yet
[00:57:15] <mhuniewicz> I'm on it...
[00:57:30] <mhuniewicz> url = archive:258f64c7-bbd1-4db0-9279-9d3b05228ddc.jar/
[00:57:35] <mhuniewicz> (with the slash)
[00:57:50] <ALR> Seriously?  archive:// should be it
[00:57:59] <mhuniewicz> Exception!
[00:58:12] <ALR> And looks like they're just trying to ge tat the root
[00:58:13] <ALR> I se.e
[00:58:16] <ALR> *see
[00:58:43] <mhuniewicz> And then it wraps it int PersistenceUnitloadingException.
[00:58:44] <ALR> Nevermind my comment before on the protocol
[00:58:54] <ALR> We knew that already :)
[00:59:03] <ALR> They're trying to openStream on the root
[01:00:10] <ALR> OK, I see exactly what they're doing
[01:01:16] <ALR> Back in a bit.  I'll think on what to do here.
[01:01:22] <mhuniewicz> k
[01:01:24] <mhuniewicz> I'm off to bed.
[01:01:27] <mhuniewicz> Good night
[01:01:29] <ALR> Thanks.  Later
[01:01:35] <mhuniewicz> Thanks for your help
[01:01:40] *** mhuniewicz has quit IRC
[01:45:34] <OndrejZizka> emanuel somewhere?
[01:47:51] <ALR> OndrejZizka: It's late for him
[02:03:57] *** bleathem_mtg is now known as bleatehm
[02:04:02] *** bleatehm is now known as bleathem
[02:47:56] *** ianbrandt has quit IRC
[03:26:27] *** bgeorges has joined #jbosstesting
[03:38:15] *** ldimaggi has joined #jbosstesting
[05:25:20] *** ldimaggi has quit IRC
[05:30:26] *** aaronwalker has joined #jbosstesting
[05:35:02] *** aaronwalker has quit IRC
[05:39:05] *** alesj has joined #jbosstesting
[05:39:05] *** alesj1 has quit IRC
[05:42:35] *** dblevins has quit IRC
[05:44:34] *** bleathem has quit IRC
[05:47:33] *** dblevins has joined #jbosstesting
[05:49:21] *** bleathem has joined #jbosstesting
[07:12:15] *** lightguard_jp has joined #jbosstesting
[07:37:23] *** aslak has joined #jbosstesting
[07:39:41] *** kpiwko has joined #jbosstesting
[07:43:38] *** kpiwko has quit IRC
[07:44:06] *** kpiwko has joined #jbosstesting
[07:59:32] *** ge0ffrey has joined #jbosstesting
[08:11:23] *** tdiesler has joined #jbosstesting
[08:14:01] *** tommysdk has joined #jbosstesting
[08:17:58] *** mgoldmann has joined #jbosstesting
[08:18:49] *** jharting has joined #jbosstesting
[08:26:16] <aslak> kpiwko, good morning
[08:26:25] <kpiwko> aslak: good morning
[08:26:52] <aslak> kpiwko, did you find any password options in jgit ?
[08:27:55] <kpiwko> aslak: there is some prompt according to the javadoc...but I was out of the laptop for the rest of the day so I haven't look at it closely yet
[08:28:13] <aslak> aa ok
[08:30:40] *** bgeorges has quit IRC
[08:41:52] *** Jaikiran has joined #jbosstesting
[09:04:02] *** bgeorges has joined #jbosstesting
[09:08:57] *** bleathem has quit IRC
[09:18:52] *** alesj has quit IRC
[09:18:53] *** alesj1 has joined #jbosstesting
[09:26:17] <ALR> aslak: I think we should have a call w/ Ralf
[09:26:32] <aslak> ALR, good morning..
[09:26:39] <aslak> ALR, sure, sup?
[09:27:37] <ALR> aslak: Well, the generated notion of SD scares me a bit
[09:27:52] <ALR> Well, "scares" is the wrong word.
[09:27:57] <ALR> It brings up some issues.
[09:28:13] <ALR> Like first of all, I don't think generated code should be kept in a repo.
[09:28:21] <ALR> Makes doing diffs impossible
[09:28:29] <ALR> Second, I think the gen should be part of the build.
[09:28:49] <ALR> And third, there's presently some manual stuff to do after things are gen'd.
[09:29:08] <ALR> So in the sake of making changes going forward, I'd like to nail down:
[09:29:18] <aslak> ALR, why are diffs impossible ?
[09:29:43] <ALR> 1) Is the gen code meant to just be a starter; we throw that in the repo and then going forward edit that and toss the gen stuff we used to kickstart?
[09:29:59] <ALR> aslak: Because they don't show line-by-line changes.  The diff comes up as the whole file.
[09:30:10] *** maschmid has joined #jbosstesting
[09:30:11] <aslak> ALR, ?
[09:30:26] <aslak> ALR, same file name, git handled the diff
[09:30:29] <aslak> handles
[09:30:46] <ALR> https://github.com/shrinkwrap/descriptors/commit/f6ac2c1e9dfc6d6efd4891257f957e9fbeb38b1e
[09:30:48] <jbossbot> git [descriptors] f6ac2c1.. ralfbattenfeld [SHRINKDESK-54] Moved generated test classes to impl
[09:31:54] <aslak> but that's a move of the generated stuff to a new folder
[09:32:06] <ALR> Yet it doesn't show as a move.
[09:32:25] <ALR> I've some experience w/ generated code.
[09:32:32] <aslak> that has to do with how the move was done, not that it is generated
[09:32:36] <ALR> It just gets icky to manage, for lack of a better term.
[09:32:47] <ALR> That
[09:33:12] <aslak> As long as we generate source, and commit the source. git will still track the content
[09:33:12] <ALR> That's just the point.  Gen-ing stuff never really does a proper move.  It'll blow away existing and rewrite.
[09:33:25] <aslak> doesn't matter
[09:33:37] <ALR> I'm thinking back to my XDoclet days.
[09:33:42] <aslak> git tracks content
[09:33:47] <ALR> That's true.
[09:33:56] <ALR> Man, did I hate those XDoclet days.
[09:33:57] <ALR> :)
[09:34:02] <aslak> and your not deleting the file between commits nor are you renaming
[09:34:11] <ALR> Generating EJB local/remote/home interfaces
[09:34:37] <aslak> git checkout .... rm file & gen file git commit.. is the same as git checkout - eidt file manually  git commit
[09:34:41] <ALR> I'm not saying we stop gen-ing
[09:34:53] <ALR> Just that the current process is a bit weird.
[09:35:00] <ALR> The gen folder isn't part of the build.
[09:35:04] <ALR> It's manually moved.
[09:35:40] <ALR> So I'd like to nail down: looking forward, are we going to dump the gen stuff and work off the existing stuff as if we wrote it ourselves?  Or are we gonna rely on gen to create the stuff, always?
[09:35:47] <aslak> that can be automated for sure.. but it still won't mess with the change log as long as you gen and overwrite as part of the same commit
[09:36:08] <ALR> There's some other issues.
[09:36:17] <ALR> Like we need to manually organize imports after the gen.
[09:36:24] <ALR> I'm not sure why that is yet.
[09:36:26] <aslak> ?
[09:36:34] <ALR> It's in Ralf's instructions.
[09:36:54] <ALR> Anyway, just a chance to sync w/ Ralf.
[09:37:05] <ALR> I'd like to get a timeframe for when we can have this stuff in the API.
[09:37:13] <ALR> And start doing alphas with it in there.
[09:37:20] <ALR> I'm thinking this:
[09:37:23] *** kevinpollet has joined #jbosstesting
[09:37:33] <ALR> 1.1.0-alpha will have the SD SPI for you
[09:37:44] <ALR> When that goes to Beta, you can rely upon it in ARQ
[09:37:45] <aslak> the gen model does make custom changes 'impossible'.. that was one aspect i was trying to address in the model i was working on
[09:37:57] <aslak> yea
[09:38:00] <ALR> 1.2.0-series will have Ralf's new API stuff.
[09:38:19] <ALR> So Ralf's stuff does not block you.
[09:38:29] <ALR> But when 1.2.0 goes beta, then we can do ARQ-517
[09:38:30] <jbossbot> jira [ARQ-517] Remove hack/copied classes introduced to protocol-servlet as part of ARQ-514 [Open (Unresolved) Task, Major, Unassigned] https://issues.jboss.org/browse/ARQ-517
[09:38:31] <jbossbot> jira [ARQ-514] Remove all compile-time dependencies upon ShrinkWrap Descriptors Impl Module [Pull Request Sent (Unresolved) Task, Blocker, Andrew Rubinger] https://issues.jboss.org/browse/ARQ-514
[09:38:39] <aslak> ALR, that's taken care of already, the desc are moved to arq remember..
[09:38:44] <aslak> yea
[09:38:53] <ALR> No, 517 is the cleanup task
[09:38:58] <ALR> To remove what was copied.
[09:38:59] <aslak> :)
[09:39:20] <ALR> aslak: BTW, did you push my last change upstream?
[09:39:22] <aslak> yea, but i'm currently not being blocked by Ralf was what i meant
[09:39:27] <ALR> Oh, no, you're not.
[09:39:41] <ALR> But you will be if I were to introduce Ralf in 1.1.0-series
[09:39:47] <aslak> ALR, no, not yet..  working on some build issues wit hmanveresolver
[09:39:50] <aslak> mavenresolver
[09:40:09] <ALR> aslak: OK.  I think the issues we saw with the previous one I have corrected now.
[09:40:24] <ALR> aslak: Also, did SHRINKWRAP-306 today
[09:40:25] <jbossbot> jira [SHRINKWRAP-306] URL.openStream on directory URLs from the ShrinkWrapClassLoader throw an NPE [Resolved (Done) Bug, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKWRAP-306
[09:40:38] <aslak> ALR, yea, saw the comments
[09:40:54] <ALR> FYI.  Might be interesting just for you to see the symptom and what I did to fix.  In case you get complaints about it again from ARQ
[09:41:00] <ALR> Cool.
[09:41:19] <aslak> ALR, yea, read it this morning.. :)
[09:41:29] <ALR> Cool
[09:42:06] <aslak> but it does bring up a question if we should support .. tho
[09:42:08] <ALR> Then I guess you also saw the comments issue in SD is fixed.
[09:42:14] <aslak> yea
[09:42:15] <ALR> I'm not sure if you were still around to see that push
[09:42:22] <ALR> You asked if I had it somewhere you could see.
[09:42:47] <ALR> It's not on master yet.  But here: https://github.com/shrinkwrap/descriptors/commit/6b12649d652cca6feb113ed0a60732fd796a5f99
[09:42:47] <jbossbot> git [descriptors] 6b12649.. Andrew Lee Rubinger [SHRINKDESC-31] Fix the failing test case by adding comment structure to Node
[09:42:48] <jbossbot> jira [SHRINKDESC-31] Descriptors clobbers file markup and formatting [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/SHRINKDESC-31
[09:43:08] <aslak> no, i asked aobut the Node spi method name changes.. :) but i found them in the sd-58 branch
[09:43:16] <ALR> Oh.
[09:43:22] <ALR> Yeah, that's still a work in progress.
[09:43:52] <aslak> but 32 is built on 58 now right
[09:43:59] <aslak> 31 sorry
[09:44:14] <ALR> aslak: Just pushes another one: https://github.com/shrinkwrap/descriptors/commit/f9ba8e18a09daae302b0a04926f60c2bebd2c94e
[09:44:14] <jbossbot> git [descriptors] f9ba8e1.. Andrew Lee Rubinger [SHRINKDESC-58] Rename to Node.getChildren
[09:44:15] <jbossbot> jira [SHRINKDESC-58] Refine SPI for locking to Beta-level [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKDESC-58
[09:44:34] <ALR> aslak: Yeah, 31 kinda intersects with other stuff I did in 58
[09:44:45] <ALR> 58 will go upstream sooner than 54
[09:44:55] <ALR> 58 == 1.1.0-alpha-2
[09:45:02] <ALR> 54 == 1.2.0-series
[09:45:12] <aslak> make sense.. hehe
[09:45:18] <ALR> So now there's a nasty fork situation brewing.
[09:45:26] <ALR> Meaning I need to get 58 done ASAP
[09:45:28] <ALR> And into master again
[09:45:56] <ALR> I'm gonna do some more stuff on that now
[09:46:43] <aslak> Ralfs gen is still xslt based right ?
[09:46:46] <ALR> Yeah
[09:46:55] <ALR> aslak: Thoughts on create() > createChild() ?
[09:47:32] <ALR> (In Node)
[09:47:39] *** lightguard_jp has quit IRC
[09:47:48] <aslak> sure
[09:47:52] <ALR> Great
[09:51:04] <ALR> aslak: I'm ready to talk Queries now, if you have some time.
[09:51:15] <aslak> sure
[09:51:21] <ALR> Let's start: What is a Query?
[09:51:37] <ALR> Meaning: "What is the design intent of Query"?
[09:51:41] <aslak> Object / Content wise it's a Tree Expression
[09:52:43] <aslak> a way of expressing a sub structure
[09:52:50] <ALR> You mean an Expression Tree, the data structure?
[09:52:54] <aslak> or a tree structure for that matter.. think XPath
[09:53:44] <aslak> so you can say, give me Nodes that match "child/sub?someattr=withvalue"
[09:53:45] <ALR> Well, I don't see anything tree-like in it.
[09:54:31] <ALR> Let me try to unwrap this in my understanding.
[09:54:38] <ALR> NodeQuery looks like a pattern
[09:55:02] <ALR> ie. You want to try to match some pattern of element name/content/attributes
[09:55:12] <ALR> NodeQuery is the value object which describes that pattern, no?
[09:55:18] <aslak> correct
[09:55:34] <ALR> OK.  So can we start by maybe renaming this to...
[09:55:35] <aslak> or, NodeQuery describes one level, Query has all levels
[09:55:40] <ALR> SomethingPattern ?
[09:55:49] <ALR> NodePattern maybe?
[09:56:13] <ALR> Or Criteria?
[09:56:23] <ALR> It's really a search criterion.
[09:56:24] <aslak> Criteria sure
[09:56:38] <ALR> Criteria is plural in English.  Criterion is the singular.
[09:57:26] <ALR> I guess the criteria are text/name/attributes, together a plural.
[09:57:27] <ALR> So:
[09:57:28] <aslak> i don't think Query is wrong, but NodeQuery could be Criteria
[09:57:41] <ALR> NodeQuery > NodeCriteria
[09:57:42] <ALR> ?
[09:57:51] <aslak> sure
[09:57:53] <ALR> Or SearchCriteria?
[09:58:04] <ALR> QueryCriteria
[09:58:15] <aslak> what about just Criteria ?
[09:58:24] <ALR> Criteria for what?
[09:58:32] *** lfryc has joined #jbosstesting
[09:58:34] <ALR> I guess it's in spi.query
[09:58:52] <aslak> yea, in shrinkwrap descriptors, and spi operate on nodes etc
[09:58:55] <ALR> So by namespace it's a query criteria.  OK.  Criteria it is.
[09:59:10] *** vtunka has joined #jbosstesting
[10:00:21] <ALR> So: Criteria == A pattern that may be executed as part of a {@link Query} upon a {@link Node} in a search.
[10:00:58] <tommysdk> good morning all
[10:02:33] <ALR> tommysdk: Morning.
[10:02:41] <ALR> aslak and I are doing some SD SPI review.
[10:03:12] <tommysdk> Alright
[10:03:21] <aslak> ALR, search is one option. but yea
[10:03:34] <ALR> aslak: Feel free to refine that definition.
[10:03:38] <aslak> you do state 'may'
[10:04:36] <aslak> tommysdk, good morning
[10:05:01] <aslak> kpiwko, i found a bug!! :)
[10:06:03] <ALR> aslak: What else could you do with it?
[10:06:27] <kpiwko> aslak: great!
[10:06:31] <kpiwko> aslak: where?
[10:07:18] <aslak> kpiwko, how is MavenResolver.goOffline() suppose to work? it is used to decide the LocaRepositoryType in localRepositoryManager(). the problem is getSession that builds this is called from MavenBuilderImpl which is called by the DependencyResolver. so by the time i can say goOffline in API, the decision has already been taken not to go offline
[10:07:31] <aslak> ALR, create
[10:07:50] <kpiwko> aslak: hmm
[10:08:29] <aslak> ALR, you have different impls of the 'query handler', one is Get which functions like a search, return all matched. getSingle will return but force it to be a unique match. and getOrCreate that will go down the structure and create the nodes matching the Criteria if they don't exist
[10:08:31] <ALR> aslak: Can "text" in Criteria be null?  Looks so.
[10:08:34] <kpiwko> aslak: it is likely that I'm creating a new session if settings.xml is reloaded but I forgot to do so for goOffline method
[10:08:46] <ALR> aslak: Yeah, I see that.
[10:09:07] <aslak> ALR, text yes.. a nodes Body is not required
[10:09:15] <ALR> In the create case, this isn't really a Criteria
[10:09:17] <kpiwko> aslak: it does two things, apart from local repository type there some other setting I don't recall without checking source first
[10:09:18] <ALR> It's more a pattern.
[10:09:21] <aslak> ALR, the only thing required is the node name
[10:09:42] <ALR> Whether the Pattern is used as a Criteria or a Create is up to the use.
[10:09:56] <aslak> ALR, true
[10:10:03] <ALR> XPath expression...that's a Pattern
[10:10:24] <ALR> So..I'm inclined to move this out of spi.query and into spi.  And call it NodePattern.
[10:10:46] <aslak> kpiwko, yea, it's being set on the DefaultRepositorySystemSession as well
[10:11:23] <ALR> tommysdk: You all set up with everything you need from me?
[10:11:49] <aslak> ALR, the spi.query package is suppose to 'join' all of this stuff as a 'group' of what ever it is
[10:12:00] <tommysdk> ALR: for sd-18 yes
[10:12:07] <ALR> Cool.
[10:12:11] <ALR> tommysdk: Cool
[10:12:12] <tommysdk> ALR: did you make any conclusions for sd-31 ?
[10:12:17] <tommysdk> shrinkdesc-31
[10:12:22] <tommysdk> SHRINKDESC-31
[10:12:24] <jbossbot> jira [SHRINKDESC-31] Descriptors clobbers file markup and formatting [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/SHRINKDESC-31
[10:12:42] <ALR> tommysdk: https://github.com/shrinkwrap/descriptors/commit/6b12649d652cca6feb113ed0a60732fd796a5f99
[10:12:42] <jbossbot> git [descriptors] 6b12649.. Andrew Lee Rubinger [SHRINKDESC-31] Fix the failing test case by adding comment structure to Node
[10:13:01] <tommysdk> ALR: Very nice
[10:16:08] <aslak> .. more coffee...
[10:19:28] <ALR> aslak: hehe, I've got an idea when you're back.
[10:22:48] <tommysdk> production clearly goes down when you're out of coffee :)
[10:23:19] <ALR> That's why we keep him in a development-only environment.
[10:23:32] <ALR> Production environments can't go down.
[10:23:44] <ALR> We're not good enough for mission-critical systems.
[10:23:44] *** bgeorges has quit IRC
[10:26:43] <tommysdk> ALR: hahaha
[10:26:53] <tommysdk> ALR: Is aslak aware of it? ;)
[10:29:04] <aslak> hehe
[10:30:17] <ALR> aslak: So now we have "Pattern", which was previously NodeQuery
[10:30:26] <ALR> Notice anything really similar about Pattern and Node?
[10:33:24] *** aslak has quit IRC
[10:35:24] *** aslak has joined #jbosstesting
[10:36:35] *** Jaikiran has quit IRC
[10:36:57] *** Jaikiran has joined #jbosstesting
[10:44:18] <tommysdk> ALR: SHRINKDESC-18 - So if we specifiy a stream to be closed, but can't close it, what do we expect to happen? Throw an IAE?
[10:44:20] 
[10:44:58] <ALR> tommysdk: Like if you get an IOE? Log a warning
[10:45:08] <tommysdk> yeah
[10:45:14] <tommysdk> alright
[10:46:35] <tommysdk> ALR: We use the java.util.logging right?
[10:47:02] *** Jaikiran has quit IRC
[10:48:24] <ALR> yeah
[10:50:33] <ALR> aslak: What's an "absolute" query?
[10:51:09] <aslak> relative to current or not
[10:51:45] <aslak> somenode.get("/web-app") will give you the root node
[10:51:55] <aslak> if that's the root node
[10:52:18] <tommysdk> I think the docs can clarify this a bit
[10:52:52] <aslak> brb, getting some food
[10:53:33] <ALR> aslak: OK.  Here when you return.
[10:53:44] <ALR> aslak: I'm still uncovering some stuff here.
[11:02:57] *** Jaikiran has joined #jbosstesting
[11:13:45] *** Jaikiran has quit IRC
[11:14:10] *** Jaikiran has joined #jbosstesting
[11:21:38] *** Jaikiran has quit IRC
[11:22:25] *** Jaikiran has joined #jbosstesting
[11:27:05] *** jep has joined #jbosstesting
[11:35:23] *** DavideD has joined #jbosstesting
[11:36:56] <jep> When i start a simple ejb test on jboss 6 remote, then i became an java.lang.ClassNotFoundException: org.jnp.server.NamingServer. I use arquillian 1.0.0.CR1, what can i do to resolve this error?
[11:36:57] <jep> hello
[11:37:53] <jep> can someone help me with this problem?
[11:50:34] <tommysdk> jep: do you have jboss as client on your classpath?
[11:53:21] <aslak> ALR, back.. :)
[11:53:28] <aslak> ALR, what you got for me ?
[11:54:05] <jep> no, only jboss-as-profileservice-client
[11:55:31] <jep> should i use jboss as client instead of jboss-as-profileservice-client?
[11:55:46] <jep> or both?
[11:58:52] *** kevinpollet has quit IRC
[12:08:14] <tommysdk> jep: i just needed to add the arquillian-jbossas-remote-6 and jboss-as-client jars and was fine, so try that
[12:08:54] <jep> tommysdk: ok, i will try that (sorry for my bad english)
[12:09:14] <tommysdk> jep: np
[12:09:53] <ALR> aslak: Back too
[12:09:55] <ALR> aslak: OK
[12:10:02] <ALR> aslak: Umm...lemme see where I'm at
[12:10:30] <ALR> aslak: Basically Node is copying a bunch of stuff from what we're now calling "Pattern"
[12:10:41] <ALR> text, attributes, name
[12:10:54] <ALR> So I'm deciding whether Node should *be a* Pattern type
[12:10:59] <ALR> Or *have a* Pattern
[12:11:41] <ALR> Things get more complex when we get into the naming of things like "CreateQuery"
[12:11:51] <ALR> Which is not in and of itself a Query
[12:11:55] <ALR> It's a QueryExecutor
[12:12:03] <ALR> So decoupling that is in order I think
[12:12:33] <ALR> Defining what it is to be a Query (basically a list of Patterns and whether or not these patterns should be matched in an absolute fashion)
[12:12:48] <ALR> ...and differentiating that from the AbstractQueryExecutor stuff and its children
[12:14:27] <ALR> aslak: ^
[12:14:57] <aslak> The Query and QueryExecutor are separate ?
[12:15:37] <ALR> aslak: At the moment yes
[12:16:23] <ALR> Also QueryExecutor is parameterized with generics, and I don't understand why.
[12:16:44] <aslak> Node or List<Node>
[12:16:49] <ALR> Every concrete impl of it passes Node
[12:16:57] <ALR> Oh, List<Node>, OK
[12:17:05] <aslak> return type is generic
[12:17:08] <ALR> GetQuery.  I see it.
[12:17:14] <ALR> I hadn't noticed that.
[12:18:02] <aslak> i don't think a Node is a Pattern
[12:18:18] <ALR> It isn't.
[12:18:18] <aslak> Even tho they currently hold the same data
[12:18:22] <tommysdk> ALR: I've commited a proposal for SHRINKDESC-18 > https://github.com/tommysdk/descriptors/tree/SHRINKDESC-18
[12:18:22] <ALR> It just has the same properties
[12:18:23] 
[12:18:29] <ALR> So Node *has a* pattern
[12:18:40] <ALR> tommysdk: Will look in one sec
[12:18:52] <tommysdk> np, no hurry
[12:18:52] <aslak> "ALR, So I'm deciding whether Node should *be a* Pattern type"
[12:19:41] <ALR> aslak: Yup
[12:19:49] <ALR> *has a" is better
[12:20:15] <ALR> The only reason I considered making it a Pattern child is because extension is just easier than delegation
[12:20:38] <ALR> But that's just CS101 shit why that's a lazy shortcut :)
[12:20:54] <aslak> but that would bind the Node to a specific Pattern matching alg.. which is a bit strange
[12:21:20] <aslak> consider e.g. /mychild.*/
[12:23:43] <aslak> take RegExp for instance, a RegExp you can apply/match against a String, but the String is not the Pattern it slef
[12:24:04] <kpiwko> aslak: btw, have you any time to look at http://pastebin.com/Ew3j6UhL I send you yesterday?
[12:24:44] <aslak> kpiwko, looks very strange. it seems it's trying to load LoadableExtension, no package ?
[12:25:53] <kpiwko> aslak: exactly...but code with it compiles just fine
[12:25:53] <aslak> kpiwko, looks like you have a SPI file somewhere that refers to a half qualified name ?
[12:26:08] <ALR> aslak: Pattern has nothing to do with matching algs.
[12:26:13] <ALR> aslak: It's just a value object
[12:26:37] <aslak> ALR, well, true..
[12:26:57] <aslak> but i don't see what it has to do with the Node either
[12:27:01] <ALR> And as a model, Node *does* have a Pattern
[12:27:17] <kpiwko> aslak: haven't found any
[12:27:17] <aslak> Node has data that could match a Pattern
[12:27:42] <ALR> aslak: The only data Node has that Pattern doesn't is children
[12:30:03] <aslak> kpiwko, debug JavaSPIExtensionLoader:109, url should tell you which file
[12:30:17] <kpiwko> aslak: ok
[12:30:46] <aslak> ALR, sure, but Pattern.name != Node.name
[12:31:05] <ALR> aslak: Explain the difference?
[12:31:16] <ALR> Node.name is a element name
[12:31:19] <aslak> ALR, not currently, because we only support direct matching..
[12:31:21] <ALR> Pattern.name is ... ?
[12:31:35] <aslak> Pattern.name=muchild.*
[12:31:44] <ALR> muchild?
[12:32:10] <aslak> Pattern.name is the value of how to store what to match with Node.name
[12:32:14] <aslak> it's not Node.name
[12:32:20] <aslak> so. Pattern.name could be a reg exp.
[12:32:47] <ALR> It's still reusable
[12:33:41] *** tommysdk is now known as tommyafk
[12:33:52] <aslak> ALR, your talking about creating a 'fake' node tree to use for matching a existing node tree ?
[12:34:23] <aslak> mytree.get(Node("root").createChild(".*child"))  ?
[12:34:33] <ALR> No, I'm just talking about reuse in modeling.
[12:35:01] <aslak> i don't understand your modeling
[12:35:09] <ALR> The Pattern used for matching/queries is not going to be the same as the one used to hold the internal state of the Node
[12:35:40] <ALR> Basically I don't see why Pattern/Node need to have basically the same stuff in it.
[12:35:41] <aslak> Pattern is conceptually something different from Node.. i don't get why you want them to be the same
[12:36:15] <ALR> A Node *has a* Pattern.  It has a text, a Map of String attributes, and a name
[12:36:51] <aslak> i don't agree with your "has a" relationship
[12:37:31] <aslak> A node has data, and can match a Pattern, but it doesn't have a Pattern
[12:37:56] <aslak> actually it the other way around, a Pattern can match a Node
[12:38:35] <aslak> that Pattern and Node share the same type of data does not qualify them to be the same
[12:38:55] * ALR thinking on that
[12:39:02] <aslak> e.g. a Pattern can match 2 of a Nodes 10 attributes,
[12:39:06] <ALR> Your last point may be true.
[12:39:30] <aslak> and still be true
[12:39:41] <ALR> A Node matches a Pattern
[12:39:46] <ALR> Patterns don't match Nodes
[12:39:57] <ALR> Let me reword that
[12:40:13] <ALR> A Node may fit the stereotype described by Pattern
[12:40:20] <aslak> sure
[12:40:47] <ALR> OK.
[12:40:51] <ALR> I can buy that
[12:40:55] <ALR> git reset --hard HEAD done
[12:41:02] <aslak> hehe :)
[12:41:22] <ALR> Onto the relationship between Query and QueryExecutors
[12:41:54] <ALR> First to define "Query"
[12:42:19] <aslak> Pattern(Query) is only a value object, the QueryExecutors can perform that pattern on a Node
[12:42:23] <ALR> Query is another value object.  It holds a list of Patterns and a flag denoting whether or not it's absolute.
[12:42:32] <aslak> mm
[12:42:47] <ALR> What's Pattern(Query) ?
[12:42:56] <aslak> ignore that..
[12:43:25] <aslak> (i thought we were calling Query Pattern now, but it was only NodeQuery.. so ignore the Pattern part of it.. )
[12:45:54] <ALR> OK.
[12:46:01] <ALR> So define Query.
[12:46:02] <ALR> :)
[12:46:10] <ALR> Without saying what it is.  What it's intended to do.
[12:46:41] <aslak> ALR, no no, your def of Query is correct as it stands now. it's just a 'join' point of the sub patterns
[12:47:05] <aslak> wondering if we should make Pattern  a tree instead tho
[12:47:25] <ALR> My definition just kinda says what it is, not what it does or is used for.
[12:47:27] <ALR> How about:
[12:47:28] <ALR> "A Query is a collection of Patterns which may be run upon a Node tree in order to find potential matches"
[12:47:59] <aslak> yea, sounds right
[12:49:03] <ALR> k
[12:54:43] <ALR> aslak: Allright.  QueryExecutor
[12:58:50] <ALR> aslak: This makes little sense to me:
[12:58:51] <ALR> T execute(Node node);
[12:59:09] <ALR> Because nowhere does it take into account a Query
[12:59:15] <ALR> Yet it's named QueryExecutor.
[13:02:51] <aslak> should probably be, execute(Query, Node)
[13:03:25] <aslak> i think i remember having it as that at some point,  but don't remember why it was changed
[13:03:49] <ALR> Oh my GOD
[13:03:56] <ALR> a do/while loop!
[13:04:03] <ALR> I don't think I've ever seen one in the wild!
[13:04:34] <aslak> heh
[13:06:26] <aslak> hmm. that doesn't even work
[13:07:16] <aslak> should have been: Node root = child; do { root = child.parent() } while(root.parent != null)
[13:07:29] <aslak> no, damn it wrong again
[13:07:35] <aslak> should have been: Node root = child; do { root = root.parent() } while(root.parent != null)
[13:09:13] <ALR> aslak: OK, so let's do CreateQuery
[13:09:17] <ALR> It's not a Query
[13:09:24] <ALR> So the naming kinda doesn't make sense to me.
[13:09:34] <ALR> Could be a CreateQueryExecutor
[13:09:40] <aslak> it's a QueryExecutor that can Create a Query
[13:09:47] <ALR> What does it do?
[13:10:00] <ALR> It doesn't create queries.
[13:10:03] <aslak> or not create the Query, but create a Node tree based on the Query
[13:10:23] <ALR> That's not a query, I don't think, really.
[13:10:37] <ALR> Creating a Node tree doesn't sound like a Query to me
[13:10:44] <ALR> That's more like an action.
[13:10:48] <ALR> A query is a question.
[13:10:49] <aslak> it uses the Query object to create a node tree
[13:11:17] <ALR> Hehe, then it's not using it as a query
[13:11:25] <ALR> It's using it as a collection of Patterns
[13:11:30] <aslak> yes
[13:11:33] <ALR> So:
[13:11:39] <ALR> Query > PatternSet ?
[13:11:52] <aslak> no
[13:12:07] <ALR> Query doesn't ask anything.
[13:12:20] <ALR> It's just a collection of Patterns.  Says so in the JavaDoc
[13:12:26] <ALR> You can use this to make a query.
[13:12:27] <aslak> if you think logically it is a PAtternSet, but i don't like th enaming
[13:12:35] <ALR> OK, sure, propose something else?
[13:12:40] <aslak> Pattern has a 'child' Pattern is better
[13:12:53] <ALR> I don't like "Set", because it's not a mathematical Set really
[13:14:19] <aslak> Pattern Pattern.child()
[13:15:12] <ALR> ?
[13:15:17] <aslak> well.. it's a bit odd to have a Pattern in the middle of the chain saying it's Absolute, but
[13:15:30] <ALR> I was thinking that earlier.
[13:15:42] *** jep has quit IRC
[13:16:08] <ALR> Really you want the "absolute" bit to be reflected when executing the query
[13:16:18] <ALR> Still need a better name for Query now though
[13:16:26] <aslak> maybe it should be removed from the Pattern 'lanauge'. and father have. Node.getRoot() instead
[13:16:37] <ALR> You lost me.
[13:17:42] <aslak> Absolute, it's represented by starting with the / right. convenience for being apple to apply Patterns from the Root node and down. it's bascially the same as Node.getRootNode().apply(Relative Pattern)
[13:17:57] <aslak> apple/able
[13:19:37] <ALR> Right, but I'm still stuck on defining these working bits :)
[13:20:22] <aslak> i'm just saying, if we remove Absolute from the Pattern language, we can make Pattern a chain of Patterns, instead of the top level Query object.
[13:22:01] <aslak> Query.patterns<List<Pattern>> vs Pattern.child<Pattern>
[13:22:15] <ALR> I was thinking kill Query
[13:22:17] <ALR> And do:
[13:22:34] <aslak> as described in my new modeling language.. following. Type.fieldName<FieldType> :)
[13:22:40] <ALR> QueryExecutor.execute(Node,Pattern...)
[13:24:01] <aslak> with no relation.. sure that will work in the current setup
[13:24:58] <ALR> aslak: Sorry, we have some disconnect I think
[13:25:16] <ALR> I'm trying to understand concretely what you're suggesting and falling short.
[13:25:55] <ALR> What I'm thinking is very simply: get rid of the class now called "Query"
[13:26:05] <aslak> same here
[13:26:13] <ALR> And in QueryExecutor.execute, accept all params we need which would comprise a Query.
[13:26:25] *** jose_freitas has joined #jbosstesting
[13:26:35] <ALR> QueryExecutor.execute(Node,isAbsolute,Pattern...)
[13:26:39] <ALR> WDYT?
[13:27:37] <aslak> i'm saying, we can drop isAbsolute, if you want to query from the RootNode and down, send the RootNode to the QueryExecutor instead
[13:27:42] <jose_freitas> morning
[13:28:08] <aslak> jose_freitas, morning
[13:28:10] <ALR> aslak: Ah, even better.
[13:28:18] <ALR> aslak: No need for isAbsolute at all then
[13:28:29] <ALR> QueryExecutor.execute(Node, Pattern...)
[13:28:33] <aslak> ALR, no, but we can add convenience Node.getRoot()
[13:28:43] <ALR> And put it on the user to pass in the root Node
[13:28:47] <ALR> Sure.  That's easy.
[13:38:18] <ALR> aslak: Tired
[13:38:49] <ALR> I think that answers all my design intent questions though
[13:38:59] <ALR> We properly identified each area and discussed naming
[13:39:15] <ALR> So I should be able to get this refactored now into something I think we'll like.
[13:40:14] <ALR> Only pending design issue I see:
[13:40:20] <ALR> Issuing queries to Node directly
[13:42:10] *** ALR has quit IRC
[13:43:57] *** galderz has joined #jbosstesting
[13:44:46] <aslak> kpiwko, did you find the issue?
[13:48:01] *** kevinpollet has joined #jbosstesting
[14:02:45] <jose_freitas> hey aslak, did you fix your coffee machine?
[14:03:21] *** galderz has quit IRC
[14:05:41] *** DavideD has quit IRC
[14:05:44] <aslak> jose_freitas, yes, working like a charm :)
[14:10:30] <jose_freitas> did you write a documentation about it?
[14:10:40] <jose_freitas> hahaha
[14:11:32] *** pilhuhn has joined #jbosstesting
[14:11:32] *** pilhuhn has quit IRC
[14:11:32] *** pilhuhn has joined #jbosstesting
[14:11:41] <aslak> jose_freitas, hehe, no plenty of docs on the subject already.. :)
[14:11:49] *** jharting has quit IRC
[14:14:10] *** alesj1 has quit IRC
[14:31:06] *** tommyafk is now known as tommysdk
[14:31:40] *** galderz has joined #jbosstesting
[14:44:23] <kpiwko> aslak: I wasn't searching for it yet
[14:44:31] *** pilhuhn is now known as pil-afk-biab
[14:44:47] <kpiwko> * I haven't searched it yet yet
[14:46:09] <aslak> kpiwko, aa ok
[14:52:03] <kpiwko> aslak: the reason found...and pretty weird...seems Eclipse caused some intermediate state to be fetched by Maven reactor instead of correct artifact from local repository
[14:52:46] <aslak> kpiwko, and one more..  when using system property org.apache.maven.offline=true, this is read and reflected correctly in MavenSettingsBuilder in enrichWithOfflineMode. MavenDependencyResolverSettings has a MavenSettingsBuilder, but the MavenDependencyResolverSettings has it's own version of offline which is still false so the wrong Repo Type is selected
[14:53:15] <kpiwko> aslak, I'll check that
[14:54:33] <aslak> kpiwko, i created SHRINKWRAP-307 for you..
[14:54:35] <jbossbot> jira [SHRINKWRAP-307] MavenDependencyResolver goOffline does not trigger the Simple Repository Type [Open (Unresolved) Bug, Major, Karel Piwko] https://issues.jboss.org/browse/SHRINKWRAP-307
[14:55:25] <aslak> kpiwko, hmm, but which state had that type of SPI file ?
[14:55:25] <kpiwko> aslak: thanks
[14:55:47] <kpiwko> aslak: drone-enhanced-selenium
[14:56:20] *** tommysdk has quit IRC
[14:56:21] <aslak> aa, ok.. enhanced ? what new ?
[14:56:23] <aslak> 's
[14:57:22] <kpiwko> aslak: actually it is a name for extented DefaultSelenium, which is used in a project I'm migrating to Arquillian
[14:57:40] <kpiwko> aslak: so nothing to be released, I'm just profiting from Drone flexibility
[14:57:48] <aslak> aa ic. .:)
[15:02:30] <kpiwko> aslak: what's expected if org.apache.maven.offline is set and somebody calls goOffline()...does it have priority?
[15:02:45] <kpiwko> aslak: to be precise, does method call have priority?
[15:03:46] *** aslak has quit IRC
[15:04:36] *** aslak has joined #jbosstesting
[15:05:01] <aslak> kpiwko, i would say no. System Prop overrides
[15:05:03] <aslak> :)
[15:06:26] <kpiwko> aslak: yep, if there're 2 ways how to implement something, it's always the more difficult one :)
[15:11:15] *** ldimaggi has joined #jbosstesting
[15:11:55] *** gastaldi has joined #jbosstesting
[15:21:01] <kpiwko> aslak: both should be fixed in https://github.com/shrinkwrap/shrinkwrap/pull/34
[15:26:18] *** gastaldi has left #jbosstesting
[15:39:03] *** pil-afk-biab is now known as pilhuhn
[15:49:35] *** tommysdk has joined #jbosstesting
[15:50:29] <tommysdk> aslak: are you in Oslo atm?
[15:55:58] <aslak> tommysdk, yea, i felt the explotion
[15:56:21] <tommysdk> tommysdk: really really scary!
[15:56:32] <tommysdk> i mean aslak ^ :P
[15:57:47] <aslak> tommysdk, yea.. still no info on what actually happened but
[15:58:04] <tommysdk> hard to imagine it beeing an accident
[15:58:19] <aslak> it can't be.. they are reporting a second now
[15:59:12] <tommysdk> aslak, is aftenposten.no a good news source or do you recommend another?
[15:59:49] <aslak> the better one i would say..  vg.no dagbladet.no
[15:59:54] <aslak> two other options
[16:00:08] <aslak> all of them have offices in the building next to where it happened
[16:00:36] <tommysdk> great thanks
[16:02:37] <aslak> tommysdk, nrk.no as well
[16:04:10] *** maschmid has quit IRC
[16:04:20] <tommysdk> aslak, looks like a war zone
[16:04:33] <aslak> second explosion was rubish..
[16:04:45] <aslak> yea it does
[16:04:46] <tommysdk> but you're ok?
[16:04:56] <aslak> i'm fine.. :)
[16:05:01] <tommysdk> great :)
[16:09:38] <aslak> multiple carbombs is one report
[16:11:06] *** bleathem has joined #jbosstesting
[16:11:46] <tommysdk> why on earth
[16:15:16] <tommysdk> aslak, what is "unna" in norwegian?
[16:15:42] <tommysdk> away?
[16:15:46] <aslak> yea
[16:15:57] <tommysdk> undan in swedish then :)
[16:16:17] <aslak> yea.. :)
[16:18:47] <aslak> lol.. they could have dropped the 15 sec commercial on the explosion videos in the news paper..
[16:19:09] <tommysdk> lol, yeah. pretty tasteless
[16:19:25] <tommysdk> don't think those companies would have minded that
[16:19:56] <aslak> i'm sure the phone company would be happy _NOT_ being assosiated with a bombing in oslo
[16:20:34] <tommysdk> exactly
[16:20:51] <tommysdk> crossing my fingers it's not associated to any terrorist attack
[16:24:48] <tommysdk> aslak, +1 on that tweet :)
[16:25:05] <aslak> :)
[16:28:14] <tommysdk> good pace on #oslo tweets atm
[16:31:44] <jose_freitas> I hope it's not any terrorist attack too
[16:32:20] <jose_freitas> I remember when london was attacked in 2005 and a brazillian guy was killed at subway by mistake
[16:32:53] <tommysdk> would be very scary if we had a second terrorist attack in Scandinavia in 7 months.
[16:33:31] <aslak> twitter is impossible to follow..  to fast
[16:33:48] <tommysdk> Swedish police have raised their security level in Stockholm as a consequence of the Oslo explosion
[16:33:58] <tommysdk> aslak: yeah
[16:34:27] <jose_freitas> I can't imagine the paranoia caused with a terrorist attack
[16:35:00] 
[16:35:22] <tommysdk> so they suspect that 1 has been killed
[16:35:32] <aslak> yea
[16:37:44] <jose_freitas> tommysdk, are you from scandinavia as well?
[16:38:05] <aslak> http://maps.google.com/maps?q=oslo&hl=en&ll=59.915397,10.745455&spn=0.004437,0.010675&t=h&z=17
[16:38:07] <tommysdk> jose_freitas, yes, I'm from Stockholm, Sweden
[16:42:49] *** mbg has joined #jbosstesting
[17:06:59] <tommysdk> http://bbc.in/o2XlDx
[17:08:15] <aslak> tdiesler, you want a CR3 ?
[17:12:29] *** kevinpollet has quit IRC
[17:35:21] *** tommysdk has quit IRC
[17:48:16] *** pilhuhn is now known as pil-dinner
[17:48:55] *** alesj has joined #jbosstesting
[17:48:58] *** alesj has left #jbosstesting
[18:06:38] *** mgoldmann has quit IRC
[18:11:59] *** lfryc has quit IRC
[18:38:05] *** ianbrandt has joined #jbosstesting
[18:51:18] *** maeste has quit IRC
[18:57:38] *** galderz has quit IRC
[18:59:11] *** aslak has quit IRC
[19:09:45] *** ldimaggi has quit IRC
[19:22:56] *** aslak has joined #jbosstesting
[19:25:21] *** ldimaggi has joined #jbosstesting
[20:06:06] *** alesj has joined #jbosstesting
[20:06:27] *** ALR has joined #jbosstesting
[20:14:47] *** ldimaggi has quit IRC
[20:17:50] *** kpiwko has quit IRC
[20:28:08] *** ldimaggi has joined #jbosstesting
[20:31:43] *** tdiesler has quit IRC
[20:34:13] *** ge0ffrey has quit IRC
[20:43:09] *** vtunka has quit IRC
[21:19:42] *** jose_freitas_aw has joined #jbosstesting
[21:19:48] *** jose_freitas has quit IRC
[21:19:54] *** jose_freitas_aw is now known as jose_freitas
[21:23:18] *** jose_freitas_aw has joined #jbosstesting
[21:24:45] *** jose_freitas has quit IRC
[21:25:11] *** mhuniewicz has joined #jbosstesting
[21:26:42] *** jose_freitas_aw is now known as jose_freitas
[21:27:31] *** mbg has quit IRC
[21:27:55] <mhuniewicz> ALR, hi, I saw you fixed that NPE?
[21:27:55] *** pil-dinner has quit IRC
[21:31:16] *** jose_freitas_aw has joined #jbosstesting
[21:32:20] *** lightguard_jp has joined #jbosstesting
[21:34:23] *** jose_freitas has quit IRC
[21:36:26] *** Jaikiran has quit IRC
[21:38:39] *** jose_freitas_aw is now known as jose_freitas
[21:44:05] <ALR> mhuniewicz: Yes.
[21:44:09] <ALR> Have you tried it out?
[21:44:34] <mhuniewicz> Not yet. Do you suggest I check it out and build it?
[22:12:33] *** ldimaggi has quit IRC
[22:56:52] *** jose_freitas has quit IRC
[23:02:45] *** alesj has quit IRC
[23:10:36] *** galderz has joined #jbosstesting
[23:10:43] *** galderz has quit IRC
[23:14:12] <mhuniewicz> ALR, are you there?
[23:31:26] <ALR> mhuniewicz: yes
[23:31:34] <mhuniewicz> ALR, I'm unable to build it.
[23:31:36] <mhuniewicz> Eclipse
[23:31:43] <mhuniewicz> gives me Maven problems, missing deps.
[23:31:50] <mhuniewicz> Console build just fails strangely.
[23:32:05] <ALR> Need to see the failures.
[23:32:46] <mhuniewicz> http://pastebin.com/nuXcfH3K
[23:33:06] <ALR> Pretty clear to me.
[23:33:13] <ALR> Line 30
[23:33:35] <mhuniewicz> It's a warning. Is that what's making it fail?
[23:33:47] <ALR> The failure report is at line 57
[23:34:17] <mhuniewicz> Hm. OK. Can it be JRE6?
[23:34:27] <ALR> It *can* be, but it shouldn't.
[23:34:43] <ALR> ShrinkWrap is compiled w/ JDK6, but is intended to run in JRE5
[23:34:48] <ALR> So we run the tests in JRE5.
[23:42:42] <mhuniewicz> ALR, ok, I got it to run and now I'm getting "Access denied to: http://repository.jboss.org/maven2/trove/trove/2.1.1/trove-2.1.1.po[...]"
[23:42:56] <ALR> Iiiiiiiiinteresting.
[23:43:12] <ALR> One sec.
[23:43:27] <mhuniewicz> In Eclipse there's a lot of things missing, even JUnit. It might be something on my side.
[23:43:55] <ALR> Don't worry about Eclipse just yet
[23:44:09] <ALR> You don't even need to bring this into an IDE.
[23:44:16] <ALR> Just to build it and try it on your system
[23:44:20] <mhuniewicz> I realise that.
[23:44:33] <ALR> So "trove" has moved to the JBoss Deprecated repository.
[23:44:37] <ALR> Which is not in the POM config.
[23:44:39] <ALR> I'll update that.
[23:46:49] <ALR> mhuniewicz: Try making your root pom.xml ave this in the "repositories" section: https://gist.github.com/1100506
[23:49:55] <mhuniewicz> ALR, did not help.
[23:50:46] <ALR> mhuniewicz: Full error?
[23:51:48] <mhuniewicz> http://pastebin.com/6ZjccDQg
[23:52:01] <mhuniewicz> ALR, there you go.
[23:52:28] <ALR> Access denied?
[23:52:35] <ALR> Uno momento por favor
[23:52:54] <mhuniewicz> Puedo esperar.
[23:54:55] <ALR> I get Downloaded: http://repository.jboss.org/maven2/trove/trove/2.1.1/trove-2.1.1.jar (353 KB at 237.5 KB/sec)
[23:55:07] <ALR> Will ping our build system guy
[23:55:20] <mhuniewicz> That's forbidden as well.
[23:55:24] <ALR> ...who is not online.
[23:55:37] <ALR> Wow, really?
[23:55:45] <ALR> I can get right to it from the browser.
[23:55:49] <ALR> Where are you?
[23:56:10] <mhuniewicz> Where as in geographically?
[23:56:14] <ALR> Yes.
[23:56:22] <mhuniewicz> In the UK.
[23:56:37] <ALR> How about some output from wget?
[23:56:55] <ALR> Or what's the browser error when you try?  403?
[23:57:00] <ALR> HTTP 403 I mean?





top