July 21, 2011 
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31
July 21, 2011 [00:01:28] *** stuartdouglas has left #jbosstesting [00:03:31] *** mbg has quit IRC [00:07:32] *** alesj has quit IRC [00:09:33] *** tdiesler has quit IRC [00:09:34] *** tdiesler has joined #jbosstesting [00:09:34] *** tdiesler has quit IRC [00:09:34] *** tdiesler has joined #jbosstesting [00:25:42] *** lightguard_jp has quit IRC [00:47:23] *** tdiesler1 has joined #jbosstesting [00:50:14] *** tdiesler has quit IRC [00:50:14] *** ianbrandt has quit IRC [00:50:45] *** bleathem has quit IRC [01:28:31] *** aaronwalker has joined #jbosstesting [02:39:14] *** tdiesler has joined #jbosstesting [02:40:07] *** tdiesler1 has quit IRC [02:46:36] *** gastaldi has joined #jbosstesting [03:24:27] *** gastaldi has left #jbosstesting [04:05:50] *** bgeorges has joined #jbosstesting [04:37:57] *** ldimaggi has joined #jbosstesting [04:40:09] *** bgeorges has quit IRC [04:43:18] *** bgeorges has joined #jbosstesting [04:43:27] *** bgeorges has quit IRC [04:49:39] *** bgeorges has joined #jbosstesting [04:56:38] *** gastaldi has joined #jbosstesting [04:58:17] *** bgeorges has quit IRC [05:00:07] *** gastaldi has quit IRC [05:01:05] *** bgeorges has joined #jbosstesting [05:05:29] *** bgeorges has quit IRC [05:05:30] *** ldimaggi has quit IRC [05:06:53] *** bgeorges has joined #jbosstesting [05:35:37] *** bleathem has joined #jbosstesting [06:14:34] *** lightguard_jp has joined #jbosstesting [06:30:09] *** bgeorges has quit IRC [07:48:37] *** aslak has joined #jbosstesting [07:48:37] *** aslak has quit IRC [07:48:37] *** aslak has joined #jbosstesting [08:15:18] *** mgoldmann has joined #jbosstesting [08:15:18] *** mgoldmann has quit IRC [08:15:19] *** mgoldmann has joined #jbosstesting [08:18:47] *** tommysdk has joined #jbosstesting [08:20:53] *** jharting has joined #jbosstesting [08:39:13] *** alesj has joined #jbosstesting [08:39:33] *** tdiesler has quit IRC [08:40:41] *** kpiwko has joined #jbosstesting [08:42:00] *** bleathem has quit IRC [08:48:06] *** ge0ffrey has joined #jbosstesting [08:52:00] *** milestone has joined #jbosstesting [08:52:04] <milestone> morning [08:53:13] *** mgoldmann has quit IRC [09:05:14] <aslak> morning [09:06:29] *** mgoldmann has joined #jbosstesting [09:09:37] *** milestone has quit IRC [09:10:11] *** milestone has joined #jbosstesting [09:15:49] *** ALR has joined #jbosstesting [09:16:01] <ALR> aslak: Morning. [09:19:00] <aslak> ALR, morning [09:19:21] <ALR> aslak: Missed ya yesterday [09:19:30] <ALR> But all set here for SPI/SD review if you want. [09:19:38] <ALR> Maybe you're still getting your coffee though :) [09:19:53] <aslak> ALR, sure, coffee's done.. :) [09:20:42] *** kevinpollet has joined #jbosstesting [09:20:43] <ALR> aslak: Cool, lemme know when you've got it brought out [09:20:48] <ALR> *brought up [09:21:11] <aslak> it's up :) [09:21:20] <aslak> and half way down.. hehe [09:21:39] <aslak> just pulled in latest sd [09:21:56] <ALR> Cool [09:22:09] * ALR letting the IDE update/build [09:22:27] <ALR> aslak: OK, ready [09:23:39] *** maschmid has joined #jbosstesting [09:24:09] <ALR> Charset. Package-private, thus not exported. So it's fine as-is. [09:24:22] <ALR> DescriptorExporter [09:24:23] <ALR> Needs docs. [09:24:49] <ALR> Signature: void to(Node node, OutputStream output); [09:25:06] <ALR> So it's not really a DescriptorExporter; it exports Nodes [09:25:32] <ALR> aslak: ^ [09:25:46] <aslak> mm [09:25:49] <ALR> aslak: Unless I should fire up Skype? Might be faster? [09:26:13] <aslak> naa, irc is fine.. [09:26:18] <ALR> k [09:26:29] <ALR> So yeah, let's first start with Node I guess. [09:26:38] <ALR> Makes more sense than doing this in Alphabetical order. [09:26:49] <ALR> What's a Node? :) [09:27:42] <ALR> I'd say it's a classic tree data structure element [09:27:59] <aslak> alesj, well, your correct, it doesn't export Descriptors. it exports Nodes, which is given by the NodeProvider [09:28:13] <aslak> alesj, back to sleep.. [09:28:15] <aslak> ALR, ^ [09:28:17] <aslak> :) [09:28:17] <ALR> Which may contain a set of elements/attributes, String data, or children [09:28:43] <ALR> aslak: Yep. So first we define Node. And might have to add some stuff to it. [09:28:45] <ALR> A Node: [09:28:46] <aslak> correct [09:29:40] <ALR> Should contain EITHER String CDATA OR List<Node> children. Correct? [09:29:50] <ALR> ie. Never will it have both String data and children [09:30:10] <aslak> it would have both, xhtml does [09:30:14] <aslak> would/could [09:30:49] <aslak> well.. in xhtml it has TextNodes [09:30:53] <ALR> Seriously? Where? [09:31:11] <aslak> <p>bla bla <b> bal</b> </p> [09:31:13] <ALR> <something element="value"> DATA <child /> </something> [09:31:42] <ALR> Oooh. Good point. OK. God that's so weird. [09:32:10] <aslak> but the parser reads that as, P -> TextNode > B > TextNode or similar [09:32:21] <ALR> Doesn't matter. [09:32:24] <ALR> XHTML is still valid XML. [09:32:34] <aslak> yea [09:33:12] <ALR> OK, so no assertions needed there. [09:33:22] <ALR> aslak: How do you feel about all the Query stuff in the Node class? [09:36:18] <ALR> aslak: Also the API is a bit inconsistent. For instance "children" returns the children instead of "getChildren()". [09:36:29] <ALR> Same for "attributes" [09:37:00] <aslak> ALR, i like it in general. very powerful. saves a bunch of code. but could be a bit un intuative, needs docs i guess [09:37:21] <ALR> aslak: Yeah, is this you or Lincoln? [09:37:33] <ALR> Because it looks like it's trying to use fluent syntax, but no fluency. [09:37:47] <aslak> Node ? [09:37:51] <ALR> Yeah [09:37:54] <aslak> that's me [09:37:56] <ALR> I know you initially did it [09:38:10] <ALR> But like "attribute" [09:38:19] <ALR> Why not "getAttribute()" ? [09:38:30] <aslak> why get.. :) [09:38:52] <ALR> Well, let's talk SW conventions. [09:39:07] <ALR> The pure standard, of course, is a simplified JavaBeans naming strategy [09:39:10] <ALR> getXXX, setXXX [09:39:10] <aslak> attributes() = all, attribute(String name) = get, attribute(String name, Object vlaue) = set [09:39:34] <ALR> In SW we have fluency by eliminating the "get" and returning the same object. [09:39:56] <ALR> In line with that article on method chaining Fowler did. [09:40:22] <ALR> Yeah, that's really weird to me. [09:40:33] <ALR> That one argument is a get, and 2 arguments is a set [09:42:08] <ALR> Also "attributes" now is exporting a mutable view. So I'll wrap that in an immutable one. [09:42:53] <aslak> yea, that should be immutable [09:44:01] <ALR> aslak: "attribute(String,Object)" I'll leave as-is [09:44:05] <ALR> Because it returns Node [09:44:14] <ALR> Thus follows the method chaining convention [09:44:42] <ALR> aslak: also, what;s with null inputs here? Probably not allowed at all for String name [09:44:56] <ALR> So we need a doc'd IllegalArgumentException and precondition check, yes? [09:45:00] <aslak> we run into the same issue here as we do with the whole of SD, method chaining combined with a Get view becomes odd [09:45:37] <ALR> aslak: The "up" solution solves that one :) [09:45:54] <aslak> yea cuz that's not strage at all [09:46:14] <aslak> :) [09:46:19] <ALR> Hehe, I like it. [09:46:27] <ALR> And I think it won the discussion in that gist. [09:46:45] <aslak> parent would be a better name tho [09:46:52] <ALR> aslak: What about the Query stuff in here? [09:47:01] <ALR> aslak: Think it belongs, or should be moved elsewhere? [09:47:20] <ALR> (Part of me thinks that the querying bit should be elsewhere, leaving Node to just be a data structure" [09:48:10] <aslak> yea i had it separated first if i remember correctly, but then you get a api like: [09:48:33] <aslak> Query.get(Node, Expression) [09:48:48] <ALR> Or: [09:48:49] <aslak> Node.get(Expression) seemed nicer [09:49:08] <ALR> Query q = QueryFactory.create(Node); [09:49:17] <ALR> q.get(expression); [09:49:24] <aslak> yea, that's not helping [09:49:39] <ALR> Well, it associates a Query w/ a Node [09:49:56] <ALR> So why is that problematic? [09:50:14] <aslak> it does, but it is to ask the Node, i operate on your tree [09:50:34] <aslak> node.getOrCreate("childNode?attribute=value").text(ba) [09:51:04] <aslak> will create <childNode attribute="value">ba</childNode> [09:51:31] <aslak> or will set ba on childNode with attribute=value if it exists [09:53:00] <ALR> I still don't see why from an API perspective, the suggestion I just made is problematic. [09:53:09] <aslak> persistence.getOrCreate("persistence-unit?name=myunit").getOrCreate("jta-data-source").text("myDataSource") [09:54:04] <ALR> Query q = QueryFactory.create(persistence); [09:54:34] <ALR> q.getOrCreate("persistence-unit?name=myunit").getOrCreate()...; [09:55:04] <ALR> Basically keep the ability to chain those construction points, because that's useful. [09:55:09] <ALR> But just not a part of Node [09:55:26] <ALR> Actually it wouldn't be on Query [09:55:33] <ALR> It'd be on something else, something new. [09:55:57] <ALR> Because "Query" is kind of a weird name I think, considering what it currently does. [09:56:20] <ALR> aslak: ^ Simple question as an aside. "name" of Node should have no spaces, correct? [09:56:40] <aslak> you could actually do that in one, ("persistence-unit?name=myunit/jta-data-source").text("MyDataSource").. but not related.. :) [09:58:35] <aslak> ALR, space in name.. hmm.. probably not allowed, not in xml atleast [09:58:55] <aslak> ALR, why is query wired? because it can create nodes as well? [09:59:30] <ALR> aslak: Query is weird because I've no idea what it represents. [09:59:51] <ALR> Or how it differs from NodeQuery [10:00:02] <ALR> So now fixing up the space assertions in Node [10:00:22] *** aaronwalker has quit IRC [10:01:37] <aslak> aa.. Query is the root, NodeQuery represent the individual sub tree representation [10:02:48] <aslak> "persistence-unit?attr=val/subnode" = Query -> NodeQuery("persistence-unit?atr=value") -> NodeQuery("subnode") [10:03:00] <ALR> Umm... [10:03:07] <ALR> Wrapping my head around that [10:03:22] <aslak> One NodeQuery pr / in the 'expression' [10:04:50] <aslak> ALR, i wrote most of it on the Airport in Copenhagen over a Carrot cake, so i'm sure we can find some wholes in it.. hehe [10:05:08] <ALR> Ha, it's no problem [10:05:15] <ALR> It's proven to be really useful and fast [10:05:26] <ALR> And I remember you did this as a prototype [10:05:34] <ALR> So all I'm trying to do now is tighten it up [10:05:35] <ALR> Document. [10:05:38] <ALR> Builk up tests [10:05:43] <ALR> Rename etc [10:05:59] <aslak> ALR, yea, the prototype was without the Query bits.. [10:06:57] <tommysdk> Hi everyone [10:07:11] <ALR> tommysdk: Hiya [10:07:11] *** milestone has left #jbosstesting [10:07:47] <ALR> aslak: As it is, I already have some work to do before we talk again. [10:07:55] <ALR> My instinct is to remove the query stuff from the Node class. [10:07:59] <ALR> But I'll dig deeper. [10:08:10] <ALR> In the meantime, I've some more tests to write. I renamed a doc'd some already. [10:08:19] <ALR> Also added assertions etc [10:09:02] <aslak> cool [10:09:08] <ALR> For instance we're missing coverage on some "remove" methods, etc. [10:09:23] <ALR> So let me continue on this and we'll regroup when I've got some commits in. [10:10:08] <ALR> aslak: Something we once talked about and since haven't is the idea of pluggable backends. [10:10:18] <ALR> Right now we have XmlDomExporter/Importer [10:10:29] <ALR> In theory we could also plug in SAX ones. [10:10:59] <ALR> This is such a cool project [10:11:46] <aslak> ALR, yea, i did some sax based ones at some point as a test, but the handling of namespace was different from dom so i dropped it for now. we wern't talking about huge xml docs anyway so [10:12:11] <ALR> I think the only DOM usage we have now is limited to those classes. [10:12:17] <aslak> yea [10:12:34] <aslak> we still have issues with comments tho.. [10:12:46] <ALR> aslak: Something that concerns me is comments, and also formatting/spacing [10:12:50] <ALR> Yes, exactly. [10:13:11] <aslak> need some special handling.. tey get sucked in as Node("#Comment") or similar, the xml exporter doe snot like that has a name of a node [10:13:16] <ALR> I think we need to provide for pluggable formatters in our toString exports. [10:13:39] <ALR> So users can set things like "line after element" or "space indents" or whatever [10:13:40] <aslak> has/as [10:13:57] <aslak> mm sure.. [10:14:03] <ALR> aslak: Got a failing testcase for that? [10:14:15] <aslak> ALR, no, but i ca get you one.. [10:14:35] <ALR> aslak: That'd be great. [10:14:47] <ALR> And an JIRA for it if you've got. [10:14:52] <ALR> OK, I'm off to bed. [10:14:55] <ALR> I'll keep on this. [10:15:13] <ALR> I don't think it'll take too long to whip this into something we can confidently support. [10:15:22] <ALR> Basically just renaming and moving stuff. [10:15:26] <ALR> The Node design is great [10:15:34] <ALR> Our own data structure under the hood [10:15:39] <ALR> Which is FAST [10:15:53] <ALR> And pluggable importers/exporters to integrate w/ STAX SAX DOM, etc [10:15:59] <ALR> Nice, I'm very happy with it. [10:16:02] <aslak> :) [10:16:18] <ALR> Just needs API love, which I can give pretty easily here. [10:16:28] <ALR> tommysdk: Need anything from me now? [10:17:02] <tommysdk> ALR: Not atm, think I managed to sort out the rebase stuff by myself. took a while though :) [10:17:04] <aslak> ALR, yea, go of to bed and have some wet api love dreams.. ha ha .. ;) [10:17:09] <ALR> tommysdk: It's upstream now. [10:17:15] <ALR> aslak: ha. [10:17:19] <tommysdk> ALR: Just noticed, great stuff [10:17:48] <tommysdk> Think I'll have some time to dig into shrinkdesc today [10:17:57] <ALR> tommysdk: Have fun. [10:18:11] <ALR> tommysdk: Get in touch with Ralf Battenfeld [10:18:17] *** galderz has joined #jbosstesting [10:18:22] <ALR> tommysdk: See if he has any work he wants to offload on you [10:18:34] <tommysdk> ALR: Sure, how do I get in touch with him? [10:18:41] <ALR> tommysdk: Upcoming stuff there is in upstream/SHRINKDESC-54 [10:18:43] <jbossbot> jira [SHRINKDESC-54] Develop/split EE Spec and JBoss-specific descriptors in API / Impl Split [Open (Unresolved) Task, Major, Ralf Battenfeld] https://issues.jboss.org/browse/SHRINKDESC-54 [10:18:49] <ALR> tommysdk: GitHub message should suffice. [10:19:02] <tommysdk> ALR: alright, thanks [10:19:10] <ALR> tommysdk: Basically the current state is this: [10:19:21] <ALR> I broke out some stuff into a proper SPI last week [10:19:34] <ALR> Now aslak and I are combing through it to put the SPI into a locked, supportable state [10:19:48] <ALR> Concurrently, Ralf has been building out APIs for the actual descriptors [10:20:00] <ALR> The models for spec and JBoss XSDs [10:20:04] <ALR> Like WebAppDescriptor [10:20:07] <aslak> ALR, another thing.. endUserViewImplType seems odd in XMLImporter [10:20:50] <ALR> aslak: Haven't gotten to that part of the review yet [10:20:53] <ALR> Looks like something I wrote [10:21:10] <ALR> Well, from DescriptorImporterBase [10:21:28] <aslak> ALR, wonder if we should split NodeImport and DescriptorImport [10:21:48] <aslak> NodeImport = XMLImporter, DescirptorImporter wraps it in the enduserview etc [10:21:55] <ALR> return endUserViewImplType.cast(ApiExposition.createFromImplModelType(endUserViewImplType, descriptorName)); [10:22:04] <ALR> That's the culprit [10:22:20] <ALR> DescriptorImporterBase.from() [10:22:39] <ALR> aslak: Let's revisit that after working out Node/Query [10:36:08] <tommysdk> ALR: alright. Is JIRA up-to-date with this? [10:37:07] <ALR> tommysdk: Kinda. [10:37:16] <ALR> tommysdk: SD-54 is a blanket development task [10:37:28] *** alesj has quit IRC [10:37:50] <ALR> Basically I keep upstream/SD-54 rebased off upstream/master [10:37:57] <ALR> And we can all sync off of it [10:38:25] <ALR> Once we have an API we feel is OK to start testing out in the community, we'll merge it in. [10:39:11] <tommysdk> ok [10:41:52] *** OndraZizka has quit IRC [10:42:10] *** OndraZizka has joined #jbosstesting [10:42:49] <ALR> tommysdk: Actually, looking for something to do? [10:42:50] <ALR> :) [10:43:10] <tommysdk> ALR: Kind of :) [10:43:20] <ALR> tommysdk: One sec. [10:43:27] <tommysdk> ALR: Have some spare hours today and tomorrow to put in good use [10:43:48] <tommysdk> ALR: On Shrinkwrap/Shrinkdesc/Arquillian/whatever :) [10:45:19] *** alesj has joined #jbosstesting [10:47:10] <ALR> tommysdk: SHRINKDESC-58 [10:47:12] <jbossbot> jira [SHRINKDESC-58] Refine SPI for locking to Beta-level [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKDESC-58 [10:47:21] <ALR> tommysdk: https://github.com/shrinkwrap/descriptors/tree/SHRINKDESC-58 [10:47:22] <OndraZizka> Hi all, javac doesn't compile package-info.java to package-info.class. 1) Is it supposed to? 2) How do I achieve that? Hibernate needs that... [10:49:40] <tommysdk> ALR: Then a git question, how do I get that upstream branch down locally? [10:51:05] <ALR> tommysdk: OK, first, here's the task which should be easy for you [10:51:13] <ALR> (But will save me some time tomorrow): https://issues.jboss.org/browse/SHRINKDESC-59 [10:51:14] [10:51:27] <ALR> tommysdk: Show me output of "git remote" ? [10:51:30] *** tdiesler has joined #jbosstesting [10:51:58] <tommysdk> projects/descriptors$ git remote [10:52:00] <tommysdk> origin [10:52:00] <OndraZizka> kpiwko, jharting, nevite? ^^ [10:52:00] <tommysdk> master [10:52:06] <tommysdk> upstream [10:52:14] <ALR> tommysdk: "git remote -v " [10:52:35] <kpiwko> OndraZizka: nevime co? [10:52:41] <ALR> OndraZizka: I thought package-info was just for JavaDoc? [10:52:57] <tommysdk> origin git at github dot com:tommysdk/descriptors.git (fetch) origin git at github dot com:tommysdk/descriptors.git (push) upstream git://github.com/shrinkwrap/descriptors.git (fetch) upstream git://github.com/shrinkwrap/descriptors.git (push) [10:52:58] <OndraZizka> ALR, me too, but hibernate does this: [10:53:06] <OndraZizka> public XPackage packageForName(String packageName) throws ClassNotFoundException { [10:53:06] <OndraZizka> return getXAnnotatedElement( ReflectHelper.classForName( packageName + ".package-info" ).getPackage() ); [10:53:06] <OndraZizka> } [10:53:15] <tommysdk> with line feeds: [10:53:16] <tommysdk> origin git at github dot com:tommysdk/descriptors.git (fetch) [10:53:20] <tommysdk> origin git at github dot com:tommysdk/descriptors.git (push) [10:53:23] <tommysdk> upstream git://github.com/shrinkwrap/descriptors.git (fetch) [10:53:25] <tommysdk> upstream git://github.com/shrinkwrap/descriptors.git (push) [10:53:43] <ALR> OndraZizka: In that case, no idea :) Sorry. [10:53:45] <ALR> tommysdk: "git fetch upstream" [10:54:21] <tommysdk> ALR: ah sweet, got new branches for -54 and -58 now [10:54:22] <ALR> tommysdk: Then: "git checkout -b SHRINKDESC-59 remotes/upstream/SHRINKDESC-58" [10:54:23] [10:54:23] <jbossbot> jira [SHRINKDESC-58] Refine SPI for locking to Beta-level [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKDESC-58 [10:54:59] <ALR> tommysdk: So that'll make you a new branch for 59, switch you into it, and it'll be based off upstream -58 [10:55:06] <tommysdk> ALR: Branch SHRINKDESC-59 set up to track remote branch SHRINKDESC-58 from upstream. [10:55:08] [10:55:09] <jbossbot> jira [SHRINKDESC-58] Refine SPI for locking to Beta-level [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKDESC-58 [10:55:09] <tommysdk> ALR: Cool [10:55:13] <ALR> tommysdk: Exactly. [10:55:22] <ALR> tommysdk: So *only* use that branch for 59 commit. [10:55:24] <ALR> commits [10:55:33] <tommysdk> great! [10:55:42] <ALR> 58 I'll work on in general. 59 is a subtask [10:55:54] <ALR> tommysdk: No, thank you, that'll save me some time in writing those tests. [10:55:58] <kpiwko> OndraZizka: ALR: package-info.java should be javadoc only [10:55:58] <ALR> Later guys. [10:56:22] <tommysdk> ALR: ! See you later :) [10:56:23] <kpiwko> OndraZizka: there was some groovy compiler issue because of it trying to compile it as well [10:56:28] <kpiwko> ALR: good night! [10:56:38] <ALR> :D Bye y'all [10:57:46] <OndraZizka> kpiwko, not true entirely - jpechane found that with annotation, it compiles into a class [10:58:19] <OndraZizka> I suggested to put it into our employee candidate exams :) [10:58:53] <kpiwko> OndraZizka: :)...because there is no other place where to put package scoped annotation [10:59:07] <OndraZizka> true, true [10:59:40] <kpiwko> OndraZizka: could you remove one of the Zizkas out? [10:59:41] <kpiwko> :) [11:00:41] <kpiwko> OndraZizka: maybe PackageBased annotations are compile time only, and thus nothing is generated [11:01:15] <OndraZizka> kpiwko: I had no annotation there. [11:01:41] <OndraZizka> kpiwko: I looked for some annotation and only one with RUNTIME retention is @Deprecated :) [11:01:54] <OndraZizka> So now my package is deprecated. But hibernate is happy :) [11:02:06] <OndraZizka> (in JDK) [11:03:43] <kpiwko> OndraZizka: what do you need that annotation for? [11:04:01] <OndraZizka> to make it compile into .class [11:11:38] *** kevinpollet_ has joined #jbosstesting [11:13:16] *** kevinpollet has quit IRC [11:13:17] *** kevinpollet_ is now known as kevinpollet [11:46:30] *** ALR has quit IRC [12:00:57] *** kevinpollet has quit IRC [12:05:46] *** lightguard_jp is now known as lightguard_jp_aw [12:06:35] *** lightguard_jp_aw is now known as lightguard_jp [12:26:55] *** lightguard_jp has quit IRC [12:28:01] <aslak> kpiwko, having some issues running the openshift examples against my own account. getting auth failure on checkout [12:28:25] <aslak> kpiwko, do you have a non password protected ssh key ? [12:28:44] <aslak> kpiwko, good morning btw.. :) [12:28:51] <kpiwko> aslak: morning/afternonn [12:29:01] * aslak a bit on off, fixing the coffee machine.. :) [12:29:24] <kpiwko> aslak: yes, I probably have non protected one for openshift [12:29:32] <kpiwko> aslak: as it never asked for a password [12:29:43] <aslak> ok, mine does.. [12:30:09] <aslak> any password input for jgit ? [12:31:20] <aslak> kpiwko, code looks great btw.. :) [12:31:49] <aslak> kpiwko, visually appealing [12:31:51] <aslak> :) [12:31:54] <kpiwko> aslak: :) [12:32:16] <kpiwko> aslak: I'm just checking if there is jgit support for asking user to provide an input [12:43:04] *** jhuska has joined #jbosstesting [13:14:20] *** pilhuhn has joined #jbosstesting [13:15:13] *** ge0ffrey has quit IRC [13:18:11] *** maeste2 has joined #jbosstesting [13:18:21] *** maeste2 has quit IRC [13:18:27] *** maeste has quit IRC [13:19:21] *** maeste has joined #jbosstesting [13:26:50] *** kevinpollet has joined #jbosstesting [13:34:28] *** stuartdouglas has joined #jbosstesting [13:40:05] *** jose_freitas has joined #jbosstesting [13:41:29] *** stuartdouglas has quit IRC [13:42:03] *** alesj has quit IRC [14:02:26] *** kevinpollet has quit IRC [14:03:09] *** kevinpollet has joined #jbosstesting [14:18:48] <kpiwko> aslak: have you ever seen something like http://pastebin.com/Ew3j6UhL ? seems utterly strange to me, it was working fine few hours ago [14:37:39] *** stuartdouglas has joined #jbosstesting [14:50:04] *** kpiwko has quit IRC [14:52:09] *** alesj has joined #jbosstesting [15:02:35] *** ldimaggi has joined #jbosstesting [15:02:55] <jose_freitas> hey aslak, g'morning [15:03:10] <jose_freitas> ["jboss.deployment.unit.\"624b14c4-2bb2-4d44-be9e-1168c4d29164.war\".CONFIGURE_MODULE missing [ jboss.module.information.service.\"deployment.arquillian-service\".main ] [15:03:24] <jose_freitas> do you know what could be cause the absence of this service? [15:03:40] <aslak> jose_freitas, aaa.. yea [15:03:43] <aslak> hmmm [15:04:07] <aslak> jose_freitas, it's not being installed since your not using the jmx-as7 protocol [15:04:26] <aslak> jose_freitas, needs to move that -add module dep- to the protocol, and not as a general handler for the as7 container [15:06:21] <jose_freitas> hm [15:07:02] <aslak> AS7-1361 [15:07:04] <jbossbot> jira [AS7-1361] Arquillian AS7 Container force a dependency on Arquillian Service in manifest.mf [Open (Unresolved) Bug, Major, Andrew Rubinger] https://issues.jboss.org/browse/AS7-1361 [15:08:23] <jose_freitas> hmm [15:08:51] <aslak> jose_freitas, doing a lot of as7 stuff today, so i'll look into it later [15:09:17] <jose_freitas> ok [15:09:37] <jose_freitas> one thing, when using servlet protocol, do we still need the conf: <protocol type="jmx-as7">... [15:09:38] <jose_freitas> ? [15:09:50] <aslak> no [15:10:04] <aslak> the need for doing that will be removed today as well. :) [15:10:14] <jose_freitas> great :) [15:10:33] <aslak> now, back to assembling my coffee machine so i can wake up.. :) [15:10:40] <jose_freitas> :) [15:10:48] *** bgeorges has joined #jbosstesting [15:21:20] *** jharting has quit IRC [15:25:34] *** galderz has quit IRC [15:25:57] *** galderz has joined #jbosstesting [15:30:26] *** lfryc has joined #jbosstesting [15:49:55] *** galderz has quit IRC [15:50:21] *** bleathem has joined #jbosstesting [16:08:22] *** galderz has joined #jbosstesting [16:08:26] *** ldimaggi has quit IRC [16:11:00] *** bdlink has joined #jbosstesting [16:19:42] *** jose_freitas has quit IRC [16:24:08] *** ldimaggi has joined #jbosstesting [16:39:59] *** bgeorges has quit IRC [17:13:04] *** jhuska has quit IRC [17:14:16] *** kevinpollet has quit IRC [17:14:55] *** jhuska has joined #jbosstesting [17:18:47] *** tdiesler has quit IRC [17:25:37] *** mbg has joined #jbosstesting [17:25:42] *** jdlee has quit IRC [17:25:52] *** pilhuhn has quit IRC [17:30:38] *** jdlee has joined #jbosstesting [17:30:38] *** jdlee has joined #jbosstesting [17:30:57] *** alesj has quit IRC [17:33:46] *** alesj has joined #jbosstesting [17:39:05] *** maschmid has quit IRC [17:39:25] *** ldimaggi has quit IRC [17:42:29] *** jose_freitas has joined #jbosstesting [17:53:00] *** ldimaggi has joined #jbosstesting [18:00:13] *** vtunka has quit IRC [18:01:49] *** lfryc has quit IRC [18:09:08] *** mbg has quit IRC [18:11:42] *** bdlink has quit IRC [18:13:25] *** bdlink has joined #jbosstesting [18:14:20] *** kevinpollet has joined #jbosstesting [18:14:44] *** kevinpollet has quit IRC [18:32:50] *** ianbrandt has joined #jbosstesting [18:47:10] *** rbattenfeld has joined #jbosstesting [18:57:11] *** jhuska has quit IRC [19:12:09] *** jhuska has joined #jbosstesting [19:20:02] *** ldimaggi has quit IRC [19:26:56] *** bdlink has quit IRC [19:29:52] <ianbrandt> aslak, CLA signed and approved. Pull requests opened. :) [19:33:09] *** ldimaggi has joined #jbosstesting [19:34:30] *** jhuska has quit IRC [19:42:27] <aslak> ianbrandt, aa, corporate.. :) [19:46:30] *** lightguard_jp has joined #jbosstesting [19:47:14] *** mbg has joined #jbosstesting [19:54:32] *** aslak has quit IRC [20:00:13] *** null4j has joined #jbosstesting [20:02:48] *** null4j has quit IRC [20:32:38] *** kevinpollet has joined #jbosstesting [20:34:08] *** kevinpollet has joined #jbosstesting [20:36:40] *** rbattenfeld has quit IRC [20:37:05] *** rbattenfeld has joined #jbosstesting [20:41:12] *** ALR has joined #jbosstesting [20:48:54] *** galderz1 has joined #jbosstesting [20:48:55] *** galderz has quit IRC [20:48:58] *** galderz1 has quit IRC [20:52:38] *** aslak has joined #jbosstesting [21:01:52] *** mgoldmann has quit IRC [21:06:11] *** kevinpollet has quit IRC [21:10:48] *** ldimaggi has quit IRC [21:16:44] <tommysdk> ALR: you there? [21:16:50] <ALR> tommysdk: Yep. [21:17:28] <tommysdk> ALR: great, I've spent a couple of hours on the sd-59 issue and i'm about to commit [21:18:09] <tommysdk> ALR: do i just commit as usual, and it'll be based of the upstream/sd-58 as specified when the branch was created? [21:18:29] <tommysdk> ALR: i mean push instead of commit [21:18:38] <ALR> tommysdk: Yep, push to your local -59 [21:18:46] <ALR> I'll just cherry-pick it probably. [21:18:51] <tommysdk> cool [21:19:11] <tommysdk> i discovered an NPE @ Node.remove(Query) [21:19:18] <tommysdk> thanks to a test case :) [21:19:27] <ALR> See? :P [21:19:44] <tommysdk> if List<Node> found = get(query) is null, we get an NPE, made it return an empty list instead [21:19:54] <tommysdk> rather than proceed with the loop [21:20:39] <ALR> Better to return an empty list> [21:20:40] <ALR> ? [21:20:51] <ALR> Or just guard the loop to not run if the list is null? [21:21:24] <ALR> According to JavaDoc of get(String name), empty list [21:21:35] <ALR> So I guess that's right [21:21:43] <ALR> And get(Query query) has no JavaDoc now (unless you've put it in) [21:21:47] <ALR> So needs that :) [21:21:49] <ALR> tommysdk: ^ [21:21:58] <tommysdk> ya, don't think we'll expect to get null in return there [21:22:01] <tommysdk> i'll add that :) [21:22:41] *** ldimaggi has joined #jbosstesting [21:22:59] <ALR> Word. [21:23:09] <ALR> See how much simple doc-ing, testing matters? [21:23:28] <ALR> Like even we aren't quite sure the expected outputs/inputs until we do this stuff. [21:23:33] <ALR> Same for precondition checking [21:23:54] <ALR> And clearly stating stuff like "Name is required and must have no spaces, else IllegalArgumentException" [21:24:07] <ALR> IMO it's just the best way to kill NPEs [21:24:10] <ALR> Lots of assertions [21:24:14] <tommysdk> Yeah [21:24:24] <tommysdk> Totally agree [21:24:52] <tommysdk> And null checks are so easily neglected [21:25:33] *** lightguard_jp is now known as lightguard_jp_aw [21:26:01] <tommysdk> the more testing, the merrier ;) [21:26:25] <tommysdk> ALR: check out the stuff at https://github.com/tommysdk/descriptors/commits/SHRINKDESC-59 [21:26:27] [21:26:36] * ALR pulls [21:28:44] <tommysdk> so 19 test cases added, got decent results off cobertura [21:29:24] <tommysdk> did pay some attention not only on overall coverage but different conditions also, so not only "happy-tests" [21:34:16] <ALR> tommysdk: Node.java on 99% coverage. Awesome [21:34:16] *** mhuniewicz has joined #jbosstesting [21:34:36] * tommysdk thumbs up [21:35:06] <tommysdk> ALR: The last percent is the hardest? :) [21:35:11] <ALR> tommysdk: Forget it. [21:35:26] <tommysdk> :) [21:35:44] <ALR> So I'm wondering about Surefire forkMode always [21:35:47] <ALR> And if we really need that [21:36:06] <OndrejZizka> Where can I find max? [21:36:10] <ALR> With Ralf's new generated tests, 1700ish of 'em, build takes 5 minutes. [21:36:20] <ALR> But if I put surefire to fork once, that goes to 18s [21:36:39] <mhuniewicz> aslak, hi there :) [21:36:44] <ALR> So do we *really* need VM isolation here? [21:37:16] <ALR> tommysdk: Upstream https://github.com/shrinkwrap/descriptors/tree/SHRINKDESC-58 [21:37:17] <jbossbot> jira [SHRINKDESC-58] Refine SPI for locking to Beta-level [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKDESC-58 [21:38:29] <tommysdk> ALR: Without deeper knowledge about the descriptors internals, I would think we do not need VM isolation [21:40:58] <aslak> mhuniewicz, heya [21:41:07] <mhuniewicz> aslak, any luck with that thing? [21:41:31] <aslak> mhuniewicz, sorry, haven't gotten to it yet. coffee machine needed fixing.. :) [21:41:31] <jose_freitas> how can I tell MavenDependencyResolver to a repo other then http://repo1.maven.org/maven2 ? [21:41:42] <aslak> all dev stops without coffee hehe [21:42:12] <jose_freitas> to use* a repo? [21:42:16] <mhuniewicz> aslak, I recommend yellow bourbon, great coffee. [21:42:20] <mhuniewicz> I discovered it lately. [21:42:29] <mhuniewicz> That's a kind, not a brand. [21:43:00] <aslak> mhuniewicz, brazil.. hmm, gotta check that out [21:43:24] <mhuniewicz> Low in acidity. :) [21:43:28] <mhuniewicz> Smells like heaven. [21:49:21] <mhuniewicz> aslak, is there some sort of deadline for that functionality so that it makes it with the next release? [21:50:22] *** kevinpollet has joined #jbosstesting [21:51:22] <aslak> mhuniewicz, the sooner the better.. :) [21:51:22] *** kevinpollet has quit IRC [21:51:51] <aslak> mhuniewicz, your getting this: java.lang.IllegalStateException: A request must be associated with the context in order to load the known conversations ? [21:52:14] <mhuniewicz> aslak, I think so. I'll open it now. [21:52:43] <aslak> mhuniewicz, i get the same, but that is in the After phase. it obviously can't handle failing in Before. need to look at that [21:53:14] <aslak> but the exception that is shown before it fails in After is: java.lang.RuntimeException: Could not inject members Caused by: javax.persistence.PersistenceException: No Persistence provider for EntityManager named null [21:53:18] <ALR> tommysdk: Know how to squash commits yet? [21:53:28] <aslak> mhuniewicz, hibernate nor hsql is defined in the pom as deps ? [21:53:46] <mhuniewicz> aslak, that's the topmost one. The root cause is IllegalStateException. [21:54:16] <aslak> mhuniewicz, no, they are two different exceptions [21:54:59] <aslak> IllegalStateException happens from here: at org.jboss.arquillian.test.impl.EventTestRunnerAdaptor.after(EventTestRunnerAdaptor.java:103) [21:55:08] <mhuniewicz> aslak, okay. I'll check the configuration. [21:55:13] <aslak> the other ones happen here: at org.jboss.arquillian.test.impl.EventTestRunnerAdaptor.before(EventTestRunnerAdaptor.java:95) [21:55:34] <mhuniewicz> How can you see that? [21:55:38] <aslak> stacktrace [21:56:00] <aslak> and IllegalStateException does not have a Caused by: IllegalState... [21:56:03] <jose_freitas> I used goOffline() so it didn't need to conf repo [21:56:22] <aslak> mhuniewicz, if that was the case, IllegalState would be the Cause of the PersistenceException, but it's not [21:57:24] <aslak> jose_freitas, you can define a pom file it wil read for, or you can set the settings.xml file to use [21:57:29] <tommysdk> ALR: Squash? No, what does it mean? [21:57:53] <ALR> tommysdk: Looky: https://github.com/shrinkwrap/descriptors/commit/22dfbf9023a397166f220de419fb831b28f22a33 [21:57:54] <jbossbot> git [descriptors] 22dfbf9.. Tommy Tynja [SHRINKDESC-59] Bulk test coverage and fix bugs in Node SPI [21:57:55] [21:58:09] <ALR> tommysdk: I took your 5 commits and made them into one. [21:58:15] <ALR> And rewrote the commit message. [21:58:25] <tommysdk> ALR: Oh, cool! How do I do that? :) [21:58:35] <ALR> tommysdk: Switch into your branch [21:58:54] <aslak> ALR, tommysdk , it's the whole.. upstream doesn't care about your oops, forgot to add.. .. ;) [21:59:29] <ALR> tommysdk: "git log" should show you your stuff. [21:59:31] <aslak> not yours personally.. bu tin general hehe [21:59:35] <ALR> You'll see 5 commits. [21:59:38] <ALR> So: [21:59:44] <ALR> git rebase -i HEAD~5 [21:59:54] <ALR> Issues an interactive rebase on the past 5 commits. [21:59:58] <tommysdk> I have 8 commits [21:59:58] <ALR> It'll launch an editor. [22:00:05] <ALR> 8?! [22:00:16] <tommysdk> ALR: Yeah, 5 of mine and three of yours? [22:00:40] <jose_freitas> aslak: thanks, I'll try loading a pom [22:00:52] <ALR> tommysdk: Oh, if you scroll, you'll continue all the way down. I just meant your 5. [22:01:05] <tommysdk> ALR: Ah, then I'm with you, cool [22:01:12] <ALR> So after it launches an editor, you'll see a list of the 5 you chose. [22:01:19] <ALR> They're defaulted to "pick" [22:01:29] <ALR> Leave the first as pick, then change the rest to "squash" [22:01:34] <ALR> And save and close the editor [22:03:04] *** ldimaggi has quit IRC [22:03:20] <tommysdk> ALR: Successfully rebased and updated refs/heads/SHRINKDESC-59 [22:03:21] [22:05:20] <mhuniewicz> aslak. I fixed the first exception and the other one magically disappeared. [22:07:00] <aslak> mhuniewicz, yea.. the second one comes because, some weld setup that happens in the Before Phase after the Injection was not executed due to the exception, so when it comes to the After Phase the state is not as it should be.. [22:07:05] <tommysdk> ALR: So how to edit the commit message? [22:07:25] <mhuniewicz> aslak, I see. [22:07:28] <aslak> mhuniewicz, tho it should not fail, i'll look into that.. [22:08:00] <aslak> mm.. now we're talking: http://rtomayko.github.com/git-sh/ [22:08:27] <mhuniewicz> However, this fix involves adding dependencies - one for the database which is fine I think, but the other one is Validation provider and Hibernate Entity Manager with the fix by ALR. [22:08:48] <aslak> Validation provider ? [22:08:58] <mhuniewicz> Yeah, a JSR one. [22:09:15] <aslak> needed by hibernate ? [22:10:04] <mhuniewicz> Yes. [22:11:48] <ALR> Sorry /me on a concall [22:11:55] <ALR> Will ping back when off [22:18:38] *** ldimaggi has joined #jbosstesting [22:24:56] <mhuniewicz> aslak, I committed this to my github. What do you think about it? [22:27:37] *** rbattenfeld has left #jbosstesting [22:28:40] *** alesj1 has joined #jbosstesting [22:28:40] *** alesj has quit IRC [22:28:41] <aslak> mhuniewicz, shouldn't really release with a SNAPSHOT dep on hibernate. they seem to have a release 3. aug which should include the fix [22:28:53] <aslak> mhuniewicz, tho, it's just for test, not exposed to the user [22:29:12] <mhuniewicz> aslak, I prefer not releasing final versions with SNAPSHOT dependencies. [22:29:36] <mhuniewicz> I don't even know if it would work for you - it's pulling my local jar I think. [22:29:49] <aslak> mhuniewicz, i agree [22:30:09] <mhuniewicz> aslak, but is it in general okay to add Hibernate and HSQLDB as test deps? [22:32:33] <aslak> mhuniewicz, yea, is there another way to test it ? [22:33:06] <mhuniewicz> aslak, not really. There must be some persistence provider, same goes for DB and validation. [22:33:39] <aslak> mhuniewicz, right [22:33:59] <aslak> we're all about real testing right.. so we use a real provider.. makes sense [22:34:00] <aslak> :) [22:34:11] <mhuniewicz> Yea. :) [22:34:21] <mhuniewicz> Okay, I will add some more tests then. [22:34:27] <aslak> cool [22:34:35] <mhuniewicz> aslak, does it mean we are waiting with this until Hibernate 4 comes out? [22:36:58] <aslak> mhuniewicz, yea, or in theory use another provider for now.. [22:37:34] <mhuniewicz> aslak, do you know of one that would not have that problem Hibernate did? [22:37:53] <aslak> mhuniewicz, no idea.. :) [22:37:58] <aslak> need to test [22:38:17] <aslak> and report issues upstream if they don't.. :) [22:40:03] <mhuniewicz> aslak, any suggestions? : [22:40:05] <mhuniewicz> :) [22:40:29] <aslak> eclipselink ? [22:40:52] <mhuniewicz> Okay, I'll play with that. [22:41:05] <mhuniewicz> Thanks for help. I'll be off now. I'll let you know once I get somewhere. [22:41:07] <mhuniewicz> Cheers. :) [22:41:21] <aslak> mhuniewicz, thanks for the work.. :) cheers [22:41:42] *** mhuniewicz has quit IRC [22:50:08] <ALR> tommysdk: Still here? [22:50:11] <ALR> Off the cust call. [22:50:22] <tommysdk> ALR: Yep [22:50:50] <ALR> tommysdk: OK, where'd you leave off? [22:51:09] <ALR> After you save/exit the first editor, it'll pop up another editor asking you to rewrite the commit message [22:51:24] <tommysdk> ALR: That's where I'm at atm. [22:51:52] <ALR> tommysdk: OK, so basically delete or comment all lines you don't want [22:52:07] <ALR> And you can edit the commit message as you like. When you save/exit, the rebase will be done [22:52:23] <ALR> And you can "git log" and see that your 5 commits are now 1, with the updated message [22:54:05] <tommysdk> OK, So I have written a new commit message, I'll do the save now. [22:54:40] <tommysdk> ALR: Yes, I see it. Cool! [22:54:48] * tommysdk likes squash [22:55:26] <tommysdk> ALR: So I guess I can pushed them up to my fork as well [22:55:33] <tommysdk> ALR: And it will be reflected on github [22:55:38] <ALR> tommysdk: Yup. Course you'll need to rewrite. [22:55:43] <ALR> And I already squashed it [22:55:46] <ALR> And put it upstream [22:55:50] <tommysdk> Ah ok [22:55:53] <ALR> So you don't really need to [22:56:01] <ALR> But yep, that's one element of interactive rebasing [22:56:07] <ALR> Very, very powerful [22:56:10] <tommysdk> But I know how to do it next time :) [22:56:14] <ALR> Yep. [22:56:21] <ALR> So what I generally do: [22:56:31] <ALR> Work on stuff....commit really frequently [22:56:44] <ALR> Push to my GitHub remote for backup [22:56:53] <ALR> Then when it's done, I squash it al down [22:56:58] <ALR> So I get frequent save points [22:57:17] <ALR> But still can keep single commits to make stuff easier to read and sort through (also it's atomic that way) [22:57:27] <tommysdk> ALR: Sounds like a very reasonable workflow [22:57:56] <tommysdk> ALR: Thats what I really like with Git, makes it much easier to work in small, incremental steps [22:58:19] <ALR> tommysdk: No Gravatar for you? [22:58:38] <tommysdk> And for each day I work with Git, the more I like it, and the more I dislike Subversion, which is the only other version control system I've worked with [22:59:06] <tommysdk> ALR: Haven't prioritized it ;) [22:59:14] <ALR> Hehe [22:59:25] <ALR> tommysdk: Let me see what else I can find for you. :) [22:59:26] <tommysdk> I like cats :P [22:59:55] <tommysdk> ALR: Sure, I have an hour or two tomorrow as well [23:01:31] <ALR> Awesome. [23:01:45] <ALR> I'll continue w/ my SPI cleanup and I'm sure some stuff will come up. [23:02:07] <tommysdk> ALR: Sounds great [23:04:51] *** ldimaggi has quit IRC [23:06:03] *** lightguard_jp_aw is now known as lightguard_jp [23:09:13] <tommysdk> Browsing through the Shrinkdesc JIRA [23:09:55] <tommysdk> ALR: Noticed I've made some work on SHRINKDESC-18 about 10 days ago, care to throw an eye on it? [23:09:57] <jbossbot> jira [SHRINKDESC-18] import should close stream [Open (Unresolved) Feature Request, Minor, Unassigned] https://issues.jboss.org/browse/SHRINKDESC-18 [23:10:10] <ALR> Oh [23:10:11] <ALR> Sure [23:10:25] <ALR> tommysdk: Was there a pull request? [23:10:45] <tommysdk> ALR: No, just a commit on my fork [23:10:52] <ALR> k, looking [23:11:25] <tommysdk> ALR: As I didn't exactly know what Dan was looking for [23:11:26] <ALR> Hm, mergey conflicts abound. [23:11:41] [23:11:48] <ALR> Silent swallow. :) [23:12:18] <tommysdk> ALR: Hehe, don't remember, but propably as I just threw it out there :) [23:12:33] <tommysdk> tommysdk: those things tend to last though [23:13:05] <tommysdk> ALR: Do you have any thoughts on SHRINKDESC-31? [23:13:06] <jbossbot> jira [SHRINKDESC-31] Descriptors clobbers file markup and formatting [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/SHRINKDESC-31 [23:13:15] <ALR> Jeez, one at a time! [23:13:17] <ALR> :D [23:13:23] <tommysdk> haha :D [23:14:04] <ALR> tommysdk: First off, for -18, I'm not even sure that's valid. [23:14:14] <ALR> Whoever opened the stream should close it. [23:14:38] <tommysdk> ALR: Agree [23:14:44] <ALR> But Dan has a point. [23:14:58] <ALR> It makes more work for the user, and try{catch} nonsense they shouldn't have to do [23:15:09] <ALR> When pragmatically they just wanna get the contents [23:20:49] *** gastaldi has joined #jbosstesting [23:20:57] <ALR> Shit. [23:20:58] <tommysdk> ALR: Yeah a decision has to be made there. As long as it is always OK/expected that the stream should be closed I think it's fine (and as long as it is documented). Losing the boilerplate try/catch is very appealing imo. [23:21:01] <ALR> I can't decide. [23:21:10] <tommysdk> Hehe [23:21:14] <ALR> Because like, look: [23:21:15] <tommysdk> Are we in a rush? [23:21:25] <ALR> We delegate to this: Document doc=builder.parse(stream); [23:21:38] <ALR> Now, how is DOM any different? [23:21:44] <ALR> It doesn't close the stream. [23:21:55] <ALR> I think we should not close it. [23:21:59] <ALR> Some streams are resettable. [23:22:05] <ALR> Meaning the client can reuse it. [23:22:17] <ALR> And they're the ones who opened it. [23:22:27] <gastaldi> Wise decision [23:22:32] <ALR> I usually opt for correctness/symmetry over convenience. [23:22:48] <ALR> So I think I'll reject 18. Cool? [23:22:53] <tommysdk> Sure [23:23:04] <ALR> gastaldi: Thanks for the input. [23:23:07] <tommysdk> Then we can just close the issue [23:23:27] <tommysdk> It's going to be a pain otherwise, if different builders treats the streams differently [23:23:53] <ALR> There is another option. [23:24:08] <tommysdk> As long as users know what to expect [23:24:11] <ALR> from(InputStream in,boolean close) [23:24:34] <ALR> gastaldi: tommysdk; ^ WDYT? Overkill? [23:24:39] <gastaldi> yeah [23:24:46] <gastaldi> unecessary [23:24:53] <ALR> gastaldi: Keep in mind it reduces user boilerplate [23:25:05] <ALR> A mission of these projects is to keep user code small/lean. [23:25:20] <tommysdk> Hmm [23:25:23] <gastaldi> hummmm so this method would only be called if the user wants to keep the stream open [23:25:29] <gastaldi> let close be the default then [23:25:40] <ALR> Yes, closed to be the default. [23:26:03] <tommysdk> Sounds good [23:26:08] <ALR> But let 'em keep it open so they can maintain the control they deserve [23:26:13] <ALR> OK, let's do that. [23:26:18] <tommysdk> Personally I hate boilerplate try/catch [23:26:22] <gastaldi> well, there always an option of creating a subclass of InputStream that close() does nothing [23:26:25] <gastaldi> to prevent that [23:26:33] <tommysdk> Actually, I don't really like checked exceptions at all [23:26:35] <ALR> tommysdk: Wanna prep a new commit for that, but base it off SHRINKDESC-58? [23:26:37] <jbossbot> jira [SHRINKDESC-58] Refine SPI for locking to Beta-level [Open (Unresolved) Task, Major, Andrew Rubinger] https://issues.jboss.org/browse/SHRINKDESC-58 [23:26:38] <gastaldi> but allright [23:26:40] <tommysdk> But thats a different story [23:26:48] <ALR> I'll rewrite the JIRA [23:26:56] <tommysdk> Sure thing [23:26:59] <gastaldi> Yeah, close the MF :) [23:27:54] *** bleathem is now known as bleathem_mtg [23:29:57] <ALR> Assigned and commented SHRINKDESC-18 [23:29:59] [23:31:23] <ALR> tommysdk: Ah good, aslak gave us a case for 31 [23:32:10] <tommysdk> ALR: Doesn't sound that complicated, or I'm I wrong? [23:32:19] *** jose_freitas has quit IRC [23:32:54] <ALR> tommysdk: For impl, I'm not sure yet [23:32:57] <ALR> I'm looking at it. [23:33:06] <ALR> Depends where we lose the comments. [23:33:26] <ALR> Perhaps via the DOM import. Perhaps because our own object model doesn't support it in Node. [23:33:43] <aslak> ALR, we don't lose them, we can't export them [23:34:22] <aslak> they are imported as a Node with name #comment, and attempted exported as a element with the same name.. but #comment is a illegal name for a element [23:34:50] <ALR> What's importing the Node name #comment? [23:34:59] <aslak> the simples fix is in the exporter, if(source.name().equals("#comment")) target.createComment(source.text()) [23:35:03] <ALR> aslak: I just brought in your test case. Am running it now and will try at a fix. [23:35:08] <ALR> That's hacky. [23:35:17] <aslak> ALR, i know.. [23:35:22] <aslak> the Importer is doing that [23:35:33] <ALR> I'm gonna pay; thanks for the test case. [23:35:36] <ALR> *play [23:35:39] <aslak> creating the name i mean.. a CommentNode has the name #comment form the XML API [23:36:14] <ALR> So then on import we should reflect that in our model I think. [23:36:32] <ALR> If #comment is illegal element name during export, but that's what we get on import... [23:37:17] <aslak> or # is a illegal char :) [23:37:33] *** gastaldi has left #jbosstesting [23:37:58] <ALR> Right [23:38:55] *** jose_freitas has joined #jbosstesting [23:39:24] <ALR> Yeah, comes from org.w3c.dom.Node.getNodeName(): #comment [23:39:31] <aslak> yea [23:41:45] <tommysdk> I'm in desperate need of sleep :) [23:43:33] <tommysdk> ALR: If sd-38 requires some more work, put a comment there and I can have a look at it tomorrow, if needed [23:45:07] <tommysdk> Good night all [23:45:11] <ALR> tommysdk: Sleep tight. [23:45:13] <ALR> SHRINKDESC-38 [23:45:15] <jbossbot> jira [SHRINKDESC-38] Support Descriptor.name [Closed (Done) Feature Request, Major, Aslak Knutsen] https://issues.jboss.org/browse/SHRINKDESC-38 [23:45:15] <aslak> tommysdk, night ! [23:45:28] *** tommysdk has quit IRC [23:45:40] <ALR> Huh? 38 is closed. [23:46:15] <aslak> ALR, probably sd-58 [23:47:13] <ALR> Yeah [23:47:30] <ALR> aslak: I think your test case is flawed... [23:48:19] <ALR> Hmm, maybe not. [23:48:21] <ALR> One sec [23:48:21] <aslak> ALR, it's not a 100% proper testcase no.. it just assumes it's not going to fail [23:48:41] <ALR> No, it's very illlustrative. I messed something up in my fix. No worries. [23:49:05] <aslak> needs to assert on the state of the export, to verify the comment is still there