[20:51] Action: Dymaxion heads for the Cafe via eir house (gotta get my power brick) [20:56] howdy, aspa [21:04] hiya! [21:07] getting ready to head to the cafe where I'm meetingg paul ^^ [21:07] bbiab! [21:10] awesome! :) see y'all soon [21:28] Action: Dymaxion just grabber eir power brick, will be there momentarily [21:36] ok, here...I forgot my power tho :( [21:36] my batt's almost dead [21:37] bummer. :( [21:38] and paul doesn't have matching power [21:39] so, i put a file up on Zettai: /usr/local/src/trike/Information-Data Formats-XSD-BEL.7.mcz [21:39] i think it will install into the current Trike release [21:39] oki [21:39] which should be set up as a dev image already [21:42] Action: Dymaxion gets online. [21:43] oh, so I was queried today about trike [21:44] and whether it could be used for production [21:44] production of what? [21:44] threat models? [21:44] it doesn't do import [21:45] I think it would be a good idea if we got the limitations of v1 documented like we were planning [21:45] so, i dunno. i might say no. [21:45] yah. [21:45] production...like a production environment [21:45] that would be good [21:46] are you suggesting that we do that tonight? [21:46] :/ [21:46] no [21:46] but we should probably do it next week [21:46] Action: Dymaxion grins [21:46] ok. let's do it then. [21:46] Dym, did you see the file to get off Zettai? [21:47] it's /usr/local/src/trike/Information-Data Formats-XSD-BEL.7.mcz [21:47] yeah [21:48] you will have to load it via the monticello browser [21:50] which package do I want? [21:51] er, right, information-data-formats; what are the information diagram packages? [21:52] we aren't using them yet [21:52] ok [21:52] they are the data model for DFDs & Entity Modeling [21:52] Action: DymAra nods [21:52] sorry i'm slow [21:52] ok, loaded [21:52] ok [21:52] so now the question is, what to work on [21:52] i still don't remember my design [21:53] you could start by reading the XSDClassFactory commene [21:53] er, commene [21:53] argh [21:53] commenT. [21:53] ok [21:54] i had to comment some stuff out of XMLNamespaceResolver class>>initialize [21:54] to get it to load [21:58] ahhh [21:58] ok. the place to start is actually XMLNamespace [21:59] read that comment next [21:59] then i would create an XMLNamespaceTest [22:00] you can use XSDClassFactoryTest as an example [22:00] to see if it worked, pop up a test runner from the open menu & go to Information-Data For... whatever I can't see [22:02] the next thing I would do is get the content of all the examples in the spec into XMLNamespaceTest so you can run some tests [22:03] this would be like the section non-normative tests in XSDClassFactoryTest (class side) [22:03] there's a naming convention in the XSDClassFactoryTest commene [22:03] T [22:06] are y'all still there? [22:09] yes, we're doing as our dictator is demanding [22:09] ;) [22:09] oh, ok. :) [22:10] i'm working on figuring out how namespaceresolver works [22:10] so don't do that part, do something else :) [22:10] generateApplicableFacetTests in XSDClassFactoryTest looks rather intimidating [22:10] yah [22:10] where did it originate from [22:10] i wrote it [22:10] oh, you mean the data [22:11] it came from the Primer [22:11] XSD part 0 or whatever [22:11] ok [22:11] hrm [22:11] we're conf00zled [22:11] ok, which part? [22:11] and it's unclear what you're expecting [22:11] all of it [22:11] ok [22:12] i tried to get more organized but it is just really big [22:12] Action: DymAra nods [22:12] so, what i'm hoping for tonight is that you can write some test cases for XML namespaces & the resolution thereof [22:13] I thought we were going to be having fun :P [22:13] meantime i'm trying to figure out which part of the namespace stuff is supposdd to go where so i can help walk you through writivg that stuff next time [22:13] ok [22:13] what would make it fun? [22:14] we'll make our own fun [22:14] ok.. [22:15] basically, if you can get the data from the namespace spec into the form of test cases, that would be swell [22:15] the test generation section of XSDClassFactoryTest is scary & should probably be ignorded [22:16] the point is to get any XML snippets into the class side somehow, so we can ask for them when we run the tests [22:16] & then the tests themselves will go on the instance side [22:16] (of XMLNamespaceTest) [22:16] did that help? [22:19] eah [22:19] yes, thanks ^^ [22:30] one test idea: XMLNamespace withUri: foo should == XMLNamespace withUri: foo [22:30] (replace foo with a URI) [22:44] ok. here are the rest of my notes about XMLNamespace: [22:44] XMLNamespace>>perElement: needs to issue the same instance every time it is asked with the same argument [22:44] see XMLNamespace class>>withUri: for a similar idea [22:44] test needed to confirm this as well [22:44] make sure you just hand them out for nil [22:45] if elementType is changed from nil to some type at some point in the future, it should get registered with the same-uri global space [22:45] what should happen if it's already there? [22:45] initializePerElementPartition should read: [22:45] contents at: #perElementAttributes put: Dictionary new [22:45] fix class comment: uri can't be nil, only '' (empty) [22:45] get a perElement XMLNamespace, then compare its allElements and globalAttributes to the one you got it from -- they should be == [22:45] write atAttribute: [22:45] (which name spaces should it look in, in what order? will require reading spec) [22:45] help! [22:45] how do you escape '? [22:45] 'if you don''t like this string, it''s too bad' [22:45] gimnee [22:46] thx ^^ [22:46] is that what you're looking for? [22:46] :) [22:46] (those are doubled single quotes) [22:46] yeah, I got it :) [22:46] and yes, that's what we were looking for [22:46] thx ^^ [22:59] ok, now i'm hacking at XMLNamespaceResolver. so don't change it. [22:59] uh, ok, we'll roll back all of our changes [22:59] ;-) [23:00] ok, we're done getting the examples in. [23:00] now what? [23:02] cool :) [23:03] earlier ii pasted my notes about other changes i think need to be made [23:03] e.g. XMLNamespace>>perElement: is wrong [23:04] that would be the next set of things to do [23:04] oh. [23:04] except, have you written tests for the examples yet, or just got the examples in as class-side methods? [23:05] so, fixing the code there, not writing tests? [23:05] now, we just have the examples in. [23:05] oh, ok. [23:05] better do the test writing part [23:05] so, pick an example method & paste it in here [23:05] not sure where to go from here; neither of us understand what you were doing with those test [23:05] & i will give you advice about how to get goivg [23:06] [23:06] [23:06] [23:06] oh. dang [23:07] you need the resolver for that one [23:07] oh! [23:07] you're instilling us with the utmost confidence, dictator-maam [23:07] hee hee [23:08] well, hopefully you knew what you were in for when you headed for this 2-bit island [23:08] Action: DymAra peels another banana [23:09] ok. do that list i pasted earlier, modifying actual code, while i figure out what those tests should look like [23:09] ok [23:11] we are hosed, all those tests from the spec require stuff that we haven't written or figured out yet [23:12] awesome [23:12] can you explain what this is doing? : [23:12] withUri: aURI a new or existing instance of me, with the given URI." [23:12] newInstance | [23:12] URI asString = '' [23:13] well, that' didn't work [23:13] what does the withUri method do? [23:13] we can't read it [23:13] :) [23:16] or at least we think it's already doing what you said you want it to do [23:18] oh. withUri: is great [23:18] it's perElement: that needs to do likewise [23:19] so, perElement: is supposed to create a child of this namespace, for a particular element [23:19] except, if that particular element already has one, it should answer that one instead of a new one [23:20] so the namespace needs to keep a list, indexed by element, of its children [23:20] this will mean adding a new instance variable to the namespace [23:20] wtf is a partition?\ [23:21] see A.2 in the spec [23:21] or are you asking about the mathematical concept? [23:21] I dunno, your comments talks about partitions [23:21] and it's unclear what that means [23:22] (in PerElement) [23:22] oh. [23:22] yah, it's the A.2 sense [23:22] basically every [23:22] ns spec? [23:22] xmlns is really 3 namespaces in the normal sense [23:22] yeah [23:23] sometimes only 2 of them are valid, sometimes (if it's a per-element namespace) they all are [23:24] so what's in perElement now is what you would do if you didn't already have such an element in your list of children [23:25] er, i mean if you had nothing for aClass in your list of children [23:32] Steph is tired and not feeling so well, so I'm going to head back to my place; bbia few. [23:32] ok.. [23:32] see you shortly [23:32] bye, ara! [23:32] feel better [23:43] ok, back [23:44] whatcha got? is it working? [23:45] just sat down again [23:45] the partition stuff is sort of breaking my head; I understand what they're trying to do, but the definition of what each set contains is ... unclear. [23:46] yah [23:46] it's pretty icky [23:46] basically, the sets are used by XSDClassFactory & XMLNamespaceResolver [23:47] do you want to just talk on the phone? [23:48] yeah, give me a minute [23:50] (getting set up to work on my desktop) [23:52] ok, call me? [23:52] (home line) [23:53] hrm. There's a newer version of the namespace spec than the one you had linked [23:54] oops; try again? [23:56] http://www.w3.org/TR/xml-names11/ [00:00] --- Thu Mar 30 2006 [00:09] oh, except it's after 10pm & i have an 8am tomorrow [00:09] so i have to go [00:09] :( [00:10] ok [00:10] can you stick your code up on Zettai? [00:13] yeah, let me clean it up a little [00:14] ugh, it's been too long since I wrote any code. [00:15] How do I file out changes again?: [00:16] monticello browser [00:16] select Information-blah blah blah [00:16] then select a repository & hit save [00:16] it'll prompt you for comments [00:19] right [00:19] done [00:20] awesome. [00:20] permission denied [00:20] oh, right [00:20] hang on, I'll fix it [00:21] ok [00:22] got it [00:22] i will integrate & put one back [00:22] yay [00:22] there may be some noise -- look at it before you merge [00:22] (but not right this instant -- i'm going to bed) [00:22] ok [00:22] will do [00:23] sleep well [00:23] Action: Dymaxion heads off to read the XSD spec^W^W^W^Wsleep [00:23] heee hee [00:23] g'night