| [03:43] <mhall119> jvrbanac: I was wondering where the method docs were being stores, I wonder why it's in retval | |
| [03:44] <mhall119> I suppose we need to figure out how to parse and linkify these gir doc strings | |
| [03:44] <mhall119> looking pretty good so far though | |
| [03:46] <jvrbanac> Yeah. I don't know. I noticed that they were getting stored in the db. So I went searching for where they were getting stored in the model | |
| [03:47] <jvrbanac> pdb to the rescue :D | |
| [03:49] <mhall119> heh, yeah, probably something funky about the GIR file spec | |
| [03:50] <mhall119> anyway, I pushed another revision for some added styling, plus breaking the methods table up based on defining class | |
| [03:51] <jvrbanac> That is much better! My eyes thank you! | |
| [03:54] <jvrbanac> Btw, was there a reason you set that template conditional to the string of "None" | |
| [03:56] <mhall119> yeah, some are hidden if I make it a string, others are hidden if it's the literal None, I'm not entirely sure why just yet | |
| [03:56] <mhall119> I suspect some of them actually have a value of "None" | |
| [03:57] <jvrbanac> weird. I didn't noticed that, ok. | |
| [03:57] <mhall119> anyway, almost midnight here, I'll chat with you tomorrow | |
| [03:58] <jvrbanac> Yep. Have a good one! | |
| [19:58] <mhall119> jvrbanac: gah, these Type records are gonna make everything difficult, they have no direct link to any Node records | |
| [20:00] <jvrbanac> mhall119, You know, I was looking at that the whole model structure and it seemed like it could be cleaned up and simplified a bit. | |
| [20:01] <jvrbanac> It could be because I'm pretty new to that code, but it seemed like it was more complicated then it needed to be | |
| [20:30] <mhall119> jvrbanac: I think that's how C/GObject do things that is making it overly complicated | |
| [20:50] <jvrbanac> mhall119, yeah, oh well. So is there a reason why Type records don't have a direct link to the Nodes? It would seem important to link the two at some point. That way you could easily correlate the return types. | |
| [20:50] <mhall119> jvrbanac: I think it's because of the way C works | |
| [20:50] <jvrbanac> lol ok | |
| [20:51] <mhall119> a Node has a Namespace, which has a version | |
| [20:51] <mhall119> so, PreviewAction is in Unity namespace, version 6.0 | |
| [20:52] <mhall119> but C doesn't really care where the Type is, so it says it's a PreviewAction type, and then it's up to the runtime to decice that PreviewAction is | |
| [20:52] <jvrbanac> got it | |
| [20:52] <mhall119> could be PreviewAction in Unity 5, could be PreviewAction in Unity 6 | |
| [20:52] <mhall119> heck, it could be PreviewAction in FakeUnityLibrary 10.0 | |
| [20:53] <mhall119> so, future work is going to need to have smarter lookups than I have in there now | |
| [20:53] <mhall119> we'll need say "Find PreviewAction in the Ubuntu 12.04 platform definition" | |
| [20:53] <mhall119> and from there know to use Unity 5, not Unity 6 | |
| [20:54] <mhall119> as verbose and tedious as Java is to write, it cerainly made tooling easy | |
| [20:55] <jvrbanac> true | |
| [20:56] <jvrbanac> Brb... moving to a different meeting room :) | |