#gemrb@irc.freenode.net logs for 9 Feb 2015 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[02:10:23] <-- turtleman_ has left IRC (Ping timeout: 245 seconds)
[03:18:36] --> Eli2 has joined #gemrb
[06:03:51] <-- Lightkey has left IRC (Ping timeout: 245 seconds)
[06:15:52] --> Lightkey has joined #gemrb
[09:00:08] <edheldil> Good Monday morning, guys
[09:15:22] <lynxlynxlynx> oj
[10:09:10] <-- edheldil has left IRC (Remote host closed the connection)
[11:13:59] --> edheldil has joined #gemrb
[11:13:59] --- ChanServ gives channel operator status to edheldil
[13:15:30] --> turtleman_ has joined #gemrb
[13:21:10] <lynxlynxlynx> http://devopsreactions.tumblr.com/post/110529123748
[13:21:14] <Pepelka> Lockess algorithm - DevOps Reactions
[13:21:15] <Pepelka> »by Lwahonen«
[14:56:27] <-- turtleman_ has left IRC (Remote host closed the connection)
[14:57:05] --> turtleman_ has joined #gemrb
[16:33:48] <-- turtleman_ has left IRC (Ping timeout: 264 seconds)
[16:34:26] --> fizzle has joined #gemrb
[18:25:18] --> kpederse1 has joined #gemrb
[18:27:49] <-- kpedersen has left IRC (Ping timeout: 265 seconds)
[19:51:49] <Pepelka> [commit] fizzet: delete MessageWindow controls before storing others https://github.com/gemrb/gemrb/commit/c1fcf052689b411c3f2ec50f43e8223be5a7f7f1
[19:51:50] <Pepelka> [commit] fizzet: convert a few more loops to using iterators https://github.com/gemrb/gemrb/commit/eba625818b12fb0cccf210fb10e189088abb277d
[19:51:52] <Pepelka> [commit] fizzet: move Interface::GetControl to GUIScript https://github.com/gemrb/gemrb/commit/9de77d0c260d274c6cf382b667f3725560abdc9a
[20:04:57] --> turtleman_ has joined #gemrb
[20:07:21] --> edheldil_ has joined #gemrb
[20:08:48] <lynxlynxlynx> hmm, is that something new? That control ids can change unpredictably
[20:17:11] <edheldil_> cre22 might have fields for songs and effects, but I can't find a cre in iwd2 using it ....
[20:17:38] <edheldil_> Unless it's in some cre embedded in another object
[20:17:52] <fizzle> lynx: no, the IDs don't change
[20:17:58] <fizzle> biut indices do
[20:18:08] <lynxlynxlynx> ok, phew :)
[20:18:27] <fizzle> and the python side uses them a lot
[20:18:45] <edheldil_> sure, indices are something we assign as we insert the widgets to windows
[20:19:06] <fizzle> I was tempted to do away with the indices entirely...
[20:19:14] <edheldil_> but do we use them a lot?
[20:19:32] <lynxlynxlynx> i don't understand that commit then
[20:19:43] <fizzle> except for the first lookup the python code uses them all the time
[20:19:58] <lynxlynxlynx> deleting a subcontrol can't break the parent
[20:20:08] <fizzle> if you do control = win.GetControl(x)
[20:20:23] <fizzle> control is the index
[20:20:32] <lynxlynxlynx> edheldil_: perhaps only for pcs (bards)
[20:21:13] <lynxlynxlynx> isn't that the index from the chu?
[20:21:45] <fizzle> well, yes, but if you remove controls (like in the commit) indices from the CHU change
[20:21:47] <lynxlynxlynx> we have so many places using sparse indices/ids, that's why i'm confused
[20:21:57] <lynxlynxlynx> ouch
[20:22:11] <lynxlynxlynx> so it's better to just hide them if you don't want trouble
[20:22:33] <lynxlynxlynx> ok
[20:22:37] <fizzle> yes, but lookup using the index is quite a bit faster
[20:22:48] <fizzle> I suppose that's why python uses them so much
[20:22:59] <lynxlynxlynx> initially i thought your index was refering to the vector index
[20:23:06] <fizzle> it is
[20:23:20] <lynxlynxlynx> yes, but it is also the same as the id
[20:23:35] <fizzle> no, not necessarily
[20:23:38] <lynxlynxlynx> initially
[20:24:03] <fizzle> I don't think controls are always in order in the chu
[20:24:21] <fizzle> and we have labels with gigantic ids
[20:24:27] <fizzle> so, no
[20:25:26] <fizzle> it's a bit confusing, that's why I was tempted
[20:25:57] <fizzle> treating the index as an internal implementation detail gives a much nicer interface
[20:26:27] <fizzle> but there's no denying it's going to be slower
[20:26:51] <lynxlynxlynx> labels just have an extra bit to demark them from others
[20:27:00] <lynxlynxlynx> since they share their ids often
[20:27:17] <edheldil_> ... for whatever silly reason ....
[20:27:48] <fizzle> indeed, but that obviously means that ID possibly != index
[20:27:48] <lynxlynxlynx> yeah, you could map controls to each other with an extra field in the structure
[20:28:17] <edheldil_> but I am sure there are some CHUs with holes in indices
[20:28:39] <lynxlynxlynx> definitely
[20:32:03] <fizzle> I think that bug was simply exposed now by brada's change to SubstituteForControl
[20:41:34] <-- fizzle has left #gemrb
[21:03:44] <edheldil_> lynxlynxlynx: pushed cre22 parser
[21:03:57] <lynxlynxlynx> thanks!
[21:15:52] <-- turtleman_ has left IRC (Ping timeout: 244 seconds)
[21:55:17] <-- Eli2 has left IRC (Remote host closed the connection)
[23:23:49] <lynxlynxlynx> i keep finding test cases i can reliably trigger outside the testsuite, but not inside it :<
[23:23:59] <Pepelka> [commit] lynxlynxlynx: iwd2: fixed the seemingly heisen obscuring of the last line of text https://github.com/gemrb/gemrb/commit/da351162635305e5943473033c74a1e12a9ec50e
[23:24:00] <Pepelka> [commit] lynxlynxlynx: iwd::cg: fixed crashing and most spacing (spaces after : still ignored) https://github.com/gemrb/gemrb/commit/ea930f46495ee6c80a5c74c5671c9b357b474277
[23:24:01] <Pepelka> [commit] lynxlynxlynx: Tests: align expected and reported output for easier comparison https://github.com/gemrb/gemrb/commit/558efcd8222f42ac9f2330dd59c15b77c589dacd
[23:34:15] <-- edheldil_ has left IRC (Ping timeout: 265 seconds)