Clonk-Karl: Provide log for Monday's meeting

2010-03-02T19:46:44Z

Provide log for Monday's meeting

New page

== Summary ==
* Isilkor: [http://hg.openclonk.org/openclonk/rev/19d8ca039e8c OGRE binary loader] nearly done, introduces Boost dependency
* Mimmo_O: Scripted a ring menu, to be used for Workshop-like buildings (selection of things to be built/crafted)
** Probably not appropriate for Magic
* ck: Investigated animation bug ([http://bugs.openclonk.org/view.php?id=142 #142]), nothing sustainable so far
* Reviewed [http://forum.openclonk.org/topic_show.pl?pid=7397#pid7397 Carli's patch] that removes shareware restriction
* Soldat-like mouse scrolling annoying for selecting inventory items with the mouse and in developer mode
** Can it be deactivated?

== Full Log ==
<pre>18:59:14-Mimmo_O:#openclonk-dev- bing bing 1900 report in!
18:59:29 * Mimmo_O !
18:59:43 * ck
19:00:33<!Isilkor> http://www.nosebud.de/~nh/clonk/openclonk-10-03-01-17-30.png
19:01:20<!ck> Isilkor: Is this using DirectX?
19:01:29<!Mimmo_O> plant people!
19:01:41<!Isilkor> no, it's loading from Ogre binary format
19:01:45<!Isilkor> the animation loading code is still screwy
19:02:01<!Isilkor> ignoring the animation, all is well: http://www.nosebud.de/~nh/clonk/openclonk-10-03-01-18-42.png
19:02:17<!ck> cool
19:03:01<!ck> I investigated a bit the remaining animation bug(s)... no results so far however
19:03:27<!Isilkor> Also allow me to remark that the binary format is rather weird
19:03:30<!ck> Guess I'll write a little tool using ogre to show the animation and then compare what it is doing different then clonk
19:03:41<!ck> Isilkor: so no 1:1 equivalence with the XML?
19:03:46<!Isilkor> ck: partially
19:04:02<!Mimmo_O> anyone wants to report his/her work done in the last week?
19:04:17<!Isilkor> But the most interesting part is that the format is chunked, and whether some values are present or not depend on the size of the chunks
19:04:28<!ck> Mimmo_O: I just did. :P
19:04:39<!Isilkor> I as well :P
19:04:57< loriel> I did not do anything, hope this helps
19:05:05<!Isilkor> Oh, also we now have a working 64-bit windows build
19:05:08< Matthi> And I don't want to report anything
19:05:09<!ck> I could also tell you a lot about particle physics I learned last week but I don't guess this is going to help for OC :P
19:05:39<!ck> Isilkor: hm. Can't there be ambiguities that way?
19:05:50<!CoN^Uni> I did not do anything and don't want to report
19:05:59<!Isilkor> ck: what do you have in mind?
19:06:26<!ck> Isilkor: How do I know which value is missing if a chunk is smaller than "normal"?
19:06:34<!Isilkor> the last ones
19:06:46<!Isilkor> also it's only a few chunks that can be smaller
19:06:47<!Mimmo_O> i scripted a ring-menu, which is currently in use for magic, but it seems a bit too slow for that. imo it would fit verry good for workshop-buildings. you click/press a button, the menu pops up, and you can quickly select what you want to build/craft.
19:07:08<!Isilkor> obviously, this is mentioned nowhere in the file format header
19:07:20<!ck> Isilkor: You could write documentation as you go :P
19:07:56<!Isilkor> I might
19:08:28<!Isilkor> But the format is mostly documented, and I've currently explained oddities in comments
19:09:08<!ck> Mimmo_O: you mean it's too slow to use in a hectic game, not the engine being too slow rendering it?
19:09:20<!Mimmo_O> yes
19:09:20<!ck> For magic I liked the combo system in CR...
19:09:41<!Mimmo_O> me too. but i also like how this ring-menu turned out
19:10:02<!Isilkor> Also I'm going to give Sven and Peter aneurysms if/when I check this in
19:10:05<!Mimmo_O> for a lesser amount of spells, this one is still as fast as the combo menu (for me)
19:10:26<!ck> Isilkor: By depending on boost? :P
19:10:35<!Mimmo_O> i think if the spellcount is eight or lesser, it might be as fast and accurate as the combomenu
19:10:59<!Isilkor> ck: By depending on boost in enough places that you can't simply remove the include and a base class, yes
19:11:14<!ck> and adding an operator overload!!
19:11:38<!Mimmo_O> OVERLOAD
19:11:38<!Isilkor> eh, well, that too ;)
19:12:03<!Isilkor> but I like boost::ptr_vector well enough to use it in quite a lot of places.
19:12:24<!ck> Mimmo_O: it does not make sense though to show one menu in certain cases and the other otherwise I think
19:12:33<!Isilkor> and those templates with boost::enable_if are just to annoy people >:(
19:13:17<!Mimmo_O> ck: what? you mean having different menus for different objects/buildings/systems?
19:13:20<!ck> Isn't boost::ptr_vector easily replacable by a normal vector and a destructor cleaning up?
19:13:33<!Isilkor> In principle, yes
19:13:33<!ck> Mimmo_O: No, a different menu for magic depending on the number of available spells
19:13:53<!Mimmo_O> it wasnt my aim to have different menus
19:13:59<!Isilkor> but I don't feel like writing a special try { } catch(...) { /* cleanup */ throw } everywhere
19:14:01<!ck> I figured, just wanted to make clear :)
19:14:05<!Isilkor> so I'm using boost::ptr_vector
19:14:20<!Mimmo_O> as the magic itself is not clear to be the same way as in CR
19:14:32<!Mimmo_O> in CR, you really had lots of spells
19:14:59<!ck> Isilkor: I think it could also be used in StdMesh in a few places
19:15:05<!Isilkor> ck: All of the boost uses are trivially replaceable, but it's quite a lot of work for almost no gain
19:15:18<!ck> But I just relied on exceptions not being thrown and went with std::vector :)
19:15:29<!Mimmo_O> i made a scenario with only 5 spells atm. its playable and for me, it is as funny as many CR-magic-scnearios
19:15:42<!Isilkor> I'm using exceptions a lot while loading the format
19:16:06<!Isilkor> And by "a lot" I mean "when there's errors", not instead of other structured code ;)
19:16:32<!ck> yeah, me too. But I just let ~StdMesh() do the cleanup, so no extra code
19:17:14<!ck> maybe still not wise to end up with an inconsistent mesh if loading fails
19:17:46< loriel> Okay, so now we can replace the c4script parser with a giant pile of boost.spirit
19:18:02<!Isilkor> Well, I'm using a few temporary arrays to buffer chunks while I'm turning it into objects that StdMesh understands :)
19:18:04< loriel> Why should Peter and Sven get all the aneurysms?
19:18:44<!ck> Isilkor: I see. The StdMesh data structures were chosen to pretty much fit the XML ;)
19:19:50<!Isilkor> I'm currently turning the binary format into a rather faithful representation in memory, which then in turn gets converted into the StdMesh stuff
19:20:15<!Isilkor> So the parser is almost self-sufficient
19:20:58<!Isilkor> I've reused the Mesh math code though, becuase there isn't really a point in having multiple vector and quarternion implementations
19:21:02<!Isilkor> -r
19:21:30<!ck> sure
19:22:05<!Isilkor> oh, since feuermonster just mentioned soldat in another channel
19:22:14<!Isilkor> I don't like the soldat-style scrolling we currently have
19:22:40<!Isilkor> Having your viewport move just because you want to select an item with the mouse sucks
19:23:10<!ck> yeah, it's also annoying in developer mode when moving the mouse out of the window to for example select the console window
19:23:32<!Mimmo_O> Isilkor: i think the same
19:23:34< loriel> Yeah, thirded
19:23:38< loriel> or fourthed I guess
19:23:47<!Mimmo_O> its also annyoing if you have a fast mouse
19:24:44<!ck> Is anyone of you annoyed enough to create a forum post about it? :P
--- Log closed Mon Mar 01 19:25:32 2010
--- Log opened Mon Mar 01 19:26:03 2010
19:26:03-!- ck_ [~ck@euirc-da126f32.versanet.de] has joined #openclonk-dev
19:26:03-!- Irssi: #openclonk-dev: Total of 29 nicks [0 ops, 0 halfops, 1 voices, 28 normal]
19:26:17< ck_> 19:24:44 <!ck> Is anyone of you annoyed enough to create a forum post about it? :P
19:26:35<!Mimmo_O> [19:26:54] <!Mimmo_O> sven said it can be disabled, but i havent found a checkbox for that yet
19:26:44-!- ck [~ck@euirc-8756b4a1.versanet.de] has quit [Ping timeout]
19:26:45-!- Irssi: Join to #openclonk-dev was synced in 43 secs
19:26:55< ck_> ah, I didn't know it was optional already
19:27:23<!Mimmo_O> i dont know if it is.
19:28:05< ck_> If Sven says it can be disabled I guess there is a Players.txt entry for it
19:29:38<!Mimmo_O> dont know
19:29:56<!Mimmo_O> ck_: do you want to test the ringmenu ?
19:32:01< ck_> hm, speaking of testing... I also wanted to have a look at Carli's patch, even if it's just to make him happy...
19:32:32<!Isilkor> oh right, I'll look at it too
19:32:36< ck_> Mimmo_O: Do you have a patch?
19:32:59<!Mimmo_O> what? no. i have it build in into a scenario
19:34:17< ck_> It seems to me that the registeration code is still not fully away... though I'm not sure whether we really want to get fully rid of it... just in case ;)
19:34:34< ck_> Mimmo_O: A new scenario that is?
19:35:02<!Mimmo_O> not in the repos and contains five spells, a spellchooser and a spellcaster-object
19:35:38<!Isilkor> ck_: the C4PlayerInfoListBox.cpp changes don't go far enough in my opinion
19:36:07<!Isilkor> he's doing a deep copy of the player image without any gain, because it's no longer changed
19:36:38<!Isilkor> also the facet should be removed if it isn't used anymore
19:36:49< ck_> yeah, I was also wondering... he is removing checks like if(pRef->isRegJoinOnly()) and if(!Config.Registered()), but those functions still continue to exist... we probably don't need them anymore either though
19:37:13<!Isilkor> also also this should at the same time remove the "no valid keyfile found" log at startup
19:38:38-!- CoN^Uni is now known as CoN^off
19:40:46< ck_> hm. Can tortoisehg's log viewer show more than 3 lines of context?
19:41:09< ck_> oh. It can't even show uncommitted changes. bah.
19:44:57<!Isilkor> ck_: I think the registration code can be removed in full
19:46:43-!- Sven2 [_ve_2@goldwipf.de] has joined #openclonk-dev
19:46:46-!- mode/#openclonk-dev [+ao Sven2 Sven2] by ChanServ
19:47:14<!Mimmo_O> hey Sven2
19:48:09<!Sven2> hey
19:48:57< ck_> Hm the [NOT REGISTERED] thing in the main menu was not removed either
19:49:19<!Isilkor> hi sven
19:49:29< ck_> And there is still a "Register now!" button in the credits :)
19:50:38<!Isilkor> heh. :)
19:51:02-!- Matthi [~a@clonk-center.de] has quit [Connection reset by peer]
19:53:16-!- Matthi [~a@clonk-center.de] has joined #openclonk-dev
19:57:58<!Isilkor> ck_: I approve of his choice of license though, because it allows us to relicense the changes under ISC ;)
19:58:28-!- B_E is now known as b_e
19:58:38< ck_> Isilkor: Indeed. I think first he used some CC license until someone pointed out that it does not fit to source code or so ;)
20:04:31< ck_> Do you want to tell him about these points?
20:05:22< loriel> I was not sure how the CC-by's requirement to give attribution would work for a patch :]
20:06:39<!Isilkor> ck_: can do. Do you have anything besides the things we've mentioned?
20:06:55< ck_> Isilkor: No, otherwise I would have mentioned it :]
20:06:59<!Isilkor> :P
20:14:26-!- AlteredARMOR [~armor@109.86.180.euirc-66893ff8] has left #openclonk-dev []
20:21:04-!- b_e is now known as B_E
20:24:23<!Isilkor> oh, and you can't set your own death message without a key
20:24:47< ck_> Interesting. I didn't even know there was a restriction
20:25:14<!Isilkor> I didn't either, I just asked VS where Config.Registered() was called :)</pre>

IRCMeeting/20100301 - Revision history

Clonk-Karl: Provide log for Monday's meeting