-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi everybody!
Last evening and today we (of course) discussed several things and planned and
tested lots of stuff, too. Here is a short summary of conclusions we have so far
(in no particular order):

* We should get some WML "experts" (pioneers) to become devs, so that we have
more expertise from those people driving Wesnoth to the maximum possible. At the
moment we don't have many people around (in the irc chan #wesnoth-dev) that
really do *use* WML. We think that it might be good to have some way to get more
"feedback" on things that should be improved so that some rough edges can be
smoothed. One possible idea was the creation of an WML library with things well
maintained and required for some "advanced stuff" but not necessarily used in
mainline. Possible candidates (rough ideas) would be Bob, teaser and ken_oh.

* Should the Kalifas be included in 1.9? Yes, this is likely to be leading to
*many* discussions but it might be nice to provide players with "something new"
in 1.10 that is not along the lines of "oh, one more new campaign".

* One huge topic of discussion was the possible migration to OpenGL. We (those
here at FOSDEM) think that the migration will eventually be needed, either this
year or next year. It should be done shortly after a new stable series is out.
Hopefully some of the benefits of this move do include making better use of the
graphics card (that is available and basically idle) on smaller systems (like eg
embedded platforms). It will also make things a lot easier at some parts of the
engine. Currently we have to think about which "areas" to invalidate. If we
invalidate too much it costs performance (all via the CPU, no acceleration from
the graphics card here!), if we don't update enough, we get artifacts.
Two years ago we had some branches trying to move Wesnoth to OpenGL. At that
time we came to the conclusion that there are many systems not supporting OpenGL
well enough (especially Linux systems with open source drivers) to make it
really run nicely. Due to the huge progress in mesa (and the drop in "well
supported platforms" (because the packagers for anything beside OSX, Windows and
various Linux distributions are basically gone)) it should be possible to now
switch over to OpenGL.
To find out the current status we tested Frogatto (a sidescrolling game Dave is
currently working on) which uses OpenGL. We tested on these systems:
- - Gentoo (unstable) with mesa 7.7, Intel 945GM graphics, Core2Duo 1.5GHz
- - Debian stable with mesa 7.6, Intel 915M graphics, PentiumM 1.6GHz
- - Ubuntu 9.10, *no* 3D hardware accell!, Intel 500, Atom Z520 1.33GHz
The result was that we were basically at 50fps (framelimit at 50fps) on the
gentoo system. The debian system (with the slower CPU and graphics) was normally
well above 30fps, most times at 50fps. The Atom based system was rather
sluggish, but it was usable. It should be possible to run Wesnoth on it, since
the sidescroller *is* playable with a framerate of something like 20fps. In
general we think that OSX, Windows as well as users of proprietary graphics
drivers should be perfectly fine. Only users without proprietary drivers are
likely to not have a "really fast" OpenGL acceleration, but even for those it
should be playable via Software renderer since even the Atom was able to do so.
More recent Mesa versions normally do bring speed improvements and all currently
supported platforms should still be supported.
No, due to this change Wesnoth will still be in 2D, it won't become some 3D
shooter. ;)
One important part about the migration to OpenGL is the plan itself. It should
be started right after a new stable series is out and we should try to convert
"everything" to OpenGL while removing the "old" layer based renderer. For this
we came to the conclusion that we should have one person feeling "responsible"
for the port in general (no, we can't just copy over the iphone port!). The
"subproject leader" will of course work together with the people having most
knowledge in the respective areas and everyone is free to help "In general". For
example boucman has the knowledge about animations, mordante has the knowledge
about the GUI, ... The aim for this conversion is that the opengl edition is a)
mandatory and b) feature equivalent to the layer based engine we currently use
once the new stable series is started.

* Boucman introduced a new proposal (which is open for discussion!!!) to
introduce some new "alliance system". With this system you could have some ways
to specify which information you do share with which player. Stuff that would be
possible are:
- - Do I capture villages from player B if I walk on them?
- - Will shroud be removed for player B if I see some new area?
- - Will fog be removed for player B if it is removed for me?
- - Can I attack player B?
- - Does player B share victory with me (when I win, player B wins)
...
In general the other player does *not* have to have the same preferences for
alliances set as the current player. Those settings are possible separately for
each player in the game (some table with checkboxes?). Those settings can be
changed *IN* a game so that some diplomacy is possible (unless WML blocks this,
of course!).
Remember this is just a proposal which is open for discussion. We will also need
someone implementing this stuff (maybe GSoC?).

* We should cleanup "old" WML content that is no longer well maintained once the
new trunk is "started". This includes for example the removal of ANL (noone
feels responsible for it and it is currently not well maintained). One
problematic case is eg Dark Forecast. We should check what to do about those WML
heavy things and consider removing things not working well anymore while also
adding some new innovative stuff (if/where possible).

* Should we go for a "Games Dev Room" with gaming related talks at FOSDEM 2011?
This will require working together with other orgs. We should talk to other orgs
about what to do and if this is desired. Planning for this should basically be
started "soon". Starting planning this some time in October will be too late!


This is basically a list of things we discussed so far. We will talk more about
the topics and probably about some other new things. Please feel free to discuss
all of the above points here on the mailing list. But please do split the mails
by topic since this is just a rough overview and it will probably be easier to
follow if we split stuff into "one thread per topic". Looking forward to reading
your ideas and opinions about the points listed above. Don't forget that *now*
is the time we should talk about the new things we want to have in 1.9 since
(hopefully) 1.8 is close and the huge changes should happen right at the start
of the development series.
Greetings from the FOSDEM,
Nils Kneuper aka Ivanovic
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (GNU/Linux)

iEYEARECAAYFAkttmIYACgkQfFda9thizwVEIgCgmB3VdXMfo7wog8+4JPyCbIUh
8SMAoIqaLvl2E6PoPvC+jTsaBsFRno5c
=0OJW
-----END PGP SIGNATURE-----

_______________________________________________
Wesnoth-dev mailing list
[email protected]
https://mail.gna.org/listinfo/wesnoth-dev

Reply via email to