Page 4 of 4
Re: mapconv - fixing for linux, and other things
Posted: 11 Jul 2014, 14:42
by Anarchid
I don't think that time will ever be ripe for something that will render most of the maps we play useless. Development has to be backwards compatible.
It's not implausible to have support for multiple map formats.
Consider SM3.
Re: mapconv - fixing for linux, and other things
Posted: 11 Jul 2014, 14:49
by enetheru
Anarchid wrote:I don't think that time will ever be ripe for something that will render most of the maps we play useless. Development has to be backwards compatible.
It's not implausible to have support for multiple map formats.
Consider SM3.
I don't see the benefit of multiple map formats. except in the case of huge improvements that make old formats depracated.. but i doubt that will happen unless fully 3d terrain or spherical maps or some such come into play. and what are the chances of that?
Re: mapconv - fixing for linux, and other things
Posted: 11 Jul 2014, 15:56
by Anarchid
I don't see the benefit of multiple map formats
The uncounted gigabytes of maps already made in "legacy" format are worth not throwing it away.
While having a hypothetical nice compilerless format of pure images would (likely) be worth having for the future.
Consider COB.
Re: mapconv - fixing for linux, and other things
Posted: 12 Jul 2014, 03:38
by enetheru
Anarchid wrote:The uncounted gigabytes of maps already made in "legacy" format are worth not throwing it away.
Nothing would get thrown away, it would just become like the rest of the mapping stuff, overriden with lua so much that having an smf/smt is no longer required. technically there is no change in "format" its just how the assets are pulled.
anyway.
I was putting together a test case for why openimageio and libsquish weren't playing nice together so i could submit a bug or get help.. and it no longer exhibited the problems I was having. so replace nvtt with squish, test compile and build and its ok..
well well I'm thinking to myself.. I wonder if windows will build.?
and ba'da'bing, ba'da'bomb.. windows binary... but... performance is shockingly bad.. and it eats up so much memory that it crashes. plus its not static so I have to copy all the mingw binaries it links to along with it.. but at least its a start.
Re: mapconv - fixing for linux, and other things
Posted: 12 Jul 2014, 03:56
by smoth
wasn't sm3 removed from the code?
Re: mapconv - fixing for linux, and other things
Posted: 12 Jul 2014, 04:47
by enetheru
Re: mapconv - fixing for linux, and other things
Posted: 12 Jul 2014, 04:52
by smoth
that is just a header. I thought that JK asked if we cared if they removed sm3... sm3 has not worked properly.. EVER.
Re: mapconv - fixing for linux, and other things
Posted: 12 Jul 2014, 05:00
by enetheru
smoth wrote:that is just a header.
Why yes it is.. and it sits next to all the other files in that folder.
but its all been commented out in the CMakeLists.txt so it doesnt get built.. it just sits there looking ugly
Re: mapconv - fixing for linux, and other things
Posted: 12 Jul 2014, 05:02
by smoth
LOL it is there, mocking you!
Re: mapconv - fixing for linux, and other things
Posted: 26 Sep 2014, 02:27
by abma
SM3 is disabled, not sure if it currently would compile:
https://github.com/spring/spring/blob/d ... ts.txt#L18
so spring doesn't support SM3 these days ~ code is close to a removal.
also see
http://springrts.com/phpbb/viewtopic.php?f=13&t=30806