I know this is going back a fair bit, but when Q3A first came out, I made a few maps for it (around the year 2000) using Q3Radiant 1.0 Build 181, but I have a couple of questions:

1. Are newer versions of Q3Radiant incompatible with those old .map files created by the version I mentioned above? I just tried to open one map with "Q3Radiant Build 202", and it returned an error about parsing entities with a "{".

2. Did the old Q3Radiant by any chance screw up the texture alignments on all the brushes in old maps? Whereas the original PK3s I created were fine, one day a few years later I opened up one of my .map files in another version of Q3R and ALL the textures were misaligned. What happened?

Any answers any of you can give would be greatfully appreciated.

The Happy Friar wrote:

I think I read in a release note somewhere that at one point texture alignment was "fixed". So I assumed it was "broken" in previous versions.

Well, it seems it was REALLY broken, HappyFriar, as I had a look at the .map file itself in Notepad (THANK you, John Carmack! :wink:

) and judging by the following sample, it seems the misalignments are in the file itself, yet old Q3R worked properly with them:

Code:

{
"classname" "worldspawn"
"message" "The Mausoleum by Foebane"
"music" "music/fla22k_03.wav"
// brush 0
{
Brush
{
( -1960 529 -640 ) ( -1931 471 -640 ) ( -1802 534 -640 ) ( ( 0.004214 0.002055 2.298723 ) ( -0.002052 0.004213 -8.065026 ) ) gothic_wall/iron01_e 0 0 0
( -1799 527 -632 ) ( -1928 464 -632 ) ( -1956 522 -632 ) ( ( 0.002055 0.004211 4.373708 ) ( -0.008425 0.004112 -5.380435 ) ) gothic_trim/demoniron01_e 0 0 0
( -1795 538 -608 ) ( -1823 595 -608 ) ( -1823 595 -640 ) ( ( 0.004690 0 2.597032 ) ( 0 0.004688 0 ) ) gothic_wall/iron01_e 0 0 0
( -1841 613 -600 ) ( -1970 550 -600 ) ( -1970 550 -632 ) ( ( 0.004688 0 -45.275341 ) ( 0 0.004688 0 ) ) gothic_wall/iron01_e 0 0 0
( -1988 533 -608 ) ( -1960 475 -608 ) ( -1960 475 -640 ) ( ( 0.004689 0 -2.597257 ) ( 0 0.004688 0 ) ) gothic_wall/iron01_e 0 0 0
( -1921 450 -624 ) ( -1792 513 -624 ) ( -1792 513 -656 ) ( ( 0.004688 0 45.274487 ) ( 0 0.004688 0 ) ) gothic_wall/iron01_e 0 0 0
}
}
// brush 1
{
Brush
{
( 1760 544 -704 ) ( 1696 544 -704 ) ( 1696 480 -704 ) ( ( 0.031250 0 57 ) ( 0 0.125000 220 ) ) base_trim/border11_250 0 0 0
( 1696 480 -696 ) ( 1696 544 -696 ) ( 1760 544 -696 ) ( ( 0.015625 0 28.500000 ) ( 0 0.015625 -27.500000 ) ) sfx/clangdark_bounce 0 0 0
( 1696 480 -640 ) ( 1760 480 -640 ) ( 1760 480 -704 ) ( ( 0.031250 0 -54.999832 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
( 1760 480 -640 ) ( 1760 544 -640 ) ( 1760 544 -704 ) ( ( 0.031250 0 57 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
( 1760 544 -640 ) ( 1696 544 -640 ) ( 1696 544 -704 ) ( ( 0.031250 0 55.000137 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
( 1696 544 -640 ) ( 1696 480 -640 ) ( 1696 480 -704 ) ( ( 0.031250 0 -57 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
}
}
// brush 2
{
Brush
{
( -1696 544 -704 ) ( -1760 544 -704 ) ( -1760 480 -704 ) ( ( 0.031250 0 57 ) ( 0 0.125000 -212 ) ) base_trim/border11_250 0 0 0
( -1760 480 -696 ) ( -1760 544 -696 ) ( -1696 544 -696 ) ( ( 0.015625 0 28.500000 ) ( 0 0.015625 26.500000 ) ) sfx/clangdark_bounce 0 0 0
( -1760 480 -640 ) ( -1696 480 -640 ) ( -1696 480 -704 ) ( ( 0.031250 0 53.000160 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
( -1696 480 -640 ) ( -1696 544 -640 ) ( -1696 544 -704 ) ( ( 0.031250 0 57 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
( -1696 544 -640 ) ( -1760 544 -640 ) ( -1760 544 -704 ) ( ( 0.031250 0 -52.999870 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
( -1760 544 -640 ) ( -1760 480 -640 ) ( -1760 480 -704 ) ( ( 0.031250 0 -57 ) ( 0 0.125000 0 ) ) base_trim/border11_250 0 0 0
}
}

And I guess it's like that throughout all my other Q3A maps, too. The non-integer numbers in the middle of each line must be the texture alignments, right? In that case, it would take ages to correct in the later versions of Q3R.

Why did this happen? How could they have made such a mistake?

But then I may have used a later version of Q3R to compile and THEN it went wrong? Luckily, the .bsps that resulted are fine.

Any ideas?