Corrupted .map and working .bsp [SOLVED]

[CorSair]

Last night trying to open my map on NetRadiant I noticed nasty suprise: my map has been corrupted somehow, thus showing only gray in camera view, and no brushes in edit view. Just nothing. And also, backup too has been corrupted.

But strange thing is, map had exactly same byte amount, (over 2 000 kb), so I decided to give a shot, and do most simple compile that Radiant has, no visibility and no lights.

And when /devmapped it, I couldn't believe what happened: it was working properly.
Only thing that gives error, is probably some misplaced shader, giving this error:

Code: [Select]

TGA file header declares top-down image, ignoring

Now first question.
What are the risks of decompiling working .bsp, that has been compiled from corrupted .map file, aside from standard issues (all being structural, texture misplacing)?

And second question.
Before thinking more, I actually tried to decompile, but only what I can do, is just open, without seeing the contents of error. I haven't done any sort change to files.
I used the standard q3map2 compiler that came from NetRadiant.
Is there a file where I can access directly to q3map2 command prompt? Or download other version? I haven't been into proper decompiling business, not until now.

[UniqPhoeniX]

If you open the map in radiant, check the brush count and View->Filter menu to see if it is really corrupted. Also make sure the 'Cubic-clip the camera view' option is disabled (it's on the toolbar).

[CorSair]

Brush count 1, all unchecked, and I very rarely use that cubic clip view.

So, I decided to move my bsp to other folder than Tremulous, opened my map, made a new similar .bsp.
It was ugly, but it worked properly.

Try explain this, I don't got any clue how, or why...

[CreatureofHell]

Now that's what I call a case. 

[Ingar]

So, is there a problem, or not ?

[CorSair]

Being literate person is good thing. And using your brains too (where mine can't do it. Otherwise I wouldn't ask if could manage it.)

Now first question.
What are the risks of decompiling working .bsp, that has been compiled from corrupted .map file, aside from standard issues (all being structural, texture misplacing)?

And second question.
Before thinking more, I actually tried to decompile, but only what I can do, is just open, without seeing the contents of error. I haven't done any sort change to files.
I used the standard q3map2 compiler that came from NetRadiant.
Is there a file where I can access directly to q3map2 command prompt? Or download other version? I haven't been into proper decompiling business, not until now.

[ULTRA Random ViruS]

Netradiant is too weird for me so i stopped mapping overall.
I had a similar case before, my compiled map worked but the lighting was maxed out everywhere.
____________________________
Weird? Here's one case of being weird: when i compiled the map - it came out .bsp like normal, but it didn't work. I renamed it to .jpg so it wouldn't be override when i compile again. But when i renamed it to .jpg - a greyscale - faint 'xray' like top view of half the map was there. I was 'scared' about how freaky this is. Don't ask me for the picture - its on the old computer which is in the bin but i might have a backup somewhere... i just don't know which portable hardware it is on... soo many...

[CorSair]

Irrelevant... As always, ViruS.

Still, having same old issues with my decompiling process, can't still get peek, why it closes. And it seems google ain't my friend anymore.

[Ingar]

Talked to Corsair on IRC, basicly there are broken brushes in the map.
This is the error I got in radiant:

Code: [Select]

32198:84: parse error at '-1.#IND': expected '#number'
brush 3243: parse error
entity 0: parse error

So, I opened the file in a text editor (yes, you can open .map files in notepad!),
searched for brush 3243 and found this:

Code: [Select]

// brush 3243
{
( -64 -498 -222 ) ( -64 -482 -238 ) ( 64 -498 -222 ) pk02/pk02_generic01b_C -1.#IND -1.#IND 90 -1.#IND -1.#IND 0 0 0
( 64 -514 -238 ) ( -64 -498 -254 ) ( -64 -514 -238 ) pk02/pk02_light02a_C 192 0 90 0.125 0.125 0 0 0
( 64 -496 -224 ) ( -64 -504 -248 ) ( -64 -496 -224 ) pk02/pk02_generic01b_C 0 0 0 0.5 0.5 0 0 0
( -64 -448 -272 ) ( -64 -472 -280 ) ( 64 -448 -272 ) pk02/pk02_generic01b_C 0 0 0 0.5 0.5 0 0 0
( -160 -512 64 ) ( -160 -544 -64 ) ( -160 -512 -64 ) pk02/pk02_generic01b_C 0 0 0 0.5 0.5 0 0 0
( -128 -560 64 ) ( -128 -544 -64 ) ( -128 -560 -64 ) pk02/pk02_generic01b_C 0 0 0 0.5 0.5 0 0 0
}

This part -1.#IND -1.#IND 90 -1.#IND -1.#IND should be numbers.

After I removed the offending parts of text, radiant would load the map again.

[UniqPhoeniX]

Why can't people look for errors first?

[CorSair]

Why can't people look for errors first?

Oh please. We are NOT perfect.

But anyway, it was slight overlook on console.
I didn't get any sort errors on session, but it is, as you can see, problem [SOLVED].