I've done it countless times with no ill results.

From a post by necros from 2010 (http://www.celephais.net/board/view_thread.php?id=4&start=10212&end=10212):

i don't know why, but overlapping brushes will cause zfighting in newer engines like q3 and d3. in quake, qbsp will manage to remove overlapping WORLD brushes without any issues. (overlapping func_ or other visible bmodels WILL cause zfighting with other bmodels or the world).
afaik, it doesn't cause any problems once the bsp is in the engine.

i'm not sure where we are atm concerning the compiling aspect of overlapping brushes. i don't overlap because i like to work clean, but it's just personal preference.

(emphasis mine) 

Now go do it in Hammer/Source ;)

Its just a bad habit that you wouldnt want to carry into other game engines. 

Had overlapping brushes for the pillar lights at the end hanging from the ceiling. Other than that it was an excellent map. *nit pick* *nit pick* 

Overlap corners, don't overlap faces.

Also, someone actually referenced something I said as advice! Neat. 

Thanks for all the help all.

I've had an error compiling my current build, not too sure what a "non convex face" is, or how to fix it, but it seems to be the problem:

http://imgur.com/EHxdeAB

I tried deleting all the recent brushes I had added, but it still seems to happen. Any idea what I should be looking for so it compiles?

The map is currently wide open, void everywhere - is there a limit to how open a map can be to compile? can I plug gaps / skies later on? or even just make a sky box enclosing the limits of the map? cheers 

"CheckFace: Face with too few (a) points at (x y z)"
"CheckFace: Healing degenerate edge at (x y z)"
"CheckFace: Healing point (x y z) off plane by a"

I have no solution for it but maybe readme and check the related polygon in the map file. 

The map is currently wide open, void everywhere

with surounding walls, bottom and sky. 

Why would you want to seal your map if you're not running vis? I tend to leave the areas I'm working on open until I'm ready to move on to the next one - I only bother sealing them then.

As to non-convex faces, I'm not sure how you managed to achieve that but an editor shouldn't let you create an invalid brush like that. If you do find the culprit, and you're using a modern editor like J.A.C.K. or TrenchBroom, I'm sure the developers would be interested in seeing it...

Which -options are supported by that compiler? I would see if there's any way to get more information out of it. It's strange that it doesn't provide any coordinates in the error. 

Any of you guys know of an extended runic/metal texture set? I've found a few more in kdmtex.wad but more choice never hurts. I'd especially like to find all the different runic tiles in both glowing red and "switched off" versions (id1 is missing some in each variant). 

I always build fully sealed and error free as I go. I only run a full vis every so often, but I like to know the map will compile properly without errors before putting too much more work into it. Fixing problems is the least fun part of mapping and I wouldn't want them to pile up. 

http://www.quaketastic.com/files/texture_wads/rmq_darkmet.wad

rmq mqade some runic addons. there are more, idk where to look atm.

rick does it right. 

Also yeah, so far I've done just like Rick. 

provide leaks in early stage.

@mugwump: Might not be the same wad but why not try Runewad? 

A workflow discussion: I can't nail down a specific time frame but a general estimate is every 10 minutes or so I do a full compile. Yes I vis everything and light everything dozens of times per mapping session. Reason being is my current map is probably about the size of one of the large id levels. Vis'ing takes well under a minute maybe a tad longer. But I work on everything in a map at once. Geometry comes first, then game play and then tweaks to lighting and texturing, then back to gameplay. It's very chaotic way to work. (I know.) I ask this here because I'd like to hear other people's workflow in the hope of getting my act together. I'm not looking to get faster per se but I have a lot of room for improvement and I am always open to new ways of working. As far as saving... I save as much less often. It seems that each new area becomes a save point and if I don something risky that could cause a leak or break logic or expose a bug in the editor, I will do a new version of the map. 

Well I managed to fund the culprit - a single brush seemed to be preventing compiling. I'm not really sure why though. I'll just keep test compiling as I go to make sure I'm staying on track.

Interestingly the non-convex face error comes up even when the map compiles and runs fine... not too sure what the issue is there but it doesn't seem to be "structural" 

If you open the map with a text editor,
you can view the structure of the map file.
A rough trade is looking at the warnings
and see if the brush consists of six sides.

If I want to make really sure why a polygon keeps "troubling"
I change all texture names behind the six polygons with a white texture.
When opening the map in an editor I can see which brush in particulary is causing the error.

Of course I use ericw compilers. 

My workflow changes as I learn more mapping techniques but to answer some of your questions:

I compile whenever I want to test something I just set up. This could be a few monster spawn triggers, to a new lighting placement, or even a single brush I want to see in action! When working with lighting, I often always do -extra4 and -soft to ensure I can see what the final product will look like but sometimes just -soft if a minor change was made.

If I do not have a leak, then I always compile with the -fast VIS just to save time. When I want to do a full playtest (Start to finish) I will do -level 4.

If by save you mean physically save the map file then I do that probably every time I add something that I know will stay for the time being. Ctrl+S is muscle memory and TB handles autosaving for the grievous mistakes.

As I stated above, my workflow changes but I am starting to see grey boxing out large sections is the way to go. Like G1ftmacher said before, if you can get through the basic layout then you can get to the more personal mapping phase.

So my workflow would be at this time:

-Brainstorming, mind layouts, inspiration.
-Basic layout of large sections
-Basic texturing of said sections
-First pass detailing
-First pass lighting
-First pass gameplay
-Detailed texturing
-Major detailing and lighting
-Then it gets to a hodgepodge of all of the above but fine tuning.

I am always curious to see what other people do. 

This method is not everyone, but so far I have liked breaking map into multiple pieces, sealing then right away, layering everything and startjJust testing and make it fun. Naming goes something like this: mapname_areaname_version. Then slowly starting to combine these already tested areas together and make compromises Just then if need to be made.

The biggest problem for me personally has been that, even if you just prototype areas, Quake is not really good at that. Everything basically has to be in certain state, before knowing will it work or not. That is why I find working with separate files more efficient. Normally when I do simple layout (ideas) on paper, I don't know about values what they are going to be at the end. Also if making a hud of your map where many paths connects. You suddenly realize you have gone too far, when focusing on something, you thought was important atm (I like to assume everyone does that at some point), it is impossible to go back and visualize everything again, if there is no backups been made.

So my current method is not to think too much about trying to connect everything right away. I am sure everyone has their own way, it would be interesting to hear more like how people layout their maps? 

So Bloughsburgh, you have a very structured workflow compared to me. Funny how I work on game play and lighting very close together per section and interchangeably but you segment those out in steps. I am all over the place! Also I meant "Save as" - like a new version of the map - I guess with TB and other modern editors auto save makes this less of a concern. I am used to the old days when one CGS operation could scrap hours of work.

Newhouse, very interesting that you break the map into smaller more manageable chucks. I am a professional video editor and I often use the same technique on an edit. If I am working on a 60 second TV spot with a 5 second montage - I bust that montage out into a separate piece so I can really focus on each frame in the editing program. This might work for me.

This is all fascinating - I hope others will jump in a share their workflow. 

Dunno anything about workflow but my method so far has pretty much been make-it-as-I-go: I work on one room until I think it looks good then I move to the next room. I'm just not the kind of guy to plan ahead, I'm better at improvising. 

Looks like I'm not the only one then. As I think about it, I'm realizing all the issues I am having in my map are because I grey-boxed and planned ahead. 

Sometimes planning ahead is impossible, when working in a situation you're not familiar yet. Maybe after, planning ahead becomes more effective. 

To be clear, there are certain global end goals that I do define beforehand, but the way to achieve them remains purposefully improvisational. This is how I work as an artist, in music as well, as I have found with experience that it allows my creativity to take me to unexpected places. I never made anything good with a very defined plan. 

theres a vine texture in AD that I want to use in my map. and I also want to try transparency with brushes, like in sock's metal monstrocity. How do I achieve this? 

Set the brush as a func_illusionary. This allows players/monsters to walk through the brush. You can use the flag "alpha" to set transparency. Honestly, just open up the metmon map and check it out yourself! :) all the authors .maps are included with AD.