IMHO do what you think is best. Whether or not that conforms to some nostalgic pre-existing name based bias is another question. 

look at the original qsrc world.qc

=================
void() worldspawn =
{
lastspawn = world;
InitBodyQue ();

// custom map attributes
if ((self.model == "maps/e1m8.bsp")|| (self.model == "maps/insertyourmapnamehere.bsp"))
cvar_set ("sv_gravity", "100");
else
cvar_set ("sv_gravity", "800");
============
that should do the trick
something like that 

I agree 100%. But it's a take off of "Doom the way ID did it". And that is not something that would ever run on the classic Doom engine. It exceeds the Doom limits in almost all respects and requires engines that exceed those limits, which is why I said that I expected that this would exceed the Quake limits.

Did I say "Exceed" enough? That is kind of the point of these types of things. Quake made as if almost no limits were there to keep the original level designers constrained to very limited .bsp and hardware constraints of the day. What would they do given the broad palette that they left behind if they would do it now?

The concept was never to run inside the limits of the original Quake engine (and apologies to FifthElephant if I ever misrepresent this, that's all on me) but to take the original levels and tweak, twist, modify or otherwise change them to be the same, yet different. All while not worrying about original limits.

He's (apologies to FifthElephant if you are a she but he works better in text) is trying to re-image the original Quake levels in some other way. From past comments made by him I was just trying to remind him and the readers of this thread of what he said his idea was months ago. Mission creep sucks and all I wanted to do was to nudge him away from it. 

The sentence is "Quake The Way Id Did It" not "Quake The Way Id Might Have Done It If This And That". 

Doom The Way id Did doesn't exceed any vanilla limits. It was designed to be as close to the stock levels as possible. 

Doom The Way ID Did does run on the classic engine. The whole point of the project, and similarly FE's Quake The Way ID Did, was to make (new) levels that look and play like they could have been part of the original games, as if made by ID themselves. It does NOT mean touching up on the stock maps or making overly detailed maps that break vanilla engine limits.

I'm also of the opinion that the QTWID levels should run in vanilla engines, otherwise it would defy the goals or the idea of the project. How can they exceed the limits anyway - one should think that smaller, less detailed maps wouldn't even come close to having any of such issues. If they do, there's clearly something wrong with the design.

At the end of the day, it's FE's decision how to proceed. However, I would be very disappointed if it turned afterwards that tweaking the levels for vanilla limits was possible after all. 

This is a lot of text.

I will probably be disappointing many people with this project. The first couple levels will very likely run on the original engine but the last two maps I made I went a bit mad with the details.
The maps are supposed to be like an extension to the game not a revisit of the originals.

I think if I went back and took out all the details just to make it run in the original exe people would be mad. I might just give it a name like the expansion packs and let it be its own thing.

People will still see the inspiration from the original game 

I am a mere lurker, but I think a proper QTWID should be a community project where detailed id style analyses are used to vet and refine maps, just like DTWID. Studying the DTWID site and Doomworld discussion threads would be a good first step. And of course all maps should be vanilla-compatible. Anything else does not follow the TWID ethos and should really use another name, IMO. I very much hope someday a proper QTWID is done, and it would be unfortunate if the name was already taken by a set of maps that follow a significantly different design philosophy. 

I agree with you on all points. 

using winquake and all the maps except map 3 run fine.

The only problem is map 3 took me quite a long time to make and is fairly elaborate. I dont think I could go back and rework the entire map. I think I would have to redo a lot of it from scratch. Honestly it may be unsuitable for redoing as it's just too detailed and too much of the map is visible from any one point.

As for making QTWID a community project? I did make an attempt to do this and the guys here ended up doing a speed map session that ended up only 1 map really feeling like it was made by id. 

Which limits does the third map exceed specifically and by how much?

Community projects don't work here anymore, unfortunatly. There's simply too little of an active mapping community left. 

The geometry is disappearing, probably too much shit visible on screen. So you can see into the void in a number of areas.

This is the only map which has this problem. I'm probably going to have a look at making some cuts to get it to run in vanilla. 

Oh. I thought you were talking about the hard limits that crash old engines. Technically, according to the rules disappearing stuff or grey flashes shouldn't occur in TWID maps, indeed, mut this is not that much of a critical problem in comparison. It can be alleviated by adjusting maxsurfs and maxedges in software engines. Of course, if you could optimize the areas a little more, it would go a long way. Still, no end-of-the-world problem as it sounded at first. 

Aren't there already tonnes of maps that are qtwid (ps shouldn't it be qtwidi, what happened to the 'it'?), just play anything from 96/97/98 and they're all within original limits and mostly id textures.

What exactly is Quake the way Id did it anyway... small, blocky, underdetailed levels? Is that really what people want to play now?

I'm all for creating some levels inspired by the style and design that id did, but "extremely under-detailed" shouldn't be a prerequisite. You can make an id themed map with detail.

I wouldn't worry about a custom progs and it not running on the original engine. Is there anyone at all that would actually attempt to run this in the original engine at 640x480? I'd still be playing it in dp. 

in standard quake. Probably still needs a tiny bit of tweaking but it's pretty much resolved. (Just chopped down on detail a bit)
The detail I have reduced isn't going to be missed, the map has still retained the same feeling. I think once I am done with this pak I will make a much more detail intensive map like I did with q-deck. This should restore balance to the universe. 

According to the DTWID site:
"The idea behind this project was to imagine that the original Doom had more maps before the release, but they were lost. The mappers tried to simulate the work of John Romero, Sandy Petersen, Tom Hall and achieve that authentic id-feeling in their maps."

Achieving that id feel entailed analyzing Romero, Petersen, and Hall's individual mapping styles/habits and refining submissions based on community feedback to better emulate those styles. You can see how the project developed and find some great design analysis in this gargantuan thread:
http://www.doomworld.com/vb/wads-mods/52046-dtwid-public-beta-5-released/ 

if the name causes so much offence (spiney was rigth re RMQ reference) call it soemthing else and still do what you think is right fifth. 

Community projects don't work here anymore, unfortunately. There's simply too little of an active mapping community left.
Sadly this is the state of quake mapping nowadays, but there are a few people still around. It really depends if the project is going to be organized properly and actually finished!

if the name causes so much offence
I get the impression people are concerned because the project name means something special to them and feel it is not going in a similar direction to the previous doom project.

After reading the first couple of pages of the Doom project it seems the doom community spent a lot of time analyzing and trying to understand the original authors building styles and design ideas. I assume this is what Fifth has been doing, but his design process has not been visible to the community, hence the comments.

Here is an awesome quote from the Doom thread that could easily apply to Quake mapping:

You do know that the main reason the stock episodes looked the way they looked is because:
1. id had to make suire the levels could run on a common computer from the early 90s.
2. Romero was a better level designer than Petersen.

I would say that to create a level that mimics Romero you would have to throw away all bells-and-whistles off today and follow the rules Romero made for himself:

* always changing floor height when I wanted to change floor textures
* using special border textures between different wall segments and doorways
* being strict about texture alignment
* conscious use of contrast everywhere in a level between light and dark areas, cramped and open areas
* making sure that if a player could see outside that they should be able to somehow get there
* being strict about designing several secret areas on every level
* making my levels flow so the player will revisit areas several times so they will better understand the 3D space of the level
* creating easily recognizable landmarks in several places for easier navigation 

being strict about texture alignment

lolololol 

Well, when you don't have fancy preview modes and auto-alignment features... 

There shouldn't be this much of a debate. It's not like Fifth is planning on reinventing the wheel with a tiny mod, he's just trying to redo e1m8's key feature. Blame Romero for being a lazy dicklord who could've easily made gravity a worldspawn key/value instead of doing a shitty hack.

The goal is to make maps that emulate id's, who cares if it's /id1/ or not. Is it 'more' like id did it if he has to use an entity hack and break save game compatiblity? id would never have shipped a map like that. They would do it right in code.

Besides, done in it's own .pak, a custom start.bsp can be made. id wouldn't have had players bring down the console and type in skill choices or which map filename they want to go to. 

Blame Romero for being a lazy dicklord who could've easily made gravity a worldspawn key/value instead of doing a shitty hack.

Actually it was American who's been responsible for the main QC coding
Just wondering why the id had never fix all of the qc bugs 

although some people may be getting a bit butthurt (and not really, even) *I* certainly can't condemn anyone for that even if they do and actually this is a good design-based discussion. I just hope 5E doesn't take any of it for discouraging since it's great to see episodes still coming out these days, episodes are easily among the most exciting releases. The idea of an e1m8-inspired map obviously knocks my socks off as well (it's true though, e1m8 is arguably cooler than e3m5, yet it hasn't got tributes). 

I just hope 5E doesn't take any of it for discouraging since it's great to see episodes still coming out these days
People are just expressing opinions, ultimately its Fifth's project and he can do whatever he wants. Its certainly not much fun to hear lots of people moaning about your project idea, but there are plenty of people who will love it regardless of what direction it is going in. 

Cause it was made with TrenchBroom, and therefore, love.