Favorite bits of [REDACTED] Doom History?

[whatup876]

Not just the "someone" part but also refering to it being made for "the Doom engine" and not a source port with alterations like GZDoom.

I think even GGGManlives is guilty of this at times.

[Scuba Steve]

4 minutes ago, Xaser said:

Speaking of journalism flubs, let's not forget the repeated assertions by various media outlets that all notable Doom mods are made by a single, hyper-prolific individual named "Someone." :P

 

I'll have you know, Various Authors is the most prolific mapper of all time with more than two dozen Cacowards to their name!

[spineapple tea]

Recently for Doom's birthday, John Romero gave cool tidbits about the game on Twitter and said "the automap almost had an Asteroids easter egg" but didn't elaborate. Nobody mentioned it in the replies and as far as I'm concerned, that was the only mention of the secret ever (almost) existing.

 

 

Edited January 21, 2021 by N1ck

[Dweller Dark]

One piece of Doom history that I think Doomkid went and explained (or maybe I'm thinking of another DoomTuber) is the mouse myth. It seems that a lot of people, even with some knowledge of Doom, never knew that you could actually use the mouse to move in vanilla. Hell, only I really knew after trying to play it via the pre-Unity DOS version and the video where the actual original settings were brought up.

[Doomkid]

2 hours ago, N1ck said:

Recently for Doom's birthday, John Romero gave cool tidbits about the game on Twitter and said "the automap almost had an Asteroids easter egg" but didn't elaborate. Nobody mentioned it in the replies and as far as I'm concerned, that was the only mention of the secret ever (almost) existing.

 

 

Don’t quote me on this..  but I think there is a comment in the source referring to Asteroids as well, if that’s true it’s strong evidence for it. I would have loved Asteroids on the automap.

[Quasar]

https://doomwiki.org/wiki/Automap_asteroids

[dew]

16 hours ago, boris said:

According to the Doomwiki PrBoom is based on Boom, and SMMU on MBF (which in turn is based on Boom). Also the first release of SMMU was about 4 months after the first release of PrBoom. I guess that's how revisionism is started? ;)

 

10 hours ago, chungy said:

https://github.com/Doom-Utils/historic-ports/commits/prboom

Oops, my bad!

[Maximum Matt]

16 hours ago, Revenant100 said:

It's not Doom, but it's not far removed: There's an increasingly growing subset of increasingly wrong people who think Quake's Shambler has fur, the notion which of course is patently wrong. So wrong have these people become that even modern day id Software has forgotten their own truth, choosing to erroneously depict their modern day interpretations of the Shambler with fur.

This is classic, Civvie11 just did a video about Half-Life comparing the vortigaunts to shamblers and he says they are furry too (at 15:20) - 

 

[Gez]

More "technical details" than history, but the thing about cyan being transparent or being at palette index 247. (The actual translucent color is a weird beige-pink, the last color in the palette; and it's only transparent in id's old toolchain, not in the Doom engine itself.)

 

Another misguided technical detail I often see is that, in GZDoom, to get rid of the ugly anti-pixel blur filter, you have to turn off anisotropic filtering. No! Aniso is there to avoid textures turning into blurry messes when seen at a strong angle. By all means, turn linear filtering off, but do keep anisotropic filtering!

 

With anisotropic filtering:

Notice all the crunchy square pixels are there.

 

Without anisotropic filtering:

Look at how the BRNGRN wall turns into some amorphous sludge in the distance.

[Doomkid]

50 minutes ago, Gez said:

More "technical details" than history, but the thing about cyan being transparent or being at palette index 247. (The actual translucent color is a weird beige-pink, the last color in the palette; and it's only transparent in id's old toolchain, not in the Doom engine itself.)

Since Slade exports everything in PNG format with support for true transparency this isn't as much of an issue anymore, but every single graphic editing utility from dmgraph/dmadds/deutex all the way up through XWE treated cyan as transparent. With older 1994/1995 wads where the sprite replacements come as big packs of compile-it-yourself GIFs or BMPs, they always used cyan as the transparent color.

 

I do remember realizing the weird beige-pink was id's choice for a transparent color when looking through some of the old alphas, which iirc have parts of the sprite (seemingly unintentionally) outlined in this color.

 

Does one of the slots in the palette actually count as "the color used for transparency by the engine, aka a lack of color"? Something I've never been sure about..

 

EDIT: Found the answer at the ZDoom wiki:

The Doom palette does not contain cyan, nor is any color in the palette used for transparency due to the nature in which graphics (save flats) are stored in a WAD. Index 247, which is often pictured as cyan (R:0, G:255, B:255) is a convention used by many editing utilities for ease of editing and to better see the outside border of sprites. Index 247 is in fact black and can be used in Doom graphics, though many editing utilities will, sadly, not let you use it. It was probably used first because index 247 is redundant with index 0 and wasn't actually used in any Doom sprites, though it is used in some in Doom II and Final Doom (e.g., patch RW33_3 used in the METAL4 texture). This is a source of problem for other Doom engine games as index 247 in the Heretic, Hexen or Strife palettes is not redundant and is used in many sprites, flats and patches. Editing tools that rely on this assumption of a "transparent color" degrade these games' sprites when used.

Edited January 21, 2021 by Doomkid

[hobomaster22]

12 minutes ago, Doomkid said:

Since Slade exports everything in PNG format with support for true transparency this isn't as much of an issue anymore, but every single graphic editing utility from dmgraph/dmadds/deutex all the way up through XWE treated cyan as transparent. With older 1994/1995 wads where the sprite replacements come as big packs of compile-it-yourself GIFs or BMPs, they always used cyan as the transparent color. 

 

I do remember realizing the weird beige-pink was id's choice for a transparent color when looking through some of the old alphas, which iirc have parts of the sprite (seemingly unintentionally) outlined in this color. 

 

Does one of the slots in the palette actually count as "the color used for transparency by the engine, aka a lack of color"? Something I've never been sure about.. 

There isn't a transparent color in the doom engine. Doom's picture format doesn't store 'transparent' pixels, rather it simply skips them by storing an offsets to start drawing the column. This meant the rendering didn't need to check for a transparent pixel, it would just draw based on the offsets.

Like Gez was saying the id tools used the pink color to mark the pixel as transparent so they would be skipped when converting to doom's picture format. Basically they would count the number of cyan pixels and store that offset. A lot of community tools use cyan. People may also find it interesting that Deepsea uses the pink color.

[Diabolución]

On 1/20/2021 at 12:48 AM, A.H. Sankhatayan said:

the missing key in TNT Map 31

 

Also, most people seem to ignore that TNT31.WAD ought to be loaded, yes or yes. Even for the fixed id Anthology IWAD, unless the used port does rebuild nodes internally.

 

https://www.doomworld.com/forum/post/356764/

 

By the way, I cannot believe no one has mentioned so far that attempt of rewriting the history. Tip: idmap01.wad refutes it.

 

Okay, last post of mine in this thread, promised.

 

Edited January 21, 2021 by Diabolución
I forgot that "does."

[Tristan]

On 1/21/2021 at 1:59 AM, Xaser said:

Speaking of journalism flubs, let's not forget the repeated assertions by various media outlets that all notable Doom mods are made by a single, hyper-prolific individual named "Someone." :P

 

The exploits of Someone is one of my favourite running gags in the community. I spotted them doing stuff with Terraria recently, it's nice to see them branching out into other games.

[ChopBlock223]

On 1/20/2021 at 4:55 AM, dew said:

and he actually disrupted the pr+ timeline with his haphazard umapinfo implementation that disregarded any demo checks and backwards or forwards planning, just dumping it to whoever might eventually pick it up

I thought it didn't change anything with demo compatibility, and that this was the entire point?

[Alper002]

3 hours ago, ChopBlock223 said:

I thought it didn't change anything with demo compatibility, and that this was the entire point?

The initial releases did not do this. It recorded demos in the same format as non-umapinfo demos.

The demo signature thing was completed only a month ago, so the current official release versions of the fork don't have the proper demo compatibility if recording for a mod/mapset with umapinfo.

That pr+ timeline "disruption" that dew mentions wasn't a thing initially though, IIRC the fork didn't get anything resembling an "official status" until later(though that official status was sorta instated while the umapinfo implementation didn't take demos into consideration lol), so I'm not angry at Graf for initially releasing it the way it was.

[whatup876]

I remember hearing stories about other types of influence that Sandy Petersen had over the development of the games.

Like keeping how the BFG currently works or the shotgun.

Was any of it ever debunked?

Because i know Romero and McGee called out Tim Willits over the creation of "multiplayer only" maps.

[ChopBlock223]

8 hours ago, Alper002 said:

The initial releases did not do this. It recorded demos in the same format as non-umapinfo demos.

The demo signature thing was completed only a month ago, so the current official release versions of the fork don't have the proper demo compatibility if recording for a mod/mapset with umapinfo.

That pr+ timeline "disruption" that dew mentions wasn't a thing initially though, IIRC the fork didn't get anything resembling an "official status" until later(though that official status was sorta instated while the umapinfo implementation didn't take demos into consideration lol), so I'm not angry at Graf for initially releasing it the way it was.

So it's not a problem?

[Doomkid]

6 hours ago, whatup876 said:

I remember hearing stories about other types of influence that Sandy Petersen had over the development of the games.

Like keeping how the BFG currently works or the shotgun.

Was any of it ever debunked?

Because i know Romero and McGee called out Tim Willits over the creation of "multiplayer only" maps.

Well, Romero was the one who I remember saying the original BFG functionality had to go because "it looked like Christmas", but no one has ever disputed the claim that Sandy basically proposed how it ended up working. I also remember scratching my chin when Sandy claimed he named the monsters, but then I remembered that Imps were "Troopers" and Demons were "Sergeants" which is pretty dumb, so I believe him when he says he suggested the (far better) names they ended up with.

[Martin Howe]

On 1/21/2021 at 5:07 PM, Diabolución said:

Also, most people seem to ignore that TNT31.WAD ought to be loaded, yes or yes. Even for the fixed id Anthology IWAD, unless the used port does rebuild nodes internally.

Fascinating; GZDoom fixes the key bug internally (it has a list of fixes for well-known WADs, matches them by MD5, then fixes them during load), but I didn't know about the node bug! (ZDoom based ports build nodes on load, so it wouldn't be apparent).

[Scypek2]

16 hours ago, Doomkid said:

Well, Romero was the one who I remember saying the original BFG functionality had to go because "it looked like Christmas", but no one has ever disputed the claim that Sandy basically proposed how it ended up working. I also remember scratching my chin when Sandy claimed he named the monsters, but then I remembered that Imps were "Troopers" and Demons were "Sergeants" which is pretty dumb, so I believe him when he says he suggested the (far better) names they ended up with. 

 

Interesting, I didn't know that! I just knew he named all the levels, so that's consistent. The monsters being named later also fits with a lot of them having different names internally. I guess without Sandy we'd just end up with a bunch of Skulls, Heads, Fatsos and Babies.

[dew]

On 1/22/2021 at 10:52 PM, ChopBlock223 said:

So it's not a problem?

It wasn't a problem, because speedrunners roundly refused to record demos on it, based on the known issues if said extra features were used, and after an unfortunate commit broke even unrelated pre-existing compatability. No one's going to risk getting their runs invalidated, because they used an unreliable janky port. The archive decided not to host such demos until dust settles anyway.

 

On the other hand, the fork explicitly called itself "prboom-plus 2.5.1.7". This was extremely invasive and reminiscent of a hostile takeover. The latest pr+ version was 2.5.1.5, and a person unfamiliar with the details would absolutely assume this is the next legitimate iteration of the same product. The longer the issue would run unanswered, the more confusion would be sowed among the laypeople. Just imagine Quasar forking off GZDoom and calling it "GZDoom 5.1.0", it would be utter mayhem.

[Orchid87]

On 1/21/2021 at 11:51 AM, Doomkid said:

Don’t quote me on this..  but I think there is a comment in the source referring to Asteroids as well, if that’s true it’s strong evidence for it. I would have loved Asteroids on the automap.

I actually remember that it was known about Asteroids way before that tweet. Might be the source code indeed.

 

One thing that was going strong in the community for years is the claim that Doom is a 2D game, as in not having a Z dimension, citing infinite monster height as an evidence. Now with the Fabien Sanglard book it is finally disproven

[ChopBlock223]

17 hours ago, dew said:

It wasn't a problem, because speedrunners roundly refused to record demos on it, based on the known issues if said extra features were used, and after an unfortunate commit broke even unrelated pre-existing compatability. No one's going to risk getting their runs invalidated, because they used an unreliable janky port. The archive decided not to host such demos until dust settles anyway.

I see. Is the DSDA fork put together more properly? I love the idea of having MAPINFO capabilities with something like PrBoom+, but I wouldn't want it at the expense of demo compat.

[Alper002]

37 minutes ago, ChopBlock223 said:

I see. Is the DSDA fork put together more properly? I love the idea of having MAPINFO capabilities with something like PrBoom+, but I wouldn't want it at the expense of demo compat.

See, the demo-compat part is done, it's just that there's no available "official" release version that includes it yet. The current "official" version of the UMAPINFO fork (i.e. the basically crippled 2.5.1.7) doesn't have the implementation, and also has a pretty major progression-related bug in vanilla doom 1.

I think if the UMAPINFO fork gets a new, proper release version without crippling issues there'll be less unrest about UMAPINFO as a whole. Maybe the thread could be renamed to really make it clear that it is a split off the "real deal" as well, but that is secondary to me.

Edited January 24, 2021 by Alper002

[dew]

3 hours ago, ChopBlock223 said:

I see. Is the DSDA fork put together more properly? I love the idea of having MAPINFO capabilities with something like PrBoom+, but I wouldn't want it at the expense of demo compat.

The 2.5.1.7 fork is getting continuously worked on by fabian and many other coders, so the fix to the UMAPINFO demo header by anotak, altaz and xaser was recently merged into master. From that point, kraflab merged the changes into dsdadoom, so technically they should be equal regarding this issue. However you need to compile 2.5.1.7 yourself. kraflab does provide binaries for dsda-doom and the latest one should cover this issue already, check this thread for details and latest stable links.

(edit: Actually, looking at the timeline, I think 0.10 doesn't include the patch just yet. Might need to wait for 0.11, which will probably add experimental Heretic support as well.)

 

Also yes, I would recommend dsda-doom. Its recent quality-of-life features are very desirable.

Edited January 25, 2021 by dew

[LordMeow]

On 1/20/2021 at 4:08 AM, Quasar said:

Both Romero and Carmack have, at times in the last 25+ years, claimed that they made significant contributions toward helping the Doom community understand the game's data formats and get editors established, and this is simply revisionist history. Carmack only released the Doom bsp code well after Raphael wrote the first working community node builder, and Romero only ever released any specs for Doom II, not the first Doom game, while they would have been useful.

 

This revisionist history has made its way to the Game Engine Black Book Doom, I wonder why they insist on it both Romero and Carmack. I always saw it as a progression: Wolf3D was cracked into, Doom provided ways to play external levels with little fuss and Quake allowed a great deal of modification outright.

 

It's even logical, since you're walking an unknown path and don't know beforehand the consequences of opening your game so much. One step at a time. 

 

Why the revisionism? I don't know...

[whatup876]

This might be a topic for a thread of its own, but was the first use of the name "Doomguy" ever documented?