DOOM II demos [-complevel 2]

[Linguica]

Watched in glorious 4K, the pixels were almost as crisp as the movement.

 

One bit that I found funny was at 17:13 in Map23 where an archvile was pissed off at a chaingunner before you even got out of the tunnel to be fully visible to it!

 

[Ryback]

Personally I'm just amazed that carrying 16 health into map22 can now be considered a viable tactic for Nightmare...

 

Also, hi Panter! Not sure how I never noticed you posting here before. Your 30nm4949 is literally the one reason I started speedrunning. So thanks for that, and also the huge work you did pushing the state of the art forward in 1997 and 1998. Also, since I know you were in touch with Steffen Winterfeldt, did he ever clarify the status of 30nm6520 with you? Always been curious about that...

[PanterD2S]

16%  + Soulsphere + Armor...and then, not sure too... but with all this wallgliding, skipping of rooms, rocketjumpingNstuff maybe just consider it che...erm..magic! :D  (According to Clarke´s 3rd law "Any sufficiently advanced technology is indistinguishable from magic.")

 

Oh and I didnt post here too often yet (and If, its way too much for people to read I guess - true WoT´s^^) -  lets blame Zeromaster!

...only doing one NM Run improvement every couple of months! :P

 

 

From SW I got about 1 answer email in which he said he didnt know about / why there are these "1 GT UT´s" throughout the .lmp too. And I think I didnt get much about the other questions I asked,.. about keyboarding and why it gotta have to look so awfully slow^^ - but really not sure about this anymore- 20 years ago folks, but I  think there wasnt too much information from him at all. Except claiming it wasnt cheated. However those people I really had conversational emails with went on saying it must be cheated. (never "checked" anything myself - maybe ask Uwe Girlich (or Sven Huth?) who was the Techguy at these early times I remember. I believe in people telling the truth. So I was lost when 1 guy saying "no I didnt cheat" and 20 on opposite saying/believing he did, kinda had to leave this discussion - still somewhere in my heart feeling for the weaker side, but just too many arguments against.  Actually there wasnt any discussion at all, but probably I should be proud I got an answer, kinda being the only one being "touched"...mhm..  lemme sleep over this...

 

The only thing I can clarify for sure after 20 years (!) ->

...that 65mins are incredible boring ("PITA") to watch, and I kinda had to go through this more than once.. ;)    

..happy people these days, can watch 3x zeromaster instead :D

 

[Ryback]

Thanks for the detailed reply... so there's no certainty on this, which is pretty much what I thought.

[PanterD2S]

"There´s never certainty on anything, anytime"  Eternal Rule #1

[Jonathan]

ZeroMaster, that really was an incredible watch. It blows my mind that it's possible to play so aggressively and incorporate so many tricks and still be successful. I remember when I found the Map09 RJ thinking that it was useless because nobody would ever be crazy enough to include it in a UV run, let alone a NM run.

 

WRT to Winterfeldt, he and Girlich were friends, attending the same university, when they did most of their Doom playing and demos. Although Girlich is the most associated with demo tech, it was actually Winterfeldt who did a lot of the initial reverse engineering. He is credited as such in the LMP format description. Obviously, we will never know for sure, but the fact that they were both skilled at programming and reverse engineering, both definitely were investigating Doom and demo playback, and both produced very suspicious demos around the same time offers strong circumstantial evidence, at least. And, as I understand it, people such Adam Hegyi, Yonatan Donner, etc. did a lot of research around detecting cheated demos and were pretty certain that a number of demos from both players were not authentic. But it is all ancient history now.

[Looper]

I checked Winterfeldt's 30NM6520 and it is 100% cheated in my opinion. A lot of save states used. Some discussion happened @ Discord and people noticed that he cheated his E1M4 UV-speed, Map02 UV-speed and the failed 19NM run. I haven't looked rest of his demos.

[PanterD2S]

If I find that old Email some time on one of the old Harddiscs Ill happily sent them over to you guys (but doubt Ill find motivation to do this - the pile has become quite big with years).

He clearly stated "he did not cheat", at some point - quite sure about that.  But it seemingly wont matter too much with these walls to run against anyway ;) - basically the same as it was back in those times when people with more Tech and .lmp knowledge were looking at his stuff. One gonna have to trust their analysis if not want to get into it yourself and findout.

 

Talking about "cheating" - what about this "SR50" stuff nowadays?  I hear people say its commonly used today, but as far as I can remember this wasnt possible to do (together with Turns in the same line or something) with Good ol Doom unless you use a special mousedriver and therefore seen as cheating already...some days ago I found this old page from times I was playing and wondered about that.

Edited May 16, 2017 by PanterD2S

[Looper]

3 hours ago, PanterD2S said:

Talking about "cheating" - what about this "SR50" stuff nowadays?  I hear people say its commonly used today, but as far as I can remember this wasnt possible to do (together with Turns in the same line or something) with Good ol Doom unless you use a special mousedriver and therefore seen as cheating already...some days ago I found this old page from times I was playing and wondered about that.

You can do SR50 even in the original doom. Either by using your mouse strafe OR by pressin 3 keys the same time: strafe left + strafe on + turn left. It is not possible to turn while using SR50, EXCEPT 

 

[Keyboard_Doomer]

Looper posted before I finished my post but anyway...

 

Yes, SR50 on turns is not possible (other than what Looper linked) and it's an obvious sign of cheating if someone has SR50 tics on turns. Automated SR50 is of course another cheat and very easy to detect but otherwise SR50 is easy to achieve in both Doom.exe and any source port and is widely used nowadays.

 

I can say I actually find it a bit silly reading the old text files how using SR50 was avoided or people felt the need to point out "I only used SR50 for that one jump" when it's a legitimate movement you can achieve in the original exe without any hacks and SR40 which is not so dissimilar was used by everyone.

 

Regarding SR50, PrBoom+ since version 2.5.1.4 has a specialty and that is two-key SR50 which is on by default and you can't even turn off this "feature" in the official builds. This is generally not considered cheating because you can achieve somewhat similar behavior in unmodified Doom.exe if you bind more actions to one key and also possibly because almost no one had an idea this was introduced at some point into PrBoom+ and the community doesn't want to retroactively declare hundreds of demos from last few years illegitimate when the speedrunners' didn't even realize they were using this. However, many people still don't want to use this and don't like that it can't be disabled so it's a reason (or at least one part of the reason) why they either use older version of PrBoom+ or a build which has this behavior disabled by default.

 

Here is the thread which lead to the "exposure" of two-key SR50 in PrBoom+ and here is dew asking directly PrBoom+'s developer why was it introduced into the official build.

 

EDIT:

Seems like my "generally not considered cheating" sparked some discussion in Discord so yeah, maybe not "generally not considered cheating" but there doesn't seem to be support to declare the old demos which used this without knowledge to be illegitimate. Should the new demos with two-key SR50 be considered illegitimate? There seems to be an agreement that using it is not OK for speedrunning but as long as it is in the official build without the option to turn it off we will probably keep seeing demos recorded with it.

Edited May 16, 2017 by Keyboard_Doomer

[Jonathan]

Yeah, I think regular SR50 got conflated with automated SR50 in the attitudes of various speedrunners in the 90s, and so there ended up being a lot of strange antipathy towards it. Quite a contrast with modern speedrunning, where any and every weakness of the engine is hugely exploited, no matter how ugly the result is to watch.

 

Does anyone still have a working example of a SR50 mousedriver? I had one, but lost it a long time ago on an old computer. Would be nice to archive one somewhere like the Doom Wiki, as it's an interesting artefact of history, and seems kinda pointless trying to keep them under wraps now, when there are far easier ways to cheat.

[TimeOfDeath666]

You can disable the two-way two-key sr50 in pr+ by also binding a turn direction to strafe on. That way you will only be able two-key sr50 in one direction like in vanilla.

[Memfis]

On 15.05.2017 at 8:38 PM, Looper said:

people noticed that he cheated his E1M4 UV-speed, Map02 UV-speed and the failed 19NM run

How?

[Looper]

3 hours ago, Memfis said:

How?

Obvious save state usage in 19NM. Map02 UV-Speed seems to have couple save states too, most noticeable right after picking the blue key. E1M4 in 0:28 by Steffen might be legit, but the mouse jittering looks extremely questionable. It is the same shit as in Uwe's cheated E1M4-029, and those two demos were done almost the same day too. If you check Steffen's E1M4-024 UV-Speed, the moment he reaches the yellow key, his movement pattern changes to extremely smooth (looks like copypasted frames) and he hits everything frame perfect: The bridge switch, the yellow key door (yes, even that was perfect), exit door + exit switch, and time ends in .97. He was a bad player who cheated.

[NinjaLiquidator]

I dont like when I watch some demo and there are those "jump turns", it is impossible to watch. Happens in TAS. But only some of them.

[PanterD2S]

Quote

Should the new demos with two-key SR50 be considered illegitimate?

Who you ask?  me?  Im quite sure I never used it, so that be a "yes" then I guess.^^

But this may "spark too large discussions" - so we better not...

 

 

Quote

There seems to be an agreement that using it is not OK for speedrunning but as long as it is in the official build without the option to turn it off we will probably keep seeing demos recorded with it.

I feel there might have been a change in how people see it, yes - Still this often is one of the main concerns and kinda puts down my enjoyment of this new freakstuff I recently saw in youtube vids - there isnt even always a .lmp at all I guess.

"How much of this piece of Art would run on old doom.exe (the "official build", no?) and be kinda totally the same considering (legal) conditions I used to play? What tricks would have been possible (me being too stupid to find back in those days) which wouldnt be possible at all?".

 

If one really want to compare speedruns you need the same basic conditions, kinda logical Imperative, no?

Incomparability with a certain impact is a huge pull down, but maybe thats just me. 

 

Advantages about different "basic settings" like weather often are limited when it comes to "real" sports .

- In a "Speedrun" of 100m, you wont beat Usain Bolt´s record if tail wind exceeds 2 metres per second...or downhill too many degrees I guess.. or drugs ...but of course you can compete against those other 9 guys around at that day. Even in a Hurricane..

So the SR50 thingy seems to be accepted to be *within* what is considered "not too much of an advantage",  erm...But actually its how many % ?

about 5-10% If I recall that right? anyway a bit much for my taste considering UV or NM Speed tables...ok...already feel Im whining too much already ;)

 

But even worse - Comparability in general became a huge problem in "modern games speedrunning" imho at least as far as I can see...

Its mainly connected to "money spend in Hardware" but with kinda really *huge* impact here. (surely not talking about 5% anymore here)

"Look up to the sky, less lags, faster running" the Fallout guy said, I lol´d and thought "or buy a new graphic card and look where u walk":D

Fortunately the demotivation fully sets in *before* even trying, simply knowing its impossible unless you spent xyz $´s.

 

 

[fx02]

oO Panter :P hi mate :D

[PanterD2S]

Hey fx !

familiar faces f-finally ;)

[eLim]

Doom II NIGHTMARE! in 22:05 [30nm / NM-Speed D2ALL]

Nobody weeps more than I at the fact this demo is purely garbage. That being said, thanks for letting me skip half of map17 HEhEheHEheheehe

 

30nm2205.zip

Edited August 18, 2017 by eLim : Added video URL

[eLim]

CAUTION: VERY L4RGE MEME ALERT

Doom II can be finished pretty fast now:

ITYTD-Speed D2ALL in 18:16

 

The lack of rocket jumps comes from skill1 denying damage, which also denies momentum from damage caused. RIP

30it1816.zip

[ZeroMaster010]

Did someone say doom 2 d2all?

 

d2all uv-speed in 18:48

 

Such a good run, except for map 10 which was my own fault and map 30 which was unlucky.

 

 

30uv1848.zip

[Ryback]

3 new movie records in ten hours... this is just an embarrassment of riches.

[eLim]

7 hours ago, Ryback said:

3 new movie records in ten hours... this is just an embarrassment of riches.

I don't know if a skill1 movie really counts :P

 

Already conversed last night, but nice work ZM! Super slick minus the catastrophic map30 spawns haha

[termrork]

congratulations to both of you guys! this is incredible!!

[j4rio]

14 tyson 6:37

15 tyson 8:47

ty14-637.zip

ty15-847.zip

[rehelekretep]

damn, you made that look easy aside from that 15hp scare around the chaingunners in m15 :D

[j4rio]

16 tyson 5:39

ty16-539.zip

[4shockblast]

Dumping demos from random time periods in the past year.

 

Doom 2 map 1 nomo in 5.09.

no01-509.zip

[j4rio]

17 tyson 7:08

ty17-708.zip

[vdgg]

1 hour ago, j4rio said:

17 tyson 7:08

Last 3 chaingunners, crazy! My way of thinking would have been: "I'm 6 minutes ahead of the previous record, why risk losing this run? Let's use pistol".