Discussion of port vs vanilla recording et al. (split from Scythe thread)

[Phml]

My point was that nobody can claim to beat or improve on Vile's record if they do it with compatibility settings that make the map much easier.

[kmc]

Phml said:

My point was that nobody can claim to beat or improve on Vile's record if they do it with compatibility settings that make the map much easier.

Let's not ignore the fact that the demo was made 8 years ago. To be fair, let's all use out of date computers as well. Can't have our lower latencies and higher dpi mice giving us the upper hand now. I'd say 90% of the new demos out there use prboom-plus; things change. Most players play with higher resolution, higher gamma, uncapped frame rate, etc. Surely those make the map "much easier" as well.

[dew]

NoobBait said:

I'd say 90% of the new demos out there use prboom-plus; things change. Most players play with higher resolution, higher gamma, uncapped frame rate, etc. Surely those make the map "much easier" as well.

you're missing the point. complevel 2 will still produce the all-ghosts bug, so this is not a case of vanilla vs. prboom+ argument. the talk is about changing the complevel to something else than 2, but using another well-defined ghostless standard like 9 will change the gameplay itself and with such a map as sc30 the demos will become incomparable.
so the other option is to use -1 with all settings of 2 except the intercepts overflow, but that means the demo would only be playable in prboom+ and you'd be surprised how many people use other ports. looking at the scythe entries from last 2 years, tatsurd and kimo recorded on 1.9 exe and graim used choco.

edit: spelling

[kmc]

I understood the point from the get-go. My point was that we already compare prboom-plus demos to vanilla demos, which is almost an equally unfair comparison to complevel 2 and 9 comparisons.

[dew]

i disagree. cl2 pr+ demos are replayable with vanilla exe, can't say that about cl9 demos. what you propose would lead on a slippery slope and soon enough we're recording zdoom demos, because it lets you surf on top of monster without getting blocked. prboom's passive improvements give a small advantage, but i'd like to believe there's a middle step between compet-n's rigidity and free for all anarchy.

[kmc]

I agree with you, but there is no real standard of demos anymore. If there is, it is entirely subjective. How many people have the kills/items/secrets HUD turned on while recording? How many use TAS features dishonestly? The community is already on a slippery slope. That middle step surely exists, but I feel clearer boundaries are a must.

[Rizera]

I think this might be the tenth time I see this discussion in these forums. Prboom is a way more comfortable way to record demos and that's the reason I record in there. If I could play in Vanilla with the same "technical quality" I do in Prboom, all my demos would be recorded in 1.9. I just can't do it.

I think that if a wad is vanilla compatible, it should always be recorded with cl2 keeping the original Vanilla Doom aspect. At least that's how I do. Allghosts is just one more thing to make that Scythe MAP30 tough :P

Of course Prboom gives the player advantages to record in comparison to Vanilla but all original Iwads records have stood there for years and no one beats them, neither using Prboom nor Vanilla.

[vdgg]

NoobBait said:

I understood the point from the get-go. My point was that we already compare prboom-plus demos to vanilla demos, which is almost an equally unfair comparison to complevel 2 and 9 comparisons.

Yes and no, in both cases (case 1: vanilla vs. Prboom+, case 2: complevel 2 vs 9). And maybe let's move this discussion somewhere [/admin hint] ?

Case 1, all is my personal opinion

1. Kill/Item/Secret/Timer HUD. TOD was defending the point there's no difference and "kind of" proved it by recording outstanding UV Max/UV -fast demos.
Kills counter does not matter in many, many maps in Max category derivatives. In some maps, it makes the things easier: less time spent on recording, as you know when to quit if you missed some monsters in the middle of a demo. I can also imagine examples of an extreme edge for HUD users: difficult -respawn maps with many monsters. With the counter you know exactly when to exit, without it you toss dice, exit now and risk <100% kills, or wait?
Timer - I listen to music to judge if a long demo is "going well". dew would say if using the timer is such an abuse, I don't think so.

2. Higher resolution. I'd say it doesn't matter in most cases again. Depends on how any of us is bothered by lower resolution (320x200). Myself, not really, but at least in m2092320 (pacifist) I'd say I exploited it (distant sergeants which are almost invisible in 320x200).

3. VPOs etc. I will never ever try Requiem MAP25 Tyson in vanilla. Period.
Respect to Paska for his inf-lutz demos in which he was facing a wall, standing "backwards" in order just not to get VPOs.

And so on, and so on. My main point is that in some maps, some categories, for some players, it is 100% equal. And demos are not only comparable, players can compete against each other, one using Vanilla, other Choco, other Prboom. "Some" means actually for me, "most cases". What kind of edge you have in, say, Doom II MAP01 in any category if you use PrBoom+?

Case 2, vanilla vs boom compatibility

Here there is no difference in very few cases I'd say. Get a map with
- no arch-viles
- no PEs
- no monsters on ledges
- no abnormal behaviour like in Vile MAP17
- no doors :) (Vile MAP01)
and you can say the demos are comparable.

[kmc]

What about mouse "smoothness" regarding vanilla and prboom-plus? I recorded a couple compet-n entries in vanilla, and I found the mouse to be terribly jerky. Is this a compatibility problem, or is the mouse support corrected in prboom versus vanilla? From my personal experience, I could not achieve similar results in vanilla as I do recording in prboom-plus on ANY of my past demos. Perhaps this is a misunderstanding based on my own bad experiences recording in 1.9, but if compet-n speedrunners were able to play with equal mouse smoothness to what I have in vanilla then perhaps my sense of "awe" at their willpower and determination was a bit less warranted.

[Phml]

Let's not ignore the fact that the demo was made 8 years ago. To be fair, let's all use out of date computers as well. Can't have our lower latencies and higher dpi mice giving us the upper hand now. I'd say 90% of the new demos out there use prboom-plus; things change. Most players play with higher resolution, higher gamma, uncapped frame rate, etc. Surely those make the map "much easier" as well.

Careful with your sarcasm ; you might just find the person you're directing it to actually believes just that and was fairly obnoxious on the topic not so long ago. ;)

That is, that all that stuff gives an advantage, not that we should all use old stuff. You're quick to make the jump from one thing to the other, whereas this is never what I implied. If your goal is to compare yourself to other players as a binary better or worse perspective and in public scrutiny, I would say that it should make sense to use the same settings as they did, but if you ask me the whole thing is shallow and not worth caring about ; that's yet another rant that I've annoyed people enough with, but what can I say, while I don't mind being considered stupid for my opinions I do mind being considered stupid for things that aren't my opinions at all.

Even then, even I would recognize a fairly wide difference between QoL improvements and something as drastic and fundamental as toggling the ghost bug in a map where its presence is likely going to block off the most efficient routes and/or force you to throw away many (most?) of your attempts.

[Looper]

NoobBait said:

My point was that we already compare prboom-plus demos to vanilla demos, which is almost an equally unfair comparison to complevel 2 and 9 comparisons.

Okay, that must be the worst statement in this year. I can't help myself from laughing :D Sorry, but I found that one ludicrous :)

1ntru said:

Prboom is a way more comfortable way to record demos and that's the reason I record in there.

Heh, I found vanilla more comfortable than Prboom, but I guess it's all about getting used to.

[kmc]

Paska said:

Okay, that must be the worst statement in this year. I can't help myself from laughing :D Sorry, but I found that one ludicrous :)

Yeah, thanks for being a douche about it too. At least the other people gave a legitimate argument.

[j4rio]

Eh, you guys could argue about pretty much everything. I just noticed map 26 max tables, and it's basically legacy demo beaten by vanilla demo beaten by another vanilla beaten by cl9 prb+ beaten by vanilla beaten by cl9 prb+ and so on. Changes are quite notable, those upper cybs may fall down from ledge and you get major infight boost with boom compat, but in any case I didn't notice any text file full of raging about previous recorder not recording in a saint way. I may be wrong and there could be a nice friendly discussion regarding this (flame summit) hidden somewhere around, it's kinda expectable after all. :P

Btw, awesome demo Vile. :)

[Memfis]

dew said:

so the other option is to use -1 with all settings of 2 except the intercepts overflow

I thought that complevel 2 doesn't override overflow settings. Am I wrong?
pppp.txt: (Peter Perv's Private Place)

Designed for PrBoom-Plus -complevel 2 with:

In General Options menu, turn Intercepts Overflow
Emulation to NO, and keep all other overflow
emulation options to YES.

[Grazza]

I think recording with -complevel 2 and intercepts overflow emulation turned off by default makes perfect sense. It doesn't make it any easier to record a vanilla-compatible demo, since if the overflow would have occurred (and destroyed the demo), then you haven't recorded a vanilla-compatible demo at all. But you have at least recorded a demo, which might be of value or interest to others, but which can't be compared directly with a vanilla demo. Of precisely what value that demos is, can be debated, but it is clearly greater than zero.

Regarding comments about demos from years ago, with differint choices of ports, etc., remember that in the early 2000s, there weren't many good options. Modern operating systems were becoming DOS-hostile, Dosbox didn't exist or wasn't very usable, and no ports emulated vanilla with any degree of accuracy before Prboom-plus (and even that took a period of trial and error before to iron out the last few significant glitches). So people understood that other players might use a port not to exploit port bugs/features and thereby "beat" vanilla demos, but because they simply wanted to record demos and were searching for the best way to do so.

Generally speaking, the demos community is and has been a friendly one, where we are most interested in seeing the achievements and ideas of others. Encouragement and sharing of ideas are the norm. Competition, if it exists, is normally of a rather friendly variety - at most we are rivals, rather than enemies.

Memfis said:

I thought that complevel 2 doesn't override overflow settings. Am I wrong?

Correct. If you choose a vanilla complevel then you can choose whether to have each type of overflow emulation enabled. I have spechits and reject overflow on, and the rest off.

[blob1024]

1ntru said:

I think that if a wad is vanilla compatible, it should always be recorded with cl2 keeping the original Vanilla Doom aspect. At least that's how I do.

same here.
btw if it's still considered "unfair" to compare under correct complevel but different ports, then I'd advice the players to help the pr+team to reproduce the original wanted behaviour correctly (if it's not done by now)

nonetheless, I agree 100% with what vdgg wrote, the hud using and the higher resolution gives some advantage, but don't compromise the comparison.

[dew]

j4rio brought up:
cl9 demos on map26

wow. tbh, i didn't even realize this, because i shy away from that map in fear (*pada psssh*). eschdoom's demo shows exactly why cl9 shouldn't be compared to cl2. at least 3 cybs fall off ledges and instigate infighting in which they perish, because their attention is diverted towards the player. huy's and richie's demos are actually a MBF complevel, so there might be even weirder gameplay changes at work like different infighting (AI helping allies and whatnot). esch's record and hui's fastest demo improve on method's vanilla time considerably, but he has a massive disadvantage - and his demo is contemporary with hui's (2003)! i feel like there should be a note in the DSDA tables warning viewers about the different complevels.

also grazza has a point that a VPO-less demo is better than no demo. :)

[Grazza]

dew said:

I feel like there should be a note in the DSDA tables warning viewers about the different complevels.

I agree - the demo format/compat is far more important than the precise details about which version of the port was used, yet that is the information that is mentioned in the table. Andy inherited this information from Opulent's site, and of course it would be a vast amount of work to change it now.

[j4rio]

Maybe not all have to be, especially demos of wads with mixed compatibility should be mentioned - like scythe2, or such.

[Looper]

Sorry NoobBait. Nothing personal. Sometimes I say straight what I'm thinking and it might sound (and is?) rude.

vdgg said:

Case 1, all is my personal opinion

Timer - I listen to music to judge if a long demo is "going well". dew would say if using the timer is such an abuse, I don't think so.

Problem with music is that it gets too boring after 10th attempt, so music is turned off :)
Although I think timer isn't big advantage because you can listen to MP3 which just whispers time like every 10 seconds or so, heh. (Unless listening to music other than doom's isn't allowed :O) Anyways it's pretty easy to say if run is going good or not after handful of attempts.

In -respawn runs killcounter obviously gives an advantage.

[tempun]

Grazza said:

I agree - the demo format/compat is far more important than the precise details about which version of the port was used, yet that is the information that is mentioned in the table. Andy inherited this information from Opulent's site, and of course it would be a vast amount of work to change it now.

Not so vast actually. If Andy was able to make site updates automated, he is also able to automate the process of determining demo format.