Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Qaatar

Compet-n and port nuances [split from misc]

Recommended Posts

myk said:

Was that due to mouse acceleration? By default, if you run DOOM on Windows 98 the standard mouse driver (at least for serial mice) uses acceleration. Installing another driver (I've always used a old Genius driver that's compatible with DOS, Windows 3.1 and Windows 9x) gets rid of that.

I think it was the lack of acceleration (mouse was PS/2), which forced me to place the sensitivity setting in the cfg way beyond the default range. I had to pick something between either not being able to make a 180 turn without going across the whole pad or not being able to make the slightest movement without turning several notches on the shorttics scale. When I found the sweet spot it still took a while for me to get used to it.

Edit: Yeah it was the lack of acceleration. I just remembered someone's suggestion of using Intellimouse drivers for accel helped things a bit.

Share this post


Link to post

Interesting technique. I'll see if I can make use of it, notwithstanding my meager musical knowledge. Regardless, it seems only slightly more helpful than an intuitive grasp of time (often I can tell I am not beating my time by how things are going) and it's clearly not as efficient as an exacting timer.

The music is exactly the same as a timer. It plays the same length every time. All you need are a couple checkpoints.

There are different ways to enjoy things. Immediately speaking, no one likes it when a demo doesn't succeed, it's a disappointment, but people do like to succeed at something by certain established rules without needing to bend them. Sure, if you ease the rules, it's easier to succeed, but it's a game, not something our daily bread depends on. Ease versus achievement: the two types of enjoyment often contradict each other.

Needing to bend the rules my ass. I learned to play on ZDaemon with some non-vanilla features. It's how I've always played.

Yes, I know there are differences between vanilla and ports, and it's hard to compare two times that are exactly the same or very close if one was recorded in a port and one was recorded in vanilla. But if the times are far apart, then I really don't think there's much of an issue at all, and people are just making lame excuses for why "port" time is much faster than "vanilla" time.

Like I said, I started playing on ZDaemon and there's a monsters stat there (even though it was broken at the time), secrets counter and timer in the automap, bindable weapon keys, etc. Toke told me about PrBoom after I did my first ever demos (which were on coopbuild) and I stumbled upon the HUD while changing screen sizes and thought it was cool that there's a monsters stat in PrBoom too. I hadn't played vanilla Doom in years, and I never played it competitively. I didn't remember what features were available in vanilla, and it never occurred to me to check because I assumed everyone used the HUD.

Does it not take more work to create a demo with the same time using 320x200 software and no HUD? There's no merit there? Remember what you said: you aren't trying to compete with vanilla Doomers, and you do know there are differences with vanilla.

Yes there are differences, but to me it seems like these differences are made to look like mountains instead of mole hills. Comparing TAS to non-TAS is a mountain. Comparing port to vanilla is a mole hill. That is my point. Does it take more work to make a vanilla demo? NOT NECESSARILY. More work without the HUD? AV map29 = NO, AV map26 = NO. I tried changing my resolution to 320x200, but it won't display correctly in prboom+ or zdoom (the display is squished way up to the top of the screen with a bunch of ugly junk underneath. The display takes up like the top 1/4 of the screen, the junk takes up the bottom 3/4 of the screen).

I play with 640x480 software but I'll use 320x200 without HUD for a new av map26 demo if someone can say how to get 320x200 to work.

Share this post


Link to post

Wouldn't it be interesting to compare what kind of times you get by recording normally and then limiting yourself to vanilla, of course trying your best for both? Ofc it wouldn't be very objective truth, but would be fun to speculate. For the sake of science I might too consider setting up prboom+ finally and pick up a map for max-run.

Share this post


Link to post

TimeOfDeath said:
The music is exactly the same as a timer. It plays the same length every time. All you need are a couple checkpoints.

A couple of checkpoints? Like you noted, it is more useful to a musician or musically gifted person than anyone else. I don't distinguish musical parts too well. A timer offers one checkpoint per second, without preparation, to anyone.

But if the times are far apart, then I really don't think there's much of an issue at all, and people are just making lame excuses for why "port" time is much faster than "vanilla" time.

The aim of beating the new time with a different port or set of features is still not comparable to the other port or set of features, regardless of existing times.

Does it take more work to make a vanilla demo? NOT NECESSARILY. More work without the HUD? AV map29 = NO, AV map26 = NO.

Given the varying circumstances in demos, it's not foreseeable whether the HUD will be useful in any one recording, in almost any level. The usefulness can be as simple as confirming whether a semi-distant monster killed around the corner with a flick of the wrist is dead.

PS: I'm not sure about getting 320x200 on your monitor... maybe the render_screen_multiply setting in the CFG can help.

Share this post


Link to post

myk said:
The aim of beating the new time with a different port or set of features is still not comparable to the other port or set of features, regardless of existing times.

This sounds ridiculous for me.
Example #1: MM2 MAP01 UV Max.
Kimo Xvirus / Vanilla: 1:16
Skepticist / PrBoom(+) (unsure if HUD used): 1:15
For me, these are two outstanding demos of comparable quality (and believe or not, they're quite different from each other for such a small map). If vice versa the vanilla record was faster by a single second, I'd still state they're of comparable quality.
But if you say a PrBoom+ demo with complevel 2 which is 30% faster than Vanilla demo is not better than Vanilla demo in terms of player skill*, hiding behind an argument of not being comparable, I can only express my disbelief. What about a Chocolate Doom demo faster by 30% than vanilla, then? Can't the Choco player say he did a better run?

*unless it's a weird, very specific map - I can't recall any example

Share this post


Link to post

Forget about better, worse or quality. TimeOfDeath said that the differences didn't matter once the time was improved by a large margin. What I said is that the new time isn't compared only to what it beat, but also to any time that could beat it, which can be a slight improvement, thus affected by minor feature differences.

In other words, since PrBoom+ has more features to help the player, how many seconds faster, per minute, does a demo have to be, to be comparable?

Since you have to factor in the additional features, it's hard to tell when, objectively speaking, a demo from one engine really beat the other.

Share this post


Link to post

I just meant that if a time was improved by a large margin, I think the differences between ports/vanilla become much less of an issue (since recording in vanilla would still most probably result in an improved vanilla time). I don't know why you're being such a nazi about comparing ports to vanilla. ;) We're not comparing apples and oranges, we're comparing two kinds of apples.

And on the two maps I mentioned, I know for a fact that the HUD isn't a huge advantage, and I gave examples for av map29. In my av map26 demo, it was the first time I entered the red door, so I didn't know how many remaining monsters would be left on the HUD. So I didn't even use the monsters stat until all were dead, which only saves a matter of seconds on that map.

I only use the timer for checkpoints anyway, so using the music instead wouldn't be a problem.

Share this post


Link to post

Ok, I posted a new av map26 demo in the misc demos thread with:

-320x200
-capped framerate
-no HUD
-no sound when you find a secret

For a few minutes it was weird using the status bar (haven't used that in almost 6 years) and the wide resolution, but everything felt fine shortly after. ZDaemon has capped framerate, so that wasn't an issue. The resolution wasn't an issue. The HUD/secret sound weren't issues. I also played with -nomusic, so no timer.

Apples and apples.

Share this post


Link to post
TimeOfDeath said:

Ok, I posted a new av map26 demo in the misc demos thread with:

-320x200
-capped framerate
-no HUD
-no sound when you find a secret

It's not such map to emulate vanilla feelings. 1.9 can't visualize 65 and more things simultaneously. It's important for av26.

Share this post


Link to post

For the love of christ.

Then how do I emulate that with prboom?

Btw, I learned to play on ZDaemon with all the online monster bugs, like invisible revenant missiles, invisible mancubus fireballs, monsters that attack while their movement sprites are displayed (not their attack sprites), etc. Beat that vanilla doom.

Share this post


Link to post

why would you want to emulate vanilla doom's crashes and overflows? that is well beyond anally retentive nitpicking.

Share this post


Link to post

Maybe it'll stop the excuses. ;)

I'll even play in Chocolate Doom with the default weapon keys that I never use. But the mouse feels different. Is there acceleration or something? How do I make the mouse feel like the mouse in prboom/zdoom/every other port?

Share this post


Link to post

TimeOfDeath said:
Then how do I emulate that with prboom?

Use Chocolate Doom. Its chocolate-doom.cfg allows one to bind weapon keys as you do and to adjust acceleration properties, so you should have no problems. It also lets you properly fake 320x200 with a screen multiply feature, if you want to use a computer with limits on lower resolution.

Share this post


Link to post

Are you sure you can bind weapon keys in choco doom? I downloaded the latest version but am not sure how to do it. I accidentally pressed Q to switch to rocket launcher and the game quit.

Share this post


Link to post

Hmm, that's messed up. I'd suggest making a Chocolate Doom feature request or bug report (you might need to create an account on the tracker) to add the "quit demo" key as one of the bindable keys, as it doesn't seem to be in the CFG. It makes no sense that that key can't be changed, while the others can, so it's a bug, considering the enhanced binding feature.

You might have noticed, but to bind keys normally hard-coded in vanilla, you have to set vanilla_keyboard_mapping to 0.

Edit: I've reported the issue. You can see the report here. Once it gets fixed, you can ask exp(x) for a new build with the fix.

Share this post


Link to post
Eugene said:

It's not such map to emulate vanilla feelings. 1.9 can't visualize 65 and more things simultaneously. It's important for av26.

Really? I've watched the 13:00 demo in DosBox and I haven't noticed a single spot where the sprite limit would've made any difference.

Edit: It's barely noticable in Vile's demo as well (shooting the revs out the window and that's it).

Share this post


Link to post
TimeOfDeath said:

Ok, I posted a new av map26 demo in the misc demos thread with:
...

hehh, your 13:00 demo partially proves both sides right.
1) due to absence of the smart kill counter in HUD you didn't beat the 13min mark. that partially confirms myk's claim the HUD matters in optimized max runs where every second matters.
2) low resolution and capped frame rate have almost zero effect, that argument is purely theoretical. you've beaten your own hi-res (and HUD using) record and shredded vile's c-n record to ribbons. your demo IS NOT comparable to his, becaus it is superior over his(time wise, route planning wise, monster/weapon handling wise, take your pick).

sooo... the smart HUD makes some difference, but there's other port features that just provide modern game standards for doom, like hi-res, key mapping, etc. i believe some common sense has to be used at this point. your demos don't qualify for a c-n record, because they obviously don't observe the vanilla exe only rule. on the other hand it can be safely declared a world record on av26, because from where i stand it all the compatability requirements are met, it's a fastest time so far and the port-related TAS claims are very weak.

unless someone accuses you of using the cheat features (slo mo, segments, 2x iddt) the features you're using are just making the game sensibly up-to-date with modern times and less annoying to play without actually cheating, imo. the hud issue is fuzzy, but i don't see a problem when not recording for compet-n.

Share this post


Link to post

I wouldn't draw such precise conclusions. Whether something is of benefit doesn't just refer to what's possible as an eventual goal or not. You can always get better times with more effort, unless you're reaching an optimal point (rare except in short speed runs.) I'm sure TOD can beat almost any time of his regardless of the PrBoom or vanilla settings, but it will tend to take a bit more time with the latter. The difference is not only in what you can do, but also in how much work that requires. If you need 1 hour to get 13:00 in PrBoom but 1:10 in vanilla, there's a difference. Benjamin Franklin said time is money, and for DOOM we can say time is time ("time spent" and "demo time").

A possible exercise to visualize this is to continually challenge one's own times alternating between the two engines (or settings standards, using only PrBoom). In the long run, which cancels out luck and moods, and keeping track of total recording times, one will likely notice that PrBoom will beat vanilla quicker by some degree.

To clarify the focus of what I'm arguing: Vdgg thought I was talking about demo quality above when I talked about demos being comparable, but I'm referring to the fact that there's no real, accountable or measurable way to compare two demos recorded in a different way, even when the difference is small, because they were produced with different tools. What is the effect of any additional feature in any and every circumstance? It varies and could only be gauged statistically with a lot of recording data. By chance the HUD could save you a minute in one attempt, nothing in the next, and three minutes in another run. Making a hyperbole of it, it's like comparing a photo and a painting of a thing. There's a measuring problem which is solved either by using the same settings or by arbitrarily deciding to accept differences. Arbitrarily, you could do it by selecting either PrBoom or vanilla as the standard. And that is what is being discussed here. Are we playing Doom or PrBoom? (Note my intentional lower case spelling; I spell the game in general as DOOM.) If both count or we say "who cares," we are using PrBoom as the standard because it includes vanilla behavior as an option, but not vice versa. Yes, apples and oranges. And I'm not telling anyone to play one or the other; my only encouragement has been a transparency as to what engine or features are used.

Share this post


Link to post
Belial said:

Really? I've watched the 13:00 demo in DosBox and I haven't noticed a single spot where the sprite limit would've made any difference.

I didn't watched 13:00 demo and I didn't watched Vile's run in 1.9 in years. But I'm sure when any player chooses route to the rk door and lets knight leave their niche, he gets the sprite limit distinctly. The revs cages are second sprite disappearance example here.

Btw. I don't understand the problem. Why someone tries to emulate vanilla doom recording process in PrBoom-plus? Wants to prove that his demo should be comparable with CN demos?
If someone is able to get a faster demo then current (often aged) CN record for av26 or av29 or any other map at 320x240, with no kills counter and so on, why he doesn't simply record a demo with 1.9 exe under DosBox?

Edit:

I've red some posts in this thread. I should do it early though.

I could agree with Myk. If someone makes efforts to optimize most of CN max records for any wad except D1 and D2, he beats these times. So it's really matter of time and motivation.

Using of advanced PrBoom-plus features is a subjective thing. Someone guesses that HUD, HiRes and so on makes no difference. The other don't think so.

I make an assumption. May be another one guesses that using 85-95% games speed or re-recording parts of a demo makes no difference. He could say I can beat these times so why I should spend hours to polish second half of a demo when I run fast enough through the first half constantly. So he may decide to optimize only second half of demo via partial recording.

Understanding what an advantage is different between players. So I think if someone wants to compare his demos with previous runs he should use the same exe.

Share this post


Link to post

I never use doom2.exe because I don't use the default weapon keys and I like the HUD. I thought DosBox was illegal for Compet-N? And it lags on maps with many monsters on older computers? I've been using Chocolate Doom and it lags sometimes too (for a couple seconds then it's ok, almost randomly). I did 14:16 first exit in choco doom with the default weapon keys, but it was slow.

I understand the sprite limit now, and I agree with Belial that it's not a big deal at all in av map26. The monsters' projectiles always appear, so you can still dodge them, and they still make sounds (so you can hear when an invisible monster dies). If you look at the monsters from a different angle they will appear.

I've played this map with prboom/HUD and with choco/noHUD and I really don't think there's much of a difference. All of the vanilla limits like lo-res, capped framerate, sprite limit, no HUD/timer haven't had a negative affect on my playing. It's just the default weapon keys that I'm having trouble using, and I'm paranoid that I'll accidentally press "Q" (my rocket launcher key in prboom) and the game will exit. Also, the mouse feels a bit different in choco. I turned off acceleration but it feels way different from turning off acceleration in prboom, so I set the acceleration to 1.000 and it feels better (default is 2.000). I don't know what the threshold means.

I agree with dew about using common sense.

Share this post


Link to post
TimeOfDeath said:

I thought DosBox was illegal for Compet-N?

No, it never has been, although admittedly this was never an issue while Compet-N was still active. It is banned at Speed Demos Archive, as Dosbox is considered an "emulated environment". But then they also ban Wine, as it's a virtualisation layer, and probably even command mode on Windows XP. Adam Hegyi has validated a number of records done in the latter two environments, so I doubt there'd be an issue with Dosbox either. From memory I think people knew about the slight slowdown under XP, and this was considered okay.

TimeOfDeath said:

And it lags on maps with many monsters on older computers?

That's a Doom2.exe issue, not Dosbox.

I don't really have a strong opinion about this debate, but I think the main vanilla limit is simply emulating 1.9 behaviour accurately. Enforce that as a standard and we're already a long way toward making port demos and vanilla demos comparable.

Share this post


Link to post

Is the XP slowdown issue with DosBox or from just using doom2.exe (without dosbox) with XP (I remember Anders Johnson saying something about skogsto's run in the compet-n thread)? Is chocolate doom close enough to doom2.exe for comparison? I downloaded dosbox (where can I get doom2.exe?) but am not very smrt. I'm trying to record one last demo on this map but am not sure what exe to use. I've been using chocolate doom and don't want to change ports again and setup dosbox, but I will if I have to.

Lagging because of many monsters won't be an issue. Some of the zdoom maps I made are barely playable on my computer because of low fps, but I'm used to it.

Share this post


Link to post

I think that HUD is unfair, but TimeOfDeath didn't deny using it, so I only wanted to blame him.

I can always ignore difference between port/no-port. Over 15 years have passed since Doom was released, so vanilla exe is tricky for some players, or some players can't stand vanilla's environment. Sometimes, I invite some youtube doomers to speedrunning. When I do, I usually recommend their to use PrBoom+. Existence of ports which have demo compatibility make people enter speedrunning easily. I'm very happy about it. Everyone shouldn't blame for using your favorite exes, your favorite screen size, your favorite key config, and so on...

HUD is very different from the original status bar. You take a lot of information by using it. HUD is acceptable in this map because HUD doesn't give huge advantage in this map, but HUD is unacceptable in that map because HUD give huge advantage in that map. Who can make clear the division? It doesn't at all make sense. Basically, HUD shouldn't be used. As a result, it is the most impartial.

TimeOfDeath, I think you need not bother to use chocolate doom, and you should use PrBoom+ again (without HUD though) if you like.

At least, when someone who uses ports beat my record recorded in vanilla doom, I will make every effort to improve my dooming skill before complaining something. I'm still a unskilled speedruner. The cause of which my record is beaten cannot be very trifling difference between port/no-port.

Share this post


Link to post
Guest
This topic is now closed to further replies.
×