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

Compet-n Resurrected

Recommended Posts

exp(x) said:

Where did I label it as evil?

sorry, i'm just projecting after a few discussions on the topic in the speed demos forum. we did cover it from all angles and came to no consensus. :) i reacted to seeing it on the same line with aimbots and wallhacks... i use boom's smart counter and timer for all my pr+ runs and frankly, one of my reasons to avoid c-n was the lack of a timer.

oh, and phml's little remark is fairly spot on. i've already discussed server synchronization for valid attempts with anders back when he was interested in reviving c-n (two years ago now?) and not being able to record on a whim came up as a big obstacle.

Share this post


Link to post
entryway said:

with «-skipsec -1» you can rerecord very fast. one or two tries are not really detectable I think

Interesting. Presumably you can tie that to a key binding in the game so that you can "rewind" the game by a second? That's probably the best attack I can think of. Still, at the very least the scheme I propose would limit how far back you could rewind, and limit it so that it could only work once, maybe twice per run.

Share this post


Link to post
fraggle said:

Interesting. Presumably you can tie that to a key binding in the game so that you can "rewind" the game by a second? That's probably the best attack I can think of. Still, at the very least the scheme I propose would limit how far back you could rewind, and limit it so that it could only work once, maybe twice per run.

no, you have to quit and restart the game, but you have to understand how fast that process is when recording ultra-short uv-speed demos... i can grind a 10s run for many hours and restart a few hundred times during the session. it's such an onslaught of restarts that i'd worry about the server actually recognizing my VALID, SUCCESSFUL 10s attempt. :)

Share this post


Link to post

Doom Ports don't replicate the input in Vanilla Doom's fast start and possibly the start of 2nd, 3rd, etc.. maps in a movie run.
In Chocolate Doom and PrBoom, if players press movement keys during the splash screen and turn the mouse, the demo recorded has actually the first tic as a Wait Tick, 2nd tick has whatever the player pressed + the turn, that was applied because of the splash screen.
How it's done in Vanilla Doom, is that during the text startup, you could move the mouse and press/hold keys, the mouse turn is applied at the first tic, and the keys are applied as if they're held for 4-9 tics (varies but in DOSBox it's nearly always 5 or 6), then you get the splash screen where you can have input again that's applied normally for 1 tic.
For example here's a typical vanilla demo with fast start, Adam Hegyi's n1m1-009

#     1  (0.2s)         GF50 SL40 TL31 * Input During Text Screen *
#     2  (0.5s)         GF50 SL40      * Carried on Keyboard input from text screen *
#     3  (0.8s)         GF50 SL40      * Carried on Keyboard input from text screen *
#     4  (0.11s)        GF50 SL40      * Carried on Keyboard input from text screen *
#     5  (0.14s)        GF50 SL40      * Carried on Keyboard input from text screen *
#     6  (0.17s)        GF50 SL40      * Carried on Keyboard input from text screen *
#     7  (0.20s)        GF50 SL40 * The game appears in this moment during the splash screen *
#     8  (0.22s)        GF50 SL40 TL15 * Splash screen tic, varies from 5->10 randomly
                                         TL15 is the mouse movement that Adam did 
                                         during the splash screen *
#     9  (0.25s)        GF50 SL40
#    10  (0.28s)        GF50 SL40
#    11  (0.31s)        GF50 SL40
#    12  (0.34s)        GF50 SL40
#    13  (0.37s)        GF50 SL40
#    14  (0.40s)        GF50 SL40 TL1
#    15  (0.42s)        GF50 SL40 TL1
#    16  (0.45s)        GF50 SL40 TL1
#    17  (0.48s)        GF50 SL40
#    18  (0.51s)        GF50 SL40
#    19  (0.54s)        GF50 SL40
#    20  (0.57s)        GF50 SL40
Doom Demo Browser 2.0b2 displays the tics like these, if you have a Compet-N Demos mirror, you can point to the directory and browse the demos with ease.
This is the norm for the start of 95% of Vanilla Demos, there are
other demo starts with a pattern of their own, which I haven't
managed to find a way to replicate them, and I don't know exactly
what causes them. They could be only replicable under certain
OSes and certain conditions.
A few of them are

Xit's pl27-048
#     1  (0.2s)         GF50 SR40 TR8
#     2  (0.5s)         GF50 SR40
#     3  (0.8s)         GF50 SR40 TR1 * This is too early for splash screen? *
#     4  (0.11s)        GF50 SR40
#     5  (0.14s)        GF50 SR40
#     6  (0.17s)        GF50 SR40
#     7  (0.20s)        GF50 SR40 TR2
#     8  (0.22s)        GF50 SR40
#     9  (0.25s)        GF50 SR40
#    10  (0.28s)        GF50 SR40
Vincent's ev28-118
#     1  (0.2s)         WT * One WT tick like in PrBoom+ *
#     2  (0.5s)         GF15 SL40 TR2 * GF15 (novert off) and TR2 is from mouse *
#     3  (0.8s)         SL40
#     4  (0.11s)        SL40
#     5  (0.14s)        SL40
#     6  (0.17s)        SL40
#     7  (0.20s)        SL40
#     8  (0.22s)        SL40
#     9  (0.25s)        SL40
Vincent's nm14-043
#     1  (0.2s)         WT * One WT tick like in PrBoom+ *
#     2  (0.5s)         GF50 TR29
#     3  (0.8s)         GF50
#     4  (0.11s)        GF50
#     5  (0.14s)        GF50
#     6  (0.17s)        GF50
#     7  (0.20s)        GF50 TR5
#     8  (0.22s)        GF50 TR3
#     9  (0.25s)        GF50 TR1
#    10  (0.28s)        GF50 TR2
#    11  (0.31s)        GF50 TR2
#    12  (0.34s)        GF50 SL40 TR3
#    13  (0.37s)        GF50 SL40 TR2
#    14  (0.40s)        GF50 SL40 TR1
#    15  (0.42s)        GF50 SL40
#    16  (0.45s)        GF50 SL40
The start of these 2 of Vince's demos are like most PrBoom+ demos, it means it maybe possible to get the Vanilla Demos' starts to be like in PrBoom, but I know of no other vanilla demo like this. I remember a few other oddities from long past, but I haven't noted them.

Share this post


Link to post

The idea of using a server to automatically try to legitimitize runs is the wrong solution. It introduces two inconveniences: requiring a special port, and requiring an Internet connection; don't take the latter as a for-granted thing. It is quite easy to think of a situation where you may entirely lack connectivity but might be trying to do a new speedrun (imagine you're on a vacation, have a laptop, and get a bit bored in the middle of nowhere). Additionally, it will introduce all of the issues that modern Always-On DRM systems have -- it will assume the user has a perfectly stable connection all the time. It may work most of the time, but when an ISP or wireless router flake out, you do not want to find that your speedrun attempt is now completely invalidated due to a situation completely outside of your control.

I believe neither demanding the DOS vanilla doom(2).exe nor this kind of automated validation is the right approach. Just accept any demo that plays back in vanilla, whether it was recorded in Chocolate Doom, PrBoom, Odamex, Eternity, or what have you. Apply usual anti-cheat double-checking measures after the fact. Legitimate speedrunners should not be subject to having to prove they aren't cheating; cheaters are likely to be caught no matter what.

Share this post


Link to post
chungy said:

I believe neither demanding the DOS vanilla doom(2).exe nor this kind of automated validation is the right approach. Just accept any demo that plays back in vanilla, whether it was recorded in Chocolate Doom, PrBoom, Odamex, Eternity, or what have you. Apply usual anti-cheat double-checking measures after the fact. Legitimate speedrunners should not be subject to having to prove they aren't cheating; cheaters are likely to be caught no matter what.


I so agree with this. It's an era where doom.exe is truly a pain in the ass, and who's to say that dosbox is even running doom that accurately? If we want to split hairs and be this picky, Compet-n is dead before it even (re)starts.

Archy said:

I don't think


As i said, this isn't time for idealist thinking, this is reality. DOS is dead and DOSBox is no where close to emulating DOS correctly (they specifically state this!).

Share this post


Link to post

DOSBox emulates the DOS environment well enough that DOSBoxers vs DOSers have completely fair competition assuming both of the players' systems have optimal hardware and no tampering of software has been done (and people have been changing system environments to cheat [such as slow mo] since the beginning of gaming computers, it's no more of a threat now then then; It's just that now people don't need to change their system's environment to TAS anymore [PRBoom]).

DOSBox for anyone with basic DOS knowledge is no pain in the ass. I just have a DESKTOP icon that instantly launches Doom from DOSBox. It's literally just as easy as using Chocolate Doom.

If people get tired of typing constantly (and you'd still need to do that for running Chocolate Doom with command lime parameters), you can easily make short cuts.

For example, If I want to record a UV -RESPAWN demo for Requiem on map 21, instead of having to type in

doom2 -file REQUIEM.wad rq21fix.wad REQMUS.wad -warp 21 -skill 4 -respawn -maxdemo 999 -record RR
I could just type in
doom2 @RR 21
by making a Doom Response file named "RR" (no file extension)with the fallowing code:
-respawn
-maxdemo 999 
-skill 4
-record RR
-file REQUIEM.wad rq21fix.wad REQMUS.wad 
-warp
Very easy and hassle free.

Share this post


Link to post

Ease of use isn't the issue. Modern ports often need command line proficiency to use, as well; no difference from DOS in that regard.

The primary rationale of Compet-N requiring vanilla Doom only was to mitigate cheating attempts. It is rather difficult to pull off slow-down hacks or re-recording hacks on a pure DOS environment and doom.exe. These days, if you say "Just use DOSBox!", there is very little to stop me from simply attaching GDB to DOSBox and stepping through the program or changing the program state (additionally, there are re-recording DOSBox forks with saved states and all). It's slightly more difficult but essentially similar to using a port with such features built-in. If Compet-N forbids ports, it should forbid emulating a DOS environment as well (and how is this going to be enforced?).

Whatl I'm saying is that restricting how demos submitted to Compet-N is silly; cheaters aren't motivated to follow the rules in the first place and burdening the legitimate players is just plain out stupid. As long as the port is vanilla-compatible, that's all that counts as far as I'm concerned. (And checking that a demo is vanilla-compatible is as trivial as loading it in vanilla doom)

I can understand rules and why certain features from ports would be forbidden. PrBoom-Plus has a built-in wall hack and slow down functions. That's great for TASing, but obviously should be disallowed for Compet-N playing, and the mere (optional!) existence of the features should not bar prboom-plus from being used for Compet-N submissions. The default game in prboom-plus is just plain old Doom and that's Good Enough.

Share this post


Link to post

If somebody pays me a visit and sets my dosbox doom to work properly, I'll gladly record something for c-n.

Out of all options stated above, closed source port would be the only real solution to get some serious competition imo, and if it isn't happenning, you can never be sure. A little more drastic solution would be to advice c-n guys to also grab prb+ and join us in the not-all-that-serious competition, where we still waste half a day recording.

Share this post


Link to post
chungy said:

If Compet-N forbids ports, it should forbid emulating a DOS environment as well (and how is this going to be enforced?).

COMPET-N players have regularly been running Doom in an emulated DOS environment since the release of Windows 95. It wasn't illegal then it is shouldn't be illegal now. We shouldn't ban emulated environments because that only makes it hard on legitimate players and like you said

chungy said:

Whatl I'm saying is that restricting how demos submitted to Compet-N is silly; cheaters aren't motivated to follow the rules in the first place and burdening the legitimate players is just plain out stupid.

, cheaters will always cheat.

However, that is not why I feel so strongly to disallow Source Ports. The reason is that, to me (and many others), COMPET-N has always been about competing in the "real" game, not an exact replica (Chocolate Doom) of the real game, but the true, authentic, original V1.9 Vanilla Doom. Running Vanilla Doom in a false, emulated DOS environment does not make the game any less, real, "pure," or authentic, it just makes the operating environment "fake."

COMPET-N isn't about DOS competition, it's about Doom competition. The fact is that a false DOS environment shouldn't (without tampering), affect Doom, but as we both agreed, cheaters will always cheat, so why make it harder for the legitimate players.

Now, I know I may be offending people by my views on Source Ports' legality in COMPET-N, but I do not at all have anything against them; just hours ago I was playing Memorial and 1monster on PRBoom-Plus. I just feel like COMPET-N should stick to it's strict (and maybe irrational) roots of not allowing anything but V1.9 Vanilla Doom.

And j4rio, I am not at all saying that competition outside of COMPET-N lacks any seriousness. Just look at what went between me and DooMer 4ever on map 8 of No Rest of The Living; hours where spend try to reduce times by just a couple seconds. It was definitely just as intense and serous as anything I've ever done in C-N. And if you live in Texas, I would gladly pay you a visit and set up a working Vanilla Doom.

Anyways, I understand that my opinions and beliefs are in many ways unrealistic and and irrational, but they are my opinions and I will state them and defend them, as they [,perhaps irrationally,] mean very much to me. COMPET-N's been a big part of my life, and I'd like to keep C-N as it always has been. If COMPET-N's old rules are just ridicules by today's standards, then let's not change C-N, but create a new organization of competition; however, I think DSDA suits that just fine.

Share this post


Link to post

COMPET-N's been a big part of my life, and I'd like to keep C-N as it always has been.


As stated before, Compet-n will always accept only Vanilla demos. I will not change that because I promised it when I asked for Compet-n source.

Second, this is year 2012 and soon Doom will be 20, but people still want to compete and I want something serious with points, not recreation, PrBoom+ and other ports are not "serious" from my point except if you want to create TAS demo. Why do you people know who Anders is, who Panter is, Adam, Sedlo, Vincent, Henning, Drew, Dev, Radek...etc.? Because they played with PrBoom+? No, because they made 1000 attempts and made Vanilla demos that we all admire.

I'd like to see that in Chocolate Doom fork since this is the only port that can offer such thing. We "might" add better video (zdoom like), timer with ms, sr50 binds, monster kill counter to make it a little bit new and attractive, but on the other side we can offer more pwads so there will be plenty records for everyone. That's just an idea. Don't say PRBoom+ has all this becouse PrBoom+ is too advanced.

If somebody pays me a visit and sets my dosbox doom to work properly

You just have to download teamviewer, send me ID and password and I'll be glad to setup DOSbox for you. I hope you have a decent PC.

Share this post


Link to post

They don't remember those people because they played only on vanilla doom.exe, they remember them because they made kick-ass demos. What the vanilla Doom strictness offers is a consistent set of rules that the game plays by -- sure demos in ports that change the game mechanics can be impressive, but vanilla strictness gives everyone a level playing field to judge against.

Also you want to fork Chocolate Doom... to basically recreate PrBoom because you don't like PrBoom? That's silly. (Granted, a Chocolate Doom fork to do such things would be awesome, but for this rationale, it's ridiculous.)

Yes, PrBoom(-Plus) has all those advanced features, including ones suitable for TASes. None of it is really that relevant in the discussion. Simply run "prboom-plus -complevel 2 -record coolrun" and you won't have any significant differences compared to playing with Vanilla Doom.

Share this post


Link to post

Okay, so I've tinkered with dosbox around a bit and need only to find out how to use parameters, how to run pwads, how to change sensitivity/acceleration, how to use autorun, how to disable movement forward and back when I push mouse up and down (that one is seriously annoying). Also I really need to test out parameters first before anything because chocolate doom freezes my PC whenever I use recording parameter. And I have ancient PC.

Also

Why do you people know who Anders is, who Panter is, Adam, Sedlo, Vincent, Henning, Drew, Dev, Radek...etc.? Because they played with PrBoom+? No, because they made 1000 attempts and made Vanilla demos that we all admire.


That's not because of amount of attempts or port used. It's because of demo quality and skill. Just doing 1000 attempts of recording doesn't guarantee that your demo won't be easy to beat for someone else more skilled without much effort.

Oh, and also, I don't even think anymore that most skilled players of all time are included in your list. :)

Share this post


Link to post
chungy said:

Also you want to fork Chocolate Doom... to basically recreate PrBoom because you don't like PrBoom? That's silly. (Granted, a Chocolate Doom fork to do such things would be awesome, but for this rationale, it's ridiculous.)

Exactly what I was thinking.

Share this post


Link to post
chungy said:

The idea of using a server to automatically try to legitimitize runs is the wrong solution. It introduces two inconveniences: requiring a special port, and requiring an Internet connection; don't take the latter as a for-granted thing. It is quite easy to think of a situation where you may entirely lack connectivity but might be trying to do a new speedrun (imagine you're on a vacation, have a laptop, and get a bit bored in the middle of nowhere). Additionally, it will introduce all of the issues that modern Always-On DRM systems have -- it will assume the user has a perfectly stable connection all the time. It may work most of the time, but when an ISP or wireless router flake out, you do not want to find that your speedrun attempt is now completely invalidated due to a situation completely outside of your control.

I disagree: comparing it to DRM is a misleading argument because it associates it with all the bad things associated with DRM (copy protection etc.) which is not the case here. You're correct that it would mean that you wouldn't be able to record verified demos without an Internet connection, but nothing's to stop you recording normal demos or practising your speedruns. It would just mean that if you want the extra confidence that a verified demo gives, you'd need an Internet connection.

To use an analogy, think about how the Guinness World Records operates. Perhaps in private while you were practising, you really did perform 150 pushups in a minute. But that isn't accepted as a world record unless there is an official adjudicator present to observe it and confirm that it really took place. In an ideal world I guess we'd have adjudicators for Compet-N as well that could confirm that no cheats were used, but that's not really something that's practical :) What this does is provide something similar.

I believe neither demanding the DOS vanilla doom(2).exe nor this kind of automated validation is the right approach. Just accept any demo that plays back in vanilla, whether it was recorded in Chocolate Doom, PrBoom, Odamex, Eternity, or what have you. Apply usual anti-cheat double-checking measures after the fact. Legitimate speedrunners should not be subject to having to prove they aren't cheating; cheaters are likely to be caught no matter what.

Yeah, in the end I think you might be right here. Some degree of trust is needed, no matter what, and perhaps it's just easier to go on the honor system like has been done before.

Share this post


Link to post

chungy said:

Also you want to fork Chocolate Doom... to basically recreate PrBoom because you don't like PrBoom? That's silly. (Granted, a Chocolate Doom fork to do such things would be awesome, but for this rationale, it's ridiculous.)

Yes, PrBoom(-Plus) has all those advanced features, including ones suitable for TASes. None of it is really that relevant in the discussion. Simply run "prboom-plus -complevel 2 -record coolrun" and you won't have any significant differences compared to playing with Vanilla Doom. [/B]


You don't understand, I don't want all prboom or zdoom features in it, and so we will have to strip it, so it's the same as adding just few things to Chocolate Doom. I think you didn't get the point. I was just hoping we could build a special port in future for "new" COMPET-N database. Anyway I like PrBoom+ and I use it for many things, converting demos to videos is the first thing.

That's not because of amount of attempts or port used.
Oh, and also, I don't even think anymore that most skilled players of all time are included in your list. :)

Of course that amount of attemts is not the reason, but to make a good one they spent hours, days or even weeks to polish it and to be proud of it. With PrBoom+ I can beat in a day for example every Sedlo's record. That's ridiculous.

Share this post


Link to post

You make it sound like recording with PrBoom+ automatically implies using TAS features. Unless you mean to say improvements like the kill counter or higher resolution are by themselves enough to beat a batch of world records in a day...

Edit:

during casual multiplayer events with people spectating.


Much as I hate "always on", such a scheme would be the only way for real competitiveness IMHO.

When I see someone spending hundreds of attempts and/or weeks to make a demo, I'm not impressed at all. Monkeys, Shakespeare and all that; along with the implied notion for every success there's been hundreds of failures. Doing a ridiculous number of attempts just seems like an awkward and inefficient way to approximate what TAS features do.

Much more awe inspiring is to see someone drop in and come out with a high quality demo mere hours after even seeing the wad for the first time.

So where do I draw the line? It's entirely subjective of course (much like most C-N rules are, for that matter), and my point would be that tracking time spent on any particular wad from a particular player would be an interesting statistic and help to put things in context.

Such a thing would be entirely impossible to set up, with nothing stopping people from practicing offline. Unless you'd generate wads on the fly and give a limited time window to participate, and jump through a number of ludicrous hoops like that.

Best to accept there is simply no such thing as competitive SP Doom unless you trust people.

Share this post


Link to post
fx02 said:

With PrBoom+ I can beat in a day for example every Sedlo's record. That's ridiculous.


That's overexaggerating. If you can beat record with prboom+, you can do so with vanilla. It's just a barrier of bullshit between those two, like an absurdly annoying and inconvenient setup of primary requirements for playing or recording, which is basically one of the reasons why prboom+ is the most commonly used port these days. Oh, and as for the "I don't even think anymore that most skilled players of all time are included in your list.", well, I'd dare you to have a watch of pretty much any demos by this, this or this guy without getting a jawdrop. And keep in mind that they managed plenty of such stuff with ports that have no TAS tools implemented or even during casual multiplayer events with people spectating.

E:

@Phml

My POV on that would be that it's entirely dependant on POV.

The high skill of player can indeed get determined by doing awesome demos quickly (just fyi, Okuplok had a 5x:xx run of sunder map 11 on his channel within 2 days since I shared beta containing map 11 with him, and keep in mind that rendering the video must have taken at least a day alltogether :) ) but I don't dismiss people willing to spend time on demos. I have no doubts that SAV or Looper who do spend long time on their demos are also very highly skilled, just their POV on the matter is different. Also you can't argue that demo quality of people who spend time on demos is generally higher than those who do it quickly. And after all, it's the doom speedrunning. If you look up the term speedrunning, you'll realise that people who do it "right" are those who do spend time on polishing demos.

Share this post


Link to post

j4rio said:
That's overexaggerating. If you can beat record with prboom+, you can do so with vanilla.


Ok, then go ahead and record them today in Vanilla :)

ps. I don't enjoy on demos recorded with PrBoom+ on some neverheardof map so I won't watch them, except if it's a complete run of tnt, pl, mm, hr, etc, trick or something of interest.

Share this post


Link to post
fx02 said:

ps. I don't enjoy on demos recorded with PrBoom+ on some neverheardof map so I won't watch them, except if it's a complete run of tnt, pl, mm, hr, etc, trick or something of interest.


You really don't know what an enormous portion of doom speedrun activity you're actually missing. Maybe that's why those names you mentioned in previous posts are the 'only' that are constantly heard of, because anything done as of late is automatically considered obscure.

fx02 said:

Ok, then go ahead and record them today in Vanilla :)


That's pretty much how I wanted to solve the argument. But

Okay, so I've tinkered with dosbox around a bit and need only to find out how to use parameters, how to run pwads, how to change sensitivity/acceleration, how to use autorun, how to disable movement forward and back when I push mouse up and down (that one is seriously annoying). Also I really need to test out parameters first before anything because chocolate doom freezes my PC whenever I use recording parameter. And I have ancient PC.

Share this post


Link to post

The high skill of player can indeed get determined by doing awesome demos quickly (just fyi, Okuplok had a 5x:xx run of sunder map 11 on his channel within 2 days since I shared beta containing map 11 with him, and keep in mind that rendering the video must have taken at least a day alltogether :) ) but I don't dismiss people willing to spend time on demos. I have no doubts that SAV or Looper who do spend long time on their demos are also very highly skilled, just their POV on the matter is different. Also you can't argue that demo quality of people who spend time on demos is generally higher than those who do it quickly. And after all, it's the doom speedrunning.


While I can entirely agree on the whole thing being a matter of perspective, I just think having no restrictions whatsoever doesn't make for a very balanced environment between players, which is the heart of anything competitive; and it is odd that in a setting where the most relevant metric seems to be "time", we only look at one side of the equation.

For example if player A has 3 hours a week to spend on demos and player B has 30 hours a week, the deck is stacked. Perhaps player A might be able to compete with player B if he is a lot more skilled, but you won't ever be able to compare these players or their demos objectively, in my opinion, unless you restrict the range to specific parameters (which is what most people, heck, perhaps everyone except me, is fine with, so again, yes, perspective).

It seems silly to me because in my mind it's akin to a very specific subgame of basketball where each player could take an infinite amount of free throws, and the final ranking would be determined by whoever has the longest streak of successes, regardless if they shot 10, 100, 1000, or to put it another way, regardless if their success rate was 50% or 0.05%...

It's not so much a question of one way being better to me, but rather the inability to properly evaluate each demo - or even to know it, save when it's informally mentioned.

I do react negatively to the idea quantity means quality when the same people expressing that opinion dismiss tools, as if mechanical repetitiveness until you've essentially broke down a level to a simple matter of executing step A at time B and so on, removing any element of thinking, was somehow worthy... Which I guess it is, just like people for example enjoy competitive gymnastics (or if this particular example isn't MANLY enough for Doomworld, those japanese things where the athlete does stylish moves with a sword), but to me limiting Doom to just this very specific aspect is disappointing.

Mostly personal nitpicks and I can understand most of the community doesn't give a damn.

If you look up the term speedrunning, you'll realise that people who do it "right" are those who do spend time on polishing demos.


Ugh, don't get me started on that. It seems the 90s were a golden age for anglosaxons to claim ownership of concepts they didn't invent themselves and put their own spin on that. All hail USA cultural dominance...

Share this post


Link to post

Phml, I actually know completely what you're saying. I've always dreamed of a competition organization where the Administrator give's out a small, simple, never before seen map and players have an an hour to compete for the best UV MAX and SPEED times.

j4rio, I'll help you out with DOSBox later today, and I suggest you take my advice and NOT others (sounds really egotistic, I know), I've seen some very bad advice given by very intelligent people. It's hard to make anyone understand why certain setups are bad, but trust me, you'll save a lot of time in the future doing it my way. (If someone is so sure they have a better method, send me a PM about it.)

Also, as for the command line parameters, I'll make a thread here later today explaining how to use the command line with ease. Would you rather type in doom2 -file REQUIEM.wad rq21fix.wad REQMUS.wad -warp 21 -skill 4 -respawn -maxdemo 999 -record RR" or just "doom2 @RR 21" to make a -RESPAWN demo of REQUIEM map 21? The latter is much simpler, and easier.

If you don't want to record vanilla demos though, just sticking to PRBoom-plus on DSDA is fine with me. Everyone except fx02 will be watching.

Really, Doom competition is for entertainment, do what is most fun and fulfilling to you.

Now, j4rio, tell me how much RAM you have as well as your OS and if it's 32 or 64 bit. I'll need this to know how to optimize your DOSBox experience (sounds like a cheesy advertisement, lol).

Also, fill out this templet with your preferred key bindings:

Forward:
Backwards:
Strafe Left:
Strafe Right:
Strafe On:
Turn Left:
Turn Right:
Use Key:
Weapon 1:
2:
3:
4:
5:
6:
7:
Fire:
I already know to fix the mouse moves moving you forward and back, and enable auto-run. I help you out with everything later today, I just can't do it right now.

Share this post


Link to post

Another question j4rio, is DOSBox installed on C:\ or some other drive? If it's another drive, specify the letter.

Share this post


Link to post

Read this rather quickly, so I may have missed something.
But, I could defeat any of the proposed validation schemes in a couple days using almost any port. If I have the source available, then it can be easily augmented with a secondary demo playback mechanism that feeds control inputs to the demo recorder being sent to C-N.
This can also be rigged to play-in canned sequences on a key press.
This can be combining with a semi-accurate automatic aiming and other aids.
I doubt even demo analysis could detect this. If it could, then a few modifications to make it less perfect would defeat that.

A canned binary just takes longer and involves more binary tools.
Any part of it can be replaced with a imitation.
The easiest is to replace the whole binary with an imitation that records using the published demo format and answers all challenges correctly.

I even consider using always-run to be a form of cheating.
Most of the menu selections, gamma settings, etc, etc, are also effective cheats.

It does not matter how much you think that it would not be effective, the mere fact that some users could be presenting demos recorded this way, would lead immediately to constant recriminations.

Unless a player is on your machine, playing your copy, with your settings and your mouse, serious comparisons are dubious.

The best you could do is constantly interfere with the random-number generator, monster placements, and present a new previously unseen level.

The incentive is that you have presented a new challenge, a new game.
To post a high score, cheating as much as possible, defeating your
system.

Share this post


Link to post

Archy said:
If you don't want to record vanilla demos though, just sticking to PRBoom-plus on DSDA is fine with me. Everyone except fx02 will be watching.


Hmmm, I never said I will not watch any demo recorded with PrBoom+, just that I'm limited with time to watch everything so I pick carefully :)

Share this post


Link to post

I agree on that you should accept Vanilla-Compatible demos from any source port that can record them, along with Vanilla Doom. But if the demo fails to play back in say a pure DOS environment that uses a P3 and a true sound blaster, then too bad!

Share this post


Link to post
wesleyjohnson said:

Read this rather quickly, so I may have missed something.
But, I could defeat any of the proposed validation schemes in a couple days using almost any port. If I have the source available, then it can be easily augmented with a secondary demo playback mechanism that feeds control inputs to the demo recorder being sent to C-N.

While this is true of the closed-source port idea (another reason why I think it's a waste of time), it isn't true for the scheme I proposed: the game changes every time depending on a random seed received from the remote server at the start of the game, so you can't record and replay later.

Share this post


Link to post

I don't mind if someone plays with chocolate doom or other vanilla compatibility port as long as it doesn't show monster through walls etc. other really unfair shit what other ports don't have, because the demos are still easily comparable. What I do mind is that there will always be cheaters and they won't go away by just resurrecting C-N and saying "please, don't cheat ^_^". If I recall correctly, C-N died because of cheaters were too hard to detect, so you think it's easier now?

IMO, only solution is to have a new source port with shit loads of anti-cheat stuff, and that probably wouldn't work either. I mean if you go play FPS online, you see anti-cheat programs being updated often, but people always find a way to fool them. That's why the servers need admins and you can't have admins in single player. Also, admins do mistakes too often or just can't decide if someone is cheating or not.

Share this post


Link to post

Even with messing with the random numbers a cheater can pre-record
critical action sequences and play them back at will using a keypress.
If they want, they could write code to adapt a playback to aim shots to where the monster actually is, with some random error so that it does not look too good. The cheating playback could have special codes that tell the playback mechanism how to adapt and what searches to apply.
It would be almost the same as writing an NPC in a role playing game.

The real point is that as long as you cannot tell, it is always a suspicion. My point is that using internet, having serious comparisons is futile. So I say, try to keep it informal and don't take it serious, then there is less incentive to go through the work needed for that level of cheating.
Or just let them cheat and just give them a separate cheaters score board. Then they can show off to each other how they did it.

Share this post


Link to post
×