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

Compet-n Resurrected

Recommended Posts

Cluttering it up with "verified" and "nonverified" seems like just that, clutter. It should either be all verified or not on there at all.

I would really like to see Fraggles idea put to work, so I'll toss my vote in on that and hope it sees the light of day.

Combined with xttl's quickstart that sounds to be a very good base for something workable.

Not sure what can be done about automated sr50, but visual policing and zero tolerance. Any ideas welcome. If we can get stitching and slowmo out of the picture, thats really more than I had hoped for.

Share this post


Link to post
dew said:

fraggle: actually some short speedruns would be unaffected by your randomization, especially in levels where monsters don't block the way or they don't get alerted to screw up the prng.

I did some simple experiments a few weeks ago to see what the effect would be from randomly moving all objects in a level by a very tiny amount. My experience was that even with only a tiny change, the demos would go out of sync very quickly (less than 10 seconds).

Are there any examples of levels where monsters don't block the way or get alerted at all?

Share this post


Link to post
fraggle said:

I did some simple experiments a few weeks ago to see what the effect would be from randomly moving all objects in a level by a very tiny amount. My experience was that even with only a tiny change, the demos would go out of sync very quickly (less than 10 seconds).


For the longest time Eternity had a problem with several boom demos. the problem was (iirc) related to teleporting. In one demo in which a single imp fireball being thrown in a different place (imp on the other side when he shouldn't have been) threw the entire demo off. the fireball didn't even hit the player, but it was enough to knock the entire demo out of sync.

Share this post


Link to post

Andy Johnsen said:
Bottom line is, I think FX attacked this the wrong way, but his intentions are good.

All he did was give what was functioning more informally, and something he appreciates personally, a full-fledged website. That fx decided to do this doesn't get in the way of any other proposals, and he wasn't in a position to start something with a new cheatless program or the like, so he's not to blame for failing to resuscitate a very competitive environment.

People can and will still record speed runs under the original Compet~n standards, which is speedrunning with the technology id released and now requires more trust in the honesty of participants than in the past. Just like open source software, it's part of the patrimony of the community and can't be closed by anybody.

Any change or new proposal starts a new chapter with its own name, so fx did what can be done about what we know historically as Compet~n itself, at this point. In addition, the site at doom2.net stands as its archive for its heyday before the source made cheating much more trivial.

Share this post


Link to post

so he's not to blame

Before I decided to take over COMPET-N I was thinking about changing almost everything, design, set it up, and then release it, but the thing is that we don't have time to spare if we want anything done. That's why I decided pick up old site and fill up database later, change design later, and finally database is up to date now without incoming ofc. :)

If you ask me, Chocolate Doom is good/great enough for old COMPET-N but I promised and decided I won't touch the legacy so that's why it will stay with vanilla and I think most will agree on that.

I'd even go with Chocolate Doom (not fork) for new database, but we *really* need some kind of protection.

Share this post


Link to post

We face the question here of which is better -- getting Compet-N back now, with a number of issues and questions yet to be resolved, and getting it back after an unknown amount of time.

Adam Hegyi has so far proven irreplaceable. He had the motivation and expertise to detect cheating and run a quality site in a professional way (he even was a top player on the side). Compet-N was one of the best things things the Doom scene has seen. Riding the wave of its popularity and prestige, Adam made the well-intended decisions of including PWAD's and the annual Golden Cyber Awards. These additions were enthusiastically received and increased the player activity and popularity of Compet-N even further. The tradeoff, however, was that they critically increased the workload of the sole administrator to the point where it became too much for a single, unpaid volunteer. It is apparently way too much work for volunteers to maintain a very high-quality site that at the same time accepts a myriad of categories and PWADs.

That is one reason I share Anders's vision of returning C-N to its roots by limiting it to the IWADs and the three original categories. This will presumably keep the maintenance workload bearable for unpaid volunteers while maintaining quality of submissions, fierceness of competition, and the prestige of the site.

As a secondary modification, I am also in favor of having the smallest time unit count, because I see it as being conducive to a further increase in demo quality and a further incentive for perfecting each record. Nothing with glitches would be safe, and everything would tend towards perfection. The track running World Records aren't rounded off to full seconds, and I don't feel like Doom World Records should be either. The policy of rounding the times down has resulted in that a few "less prestigious" -nomonsters records (such as map01 and e1m1) are of much higher quality and much more polished than the corresponding, "prestigious" UV Speedruns, since the latter are impossible to beat by a whole second. Introducing millisecond times would remedy this. However, I must emphasize that this question is way secondary, as it does not tie directly to the core problem stated above, and it could easily be debated at a later point, even after the hypothetical replacement site was launched.

The way I see it is as follows: getting Compet-N back immediately, but with unresolved issues, has the advantage of a possible resurge of activity that might be significant. Its disadvantage is a greater workload of maintenance and a greater risk of a cheater-infestation that might dissuade some of the most hard-core and passionate players (this resonates with me) from participating in it. Waiting until the issues are resolved has the advantages of reducing the workload of maintenance, attracting the aforementioned player type, and maintaining the prestige of the name Compet-N. Its obvious disadvantage is that the initial challenges of setting such a site up might deter the whole thing from actually materializing, which would portend more years wasted waiting in vain for a replacement and the possible resurge of player activity that such a thing might entail -- and we have been waiting for 7 years already. However, I personally would rather keep waiting for a good replacement than launch something incomplete and issue-ridden bearing the name of Compet-N. I feel Adam did such an admirable job with Compet-N that his legacy should be honored by withholding that name from anything falling short of its quality. They are huge boots to fill.

Just my two cents, though.

Share this post


Link to post
Andy Johnsen said:

I would really like to see Fraggles idea put to work, so I'll toss my vote in on that and hope it sees the light of day.

Well, I did some work on the master server last night to add the necessary signing support. As it's running now, the live master server will now send GPG-signed messages in response to requests. That's the necessary first step - next I need to make the necessary changes to Chocolate Doom itself.

The public signing key is published on the MIT public PGP server.

Share this post


Link to post

Chapter 4: In which Fraggle becomes a super boss and somehow continues to make dooming better for everyone.

Share this post


Link to post
×