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


  • Content count

  • Joined

  • Last visited

Posts posted by Fredrik

  1. Thank you Linguica for keeping this site going and making necessary upgrades. Some pixels are not what I'm used to which makes you literally the worst person since Genghis Khan, but apart from that you're a pretty cool guy.

  2. MAP30 could be done by interpreting the rules for Pacifist more creatively (or just more literally?). The player is not allowed to hurt monsters. Romero's head is present with -nomonsters. Therefore, it is not a monster, and the player is allowed to damage it. Why would it count as a monster, anyway? It's clearly just a piece of decoration, much like a barrel.

  3. I'm super serious. It has sick drums and bass, a sicker organ, and the sickest guitar solo. It gets you pumped but goes beyond that and evokes a unique, nuanced atmosphere. It's the perfect track for thousands of hours of cumulative play testing while mapping. In fact, it's so inspiring that I always wasted time running around after testing whatever I was supposed to test (usually until the end of the solo). Besides, together with the melancholic D2E1 sky (which is not unlike a painting by one of the great masters), it makes me nostalgic for playing 90s shovelware levels.

  4. It was possibly intended just to annoy Lüt, heh.

    In any case, the appropriate level of detail depends on what you're trying to accomplish. Detail for its own sake isn't impressive these days, and it's certainly not a substitute for good basic architecture. I rather enjoy playing maps that do a good job with minimal use of detail. However, high detail can also be used very effectively.

    Sunder is an interesting example. It uses detail in a completely obsessive that way that makes it feel like you're trapped in some kind of fractal of evil. It just wouldn't work if you took it out. In some other contexts, that amount of detail would be misplaced.

  5. Very nice run. Pity about the ending.

    I wish there was more competition in this category (big props for your persistent work single-handedly pushing down the time more than 16 minutes compared to the previous record holder). The routes look good, but it seems there's still a lot of time to be saved by optimizing weapon use (you have a lot of spare rockets and cells most of the time).

    Has anyone attempted a "theory TAS" (basically a segmented run of the normal route with normal play but no mistakes) to see what kind of time might be possible? It would also be nice to see an optimized built TAS, though obviously that would be an absurdly hard project. I guess Azuruish's 55:57 TAS falls somewhere in between...

  6. Another certainly doable one is ksutra map14. If you camp around the central horseshoe structure, pretty much everything dies to infighting. Two cyberdemons and a spider mastermind need to be killed with berserk + 900 bullets. One cyberdemon and the spider can both be killed safely with the pistol. The last cyberdemon is in a decent location for punching (perhaps it makes sense to kill it first of all?). If you get very lucky with the infighting, bullets might last all the way (e.g. arachnotrons might kill the spider).

    Gusta said:

    I think that all ksutra maps have berserk (+ lot of power up) but I am afraid that it's too much of luxury and those maps are way too easy for these crazy guys :-)

    Yeah, lots of turret monsters and blocking lines make life hard on the remaining maps though...

  7. ksutra map18 looks doable. There are something like 700 bullets available to begin with, which is enough to take out the two annoyingly placed cyberdemons and the two tower revenants, provided that the cyberdemons have been softened up by infighting and some punches with the invulnerability. The third cyberdemon usually dies to infighting.

    On top of that, the archvile + zombieman inside the train can be farmed forever, if needed. The archvile near the blue skull switch also potentially allows farming, but I'm not sure if there's a consistent setup for it.

    The nukage arachnotrons can potentially walk into the teleporters which allows telefragging them. However, bullet farming seems faster than waiting for that to happen...

    Everything else looks reasonable, considering that there's a berserk, plenty of health, and corners to hide behind.

  8. Well, it wouldn't necessarily be a waste of time.

    First, I think finishing each segment just a few times should give a reasonable order-of-magnitude estimate. Assume that you pick three segments that are roughly equally hard, and you play until you've finished each one 10 times. If finishing a segment takes N tries on average, you've played each segment 10N times, for 30N attempts in total. The estimated number of attempts to beat the whole level will be N^3. Most likely 30N will be small compared to N^3.

    Second, you might want to practice segments of the run separately anyway, to improve your preparation for the whole run. Doing so systematically might actually be the most efficient use of time.

    It's true that the segments wouldn't be quite independent though. If nothing else, your hands will probably be shaking from adrenaline if you ever reach the end of the tunnel (literally) in a real run :)

  9. Why not try to estimate the chance of success? Let's say that you can do it with two checkpoints A and B. Measure your success rates getting from start to A, A to B, and B to exit. Multiply the probabilities. If it looks doable in 10^4 tries, persistence should be enough. If it's going to take 10^7 tries, better find some more monkeys or keep working on strats.