abyrvalg

Holy Hell Map03 UV-Max, 27:00 hh03-2700.zip

Holy Hell Map02 UV-Max, 25:10 hh02-2510.zip

Holy Hell Map 01 UV-Max, 47:25 Also known as "let's derust a bit". Not even sure it's a valid max, but seems to be, when comparing the final stats to other demos. hh01-4725.zip

No, you're not! I was simply trying to list the most common ways of breaking the playback that came to my mind asdlkjhasjkldfhja.

All demos sync fine so far. Are you guys trying to play them back through console, or pausing/alt-tabbing while watching or something? No idea how else to make it desync consistently. Also, can't help but express how glad i am to see this wad finally get some proper treatment, about freaking time! Hyped for more demos!!1

so where to download this beast?

Or, if you prefer something that is actually entertaining, you can proceed to watch this vod instead :^) Features savestate abuse due to bullshit bad luck moments (e.g an archvile running past mancs and then falling down into lava), fun strats for every room, and lots of long and hard grindy parts. Long and hard!

To me this feels exactly like pause buffering from OOT, with one exception though - pause in doom does not get accounted for in the final timing if you're doing a single segment run. But, for example, in RTA runs i'd say it would make sense to allow this sort of stuff.

I suppose Memfis would hate watching speedruns of modern games like Skyrim, Amnesia, Soma or, say, DXHR, where they spend months figuring out the best items, fps values, timings and angles to use in order to clip through walls and out of bounds. And it ends up being the same crap as glides :D Namely, a luckbased, yet partially controllable fiddly weird trick that is annoying to perform in a run. More on topic - the only thing i kind of dislike is when mapmakers make lazy shortcuts that are too obvious and generic. For example, a 32-unit hole leading straight to the exit or to a key. It doesn't bother me much at all though, but it's not very interesting to discover such stuff. Which leads me to the next point: i really like to invent routes rather than discover them. The difference is subtle and most of the time exists only in my brain. If i successfully manage to trick myself into believing that i've found something that was unintended, it feels amazing. In all of your examples this feeling gets ruined by a creator saying "Hey! I left you a linedef right here, enjoy" instead of being subtle :p So maybe we're in the same boat here. From this perspective, Valiant is amazing, i really loved va27-127.lmp for example, the route looked quite out-of-the-box'y. But sadly, wads of such quality take too long to make, which is why i nowadays play with modern ports instead, and record shitty first finish runs like this one: http://www.youtube.com/watch?v=aQxyVuF076U since it gives more freedom and allows to break maps in the ways they were not designed for. Quite refreshing. But probably incredibly boring for the majority of community.

random incoherent post asjhldjasklfhjla: I'd say there's definitely enough ammo for everything and the traps aren't very difficult in the first half of the map, especially if you know what's coming. Wouldn't call it easy either though. Have no idea about 2nd half though. Died in the room with 5k chainers, which seems quite undoable when you're playing blind, especially since it ruins your fps so much. Despite my pc being relatively good, it was a pathetic slideshow. But for some reason, no one lagged out and the server we played on survived. Monstercount tricked me into expecting it to be a slaughtermap, but in fact it appeared to be a huge puzzle map, it's like eternall.wad on crack! Almost every step you make involves finding some cryptic, barely visible button, and it would've taken us way longer than 1h 20min to get to the chainers room (you know the one) had i not used Doom Builder while playing. Surprisingly, we encountered only one area that seemed to break in coop, and only two inescapable pits (unless jumping and crouching is required, but i believe they're not). All in all, i weirdly enjoyed the experience, - it's fun hunting for the next hidden button and exploring this hard and long beast deeper and deeper. I don't think this map is for everyone though, nor i think it has high replayability value, but it's nice and fun if you like it hard and long.

Heh this thread is a complete trainwreck, just like any of the previous threads where the same question was asked. As you can clearly see, there's absolutely no consensus, even among very skilled players who have actually beaten most of the mentioned titles. I would just recommend not dismissing any of the wads, and check them out by yourself and form your own opinion. As of more chillax demos, Ancalagon has some on his youtube channel, and i've recently made a few crappy maxes. Link to my youtube can be found in teh profile. I'd say okuplok's untitled2 holds the cake for being theoretically possible, playtested and balanced (EXCEPT FOR THAT ONE FUCKING ROOM - http://www.youtube.com/watch?v=xd8e6zS5Zuo), yet completely undoable in practice. Even with itemrespawn and 3 lives it took us with WaTaKiD several months to "almost beat" the entire thing. And strats used offline are mostly completely different and much more difficult to pull off, with only a few exceptions. You should check the freshly released Sunlust out then, since it was made by the same author in collaboration with another great mapper. Ribbiks includes sources of his music in readme files btw, which is convenient.

heeyyyyy, here's for some controversy material! Back in 2012 the now-banned Hock worked on a PRCP built run, and has actually almost finished it. Up to this day, the movie stayed in that same unfinished state, so i decided it's time to unearth it and just show it to everyone rather than letting it get lost forever. The last time i talked with Hock, he was willing to publish it after building map29+map30, but that was couple of years ago, so i guess he won't object to me uploading it. Besides, i have been heavily involved in the project (mostly in the routefinding/criticizm department), and the lmp contains my own input as well (albeit only a tiny fraction of the movie was actually made by my hand), so i have at least some say in this question. It's over than 2minutes faster than odysseyofnoises' non-built TAS, and imo contains some neat tricks and ideas, but lacks the final two maps. There are few easily improvable maps, but the overall quality is in my opinion quite above average. So, enjoy this dusty gem. Oh, and i haven't checked for turbo tics, but i believe there shouldn't be any. 28prcptas.zip

Always a pleasure to see new Akse's works, this one was not a disappointment either. Everything, especially the 3rd episode, looks greatly and carefully planned and executed :> And, dwelling into an offtopic area: do you guys only play in the zdaemon lands? I've dropped off that source port couple years ago in favor of zandro, but would like to play with you every now and then if that's possible to arrange, eheh.

Y'all guys have learned to read between the lines it seems xD

You can just use saves at AGDQ, imo the only concern should be consistency (in order to avoid loading every 15seconds in hard spots), there's no real need to practice pistol starts unless you feel like adding a self-imposed challenge. And yeah, i support this choice too, would be awesome to see you performing there, i'll definitely donate if that happens.

Hey Azuruish, since you clearly dont want to become a good TASer by most other people's standards and have your own goals, i'm curious - what are your own goals in TASing? Because right now i can't even imagine what you're trying to do. Not trying to judge you though, it's your choice what you want to do with your time :) There's one relatively offtopic fact: Make one very good TAS movie, and you'll be able to make quite good tases extremely fast. There's not much skill needed in producing high quality tases in doom (but there's a LOT of skill required to make Looper-like tases though), it's an easy and straightforward task, you just have to struggle through it once or twice and you're done. After that you will learn a lot about efficient TASing and optimization and will become much faster at producing new runs, with an added benefit of them all being higher quality. But if you choose to take babysteps (like you're doing right now), you'll lose out in a long run, even if your goal is quantity over quality. The whole quantity over quality paradigm doesn't really fit tools-assisted speedrunning though, usually their whole idea is about pushing the limits far beyond human abilities, instead of barely beyond human abilities.

Dat ancluck at 17:40, haha. Nice vid, and holy heck what a fitting grand ending for such an awesome wad, great job Ribbiks and danne. To think that a 20+ years old game can leave a man in awe not only via mechanics and sheer difficulty, but also with the looks, style and atmosphere (and so on, i'm no wordsmith like Krypto)... that's kinda impressive and unlikely, but that's exactly what it has done to me :p

Heard a lot of similar comments about buttsex. Why was it a disappointment? cba to spend 15seconds to download the wad and check it out myself.

Whoa, never seen that one happen before, interesting. Reminds me of a similar thing happening in tartaru4.wad (and previous versions of it i believe, even though it has been fixed in tartaru5) on two of the northeastern linedefs in the final room of map26, both in prboom and zandro, and most likely in other ports. In that map it's trivial to get partially inside of the wall, rendering all monsters' attacks useless against you, and also making your own explosions do nothing to you.

The beginner scenario from scifista42's drawing is a good way to start learning, such glide is usually referred to as a "guided" glide. It's probably the best to just use doom builder or whatever to create a 32-unit gap with wall on one side and a pillar on the other one, that way you can hug the wall while attempting the glide. Direction of glide matters in vanilla, you can test it yourself by gliding in one direction and then trying to glide back from the other side, and then repeat until bored. I would suggest trying the simplest technique at first, which is as follows: get close to the gap, look straight forward, increase your vertical sensitivity a bit, throw away your keyboard and from that point only use mouse to move forward, as slowly as you can. After a while move backwards a tiny bit and repeat the process. That way it's possible to get the general feel of the trick, and that will make it easier to understand more technical stuff later on if you will feel like it. Creaphis' thread indeed contains a great lot of analysis. And of course, for some of the classic IWAD glides specific approaches have been found, which work exceptionally well, but those wont help understanding the general idea i think. Also, i find guided glides to be even easier in zdoom-based ports, the slow speed approach also works there reliably, even though it's a tiny bit different. Unguided gliding is extremely luckbased and difficult in zdoom though, while it can be 100% reliable (depending on direction) in vanilla. The reasons for the latter can be found within the aforementioned thread (i believe it's called "two types of glides"), but if i recall correctly it's based on doom's sine table being imperfect and thus leading to player getting pushed west if he walks north.

Yeah, purposefully misused because i found it funny and fitting well enough due to recent tasvideos-related discussions.

lazy to check now, but iirc there's a demonstration of a broken lift in a tvr demopack, but maybe my memory fails me. If there is, there might also be some description of it. Or it could've been some random demo :D Okay i shut up.

lol, it keeps getting better, i wish i had one of those now >:D

Checkmate atheists! This is just amazing, thank you for this research. While we're at it, are there any caveats to this approach? You've already mentioned that one could have turbo tics without the -turbo command, but what about the other side of the spectrum, - is there anything we can't do that otherwise can be done with mouse/keyboard/joystick? Also, it seems like none of the "official" drivers allow for SR50 on turns, this can only be achieved by modyfing the ini (or creating your own), thus creating your own homebrew driver, is that correct? In that case from a complete purist's perspective it'll probably be still reasonable to be against SR50 TRXXX tics in the regular TAS runs, and create a separate category for the API-TASes and allow turbo and sr127 on turns within that category.

He's overreacting a bit imo, but he has some valid points. SR50 on turns is not an acceptable input and should not exist, for it's an artifact of a sourceport developer's mistake and has since found its way into pretty much every major TAS product. It's a pretty unique and paradoxical thing to happen within TAS community, since we basically try to optimize our runs as much as possible and push the game to its limit, but in the sr50's case we let ourselves to instead push the game beyond its limits and act like it's all fine and dandy, even though we're basically cheating. And it's been happening for so long that no one bothers to fix this mistake anymore. MESHUGGAH has spotted this error and is being vocal about it, and technically he's right. At least his position is understandable and reasonable. His criticizm about doom tases being suboptimal is also understandable, since we count our results in seconds, and he counts in frames, thus our optimization goals differ. That's how i view it at least.