almostmatt1

Happy Birthday Kvo! UV-Speed D2All in 1:01 - 4kvoallx-101.zip YouTube: Map 15 UV-Speed in 0:08.69 - 4kvo15x-869.zip I'm an idiot and spent a while going for the wrong route but didn't want to scrap this so I'm posting it as an I.L. to show a TAS that doesn't do the void glide. The map 15 in the D2All does the void glide and is 6 seconds faster. YouTube:

Thanks as always for your hard work! I'd also like it if opl could be put back in, please.

Loved this!

Demonfear Map 15 No Monsters (... NoMo100s I guess if you wanna be pedantic) in 0:02.23 - df15xo223.zip YouTube:

Master Levels - Paradox Pacifist in 0:02.26 - paradox-x226.zip YouTube:

Could you please tell me if you're using a pwad or just doom.wad, what map you want to make a TAS for, and what category you want to do? Like UV-Speed/No Monsters/UV Max etc. The more information the better. Also, just so you know when you make a demo with XDRE and try to play that demo using a source port, it will only work with vanilla-compatible source ports. This means it will work with DSDA-Doom, PrBoom+, Crispy Doom, Chocolate Doom etc, but not source ports like GZDoom, Zandronum etc. If you're not confident you know what the difference is, just grab DSDA-Doom from here and I'll show you what to do once you have a demo to play.

Sorry that that XDRE guide is so wordy and indirect. I intended to be as unambiguous as possible but that absolutely came at the expense of it being way too long to get through, and it being quite inefficient to get usable information out of if you just want to get up and running. It isn't how I'd write it if I had to do it again, but it probably will still answer many questions you have. The basics are this. TASes can be made by individually controlling every single input of every single frame of a demo, or they can be made by playing the game with slow motion and re-recording segments which is commonly done nowadays with DSDA-Dooms extremely valuable rewind button. Both XDRE and DSDA-Doom allow you to create a TAS frame-by-frame, but only DSDA-Doom allows you to make a demo using slowmo+segmenting. Building frame-by-frame gives you a LOT more control over what is going on, but is more time consuming and difficult to do well. The beginning of the XDRE guide has been edited to discuss some advantages and disadvantages of either program. I don't have enough experience with slowmo+segmenting or using DSDA-Doom to be much help but I've spent much more time TASing with XDRE than I have actually playing Doom, so if you decide that you would like to make a TAS with XDRE feel free to reach out any time with specific questions about what you want to accomplish and any issues you are experiencing and I'll do my best to help you out. :)

I spent most of my childhood in a rural area with an extremely poor internet connection with parents that didn't really want me using it much even when it did work (something I'm now quite grateful for) and didn't get a smartphone until I was 17, I'm 27 now. I miss how deliberate it felt to use the internet to look something up or to talk to someone, it felt like an exciting event and an opportunity. My habits absolutely changed around the end of high school and some time later I eventually realised I was addicted to being on social media, and some of the more interesting parts of that realisation were confronting the fact that I was constantly finding myself on a site/app when I had made absolutely no conscious decision to set out to do so, and the fact that I wasn't remotely enjoying myself or feeling like I was benefiting, I was really just mindlessly doing it out of habit for no real reason I could name, and I was doing it for a really dumb amount of time. I couldn't believe how much mentally better I felt when I reduced how much I used social media by about 95%, and also how unexpectedly difficult it was to do that. It's kinda like a supermarket, it's mostly filled with complete bullshit that is horrible for you and conveniently at your fingertips but if you ignore the mass of garbage in the middle and hang out around the edges you're likely to find all the healthy beneficial stuff and you're way better off, and there's some really cool stuff there. The internet is a wonderful tool but most of it completely sucks, and it's frustrating how much it feels like you have to push it to make it work for you. Kinda but not really an aside, over the past 12-18 months or so I've been playing a bunch of 1994 wads and I have unexpectedly really loved reading through a bunch of the old text files. I wasn't around to live through this myself but using computers and the internet around that time period seems to have been so much more of an exciting and special experience, and occasionally a bit of that magic shines through in what someone wrote back then that I get to read today, and does make me sad that I'll never know what that was like. Computers and the internet are technological miracles and it really sucks how much ridiculous stuff like dumb internet drama things being essentially engineered to addict you and advertise to you and as much as possible have poisoned so much of it for so many people, it ought to be way better than it is IMO.

Currently there unfortunately is not. Kraflab has mentioned that adding a brute force check for whether the player possesses health at or exceeding a certain amount may be possible in the future, but this isn't currently implemented. However, depending on the specifics of your circumstances brute forcing is probably not an advisable approach for trying to get an outcome spanning such a long time. Really it's impossible to say without knowing more about what you're doing, but as you add more and more inputs across more and more frames the amount of input combinations that need to be calculated just exponentially explodes. A few weeks ago I did a brute force across just 8 frames with reasonably small input ranges for each frame and that ended up having a little over 58 billion input combinations that needed to be tested. Even splitting up that brute force range across 32 individual brute forces ran in parallel it ended up taking almost 4 straight days, so it'd take roughly 4 months if you just ran that as the one brute force. That was for just 0.23 seconds of gameplay so brute forcing anything significant enough to realistically affect RNG across such a large amount of time probably isn't practical. Again it's hard to say without knowing exactly what you're doing but IMO it's a safe bet that, unfortunately, testing stuff out manually would be the way to go. As a general primer on this you probably can't do better than watching decinos excellent video on the topic :) Knowing what the RNG Index is and being aware of how the game deals with RNG can be definitely useful, but attempting to manipulate the index to match a certain value isn't able to get enemies to do certain things, unfortunately. Lets say you're at 10.74 seconds, and the RNG index will equal 100 when gameplay reaches 11.00 seconds. There are many different things that could have happened between 10.74 and 11.00 seconds while still having the RNG index equal 100 at 11.00 seconds, and so there is a wide variance of things that could happen at or after 11.00 seconds. For example a shotgunner could be stunned, or walking, or shooting, or dead and the RNG index could still be 100 depending on what led up to that point, so getting an exact RNG index has fairly limited significance or usefulness. As decinos video shows the game engine itself DOES think in terms of "I need something 'random' to happen, I'll look up the RNG index to choose what to do" so in this way it is deterministic of in-game outcomes, but we simply can't interact with the game in such a way as to reliably reproduce desired outcomes based on manipulating RNG to certain values. While "I want the RNG index to match X so Y outcome occurs on tic Z" would not be useful and so it isn't correct to think in terms of "X index = Y outcome", it's useful to be mindful of the fact that the RNG index is affecting everything all the time and everything is affecting it all the time, so you have a lot of options in terms of how to alter the games RNG, it'll just generally take a lot of manual work to work out how to get what you want to happen. There are exceptions to this regarding matching the index at the end of a map so it *starts* at certain values on the next map but that's a little outside the scope of your question. I finally have good news for you, there is :) launch the game with "dsda-doom.exe -iwad doom2.wad -file YourWad.wad -complevel 2 -tas -recordfromto PreviouslyMadeDemo.lmp NewDemo.lmp". When the first demo finishes being played through, you take control and PreviouslyMakeDemo.lmp + whatever you do from that point gets turned into NewDemo.lmp. You can also launch with the "-build" parameter to start the game up in build mode, but that means you'll need to skip through the existing tics of PreviouslyMadeDemo.lmp with the advance frame button. Instead, you can just hit the toggle build mode button at any point when the game is running. You can also launch with the "-skipsec x" parameter with x being the number of seconds to skip ahead if you don't feel like sitting through watching all of the first demo file.

Junk Food Map 18 NM-Speed in 0:01.49 - junk18xn149.zip YouTube:

Good thread idea, it's interesting to read peoples thoughts on this :) My favourite of my TASes basically has to be the Doom 2 UV-Speed D2All I suppose, given the amount of time and effort it took. I'm not sure where I'd even start with my thoughts about the process of making it. There were a LOT of days of putting quite a few hours into it to make very little progress, quite a few times I scrapped hours of work because I made a mistake or found something small to save a handful of frames, and countless hours waiting for XDRE brute force to do its thing. A memory that sticks out is when I got to map 13, realized I could go back to the end of map 11 to save a second because I stuffed something up, did that, had to rebuild some big parts of map 12 due to rng differences, and after getting back into map 13 I vividly remember how bad the realization felt that I hadn't really gotten any further and was still less than halfway through the overall project. There was also the fact that I knew there were some tricks that I was going to have to do but I had no idea how to do them when I started, the most notable examples being the void re-entry on maps 23 and 28 and having no clue if the jump on map 29 was possible at all, so knowing I'd inevitably need to work it all out was stressful but putting myself in a position where I was forced to learn new things ended up being quite positive for me. Getting through the last third felt like it just about killed me haha, I got a bit unhealthily obsessed with getting it done and spent way too much time on it. Including viddumping it I spent around 19 hours on it on the last day of working on it, from about 9:00 - 9:30 AM to 4:30 the next morning. I'm glad the TAS is done but I'm even more glad to be out of that state of mind! From a raw speed standpoint, I think no TAS is unbeatable and there'll always be something faster eventually but my E1M1 7.69 second UV-Speed is the closest I've gotten to feeling like there's just nothing left to squeeze out. I'm sure someone will do it some day, but it's the only demo of mine I look at and think it'd be futile for me to try to improve it further. A few other TASes stick out where I finished them and felt at the time as though I'd made large steps forward in my understanding of how to improve TAS movement, such as E1M1 Max in 0:19.97 and Doom Episode 1 in 3:23, even though these have both been beaten now. More recently I'm fairly happy with my Eviternity Map 8 NoMo TAS 0:05.40, not necessarily because of the run itself, but because a couple of years ago Vita made a NoMo TAS of that map in 0:05.66. My best, which I was quite happy with at the time, was 0:07.23. I thought I'd done a good job, but looking at Vitas demo made me realize "Oh my god, I am actually unbelievably horrible at this". It totally destroyed any sense that I felt like I knew what I was doing, and I looked at that demo as being of a quality that was just fundamentally unattainable to me and ahead of anything I'd ever accomplish, and that demo has always been personally significant as a result. Beating it was a nice moment that made me feel good about how much I'd worked to improve and learn how to get better at this stuff over the last couple of years. With goals... I've got a few big unfinished TASes I should focus on before starting anything else, but I've been thinking more and more lately that a Sunlust TAS looks fun. ¯\_(ツ)_/¯ Similar to Maribo's mention of not-strictly-playing-the-game stuff, I'm also making a tool with Rocky to split up and parallelize brute forces that I just can't wait to finish, even in an incomplete state it's now quite quick and easy to get a 10x speed increase in brute force times, so getting that done is my most prioritized goal when it comes to Doom stuff.

I also found that this was the case with several of Andy Oliveras TAS demos. Despite being marked as TASDoom demos, some desynced with complevel 6 but played fine with complevel 5. Apparently TASDoom was based on DOSDoom (which is what -cl 5 is meant to emulate), so yeah, it's a bit weird that TASDoom demos occasionally work on the complevel for the source port TASDoom is based on but not the complevel for TASDoom itself.

Many thanks to the three of you! <3 I've generally been shamefully slack on properly checking out new wads but whenever I see screenshots, gameplay videos/speedruns, and general chatter about the ever-present stream of incredibly high quality wads that are constantly being released, it breathes such a cool amount of life into this game that I love so much. I pretty much spend most of my Doom time TASing and to me that feels like a world I'll never properly fully understand, and knowing that that that big world is just a tiny tip of the iceberg that is Doom is really, really cool. General shoutout to all the mappers in this community, the life that this game still has largely because of all of you is truly something special. It's been great having @RockyGaming4725 around. Over the last 18 months he has rapidly become one of the most technically proficient TASers ever, and it's been really exciting for me seeing him get better and better and create some really amazing demos. There's not many people you could have conversations with about something like, I dunno, using XDRE to effectively diagonally wallrun to generate otherwise impossible momentum values before immediately brute forcing an optimal redirection of that momentum via an elastic collision, but Rocky is up for a good technical chat any time. He's a good dude that really knows his stuff! I truly appreciate your presence and I look forward to continuing to collaborate with you on demos through 2023. :) I'm also very happy to see @Dimon12321 continuing to create demos! Earlier in 2022 it appeared that Dimons TASing days may have been behind him due to the conflict in his country. Not that me no longer getting to see new TASes is remotely important in the greater context of that situation, but Dimon has always been one of (if not #1) my favourite TASers (I mean come on, this is just amazing! And so is this!), so Dimon, the fact that you're still putting time and effort into these demos despite the situation you are enduring is really something I personally appreciate. I hope you are well. @dsda-dev and the work that they have put into DSDA-Doom is also really sick. 0.25 was a huge update, and the TASing features that exist and are planned for the future are super exciting. When scripting capabilities get fleshed out a little more it's going to blow the doors open on the amount of stuff that's going to be possible with TASing. Can't wait! @4shockblast (and the rest of the folk responsible for the DSDArchive work), massive thanks for your time and effort spent working on and maintaining the demo archive. Also shout out to shock for recording what is probably my personal favourite demo of the year. Simply too many speedrunners to list, really. Y'all are sick, keep it up!

The channel: https://www.youtube.com/channel/UCnOJXrNSVsxzN3xaRRFZbWQ Earlier this year I made a YouTube channel where I am posting every TAS demo on DSDArchive. Videos are in 4k and their descriptions include demo download links, pwad download links, some viddump details, creation date of the demo (for a while I was manually checking every txt file and only including this if the author did not write it themselves but I recently decided to automate this and just include it for every single video), and a copy of the txt file submitted with the demo. I want it to be as accurate to the original as possible so the txt file contents are untouched, including now-dead links and factually inaccurate information written by the author, unless it exceeds YouTubes 5,000 character description limit in which case I will manually prune away some text until it fits. This is the only time that txt files are changed, and it's mentioned if so. Videos are put into playlists based on player and wad and there's also a playlist for all movie runs. I have mentioned this channel already on Doomworld and on Discord so I apologise for repeating myself, but just in case it slipped past other tasers I wanted to make this thread so hopefully they're more likely to see it for two reasons: - If you would specifically like me to not post your TAS demos please let me know. I hope noone minds me taking an opt-out approach instead of an opt-in one, I figure that if you have publicly posted a demo then you are probably just generally okay with it being public, but I guess it's possible that some people are fine with a demo on DSDArchive but not a video on YouTube for whatever reason. It's also of course perfectly reasonable and understandable that after your hard work on a TAS, perhaps you might only want views on your own channel. This is retroactive too, if I've already posted your stuff and you're unhappy about that I'll take it down. I'd love for the channel to have every single TAS demo but ultimately if you made the demo it's your call. - If you'd like, I can viddump and upload a high quality video of your TAS for you before you post the demo to Doomworld so that you have a video alongside a demo to post, which should also mean that your demo will have a video linked to it when it makes its way to DSDArchive. Some folks are not in the fortunate position where their computers and/or internet can happily deal with making and uploading videos, especially 4k, so I'm happy to help you out! Reach out any time if you'd like this done and I'll sort it out as quickly as I can.* I do want to make it clear though that this only applies to TAS demos, I'm not doing anything with non-TAS demos with this channel. As of writing this I've posted 461 videos but there are a little over 2,300 to do so there's quite some work to go yet. A decent amount of the work is thankfully automatable but with all the info I'm trying to include there are bound to be a few errors so I would also be quite appreciative if you could let me know if you notice any. A minor exception to this, sometimes the dates that are automatically detected and included in the video description differ from what the author defines in the txt file by a day which I assume is due to time zone stuff. Honestly, being off by only a day some of the time is a minor enough thing that I'm not going to worry about it. Currently I'm working through the demos of notable TASers that appear to be retired or haven't been active for a significant amount of time such as Andy Olivera and Azuruish. After that I'll go back and work through each player alphabetically while also uploading newly posted demos. I initially planned to upload new demos as they were posted from day 1 but this approach instead lets me get a large amount of the work done before needing to worry about juggling uploading already existing demos while not missing the new ones. Doing all this involves a fair bit of time consuming work so if you've already posted a TAS in the past and want to see a video of it it could be some time before I get to it, particularly if your username is alphabetically later. *Even though I'm starting with some older demos from retired TASers before uploading new stuff, if you do want a specific demo viddumped and uploaded and request this it this will take priority over everything else.

Rocky informed me that this synced immediately in UV, so it was luckily very little effort to make this work :) All the same inputs but I think it's a different enough demo to warrant posting again. Eviternity Map 08 NM-Speed (Also Pacifist + Reality) in 0:05.40 - evit0xn540.zip YouTube: