Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Looks like a real aimbot! I don't know if visitors would like this style of turning. If I were making that, I would at least try to imitate fast turning, like I do in Doom.
100% will really be insane. You'll need to take everything into account. Backtracking is not that common in this game, but I'm sure you'll have to leave some enemies behind and hope they won't run too far away trying to follow you.
A wallhack or a map with enemies shown on it would really insure you won't have to redo a piece of work because you had forgotten something.
BTW, how did you build that TAS? I'm not aware of libTAS interface. Can you set mouse coordinates for each frame?
Shadow Warrior is not really on the list. VoidSW doesn't have a Linux build yet (as far as I know) and I don't see much progress for the engine to get into a release.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
If I have a weak GPU (Intel HD Graphics 2500), does libTAS run, let's say, Shadow Warrior (Build engine) via DOSbox at 60 fps internally?
I know I have to do some tweaks to resolutions for both DOSbox and the game, but does libTAS provide hardware independence?
I don't mind running libTAS at <10 fps, but I want the video to be recorded at 60 fps for a game running on DOSbox.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Frankly speaking, SM 1 was way more entertaining than SM 2. Various sections just made me skip them, like those with Black Cat. Beating the boss (38:15) in 1 hit was great though.
Weak Yes vote!
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
I check "Resize Video" on Step 1 and calculate the resolution I want to get. If you simply upscale a native resolution in 4 times, it won't usually match YouTube patterns.
For example, 320x240 * 4 = 1280*960 will be downgraded by YT to 720p. I don't know why YT downgrades the quality as well, but such a video doesn't look much better after YT's processing then the same one in 960x720, so YT doesn't resize it on its own.
I do like that:
1) 1080 / (native height) = X,1562
2) X,1562 * (native width) = Y,8645. Rounding up to Y + 1.
That's your output resolution for resizing.
Example for 320 x 224.
1) 1080 / 224 = 4,82142857
2) 4,82142857 * 320 = 1542,85714. Rounding up to 1543.
So, your resolution is 1543 x 1080.
Maybe nowadays it makes sense to simply upscale it in 5 times, so it is more than enough for YT's 1080p, but I don't know if it's still compresses custom reses so dramatically.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
For God's sake, just make xipo a co-author in any submissions if he asks you to do that. That stream-sniping (or more like "submission-sniping") stuff can go on and on, because who knows how many unfinished WIPs he has behind the scenes.
Dealing with each individual "plagiarism" of that kind might take long and burn much nerve cells. If we dig dipper, an input file's metadata/header can be edited with appropriate tools or, at least, with a HEX editor. I can have hundreds of movies "from 2011" and blame everyone that its input is actually my input.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Meh. I got used to YouTube's hotkeys, so skipping that afterlevel music and level observations is not a problem.
Off the record, would be cool to see "max speed" TAS as well.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
No, unfortunately, it's not possible. However, there is a source port called TASDOOM which is basically a wrapper over the original DOS Doom which lets you slow the game down and use game saves like savestates. Still, it's not as effective as frame-by-frame building.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Plutonia: Revisited Community Project (2011) UV-Speed in 11:56.
Link to video
Even though it can be submitted, as it's fully built frame-by-frame and aims for the fastest time, most of the TAS isn't made by me. I just finished it (maps 28-30).
BTW, Hock told me he had had an even faster run of Episode 1, but lost it due to the computer failure.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Some unknown guy named Erik made a UV-Max TAS of this map in 2 hours 11 minutes just 1,5 weeks before Zero-Master. https://www.youtube.com/watch?v=i_g1KyRNHDg
So you can consider it "sniped".
Nothing unbelievable is in here. Zero used DSDA-Doom which is a PrBoom+ folk that provides rerecording from keyframes (kinda like savestates) which speeds up slow-motion TAS production because you don't have to playback all your progress in order to continue recording from the moment where you left off.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Is it possible to transfer regular input with TAstudio window opened? It's not that convenient when you have to use either only mouse to mark specific cells in the input grid, or only console input and then open TAstudio to perform various manipulations on it.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
And yeah, I agree that I showed... not the best part of me. I apologize, but a negative feedback is still a feedback. I would take it into the consideration, if I received a similar feedback or worse.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
If the rest of the TAS will look the same as in the 2014 WIP, I don't think the world will be that happy to see it. It got 6.2 thousand of view so far, which is not a good result.
Also, I'll need a hundred cups of coffee to read all the comments to figure out why each turn was made in that way. Probably I'm too young to enjoy such fast-pace games.
Given that 10 years have already passed and the full TAS hasn't been built yet, I would just abandon it because procrastination usually doesn't give a positive result or it won't live up to the expectations.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
I came up with the idea of assigning the end of the range to each of threads and then checking each 'end = end - total number of threads' number if it's prime. Now, the range division looks pretty simple. Thank you!
Language: java
int currentThread = 3; // this is a serial number of the thread (0..N)
int begin = 1;
int end = 30 - currentThread;
int totalThreads = 4;
for (int i = end; i >= begin; i-=totalThreads) { // i = i - totalThreads
System.out.println(i);
}
//27
//23
//19
//15
//11
//7
//3
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
Guys, I need your help.
How to divide a range of positive numbers (from 1 to 1000, for example) between several threads, so each one would be evenly loaded?
Each thread has to check each number of its range segment in a loop (from 2 to number/2) if it's a prime number. If so, than it adds it to a concurrent collection.
The easiest way to spread the load is to allocate several arrays (lists) and fill them sequentially in a for loop from start to finish. But unfortunately, I'm not allowed to do that. By any means, but without arrays or other containers.
Since processing if 2 is a prime number is much quicker than processing 997, I'm thinking about some division like:
Thread 1 takes sequence 1:500. Thread 2 takes 501:750 and so on, until thread 10 (20, 30, ...). I can't pick right proportions.
TASing is like making a film: only the best takes are shown in the final movie.
Experienced Forum User, Published Author, Active player
(481)
Joined: 4/5/2014
Posts: 1134
Location: Ukraine
I didn't expect to see so many posts here. I've watched the whole TAS and now I have mixed feelings. The game seems to be restricted at times or else I see a lot of tiny possible time savers. For example, at the beginning when the game tells you how to climb up the wall, you keep moving left even after crossing the gap. Could you jump down earlier? IIRC, Catwoman can jump down towards the direction she was moving, but you jump straight down.
TASing is like making a film: only the best takes are shown in the final movie.