RuffledBricks

Yes, it's happening, I'm generating the walkthrough for Jet Set Willy, similarly to what I did for Manic Miner. \o/
I'll describe the technical part of it in a later posts, for now I just want to announce that I've started it and start this thread.
Initially I thought that I'd make a website where the progress could be tracked, and possibly where people could help with computing power, but:
I was lazy to implement it, Single room generation turns out to be much faster than I expected (around 5 minutes instead of 5 hours per room), thanks to the optimizations suggested on this forum and much easier level structure than Manic Miner. I'm still finishing the "strategical routing" algorithm (e.g. order of rooms rather than in-room routing; I call it "journey" rather than "route" just to have a different term for it), during the next days while the "journey planner is not yet ready" I'll use @RuffledBricks speedrun route and will run the script to micro-optimize it.
Aaaand to get started with something concrete, here is the speedrun of the very first room. 😛 Enjoy!
jsw_init.mp4

Quantum Tunnelling

Willy passes through objects.

 

The program I provide here (Was) a fully functioning game. I have deliberately destroyed the part that can be played. The file provided is only for the purpose of demonstrating Manic Miner being auto played.

To auto play this version of Manic Miner press "K" while on the title screen. A sub menu will appear.

Use the keys "Z" and "X" to select a cavern.
Press ENTER to play the cavern --- (The cavern can not be played- in this version)
Press "L" to Auto play the cavern.

If the cavern selected is the first cavern, then pressing "L" to auto play, will play the whole game.

Normal keys for pause.
Normal keys to abort.
Normal keys for movement. - ( the cavern can not be played )
 
The auto play will complete the game with a score of 39526
 
 
AUTOplay.tap

Yeah, it's the same with On The Roof. Rope warp doesn't really improve how fast you execute the main bulk of the level. Its main benefit is simply in speeding up the rope.
I checked to see if it was possible to do the jump into the ceiling from the left, i.e. jumping onto the rope then immediately going to the top to do the jump, but it doesn't appear to be possible. You can't get to the top of the rope before its cut-off point becomes too low to reach the ceiling.

I already got beaten in an MM speedrun race against someone playing the PC version. Why is everyone so eager to see me do Miner Willy races that I'm blatantly going to lose?

I already got beaten in an MM speedrun race against someone playing the PC version. Why is everyone so eager to see me do Miner Willy races that I'm blatantly going to lose?

It was a bit of an overstatement from my side to say JSW will be easier. There is surely much more work to do and also much more computation time is needed. So overall JSW is surely harder, but the missing crumbling platforms aspect makes at least one aspect easier.
Well, the entire effort is/was intended to help speedrunners to find new strats. For MM I wasn't even going to generate the best possible score, it was to find new ideas. It's a scoreboard on this forum that turned me a bit into the "beat the score" direction 😛, but I didn't feel it like a "competition" either.
Similarly, with JSW I mainly want to help with finding new strats for human speedrunners. Actually @RuffledBricks recently said that he considers JSW done and there's no point in running it again unless there are new strats. So I'd consider a project a success if it will be possible to make RuffledBricks run JSW again. 🙂
As for human vs script run, as with any human vs TAS run, there's no really point in doing that. Script will know the optimal route and will execute it perfectly, it's not something that people can (or should) compete against.

Yeah, it's the same with On The Roof. Rope warp doesn't really improve how fast you execute the main bulk of the level. Its main benefit is simply in speeding up the rope.
I checked to see if it was possible to do the jump into the ceiling from the left, i.e. jumping onto the rope then immediately going to the top to do the jump, but it doesn't appear to be possible. You can't get to the top of the rope before its cut-off point becomes too low to reach the ceiling.

Yeah, it's the same with On The Roof. Rope warp doesn't really improve how fast you execute the main bulk of the level. Its main benefit is simply in speeding up the rope.
I checked to see if it was possible to do the jump into the ceiling from the left, i.e. jumping onto the rope then immediately going to the top to do the jump, but it doesn't appear to be possible. You can't get to the top of the rope before its cut-off point becomes too low to reach the ceiling.

Did some messing around and turns out the rope warp in The Beach actually is a little bit useful:
 
Looks like I'm gonna have to redo my Any% run now to get that elusive 19:47 time in...
 

Way ahead of you on this one:
 
 

It's nice to be talking to a pro 🙂.

Way ahead of you on this one:
 
 

On second thoughts: 
But then you would have to cross the room again from the top to go to the Yacht, so it's probably no good...

Thanks for the compliments Danny, much appreciated.
You would get killed from walking right, but it would be from falling too far rather than hitting Yellow Maria. I never bothered doing a comparison of which was faster because quite frankly I think jumping into Maria looks cool and I didn't want anyone to take that away from me.
We did look into this, and the OOAL deathwarp seemed like a good idea right up until we discovered you could shortcut in OABOTD by dropping into it from above. Therefore, because we're exiting OOAL via the bottom rather than the right, the deathwarp doesn't actually save you as much time as it would in UTM or Cuckoo's Nest, especially since we now have the quicker way of jumping past the skull.
"Any percent."

So here's why it's not really worth it.
Firstly, looking at this comparison here, the amount of time saved with the OOAL deathwarp is ~8 seconds:
 
By contrast, if you look at deathwarps in either UTM or Cuckoo's Nest, both of those save you closer to ~9 seconds:
(The UTM deathwarp is ever so slightly quicker than CN, although in terms of what makes the run easier on a practical basis I definitely prefer the CN one.)
They are also earlier on in the route, so you gain slightly more overall benefit from having one life less between Cuckoo's Nest and OOAL. As mentioned before, you only get six deathwarps across the run, so you want to use them where they're going to provide the most benefit.
Secondly, getting the OOAL deathwarp as close to an 8 second timesave as it is requires a pretty much pixel-perfect jump onto the leaf below the first item, so that you can then get both the items in one leftward jump. If you don't hit that, two things are likely to happen:
1) You do the leftward jump, miss the first item, and have to go up and collect it again, negating the timesave completely.
2) You switch to doing a vertical jump instead to grab the first item, then do a left jump to get the second item. The addition of a vertical jump would add between 1-2 seconds to the room, making the timesave on the deathwarp less efficient.
So not only is it not as good a timesave as UTM or CN, but it also is much harder to implement for not as much gain (other than looking cool). And that's why we don't use it.
 

To address your other points:
Honestly, I would've agreed with you on this one for a long time. For ages, I did Tool Shed before the Yacht because I assumed the timesave from the lost life would make the overall experience quicker. However, Mick_jo told me his split for that section seemed to be quicker when he did the Tool Shed last, so I did a comparison:
And he seems to be right. It seems like that back-and-forth into the Tool Shed if you do it first negates any speed benefits you get from running the Yacht section with one fewer life. So for this last run I decided to revert back to doing the Tool Shed last.
If you can explain to me how the rest of the route could be adapted to accommodate this switch in warps AND save time then I'd love to hear more! But logistically I can't see any way of using Offy Warp that wouldn't add more distance to the run overall than what we currently have.

So here's why it's not really worth it.
Firstly, looking at this comparison here, the amount of time saved with the OOAL deathwarp is ~8 seconds:
 
By contrast, if you look at deathwarps in either UTM or Cuckoo's Nest, both of those save you closer to ~9 seconds:
(The UTM deathwarp is ever so slightly quicker than CN, although in terms of what makes the run easier on a practical basis I definitely prefer the CN one.)
They are also earlier on in the route, so you gain slightly more overall benefit from having one life less between Cuckoo's Nest and OOAL. As mentioned before, you only get six deathwarps across the run, so you want to use them where they're going to provide the most benefit.
Secondly, getting the OOAL deathwarp as close to an 8 second timesave as it is requires a pretty much pixel-perfect jump onto the leaf below the first item, so that you can then get both the items in one leftward jump. If you don't hit that, two things are likely to happen:
1) You do the leftward jump, miss the first item, and have to go up and collect it again, negating the timesave completely.
2) You switch to doing a vertical jump instead to grab the first item, then do a left jump to get the second item. The addition of a vertical jump would add between 1-2 seconds to the room, making the timesave on the deathwarp less efficient.
So not only is it not as good a timesave as UTM or CN, but it also is much harder to implement for not as much gain (other than looking cool). And that's why we don't use it.
 

Congratulations on your SUB-20 run, RB! 👍 You've just made history 🙂. You can be very proud of it, indeed!
Your masterful delivery of JSW combined with your idle chat make for an intriguing combination 🙂. I think they just prove that such mastery can only be achieved when your actions are so automated that you are able to do something else besides executing such a lovely run. I've heard some comments to that effect from one of the top ski jumpers recently (ski jumping is a sport I particularly like to watch) and I couldn't agree more.
A couple of comments:
I think that when you lose a life in the Nightmare Room, you don't have to jump; if you just walk right you will still get killed, I would think. The lack of jump could save you a split second, perhaps.
Wouldn't losing the last life in "Out on a Limb" (by kamikaze'ing the second item) be more efficient than losing it in "Under the MegaTree"? Just wondering.
Thanks also for your explanation. A question: how do you pronounce Any%, I mean the % symbol?

Also, I managed to finally get a sub-20 min Any% run today. Hallelujah.
 
Still contained one slightly irritating mistake in On A Branch Over The Drive, but the rest of this I'm very proud of.

Yes, as IRF says, losing lives isn't mandatory in Any% or Warpless runs (apart from Conservatory Roof of course). However, you're going to struggle to get a particularly competitive run in those categories without doing so. Partly because of the increased running speed of the game, and partly because some rooms are just quicker to complete by losing lives.
The definitions of these categories can be found in the opening post of this thread, but basically:
Any% = just complete the game as quickly as you can (excluding cheat codes).
Warpless = complete the game without utilising any of the unintentional warps around the map (i.e. Rescue Esmerelda ceiling, Watch Tower ceiling, and rope warp in On The Roof). This doesn't include deathwarps, which are still permitted.
Both of these are common categories in speedrunning. Another common category is 100%, but this only really applies to JSW2 as you can complete that game without collecting all of the items, which you can't with JSW.

It's also more perilous as you might hit one of the crabs embedded in the sand.

Yeah, I never managed to find a use for it though, so I don't really tend to mention it out of saltiness.
 

Also, I managed to finally get a sub-20 min Any% run today. Hallelujah.
 
Still contained one slightly irritating mistake in On A Branch Over The Drive, but the rest of this I'm very proud of.

From RB's previous posts, I don't think losing all lives (bar one) is required by the rules, but that it's a tactic employed in order to speed up the game (when measuring a 'real time' completion time), because the program drawing the remaining lives slows down the game (only slightly, but it adds up over the course of a whole game).
(I'm not sure about your terminology queries.)
From your perspective Danny, of basing the speed of completion on the in-game clock time, the running speed of the program is not a factor so I think you should definitely stick with your tactic of 'Top Landing item first'.