The following is a poll on the Proposal: "Integrate All Levels into Any% subcategory as variable instead of a category."

The choices are:

A: Reorganise the leaderboard, so that "All Levels" is now a variable which appears under the "Any" category. All Levels runs will not obsolete Any% runs.

B: Leave the leaderboard as is, so that "All Levels" remains its own category.

How to Vote:

• Only those who have verified, full game, main board SuperTux runs can vote.
• Vote by replying to this message with your choice only and once only.
• Please take notice of the original Proposal and discussion on the Tux Speedrunning Community Discord.

Vote Below by replying with either "A" or "B" in place of "X" as your message:

I vote for "X"

The moderators have discussed a potential category reorganisation for the main leaderboard:

• The proposal is to convert the sub-category All Levels into a variable which appears under any%
• The proposal won't apply to CEs: All levels often serves a significant difference to any% in most/all CE worldmaps.

We do not wish to undo any accomplishments or simply remove times from the leaderboard. Instead, we propose:

1. The sub-category All Levels is changed into a variable, which would appear under any% runs.
2. Any% means "any method of completion", so all levels could appear under this definition, defined by its variable.
3. If wished, you can sort any% by what completion method was used using src's sort feature.

Arguments in support the conversion could be:

• 3 of the 4 All levels categories are variations of any% to a small extent, only allowing for roughly 4-5 additional levels over the standard route. This makes All levels, mostly serve as a category done for the sake of getting a WR time or master role. It is not necessarily a category that is done for enjoyment.
• This reduces the main category count to 8 which increases competition in other main categories, time could spent doing more enjoyable categories.
• This also would deflate the main category competition for when v0.7.0 comes out. Instead of 18 main categories, we would have 12.

Arguments against the conversion could be:

• 1 out of the 4 All levels categories (All levels, Forest, 0.5.x), is distinctive from any% - the route is different, more skips are used, more levels are played. This category is fun and could benefit from competition and therefore appearing as a main category and not a variable.

Please discuss this proposal in the Community Discord to form your opinions and share your thoughts. In case we have missed any key advantages/disadvantages, additional arguments for/against the decision will be summarised based on discussions and added above.

This post and poll will be announced on the SuperTux SRC main leaderboard forums and Discord.

• The poll will be conducted on speedrun.com/st on: 1st August, 2024 at 12:00pm BST
• Only those with verified, full-game supertux runs will be allowed to vote.
• Voters will reply under a thread with their choice
• Voting will close on: 8th August, 2024 12:00pm BST
• Votes will be counted

The decision will be to either:

• A: reorganise the leaderboard so All levels is a variable under any%,
• B: leave the leaderboard, so All Levels is its own main category.

Hey guys. With the upload of my first TAS, there are some obvious timesaves throughout the place, as well as some minor theories i have to save a bit more time and get it under 15 minutes:

1: early game boosting.

there is a lot of room for improvement in the early game when it comes to boost routing, finding the best boosts and maximising the amount of boosts you can get. some boosts give better pixels than others. imagine it like this, there are two blocks in front of tux, if tux boosts off the first he cant boost off the second because its too close. but if tux's subpixel position lines up on the first block such that he only gets a 1px boost, he technically could get up to 7. 1px may be enough to save a frame, but if tux doesnt boost of the first and boosts off the second one he may be able to boost up to 7px because of how the subframes, pixels, line up. Tux moves at like 5.3333px/f which causes these subframe positions to happen, and this is what makes boost routing difficult. if this was done in the TAS throughout, it could save 3-4 seconds over time, but its very difficult.

2: Faster egg grab.

I have a theory that if you grab an ice cube and throw it at a powerup block, creating the egg to spawn and move and fall, tux doesnt need to slow down to grab the egg, and can just run into it and move on. This would save about 1.2 seconds, but lose about 0.1 to not killing the badguys in 23rd airbourne by doing a cycle skip. However, there might be another place to get the egg and also still do the cycle skip, but more ideas need to be presented.

3: fix up the mistake in "A Fork in the Road".

There is a moment in fork in the road where the TAS loses 0.2 to a mistake, which wasnt noticed during the production.

4: potential better RNG in "Living in a Fridge".

better RNG allowing for a fast acceleration kaizo jump after the ladder section, might be able to save 0.3s

5: "Ice in the Hole" Double global cycle skip.

Probably going to be the hardest thing to TAS ever, including multiple subframe manips, insane fast accels, manouveres, but there might be a way to skip both cycles with the global elevator platform at the end of the level. I've actually gotten a very low platform in RTA before by doing some extremely clean movement, I think with the TAS, this can be amplified to a much larger extent, and maybe even skip the cycle more. This could save about 4 seconds? This is the last huge theoretical timesave out there.

6: Faster Nolok's castle ladder section

might be a way to not lose the speed at the ladder section, but its kinda uncertain.