Author Topic: ClamLix  (Read 90858 times)

0 Members and 2 Guests are viewing this topic.

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #210 on: January 13, 2018, 03:24:18 AM »
The proof collection doesn't seem to be updated with the level swap.

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: ClamLix
« Reply #211 on: January 13, 2018, 03:53:03 AM »
Thanks, forgot to push last week. Done now!

-- Simon

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #212 on: January 13, 2018, 05:03:02 AM »
I just looked at "Just drop in" because of potential new batter physics. And indeed Clam's intended replay seems to break but it doesn't seem because of the changed batter physics but because of the changed blocker physics.
I've attached a repaired replay.

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #213 on: January 13, 2018, 07:39:24 PM »
Since I probably can't really care for ClamLix from February on:
If there are any physics changes that break replays, please don't only look if there is a solving replay at all but also that Clam's intended route (Clam's replays) gets a replay as well.
The levels where that is relevant are:

Arcane:
- "Beaten into shape"
- "Builders Without Borders"
- "Over and Under and Out"
- "Send in the Square Squad"
- "Turning the Screws"

Insane:
- "Dig This!"
- "Lix Recycling Plant"

Flingsane:
- "Lost?"
- "Polar Pop"

Outtakes:
- "Just drop in!"
For "Just drop in!" it would be nice if in case my alternative route would get a repaired replay as well since the route is significantly different/ a completely different way to solve the level.

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #214 on: January 14, 2018, 02:25:35 AM »
Simon, how serious is it with the tumbler rewrite branch? Are there major changes expected? I could in advance make repaired replays for that branch.

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: ClamLix
« Reply #215 on: January 14, 2018, 02:37:44 AM »
The tumbler rewrite has a reasonable arc and the code runs much faster than the 0.9 tumbler.

The problem is different positioning horizontally, during a single frame of movement. Jumpers in tumbler-rewrite often land next to the pixel where they would land in 0.9. I should investigate this and decide whether (a change for more 0.9-like horizontal sub-frame movement) makes the the math annoying.

cThe sub-frame movement is important and will affect every tumbler path, this causes most of the replay breakage between 0.9 and tumbler-rewrite.

I think it's wiser not to make replays for that branch yet. I don't expect a physics release within the next 6 weeks. But thanks for the offer!

-- Simon

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: ClamLix
« Reply #216 on: January 14, 2018, 08:45:02 AM »
I've slept about this. If you have the time, you can take a shot at covering those -- if you're okay with doing potentially unneeded work.

Let's denote by "version A" the 0.7.9 physics plus (current branch phys) plus (current branch tumbler-rewrite).

With 0.9.7, all 113 levels are solvable, and in A, only 13 of the 113 break. It's likely, even though not guaranteed, that the final tumbler's sub-frame movement will be a mixture of the 0.9.7 movement and the A movement. Therefore, it's likely that either the old replay or your cover using A will solve it.

If you'd like to give it a go, and are okay with creating 13 replays that might not be useful later: I've rebased tumber-rewrite onto phys, you can thus build A from tumbler-rewrite.

-- Simon

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #217 on: January 14, 2018, 04:58:36 PM »
I didn't provide a replay for "Lazy Gypsy Bums" but the other 12 I've covered. I've repaired as well the failing Clam replay to "Lix Recycling Plant". Most of the replays should break with the stable branch.

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: ClamLix
« Reply #218 on: January 14, 2018, 05:22:54 PM »
I confirm that these replays solve in version A. Thanks a lot!

-- Simon

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #219 on: January 15, 2018, 12:42:06 AM »
Just for documentation:
"Flower Towers" is mainly a jumping level and one where I fiddled around until it worked in tumbler-rewrite and the stable branch. I've made a comparision picture from a certain (late) point in the level to give an impression of the differences between the two branches. Most of the level is done by this point and it's after having done 13 jumps.
Left is the stable 0.9.7 and right the tumbler-rewrite branch. I've put the picture in a zipfile to avoid spoilers but I don't think it is spoilery.

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #220 on: April 13, 2018, 03:38:45 PM »
A small probably backroute-fix to BEHIND YOU.

Spoiler (click to show/hide)

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: ClamLix
« Reply #221 on: April 14, 2018, 02:04:56 PM »
Earmarked for next release. The old proof still covered Behind You, but I've added your extra proof to the collection too; thanks.

-- Simon

Offline joshescue18

  • Posts: 147
    • View Profile
Re: ClamLix
« Reply #222 on: July 16, 2018, 02:55:02 AM »
I finished Humane.

Offline Simon

  • Administrator
  • Posts: 3860
    • View Profile
    • Lix
Re: ClamLix
« Reply #223 on: July 16, 2018, 10:45:33 AM »
Nice!

Everything looks very much intended. In particular, Square Dancing is intended (last year, we had a backroute here).

-- Simon

Offline Forestidia86

  • Posts: 721
  • inactive
    • View Profile
Re: ClamLix
« Reply #224 on: December 07, 2018, 03:40:48 PM »
A bit aesthetically improved version of "Time is Running Out".