Skip to content

Alter mimic scaling to instead of creating new items it now swaps mim…#22

Open
Dragion147 wants to merge 4 commits intogstlafrom
gstla-scaling-mimic-changes
Open

Alter mimic scaling to instead of creating new items it now swaps mim…#22
Dragion147 wants to merge 4 commits intogstlafrom
gstla-scaling-mimic-changes

Conversation

@Dragion147
Copy link
Collaborator

Rather significant changes to Mimic Scaling to now swap mimics instead of create new items It does this in 2 phases. 1st one locks mimics in place that were assigned to correct spheres by chance, the 2nd pass swaps mimics from earlier spheres to later ones prefering to swap with the weakest mimic gradually becoming stronger which forces all the weaker mimics to the front of the spheres and tougher mimics to later spheres.

Done numerous generations and iterations to observe results of generation attempts including with other games (with or without other golden sun). Generation is lengthened a bit as we do quite some extra looping but overall it will always be up to a static amount as we only loop through mimics from our own world which at most can be +- 100 items.

Dragion147 added 4 commits December 25, 2024 10:32
…ics from future spheres. It does a best effort to try and scale the mimics according to sphere depth with some variety in each sphere so it's not all just the same types over and over again.

As with all random based solutions it's hard to find a good middle ground where it is diverse enough but still fair, about 50% preplaced mimics in correct spheres will remain in that sphere, the rest will get swapped up to a certain sphere depth.
…s in the spheres that just happen to be correctly placed by the randomizer. Therefor we only swap mimics that really do not belong in their spheres and swap them with whatever the weakest mimics are left that we can find and aren't locked
Dragion147 pushed a commit that referenced this pull request Mar 8, 2025
* Fully updates requirements.py to live LADXR (#19)

* Updates dungeon2.py to LADXR-Live (#20)

No logic changes or bugfix are in this file. It is only code cleanup.

* Update dungeon1.py (#21)

- The Three of a Kind with Bomb is moved from Normal to Hard Logic

The rest is code cleanup.

lines 22-25 | 22-26 & 33 | 34 remain different in AP | Upstream with no effective difference

* Fully updates dungeon3.py to LADXR-live (#22)

Logic Changes:
- Hard mode now considers killing the enemies in the top room with pot

Everything else is cleanup.

* Fully update dungeon4.py to LADXR-live logic (#23)

Logic Changes:
- Hard Logic: Removes Feather requirement from grabbing the Pit Key
- Hell logic: new hookshot clip (line 64)
- Hell logic: hookshot spam over the first pit of crossroads, then buffer down (line 69)
- Hell logic: push block left of keyblock up, then shaq jump off the left wall and pause buffer to land on keyblock.
- Hell logic: split zol for more entities, and clip through the block left of keyblock by hookshot spam

The rest is code cleanup

* Updates dungeon5.py mostly to LADXR-Live Logic (#24)

Logic Changes:
- Hell logic: use zoomerang dashing left to get an unclipped boots superjump off the right wall over the block. reverse is push block (line 69)

The rest is cleanup.

The upstream splits the post_gohma region into pre_gohma, gohma and post_gohma. I did not implement this yet as I do not know the implications. To port this the following lines need to be changed (AP | LADXR):
18 | 18-20;
55 | 58;
65 | 68-69

* Fully update dungeon6.py logic (#25)

Logic Changes:
- Hard logic: allow damage boosting past the mini thwomps
- Glitched logic: bomb triggering elephants in two cases

Everything else is cleanup

* Fully update dungeon7.py to LADXR-live logic (#26)

Logic Changes:
- Hard logic: Three of a Kind is now possible with bombs only

Everything else is code cleanup

* Fully updates dungeon8.py to LADXR-live (#27)

Logic change:
- Hard logic: allows to drop the Gibdos into holes as a way to kill them
- Glitched logic: underground section with fire balls jumping up out of lava. Use boots superjump off left wall to jump over the pot blocking the way


The rest is code cleanup

* Fully update dungeonColor.py to LADXR-live (#28)

Logic changes:
- Normal logic: Karakoros now need power bracelet to put them into their holes
- Hard logic: Karakoros without power bracelet but with weapon
- Hell logic: Karakoros with only bombs

Everything else is code cleanup

* Updating overworld.py (#29)

* Updating overworld.py

This tries to update all logic of the Overworld.

Logic changes include:
- Normal logic: requires hookshot or shield to traverse Armos Cave
- Hard logic: Traverse Armos Cave with nothing (formerly normal logic)
- Hard logic: get the animal village bomb cave check with jump and boomerang
- Hard logic: use rooster to go to D7
- Lots of Jesus Rooster Jumps

I stopped counting and need to go over this again.

Also, please investigate line 474 AP because it's removed in LADXR-Upstream and I don't know why.

* remove featherless fisher under bridge from hard

it was moved to hell upstream and its already present in our code

---------

Co-authored-by: Alex Nordstrom <a.l.nordstrom@gmail.com>

* fixes

* add test messages

* Adds Pegasus Boots to the test (#31)

* Fix d6 boss_key logic (#30)

* restore hardmode logic

* higher logic fixes

* add bush requirement to the raft
in case the player needs to farm rupees to play again

---------

Co-authored-by: palex00 <32203971+palex00@users.noreply.github.com>
Dragion147 pushed a commit that referenced this pull request Mar 8, 2025
* Fix certain items not being added to slot data

* Change where items get added to slot data

* Add initial grass randomizer stuff

* Fix rules

* Update grass.py

Improve location names

* Remove wand and gun from logic

* Update __init__.py

* Fix logic for two pieces of grass in atoll

* Make early bushes only contain grass

* Backport changes to grass rando (#20)

* Backport changes to grass rando

* add_rule instead of set_rule for the special cases, add special cases for back of swamp laurels area cause I should've made a new region for the swamp upper entrance

* Remove item name group for grass

* Update grass rando option descriptions

- Also ignore grass fill for single player games

* Ignore grass fill option for solo rando

* Update er_rules.py

* Fix pre fill issue

* Remove duplicate option

* Add excluded grass locations back

* Hide grass fill option from simple ui options page

* Check for start with sword before setting grass rules

* Update worlds/tunic/options.py

Co-authored-by: Scipio Wright <scipiowright@gmail.com>

* Exclude grass from get_filler_item_name

- non-grass rando games were accidentally seeing grass items get shuffled in as filler, which is funny but probably shouldn't happen

* Update worlds/tunic/__init__.py

Co-authored-by: Scipio Wright <scipiowright@gmail.com>

* Apply suggestions from code review

Co-authored-by: Exempt-Medic <60412657+Exempt-Medic@users.noreply.github.com>
Co-authored-by: Scipio Wright <scipiowright@gmail.com>

* change the rest of grass_fill to local_fill

* Filter out grass from filler_items

* remove -> discard

* Update worlds/tunic/__init__.py

Co-authored-by: Exempt-Medic <60412657+Exempt-Medic@users.noreply.github.com>

* change has_stick to has_melee

* Update grass list with combat logic regions

* More fixes from combat logic merge

* Fix some dumb stuff (#21)

* Reorganize pre fill for grass

* Update option value passthrough

* Update __init__.py

* Fix region name

* Make separate pools for the grass and non-grass fills (#22)

* Make separate pools for the grass and non-grass fills

* Update worlds/tunic/__init__.py

Co-authored-by: Scipio Wright <scipiowright@gmail.com>

* Fix those things in the PR (#23)

* Use excludable property

Co-authored-by: Exempt-Medic <60412657+Exempt-Medic@users.noreply.github.com>

---------

Co-authored-by: Scipio Wright <scipiowright@gmail.com>
Co-authored-by: Exempt-Medic <60412657+Exempt-Medic@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant