Negative check counts

[FelicitusNeko]

I'm running into an issue with OT 1.6.1.0 where Dark Palace (so far) is showing a negative number of checks. This is a bog-standard Triforce Hunt, as (mostly) indicated by the options panel.

I've also noticed some issues posted by other people with screenshots that indicate negative numbers of checks for some dungeons.

[FelicitusNeko]

I've finished this run, and noticed that a few other dungeons went negative as well. In addition, some regularly accessible dungeons were yellow when they probably should have been orange (PoD in the above screenshot, for instance), and I got a couple of collectible items (eg. not map, key, etc.) in SP, but the number there didn't go down at all.

I've enclosed the log file, if that helps at all.

OpenTracker.Avalonia20210121.log

[trippsc2]

The other reports of this issues are from modes that are not yet supported by OpenTracker. To my knowledge, this is the first report of an issue in a supported mode.

Can you provide me with a few things? What randomizer did you roll the seed from? Do you have the identifier string for the seed, so I can generate it for myself for testing. Can you upload the .srm file with your save data?

[FelicitusNeko]

The randomiser I used is a bot on a Discord server I'm in. I believe it uses a version of the Bonta rando. Here's the top of the spoiler file, which includes the seed and all info:

ALttP Entrance Randomizer Version 0.6.3-pre  -  Seed: 490390492

Filling Algorithm:               balanced
Players:                         1
Teams:                           1
Hash: Hookshot, Powder, Key, Bug Net, Book
Logic:                           noglitches
Mode:                            open
Retro:                           No
Swords:                          assured
Goal:                            triforcehunt
Difficulty:                      normal
Item Functionality:              normal
Entrance Shuffle:                vanilla
Crystals required for GT:        1
Crystals required for Ganon:     6
Pyramid hole pre-opened:         Yes
Accessibility:                   items
Map shuffle:                     No
Compass shuffle:                 No
Small Key shuffle:               No
Big Key shuffle:                 No
Boss shuffle:                    none
Enemy shuffle:                   none
Enemy health:                    default
Enemy damage:                    default
Hints:                           Yes

Here is the completed SRM: ER_custom.zip

[trippsc2]

From what I can tell, that randomizer doesn't implement the ROM template changes that make it possible to track the number of keys for the dungeon found in chests. As such, the dungeon tracking will not work correctly. Whoever maintains that code will have to implement those ROM template changes for this to work correctly.

I will put a tooltip note next to that checkbox so that other users are aware that it may not work in some randomizer versions.

[marcmagus]

I ran into a similar issue with this seed https://berserkermulti.world/seed/bm66fmWZQBu9r_Bmt_752A

Also running OT1.6.1; not sure if this helps.

[marcmagus]

Per our discussion on discord: I have continued to experience this issue with all Archipelago keysanity seeds I have generated this week: https://berserkermulti.world/seed/77Be41N2TlaoYmeDbtN41w https://berserkermulti.world/seed/tM59t43CRO2gw1zuwIEwiw https://berserkermulti.world/seed/ZFhqlAk2RTyCTWCeu_e4TA https://berserkermulti.world/seed/hWhgxYH0ROGcvH9d-VW8PA https://berserkermulti.world/seed/bm66fmWZQBu9r_Bmt_752A

My most recent yaml is attached: marcmagus.yaml.txt

[trippsc2]

Changes to autotracking to address these issues will be pushed to v1.8.0, since an overhaul will be necessary for Key Drop Shuffle.

[trippsc2]

Would it be possible for you to send me your completed .srm file for one of these seeds? That way I can test without having to play through the seed.

[marcmagus]

Sure, attached is the first one I could grab (I assume attaching a gzip is ok? I'm slightly unclear on details of github attachments). Let me know if there's an issue/more would help.

T1_P1_marcmagus_77Be41N2TlaoYmeDbtN41w.srm.gz

[trippsc2]

This is mostly fixed in my dev build. The memory addresses for the number of dungeon checks made have been moved in Berserker's project. In addition, the former location for that data, that is currently used by VT randomizer and others, is being used for other things in Berserker's project. What this means is that some dungeons will show strange results until you make your first check in them in Berserker ROMs.

I have an issue open (https://github.com/Berserker66/MultiWorld-Utilities/issues/171) to hopefully keep these addresses empty, but I haven't heard anything back yet. If that isn't a feasible option, I have a few other ideas on how to fix the issue, but I'll work closely with Berserker's team on it.

I plan to release v1.8.0 once I have a more optimal solution to that problem.

[CaitSith2]

Corrected with Berserker Main and Berserker Doors.

Specifically documented that the memory range formerly used by compass counters is NOT TO BE REUSED for anything else. https://github.com/CaitSith2/z3randomizer/commit/a83315e12241c55f2128b1e6803a6e1f1637f542

[trippsc2]

These bugs should be fixed in v1.8.0+. We'll either reopen this or open a new issue if they pop up again. Closing this.