Development talk:Super Mario Kart/November 29, 1991 Proto

On the subject of the repair (by MrL314)

To do: still a lot more information I need to add. -MrL314

Heyo, I'm MrL314. I'm one of the few people that were involved in the repair of this prototype. Before I start, special thanks to Sirgren, Catador, R4M0N, Dirtbag, atomicdesolation, and SmorBjorn. There is a lot of technical background that I will be going over with regard to repairing this prototype to as accurate as we could make it. Lots of decisions were made along the way, and a lot of cool stuff was found! However, I figured that a lot of the info I will be sharing here is not particularly suited for the main article, but does give some background as to some of the what and why with regards to the data in the repair.

If there is anything in this discussion that warrants being placed in the main article, feel free to move it there.

Alright. On to discussion of the repair!

Repair download link

The link to the latest version of the repair can be found here.

Track Layouts

The track layouts used in the repair do not appear as shown in any files from the leaks unfortunately. However we have worked to recreate the tracks as accurately as we could. As they cannot be confirmed to be accurate to what would have been used at the time, these have been omitted from the main article. HUGE thanks to Sirgren and Catador for their amazing work in recreating most of the maps based on the drive-data file, and files based on the F-zero map format!

• Great work on the recreations! The mockup tracks could be shown on the main page by giving the uploaded pngs the comment "mockup|SNES" (with the {{ }} brackets) and by giving the tracks on the page the heading "mockup". Something like this:

"Below is a recreation of the track, how it might have looked:"

Mockup	Final Map
File:SMK nov91 mc3 track Mockup.png

Kowski 12:05, 2 August 2021 (UTC)

• • I’m actually not too sure on what the ruling about that is, so I didn’t include it in the main article, since it might cause some confusion. I guess it could work being listed here for now? Probably best to keep the mock-up info here for now I’d say. Not sure. I personally am indifferent about having them shown or not, but I feel like having them on the main page might be odd.

--MrL314 (talk) 13:41, 7 August 2021 (UTC)

• • • Yes, you are right. Keeping the mock up info here is a much better idea, so i created a sub-page where some of the mockups can be shown. But I don't think that all tracks have to be listed here, only those that differ drastically, such as Mario Circuit 2:

Early Map Mock Ups Some recreations of the tracks based on the early drive-data

--Kowski (talk) 17:23, 11 August 2021 (UTC)

General Info

Confusion about the February 13 date

A lot of the data and references you will see when speaking about code will refer to files that have a last modified date of February 13. This is misleading as a lot of this data cannot possibly be from Feb13. It seems like a mass amount of data was moved from one place to another on February 13, 1992, and caused a bunch of files to have a last modified date of February 13. In fact, some of the .rel files dated February 13 actually are from January 13, based on the header data of the .rel files. This made a lot of this repair very confusing, so if there is anything that I got wrong, PLEASE tell me so that I may fix the repair.

To do: other evidence?

drive-data

Dated February 13, 1992 but has data that contradicts the map .SCR files being updated in December 1991 that match final maps. This file must have been used in a prototype previous to the February Prototype, and previous to December 1991, meaning it was most likely used in this prototype. There are a few inconsistencies however that should not be overlooked. One of these is the finish line data, which you can read about in the main article.

To do: other proof

Unused Gate object

Code for these unused objects can be found in kimura/kart/mak/Scene.asm and kimura/kart/mak/Pole.asm. Based on the code, these seem to be semicircle arcs made out of coins. Interestingly, these objects use the DSP-1 Formate and Attitude commands, which don't appear anywhere in the final game's code. I have made a couple of mock-ups of what these gates would look like, which you can find here. As of right now though, I am unsure if these gates are actually implemented in this version of the game, since not all of the data in Scene and Pole matches with the compiled hex.

To do: pictures

Unused Logo Arch object

Code for these unused objects can be found in kimura/kart/mak/Scene.asm and kimura/kart/mak/Pole.asm. These seem to be loosely based off of the infamous DUNLOP bridges! They appear as an arch of letters over some line on the field. I am unsure at this time if the other two bridges are meant to line up with a start/finish line, however you can see video of my recreations in action for the NINTENDO arch and the CASTEL arch.

To do: pictures?

Breakable Block Tiles

To do: HATA.SCR picture tile properties picture

Multiple files that contain remnants of Ghost Valley 2 show the bottom left quadrant of an early version of the map. However none of those files show Ghost Valley 2 with breakable block barriers. They instead show smoothed out diagonal edges without walls. sugiyama/CAR/HATA.SCR (dated Oct30,1991) at least shows that this version of Ghost Valley 2 existed before then. However, there is stronger evidence when we look at the pre-compiled code. Each track theme has a table of tile-interactions that indicate what type of tile each tile in the tileset is. It directly corresponds to the tileset visually if looking at a 16x16 table of the tiles (see picture to the right). In kimura/games/temp/kart-data.asm, starting at line 178 is the tile properties tables for each theme, with line 196 starting the Ghost Valley (Obake) tile properties. Here we can see that Ghost Valley's tileset only has 2 rows that have tile properties, meanwhile having breakable blocks would have to include a 3rd row (as well as those secondary floor tiles you see in the final game). Thus I can confidently say that the breakable blocks were not implemented at this time.

To do: GV tileset here

HATA.CGX

Last modified on October 30, 1991. Since Lives and Coins were not implemented yet, files containing sprites for the coin and lives meter could not have been used. Since all of the other Lakitu files contain the lives and coin counters, they could not have been used at the time. The only remaining candidate is sugiyama/CAR/HATA.CGX, which does not contain those sprites, and is in fact formatted differently than the final sheet.

To do: finish pls

MrL314 (talk)