Joined: 12/8/2012
Posts: 706
Location: Missouri, USA
So I've got to thinking about looking deeper into RAM addresses, and such, for a current run I'm working on and entertained the following thought:
Has anyone ever contacted a former game programmer (e.g. Sega, NES) and picked their brain about how a specific game functions?
For instance, Mega Man 2 is one of the most impressively optimized runs I've seen on this site. What could be the possibility, say, of hunting down Nobuyuki Matsushima, calling him up on the phone and saying "Hey, could you share any programming knowledge/tricks about this game that we haven't yet figured out"?
If anything, imagine the cool factor in having one of these pioneers as a co-author:D
"But as it is written, Eye hath not seen, nor ear heard, neither have entered into the heart of man, the things which God hath prepared for them that love him." - 1 Corinthians 2:9
Ha, you got a point. I can only imagine the reaction of some of the old-timer programmers being something like this:
o_0
That Cave Story run is impressive, so far. Kind of reminds of Milon's Secret Castle for the NES.
"But as it is written, Eye hath not seen, nor ear heard, neither have entered into the heart of man, the things which God hath prepared for them that love him." - 1 Corinthians 2:9
I suspect the big problem you'll run into is that after 20+ years, most programmers are not going to remember much at all of how their old programs work. Hell, I have enough trouble remembering the stuff I wrote 5 years ago. What they do remember will probably be the tricky problems they had to solve, and the elegant hacks; it might be useful to know that stuff, but statistically speaking, you have pretty low odds of it revealing the potential for useful glitches; there's just too much code in your average game.
Now, if you could get your hands on the old source code, that'd be a different story!
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Lidnariq (who ported the [2340] NES Driar by MESHUGGAH in 05:14.45 to NROM) was kind enough to verify my L+R glitch that makes a "fixed left step" which saves 1 frame when picking up stars on the left side and also explained it to me.
NES MC Kids has a detailed source code if I'm right and the developers also made tutorials on different approaches of implementing physics in to platformer games and how the achieved it and what they had to look after.
Regarding asking the developer to check a video where they game is being destroyed by someone (I mean heavy glitches), I'm not sure that everyone will react with "Wow, very nice job" but "this isn't the way it's meant to be played".
For example, Valve/Gearbox/Turtle Rock Studios and mostly those companies that used to buy other game engine sources and modify it to their tastes are not that happy when there's a game breaking glitch. There are many 0 day exploits that I've already explained to them and even wrote solutions but the answer is either "Soon™" or a generic "Yes we know about that" and end of story. One of my 0 day exploit is 9 (!) years old and it's still working (a simple DoS to crash Half-Life TV).
However most of the times the developers are kind enough and actually gives out unreleased informations if you tell them why do you ask it and how will you use it. And if they still have that information of course.
PhD in TASing 🎓 speedrun enthusiast ❤🚷🔥 white hat hacker ▓ black box tester ░ censorships and rules...
Developer time is not free, of course, so every bug that exists in a game has to be weighed:
* How long will it take to fix? This could be a very long time indeed depending on what causes the bug.
* How bad are the consequences of the bug occurring?
* How likely is the bug to be triggered?
Bugs that are likely to happen and bugs that have bad effects tend to get priority. If "your" bug is unlikely to occur or is relatively benign then you'll probably just get ignored, especially for singleplayer games where exploiting the bug can't harm other peoples' enjoyment of the game.
Pyrel - an open-source rewrite of the Angband roguelike game in Python.
Joined: 12/8/2012
Posts: 706
Location: Missouri, USA
Derakon wrote:
I suspect the big problem you'll run into is that after 20+ years, most programmers are not going to remember much at all of how their old programs work...there's just too much code in your average game.
Very good point. Also, I was amazed at the sheer amount of code that popped up the first time I used RAM Search (70,000+ possibilities).
Now, if you could get your hands on the old source code, that'd be a different story!
That would be the $1,000,000 to ask, then:D
MESHUGGAH wrote:
Regarding asking the developer to check a video where they game is being destroyed by someone (I mean heavy glitches), I'm not sure that everyone will react with "Wow, very nice job" but "this isn't the way it's meant to be played".
I'm sure that's possible. The glitched runs of Earthbound and Chrono Trigger immediately come to mind.
However most of the times the developers are kind enough and actually gives out unreleased informations if you tell them why do you ask it and how will you use it. And if they still have that information of course.
I think that would be the coolest to give a shout out to the actual person that programmed the game you made a TAS for!
"But as it is written, Eye hath not seen, nor ear heard, neither have entered into the heart of man, the things which God hath prepared for them that love him." - 1 Corinthians 2:9
"But as it is written, Eye hath not seen, nor ear heard, neither have entered into the heart of man, the things which God hath prepared for them that love him." - 1 Corinthians 2:9
