5 ms·

Elite on the 6502: The original 6502 assembly source, heavily commented

212 points by CharlesW 2y ago

cookiengineer 2y ago
The map overview is so well done. The comments are so nice and thoroughly explaining what is stored in which register and what's being manipulated. I wish all codebases were like this. This must have taken months to document it. [1] https://elite.bbcelite.com/c64/articles/map_of_the_source_code.html https://elite.bbcelite.com/c64/articles/map_of_the_source_co... [2] https://elite.bbcelite.com/c64/main/subroutine/tactics_part_1_of_7.html https://elite.bbcelite.com/c64/main/subroutine/tactics_part_...
- MarkMoxon 2y ago
  Glad you like it! I’ve been working on this for nearly five years, along with my other software archaeology projects - I started in May 2020, back in lockdown. It’s a true labour of love. Here’s a timeline of the project so far: https://elite.bbcelite.com/about_site/site_history.html https://elite.bbcelite.com/about_site/site_history.html I’m sure it will keep growing, there’s always something more to write about…
hinkley 2y ago
If I were king of the forest, there would be a degree program in Comparative Coding (comparative lit for code). We don’t treat this as a creative discipline and thus we don’t spend time looking at masters, how they were successful and the ways they were just as human as anyone. Someone would get paid for writing stuff like this, same as people getting paid to write about Poe, Cummings, Hemingway, Thoreau. But I guess the problem is that the way we price teaching and coding, the gap is far too wide between talking about and doing, unless you do it as a side hustle/hobby.
- atan2 2y ago
  Some people do teach this type of content and do a pretty good job at it. DisplacedGamers on youtube is a great channel, Gustavo Pezzi from pikuma.com is excellent, and there is also Ben Eater also covering 6502 programming, just to name a few.
  - akoboldfrying 2y ago
    Having greatly enjoyed Ben Eater's content (especially the build-your-own-VGA!), I'll definitely check those others out. Thanks!
- nradov 2y ago
  Should people get paid to write about Poe, Cummings, Hemingway, Thoreau? The original works themselves are pretty accessible. People can just read them and form their own interpretations. I'm skeptical that I really gained anything from reading literary criticism, at least it didn't make me appreciate the original works more or make me a better writer. I understand that there's a long tradition of scholarship in comparative literature and I'm not trying to be anti-intellectual or dismissive of an entire field of study, but maybe we should examine our assumptions and consider whether we're getting a good value from paying academics to do that work? Is it possible that society would be better off if we paid them to do something else, like maybe write new literature?
  - hinkley 2y ago
    My English Lit friend definitely was aware of a trap of learning to pick apart literature before you can create your own. And then fell into it anyway. I suppose it’s like medical students and the DSM, sometimes it just freaks you out instead of informing. It should probably be an MS class or a senior level class.
  - KerrAvon 2y ago
    I don't actually care one way or another about that particular subject, but if you start making judgements about academic fields of study are worthwhile to pay for on the basis of value to society, you allow other people -- more ignorant people, let's say -- with different concepts of societal value to make somewhat unfortunate decisions. You might end up without any new literature at all.
    - nradov 2y ago
      Very little literature is written by academics. People who want to write new literature can just go ahead and do it. In a world of limited time and resources we always make judgements about which academic fields of study are worthwhile. At some level it's a zero-sum game: time spent on comparative literature is time not spent on philosophy or history or creative writing. So, given that we have to prioritize which people to pay and which courses to require, how should we make those decisions?
- wkat4242 2y ago
  I was talking to a friend who works in development and he says that they even frown on recursion now. They think it's too complex and causes junior programmers to footgun. It's a shame because recursive code is often the most clean, elegant and fast for things like traversing trees.
  - hinkley 2y ago
    One of my peeves with JavaScript is that I feel the reduce() function has its arguments in the wrong order and increases cognitive load. Elixir’s is “right”. I’ve never been that much of a fan of recursion myself. Iteration and list comprehension for me.
  - fifilura 2y ago
    I don't really write code that traverse trees that often these days. I do use libraries or services (e.g. SQL databases) that do though. And I have no idea whether the implementation of map/reduce I use is implemented as a recursive function or not, I just use it. If your friend works with implementing libraries or database engines I would be a bit shocked. If he implements applications and services I could agree that keeping things simple and not inventing your own frameworks could be a good idea.
- pjc50 2y ago
  The MIT "platform studies" series is quite close to that, http://platformstudies.com/ http://platformstudies.com/
- therealcamino 2y ago
  There's a famous annotated version of Unix V6 that falls into this category. I heard that it had copyright issues for years, though. It'd be interesting to see people write more of these about open source code. https://cs3210.cc.gatech.edu/r/unix6.pdf https://cs3210.cc.gatech.edu/r/unix6.pdf
fredoralive 2y ago
The site also now has a disassembly of !Lander, which in hindsight is an amazing demo (that begat Zarch / Virus) - super smooth 3D in 1987 on a “micro”. As a kid in the early ‘90s I didn’t appreciate it, it was that exploding spaceship thing you could run on the computers at school. But in hindsight, wow that ARM and VIDC could be amazing.
- yzydserd 2y ago
  Thanks for the Lander pointer. Virus blew my mind when I first saw it, those decades ago. Now to restrain myself from hours in the source code.
dekhn 2y ago
The game elite (on my apple iie) played a big role in me being interested in 3d graphics (most importantly, hidden line removal). I was really curious as a high school student how the game achieved what it did, and it wasn't until later that I read a few interviews with the authors (and later, the source code) that I began to appreciate just how unprepared I was back then to write a first-class game. Beyond that it was a great game to play. I used to buy drugs on poor agricultural planets, then sell them for massive profits on rich industrial ones, then buy illegal weapons and sell them back on the poor agricultural planets. Starfighting was quite good in terms of frame rate and enemy AI. At some point, hundreds of hours into the game, I was offered a unique mission to recover a spacecraft...
kmoser 2y ago
Related: https://news.ycombinator.com/item?id=42436783 https://news.ycombinator.com/item?id=42436783
mrlonglong 2y ago
The Teletext variant is amazing
markus_zhang 2y ago
I think Computer Architecture can be designed around legacy but fun hardwares. So instead of using some fantasy machines such as LC-3, students can directly get into the world of 6502. First semester: mostly concentrated on 6502 with a full 6502 CPU emulator as the final project. Second semester: concentrate on a real machine such as the NES. Students need to learn to improve their emulators to program in it (so, do you need a debugger? a decompiler?), as well as hacking roms, burning roms, using carts to load the games into a real NES, and programming new games for it. The game must run in the emulator as well as a real NES. Third semester for honour: upgrade to a 16/24/32-bit machine, e.g. 80286, SNES, 68K, whatever, and get some serious project done. Can use C if applicable. The project can even be a simple OS or a BASIC interpreter to pave the way for advanced classes. Or maybe a computer virus, anything that is fun and creative. OS and compiler classes can also be designed around real hardware and projects. e.g. Porting an early version of Linux to a once popular architecture and add a few small functionalities; Write a compiler for a LISP-like language for that architecture and OS. You don't even need Data Structure and Algorithm and those BS Java programming classes - they learn them on the way. And if they miss anything they can pick them up on the job.
- mrandish 2y ago
  I like your idea, though I'm probably biased because I learned computers and computer programming starting with zero experience, zero knowledge and nothing but a 4K, sub-1 Mhz Radio Shack Color Computer. Unfortunately, in 1980 there were virtually no real instructional materials available on microcomputers. Learning consisted mostly of just trying things, typing in listings from low-budget hobbyist 'zines and comparing notes at user's group meetings. One great advantage to your approach is that emulators are free and some of them, like MAME, have extremely powerful debugging and exploration tools built right in. Plus there's extensive developer documentation, SDKs and instructional materials available for many of the most popular platforms along with lots of commented code disassemblies. Having an actual course structured around curated sets of these materials would be pretty amazing.
  - markus_zhang 2y ago
    Yep, and I know some instructors actually tried that (from Carnegie-Mellon): https://bobrost.com/nes/ https://bobrost.com/nes/ I think it's a good idea to go further to work on more difficult retro platforms. I know teachers like LC-3 because it is simple, but I think they underestimate the devoted students.
leoc 2y ago
The "Advanced User Guide" referred to several times in the comments is The Advanced User Guide for the BBC Micro by Bray, Dickens and Holmes, 1983, ISBN 0 946827 00 1 . https://stardot.org.uk/forums/viewtopic.php?f=42&t=17242 https://stardot.org.uk/forums/viewtopic.php?f=42&t=17242 https://archive.org/details/bbc-micro-advanced-user-guide/ https://archive.org/details/bbc-micro-advanced-user-guide/ I can't find a source for this now, but IIRC Bell and Braben themselves used this when writing Elite. (I am not an expert.)
- MarkMoxon 2y ago
  Thank you, that's a good point. The site started off with just the BBC Micro version, and I never expanded the book references to be clearer. Anyway, I have reworded every reference to books in the commentary so they now contain the full title and author list, so hopefully people will be able to find the relevant tomes. I've also linked to the books from my "useful links" page in a more prominent manner: https://elite.bbcelite.com/about_site/useful_links.html https://elite.bbcelite.com/about_site/useful_links.html I hope that helps!
nejsjsjsbsb 2y ago
Past discussions https://news.ycombinator.com/from?site=bbcelite.com https://news.ycombinator.com/from?site=bbcelite.com
vardump 2y ago
Have to mention the new Commodore VIC-20 version of Elite: https://www.youtube.com/watch?v=3J0xRcux1FI https://www.youtube.com/watch?v=3J0xRcux1FI
throwaway81523 2y ago
"This site contains the original 1980s source code for the classic space game Elite, with every single line documented and (for the most part) explained. It is literally the original 6502 assembly source code, just heavily commented." For those wondering what Elite was.
hyperbrainer 2y ago
Alexander the Ok, on youtube, has an amazing video on this. It is not quite as technical as the webiste for obvious reasons, but great overview and historical facts as well. https://www.youtube.com/watch?v=lC4YLMLar5I https://www.youtube.com/watch?v=lC4YLMLar5I

212 points by CharlesW 2y ago

The map overview is so well done. The comments are so nice and thoroughly explaining what is stored in which register and what's being manipulated. I wish all codebases were like this. This must have taken months to document it. [1] https://elite.bbcelite.com/c64/articles/map_of_the_source_code.html https://elite.bbcelite.com/c64/articles/map_of_the_source_co... [2] https://elite.bbcelite.com/c64/main/subroutine/tactics_part_1_of_7.html https://elite.bbcelite.com/c64/main/subroutine/tactics_part_...

MarkMoxon 2y ago
Glad you like it! I’ve been working on this for nearly five years, along with my other software archaeology projects - I started in May 2020, back in lockdown. It’s a true labour of love. Here’s a timeline of the project so far: https://elite.bbcelite.com/about_site/site_history.html https://elite.bbcelite.com/about_site/site_history.html I’m sure it will keep growing, there’s always something more to write about…

hinkley 2y ago

If I were king of the forest, there would be a degree program in Comparative Coding (comparative lit for code). We don’t treat this as a creative discipline and thus we don’t spend time looking at masters, how they were successful and the ways they were just as human as anyone. Someone would get paid for writing stuff like this, same as people getting paid to write about Poe, Cummings, Hemingway, Thoreau. But I guess the problem is that the way we price teaching and coding, the gap is far too wide between talking about and doing, unless you do it as a side hustle/hobby.

atan2 2y ago
Some people do teach this type of content and do a pretty good job at it. DisplacedGamers on youtube is a great channel, Gustavo Pezzi from pikuma.com is excellent, and there is also Ben Eater also covering 6502 programming, just to name a few.
- akoboldfrying 2y ago
  Having greatly enjoyed Ben Eater's content (especially the build-your-own-VGA!), I'll definitely check those others out. Thanks!
nradov 2y ago
Should people get paid to write about Poe, Cummings, Hemingway, Thoreau? The original works themselves are pretty accessible. People can just read them and form their own interpretations. I'm skeptical that I really gained anything from reading literary criticism, at least it didn't make me appreciate the original works more or make me a better writer. I understand that there's a long tradition of scholarship in comparative literature and I'm not trying to be anti-intellectual or dismissive of an entire field of study, but maybe we should examine our assumptions and consider whether we're getting a good value from paying academics to do that work? Is it possible that society would be better off if we paid them to do something else, like maybe write new literature?
- hinkley 2y ago
  My English Lit friend definitely was aware of a trap of learning to pick apart literature before you can create your own. And then fell into it anyway. I suppose it’s like medical students and the DSM, sometimes it just freaks you out instead of informing. It should probably be an MS class or a senior level class.
- KerrAvon 2y ago
  I don't actually care one way or another about that particular subject, but if you start making judgements about academic fields of study are worthwhile to pay for on the basis of value to society, you allow other people -- more ignorant people, let's say -- with different concepts of societal value to make somewhat unfortunate decisions. You might end up without any new literature at all.
  - nradov 2y ago
    Very little literature is written by academics. People who want to write new literature can just go ahead and do it. In a world of limited time and resources we always make judgements about which academic fields of study are worthwhile. At some level it's a zero-sum game: time spent on comparative literature is time not spent on philosophy or history or creative writing. So, given that we have to prioritize which people to pay and which courses to require, how should we make those decisions?
wkat4242 2y ago
I was talking to a friend who works in development and he says that they even frown on recursion now. They think it's too complex and causes junior programmers to footgun. It's a shame because recursive code is often the most clean, elegant and fast for things like traversing trees.
- hinkley 2y ago
  One of my peeves with JavaScript is that I feel the reduce() function has its arguments in the wrong order and increases cognitive load. Elixir’s is “right”. I’ve never been that much of a fan of recursion myself. Iteration and list comprehension for me.
- fifilura 2y ago
  I don't really write code that traverse trees that often these days. I do use libraries or services (e.g. SQL databases) that do though. And I have no idea whether the implementation of map/reduce I use is implemented as a recursive function or not, I just use it. If your friend works with implementing libraries or database engines I would be a bit shocked. If he implements applications and services I could agree that keeping things simple and not inventing your own frameworks could be a good idea.
pjc50 2y ago
The MIT "platform studies" series is quite close to that, http://platformstudies.com/ http://platformstudies.com/
therealcamino 2y ago
There's a famous annotated version of Unix V6 that falls into this category. I heard that it had copyright issues for years, though. It'd be interesting to see people write more of these about open source code. https://cs3210.cc.gatech.edu/r/unix6.pdf https://cs3210.cc.gatech.edu/r/unix6.pdf

fredoralive 2y ago

The site also now has a disassembly of !Lander, which in hindsight is an amazing demo (that begat Zarch / Virus) - super smooth 3D in 1987 on a “micro”. As a kid in the early ‘90s I didn’t appreciate it, it was that exploding spaceship thing you could run on the computers at school. But in hindsight, wow that ARM and VIDC could be amazing.

yzydserd 2y ago
Thanks for the Lander pointer. Virus blew my mind when I first saw it, those decades ago. Now to restrain myself from hours in the source code.

dekhn 2y ago

The game elite (on my apple iie) played a big role in me being interested in 3d graphics (most importantly, hidden line removal). I was really curious as a high school student how the game achieved what it did, and it wasn't until later that I read a few interviews with the authors (and later, the source code) that I began to appreciate just how unprepared I was back then to write a first-class game. Beyond that it was a great game to play. I used to buy drugs on poor agricultural planets, then sell them for massive profits on rich industrial ones, then buy illegal weapons and sell them back on the poor agricultural planets. Starfighting was quite good in terms of frame rate and enemy AI. At some point, hundreds of hours into the game, I was offered a unique mission to recover a spacecraft...

kmoser 2y ago

Related: https://news.ycombinator.com/item?id=42436783 https://news.ycombinator.com/item?id=42436783

mrlonglong 2y ago

The Teletext variant is amazing

markus_zhang 2y ago

I think Computer Architecture can be designed around legacy but fun hardwares. So instead of using some fantasy machines such as LC-3, students can directly get into the world of 6502. First semester: mostly concentrated on 6502 with a full 6502 CPU emulator as the final project. Second semester: concentrate on a real machine such as the NES. Students need to learn to improve their emulators to program in it (so, do you need a debugger? a decompiler?), as well as hacking roms, burning roms, using carts to load the games into a real NES, and programming new games for it. The game must run in the emulator as well as a real NES. Third semester for honour: upgrade to a 16/24/32-bit machine, e.g. 80286, SNES, 68K, whatever, and get some serious project done. Can use C if applicable. The project can even be a simple OS or a BASIC interpreter to pave the way for advanced classes. Or maybe a computer virus, anything that is fun and creative. OS and compiler classes can also be designed around real hardware and projects. e.g. Porting an early version of Linux to a once popular architecture and add a few small functionalities; Write a compiler for a LISP-like language for that architecture and OS. You don't even need Data Structure and Algorithm and those BS Java programming classes - they learn them on the way. And if they miss anything they can pick them up on the job.

mrandish 2y ago
I like your idea, though I'm probably biased because I learned computers and computer programming starting with zero experience, zero knowledge and nothing but a 4K, sub-1 Mhz Radio Shack Color Computer. Unfortunately, in 1980 there were virtually no real instructional materials available on microcomputers. Learning consisted mostly of just trying things, typing in listings from low-budget hobbyist 'zines and comparing notes at user's group meetings. One great advantage to your approach is that emulators are free and some of them, like MAME, have extremely powerful debugging and exploration tools built right in. Plus there's extensive developer documentation, SDKs and instructional materials available for many of the most popular platforms along with lots of commented code disassemblies. Having an actual course structured around curated sets of these materials would be pretty amazing.
- markus_zhang 2y ago
  Yep, and I know some instructors actually tried that (from Carnegie-Mellon): https://bobrost.com/nes/ https://bobrost.com/nes/ I think it's a good idea to go further to work on more difficult retro platforms. I know teachers like LC-3 because it is simple, but I think they underestimate the devoted students.

leoc 2y ago

The "Advanced User Guide" referred to several times in the comments is The Advanced User Guide for the BBC Micro by Bray, Dickens and Holmes, 1983, ISBN 0 946827 00 1 . https://stardot.org.uk/forums/viewtopic.php?f=42&t=17242 https://stardot.org.uk/forums/viewtopic.php?f=42&t=17242 https://archive.org/details/bbc-micro-advanced-user-guide/ https://archive.org/details/bbc-micro-advanced-user-guide/ I can't find a source for this now, but IIRC Bell and Braben themselves used this when writing Elite. (I am not an expert.)

MarkMoxon 2y ago
Thank you, that's a good point. The site started off with just the BBC Micro version, and I never expanded the book references to be clearer. Anyway, I have reworded every reference to books in the commentary so they now contain the full title and author list, so hopefully people will be able to find the relevant tomes. I've also linked to the books from my "useful links" page in a more prominent manner: https://elite.bbcelite.com/about_site/useful_links.html https://elite.bbcelite.com/about_site/useful_links.html I hope that helps!

nejsjsjsbsb 2y ago

Past discussions https://news.ycombinator.com/from?site=bbcelite.com https://news.ycombinator.com/from?site=bbcelite.com

vardump 2y ago

Have to mention the new Commodore VIC-20 version of Elite: https://www.youtube.com/watch?v=3J0xRcux1FI https://www.youtube.com/watch?v=3J0xRcux1FI

throwaway81523 2y ago

"This site contains the original 1980s source code for the classic space game Elite, with every single line documented and (for the most part) explained. It is literally the original 6502 assembly source code, just heavily commented." For those wondering what Elite was.

hyperbrainer 2y ago

Alexander the Ok, on youtube, has an amazing video on this. It is not quite as technical as the webiste for obvious reasons, but great overview and historical facts as well. https://www.youtube.com/watch?v=lC4YLMLar5I https://www.youtube.com/watch?v=lC4YLMLar5I