Heh, I didn’t spoiler alerts could be nested. TIL

Hmm, works for me on Firefox, though I'm on Windows. You're on MacOS right? Could be an emscripten bug.

There are still a few little unported things, such as the fullscreen option.

You can also upload files to the browser's IDBFS database file system, although IVAN-wasm won't load savefiles from the original build of vanilla IVAN. I think this is because it was compiled so long ago with an older generation compiler, some of the data types might be assumed to be of a different size (in bytes). When these old data types are read into IVAN-wasm, there must be some sort of numerical under/overflow. I think it would take some serious CS expertise to patch it up.

There's also all the old bugs... 0.o All the same, it's progress towards porting Attnam/ivan to the web.

I think one of the potential possibilities would be to run this on node.js, and then the server could verify a playthrough and maintain a global scoreboard of "legit" hi-scores.

You can play vanilla IVAN in your browser

I guess this is what I was really intending to do before I got side-tracked with the CVS-to-git thing.

The source repository is here: https://github.com/Attnam/ivan-050-wasm

Have fun!

Heh, thanks. I was actually trying to do something else, but then I got sidetracked and did this instead

Lately I've been poking at retrieving the IVAN CVS repository, and I've managed to port the CVS repo to git.
I put the files up here under the Attnam GitHub org: https://github.com/Attnam/ivan-cvs-git

The earliest commit is from the 29th of July 2001. There's at least five years of history just in this one repository.
There is evidence in this repository (...according to Doc/Data/Lines.txt) that IVAN has been in development since at least 1 May 2001, with IVAN 0.23 alpha, and also earlier, in a project known as RL, which began on 8 April 1999.

I wrote up an explainer for porting from CVS to git using repotool, with IVAN CVS as an example.

If you want to see the whole history together with the Attnam/ivan history, you can write your own local "graft" by making Attnam/ivan-cvs-git a remote, and apply the graft to your local filesystem repository, just
from a bunch of git commands:

$ git remote add ivan-cvs-git https://github.com/Attnam/ivan-cvs-git.git
$ git fetch ivan-cvs-git
$ git replace --graft 3e507676261e9c1d6e08f257dd205e94cf0e35c6 870a5776defca8f860517cd3434430d8de8fcbaa

This basically says that the root of Attnam/ivan is the head of Attnam/ivan-cvs-git. Then you can print the whole contiguous log of 20+ years of IVAN development. Or Gource the repo and chill...

I think there is a way to attach Attnam/ivan-cvs-git to Attnam/ivan permanently, but not using the above method. I think it has to be more permanent, and needs coordination. I'm hoping someone with more experience can come along and do that.

At any rate, we have disinterred some more cool IVAN history from SourceForge, and the risk of that history disappearing has been lessened somewhat.

I guess like this?
I think you want the chain-link symbol to make urls.

Anyway, merged, thanks

Hey, I just spotted this: https://attnam.com/issues
Very nice! I like the broken bananas

Lol, my bad.

EDIT: There's a PR to fix this

Good work so far, keep it up Silver. Looks like it has the makings of the city of Bazaria.

Any time. Christmas is a good natural deadline though.
It's a good thing our INSTALL file is so comprehensive, and props to the guys who set up all those macos build scripts in the ci folder. Those are definitely still actively useful.
I was stuck conceptually the day before trying to build on ubuntu-latest. It later occurred to me that I should simply select macos-latest. Silly really.
In that vein, Microsoft's ci tooling is pretty impressive. I wonder what the catch is though? I know they came out with GitHub Copilot based on everyone's code (we should run some tests with it and see if we can get it to spit out some IVAN source code - we have some notoriously unique class and function names, hehe).

The windows build is the first to be built using proper GNU tooling for a long time. More sensible than that whack-ass MXE compiler, even if it gave a fully static binary exe. I really had no idea where those dependencies came from. At least with msys2 it's easy to trace it back to trusted maintainers.
Also easy now is finding dlls using cmake, definitely solved that headache once and for all.