I’ve gotten the double-precision feature (Double-precision math in Glulx) and the hasundo/discardundo feature (@hasundo, @discardundo opcodes in Glulx) up to the point where I can call it a spec update.
I’ve also updated Glulxe (the interpreter) to 0.6.0, with support for these features.
(A Quixe release is coming soon. I’ve updated Quixe to support these features, but I also want to wrap in some GlkOte fixes which are not quite figured out yet.)
The I6 compiler update to support the new opcodes is in the pipeline: Double-precision opcodes by erkyrath · Pull Request #191 · DavidKinder/Inform6 · GitHub
If I’ve missed anything, please let me know. The last spec update was 2010, so I’ve probably forgotten something…
Oh, right, test cases. I’ve updated Glulxercise to test all the new features.
If you run this version with an old interpreter, the new test cases will just print “Interpreter claims to not support (double-precision / extended undo). Skipping test.”
Probably should redirect the fyrevm link to the ifwiki.
Yeah, good point. (Those were old versions.)
Is it worth putting the old versions in the repository? I removed glk-dev/oldspec too, but now those files don’t exist online outside of the git history.
You could rewrite the git history of the ifarchive-if-specs repo, but probably just linking to the files’ histories is enough.
To update IFWiki, will it be a simple case of adding new dates and version numbers to Glulx - IFWiki and Glulxe - IFWiki, or is there more to it that that?
Just dates and version numbers, I think.
There will also be intepreter updates to Quixe and Git. (I hope to get the Quixe update released today.)