> As I'm ill at home, I've given a look to Tea, and I like it :-)

Since my assistant stays home, I don't see why I should go to school either.

> Martin (V.), do you intend to finish your scripting system? Because if you
> do
> not think so, I'm reading to use tea, if it meets all the requirements (and
> I
> think that, with minor changes, it does) below:

I do think so. But holidays seem blurry since I got my travel ticket (called
'ordre de marche').
But if tea is ready (pun intended) then you should go ahead and use it.

> - Parser in the preparse phase creates a context. The context should be able
> to be serialized with glob2's stream and reloaded to be executed.
> - Parser in the preparse phase should be able to parser several scripts
> independently without side effects to the parser (so that we can parse
> multiple stories).
> - Execution should be able to be limited to a certain number of instruction,
> to prevent infinite loop in case of buggy scripts

The number of these (pseudo-)instructions has to be deterministic (no
time-based) so it executes exactly the same way on all the machines in
multiplayer games.

> - Some instructions, known as waits, should suspend one thread execution
> (return from execute) and reevaluate the condition on next call to execute.

Additionally, suspended execution should be storable/reloadable including stack
and heap.

> If this works, tea is ready for glob2.

Yep.

Martin


_______________________________________________
glob2-devel mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/glob2-devel

Reply via email to