[z-machine] Small contribution to the ZMachine specs

[Chris Pickett]

Ben Collins-Sussman wrote:
> On 3/26/06, Chris Pickett <chris.pickett at mail.mcgill.ca> wrote:
> 
> 
>>Have you looked at frotz?  I suppose it can be considered as a kind of
>>reference implementation.  It sounds like you're trying a clean-room
>>approach though, so maybe you don't want to do that...
>>
> 
> 
> Yeah, we're "spiritually" trying to do a clean-room implementation --
> not for any legal reasons.  Our hope is that one would be able to
> write a z-machine by using nothing but the spec.  I suppose we can
> always look at frotz, malyon.el, or any number of other
> implementations.  We'd still like to see the spec clarified, though. 
> :-)

Okay, since you are two, here's the perfect chance.  One of you looks at 
several existing implementations and updates the spec, the other 
implements the spec.  The resulting diff should be interesting.  Of 
course you already know about this technique, but the point is that it 
(can be) fun to do (especially since you get to skip the lawyer bit).

I have no idea about actually applying your patches, but my feeling is 
that a big diff saying, "this is what we needed to do it right, and 
we've looked at what everybody is already doing" might get approved 
faster in the long run than a bunch of little patches.

-- 
Chris Pickett
http://www.sable.mcgill.ca/~cpicke/