Re: cool rant on linear objects

Paul Rubin <http://phr.cx@[EMAIL
PROTECTED]
> writes:
> 'Use-Once' Variables and Linear Objects -- Storage Management,
> Reflection and Multi-Threading
> Henry G. Baker
> ACM Sigplan Notices 30, 1 (January 1995), 45-52.
> http://home.pipeline.com/~hbaker1/Use1Var.html
> ================
> I just came across this.  I would have thought it is well known (maybe
> it is), but I've never seen it before, and probably have not seen it
> referenced anywhere (since I would have looked at it).

It has been referenced, but not much.  That may be because it wasn't
the first paper on linearity in programming languages, it didn't
address how a static type system would enforce linearity and it only
hinted at dealing with it dynamically via pointer bits.  Without
either of these you have the equivalent of C programs with explicit
malloc/free.  C programs contain plenty of linear objects,
unfortunately if you make a mistake with the linearity you typically
get a SIGSEGV/SIGBUS or a corrupted heap :-<  Consequently papers
about linearity more commonly try to create a static type system that
can enforce linearity without egregious limits on what you can write.
Various examples can be found at
http://homepages.inf.ed.ac.uk/wadler/topics/linear-logic.html.