[MLton] cvs commit: mlb integration

[Stephen Weeks]

> > Agreed.  allowImport and allowExport should be normal (and I still
> > wouldn't mind their defaults becoming false :-).
> 
> Sounds fine.  We might consider making their defaults true when compiling
> a single .sml file, or providing a command line option to set the default.

Or possibly even both.  But at least the command-line option.

> > Sounds trivial.  In fact, that's what I would expect the behavior of
> > "mlton -stop f" to be when the input file is an mlb.
> 
> Maybe.  The Basis library is always a tricky thing -- do you really want
> to pull-in all of those files.

Maybe not, if the idea is to use it like cmcat.  But it's easy enough
to take them out with grep afterwards.

> > Although, one advantage to having these, if they truly work, is that
> > we could use them for self compiling, and hence develop purely with
> > mlbs.  But I imagine there will be a lot of time-wasting gotchas in
> > writing these.
> 
> Yes, and they aren't complete solutions.  Because cm2pld will still leave
> .cm files in the imports, and likewise with mlb2pld.  So, you still need
> to fiddle with file extensions, etc.

Blech.  I think what I'd want is a whole-program mlb2pld that produces
a single pld for a whole program suitable for consumption by SML/NJ.
Don't worry about converting individual mlbs/cms.