[MLton] using MLBs for scripting

[Daniel C. Wang]

This seems like an uncessary complication that doesn't buy that much.
What's the real motivation? Just to take advantage of the byte-code interpreter?

However, if you do want to go this route, should there be some escape 
mechanism, so you can tell when you are including a snipet of code inline vs 
   an mlb directive.

----------------------------------------------------------------------
#!/usr/bin/mlton

$(MLTON_ROOT)/basis/basis.mlb
<<EOF
val () = print "hello\n"
EOF
----------------------------------------------------------------------
which is just an abbrevation for having a file hello.sml and mentioning it 
there.

Still, I really don't see the need.

Stephen Weeks wrote:
{stuff deleted}
> 
> Thoughts?
>