GHC 5.00 release update
Thu, 22 Mar 2001 02:30:33 -0800
We decided to do a no-op on this one, at least for GHC 5.0.
We don't want to steal a H98 keyword.
Adding a new -ffi flag is ok in principle, but it complicates
matters in the parser (now two groups of extra keywords).
So it's a bit of extra hassle; but we could be pressured if
it was important.
| -----Original Message-----
| From: Manuel M. T. Chakravarty [mailto:email@example.com]
| Sent: 24 February 2001 08:46
| To: Simon Marlow
| Cc: Simon Peyton-Jones; firstname.lastname@example.org
| Subject: RE: GHC 5.00 release update
| Simon Marlow <email@example.com> wrote,
| > > | * don't lump the FFI in with the rest of -fglasgow-exts and
| > > | have a -fffi or so.
| > >=20
| > > I wonder if the best thing is to leave the FFI enabled
| > > permanently (i.e. not with -fglasgow-exts). The FFI
| > > is steadily becoming H98 in effect, and it would remove
| > > a flag.
| > >=20
| > > Opinions?
| > s'fine by me - I'd rather do this than add a -fffi flag,=20
| because that
| > would require us to plumb more state around the parser.
| I don't mind much, too. However, we are taking the varid
| `foreign' away.