[MLton] starting builds

Fri, 9 Sep 2005 12:12:39 +0200

--Apple-Mail-4--222812034
Content-Transfer-Encoding: 7bit
Content-Type: text/plain;
	charset=US-ASCII;
	format=flowed

On Sep 9, 2005, at 11:52 AM, Wesley W. Terpstra wrote:
> I still don't have time to really look into this...
> Eventually, I'll get there, though. :-)

Well, I just looked around for the floor problem:
bugs.debian.org/153022

It seems that not all processors can do this (arm).
On some processors it doesn't work (hppa), but may be fixable.

I checked libc's 'floor' function. It worked on hppa.

Checking the basis standard, for once it covers this problem! :-)
We should raise an exception if the rounding mode is missing.

For the toInt/toLargeInt/etc, maybe it makes sense to use
libc math functions like 'floor' / 'ceil' / ... instead? It seems that
one can expect these to work on nearly all platforms.

It is possible that in the future, hppa will have this bug corrected,
but since some architectures will always have this problem, then
it makes sense (to me), to solve it on the MLton end anyways.

--Apple-Mail-4--222812034
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=US-ASCII

<HTML><BODY style=3D"word-wrap: break-word; -khtml-nbsp-mode: space; =
-khtml-line-break: after-white-space; "><DIV><DIV>On Sep 9, 2005, at =
11:52 AM, Wesley W. Terpstra wrote:</DIV><BLOCKQUOTE =
type=3D"cite"><DIV><FONT class=3D"Apple-style-span" color=3D"#000000">I =
still don't have time to really look into =
this...</FONT></DIV><DIV>Eventually, I'll get there, though. =
:-)</DIV></BLOCKQUOTE></DIV><BR><DIV>Well, I just looked around for the =
floor problem:</DIV><DIV>bugs.debian.org/153022</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV>It seems that not all =
processors can do this (arm).</DIV><DIV>On some processors it doesn't =
work (hppa), but may be fixable.</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV>I checked libc's 'floor' =
function. It worked on hppa.</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV>Checking the basis =
standard, for once it covers this problem! :-)</DIV><DIV>We should raise =
an exception if the rounding mode is missing.</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV><DIV>For the =
toInt/toLargeInt/etc, maybe it makes sense to use</DIV><DIV>libc math =
functions like 'floor' / 'ceil' / ... instead? It seems =
that</DIV><DIV>one can expect these to work on nearly all =
platforms.</DIV><DIV><BR class=3D"khtml-block-placeholder"></DIV><DIV>It =
is possible that in the future, hppa will have this bug =
corrected,</DIV><DIV>but since some architectures will always have this =
problem, then</DIV><DIV>it makes sense (to me), to solve it on the MLton =
end anyways.</DIV><DIV><BR =
class=3D"khtml-block-placeholder"></DIV></BODY></HTML>=

--Apple-Mail-4--222812034--