On Wed, 28 Apr 2010 02:20:58 +0900, Lars T. Kyllingstad <public@kyllingen.nospamnet> wrote:

> For systems where real is 64 bits it may be possible to use an 80-bit std.numeric.CustomFloat as an intermediate step between the stored data and the 64-bit type.
Thanks for your advice!

On Wed, 28 Apr 2010 02:20:58 +0900, Lars T. Kyllingstad <public@kyllingen.nospamnet> wrote:

> For systems where real is 64 bits it may be possible to use an 80-bit std.numeric.CustomFloat as an intermediate step between the stored data and the 64-bit type.
I tried std.numeric.CustomFloat:
-----
alias CustomFloat!(1, 64, 15) RealRep;
-----
But this code doesn't comile...

On Tue, 27 Apr 2010 21:27:35 +0900, Lars T. Kyllingstad <public@kyllingen.nospamnet> wrote:

> It doesn't seem to support the real type.  It this a limitation of the MessagePack format or just an oversight?  Even if the format doesn't directly support 80-bit floats, it should be possible to wrap them somehow.
Finished.

http://bitbucket.org/repeatedly/msgpack4d/changeset/87954e46548e

If real type equals double type:
- Packer uses double type for real type.
- Unpacker raises UnpackException when receives real type.
  The reason for this, To use real type means that precision is important.

How does it look?

On Tue, 27 Apr 2010 15:37:56 -0300, Masahiro Nakagawa <repeatedly@gmail.com> wrote:

> On Wed, 28 Apr 2010 02:20:58 +0900, Lars T. Kyllingstad <public@kyllingen.nospamnet> wrote:
>
>> For systems where real is 64 bits it may be possible to use an 80-bit std.numeric.CustomFloat as an intermediate step between the stored data and the 64-bit type.
> I tried std.numeric.CustomFloat:
> -----
> alias CustomFloat!(1, 64, 15) RealRep;
> -----
> But this code doesn't comile...

CustomFloat currently is currently broken. There are some partial patches (see http://d.puremagic.com/issues/show_bug.cgi?id=3520), but I don't they will address this issue.

On Wed, 28 Apr 2010 10:02:52 +0900, Robert Jacques <sandford@jhu.edu> wrote:

> On Tue, 27 Apr 2010 15:37:56 -0300, Masahiro Nakagawa <repeatedly@gmail.com> wrote:
>
>> On Wed, 28 Apr 2010 02:20:58 +0900, Lars T. Kyllingstad <public@kyllingen.nospamnet> wrote:
>>
>>> For systems where real is 64 bits it may be possible to use an 80-bit std.numeric.CustomFloat as an intermediate step between the stored data and the 64-bit type.
>> I tried std.numeric.CustomFloat:
>> -----
>> alias CustomFloat!(1, 64, 15) RealRep;
>> -----
>> But this code doesn't comile...
>
> CustomFloat currently is currently broken. There are some partial patches (see http://d.puremagic.com/issues/show_bug.cgi?id=3520), but I don't they will address this issue.
Yeah. CustomFloat uses std.bitmanip.bitfields for internal state, but
bitfields can't deal with large size :(