Re: [boost] [serialization] portability
----Original Message---- From: troy d. straszheim [mailto:troy@resophonic.com] Sent: 18 November 2005 11:41
But floats and doubles are no problem. They have the same size and layout whereever you are, and since you're just flipping bytes and writing/loading binary, you don't have to worry about NaN or inf or any of that, the archive remains blissfully ignorant.
IBM format? VAX floating point? (Not forgetting that the VAX had TWO 64-bit floating point types :-) -- Martin Bonner Martin.Bonner@Pitechnology.com Pi Technology, Milton Hall, Ely Road, Milton, Cambridge, CB4 6WZ, ENGLAND Tel: +44 (0)1223 441434
On Fri, Nov 18, 2005 at 12:50:16PM -0000, Martin Bonner wrote:
----Original Message---- From: troy d. straszheim [mailto:troy@resophonic.com] Sent: 18 November 2005 11:41
But floats and doubles are no problem. They have the same size and layout whereever you are, and since you're just flipping bytes and writing/loading binary, you don't have to worry about NaN or inf or any of that, the archive remains blissfully ignorant.
IBM format? VAX floating point? (Not forgetting that the VAX had TWO 64-bit floating point types :-)
You know, I think that text_oarchive is inappropriately named. Clearly this is an ascii_oarchive, not to be confused with the ebcdic_oarchive. :) -t
troy d. straszheim wrote:
On Fri, Nov 18, 2005 at 12:50:16PM -0000, Martin Bonner wrote:
----Original Message---- From: troy d. straszheim [mailto:troy@resophonic.com] Sent: 18 November 2005 11:41
But floats and doubles are no problem. They have the same size and layout whereever you are, and since you're just flipping bytes and writing/loading binary, you don't have to worry about NaN or inf or any of that, the archive remains blissfully ignorant.
IBM format? VAX floating point? (Not forgetting that the VAX had TWO 64-bit floating point types :-)
You know, I think that text_oarchive is inappropriately named. Clearly this is an ascii_oarchive, not to be confused with the ebcdic_oarchive. :)
I would expect that if one want's an ebcidc-?archive one would use the text archive with a stream that in turn uses an ebcidc code_cvt facet. Note this is how unicode (UTF-8) is implemented in the xml_?archives. Robert Ramey
On Fri, Nov 18, 2005 at 07:44:13AM -0800, Robert Ramey wrote:
troy d. straszheim wrote:
On Fri, Nov 18, 2005 at 12:50:16PM -0000, Martin Bonner wrote:
----Original Message---- From: troy d. straszheim [mailto:troy@resophonic.com] Sent: 18 November 2005 11:41
But floats and doubles are no problem. They have the same size and layout whereever you are, and since you're just flipping bytes and writing/loading binary, you don't have to worry about NaN or inf or any of that, the archive remains blissfully ignorant.
IBM format? VAX floating point? (Not forgetting that the VAX had TWO 64-bit floating point types :-)
You know, I think that text_oarchive is inappropriately named. Clearly this is an ascii_oarchive, not to be confused with the ebcdic_oarchive. :)
I would expect that if one want's an ebcidc-?archive one would use the text archive with a stream that in turn uses an ebcidc code_cvt facet.
Note this is how unicode (UTF-8) is implemented in the xml_?archives.
Robert Ramey
I should know better than to try to make jokes about this kind of thing.
On Nov 18, 2005, at 1:50 PM, Martin Bonner wrote:
----Original Message---- From: troy d. straszheim [mailto:troy@resophonic.com] Sent: 18 November 2005 11:41
But floats and doubles are no problem. They have the same size and layout whereever you are, and since you're just flipping bytes and writing/loading binary, you don't have to worry about NaN or inf or any of that, the archive remains blissfully ignorant.
IBM format? VAX floating point? (Not forgetting that the VAX had TWO 64-bit floating point types :-)
In addition, to give just one example the Cray SV1 has 8 byte floats. Matthias
participants (4)
-
Martin Bonner -
Matthias Troyer -
Robert Ramey -
troy d. straszheim