On Tue, 4 Mar 2014 12:24:41 -0500, Tony Harminc wrote: >On 3 March 2014 20:54, Farley, Peter x23353 wrote: > >> GIYF. I refer you to these Wikipedia references, the first of which makes >> it quite clear that iso-8859-1 is definitely NOT >> Windows, though it does call it "ASCII-based"; and the second of which is a >> nice reference for IBM-1047, from which >> you can see that there is a reasonable chance to convert between the two >> encodings, though not completely >> transparently, whatever IBM chooses to call the iso-8859-1 encoding. > >More than a reasonable chance; since IBM-1047 and ISO-8859-1 encode >*exactly* the same set of displayable and control characters (what IBM >calls Character Set (CS) 697, and ISO calls (duh) Latin 1)), >converting without loss is trivial. > >Windows, when it uses them at all, uses code pages that contain >displayable characters in places where 1047 et al have control >characters, so it is not in general possible to convert from e.g. >Western Windows 1252 to IBM-1047 without loss. > Actually, it depends on your definition of "loss". If both pages contain 256 code points, it may be posible to convert from one to the other, then back, and have the same thing you started with. The interim may not look the same, but that may not matter.
I think I'll go with "ASCII-based" until Shmuel objects. And long thereafter. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
