Re: XyShell Clipboard and conversion of ANSI characters

["Robert Holmgren" holmgren@xxxxxxxx]

** Reply to message from fenda@xxxxxxxx on Fri, 23 Aug 2002 14:30:59 +0200

> Robert,
> I am using 850 codepage. For example, if I select this piece of Notepad
> characters: á é í ó ú ñ Ñ Á É Í Ó Ú and I clip them to a windowed
> XyWrite screen, I get wrong characters (the equivalent ones in 437
> codepage)... It is the same
> if I am working in a full screen XyWrite session.

Manuel:

I performed this experiment: First, I used CLIP to copy the "á é í ó
ú ñ Ñ Á É
Í Ó Ú" characters above from my mailer into XyWrite. And indeed, I got
different-looking characters (mostly drawing characters), undoubtedly identical
to the ones you are seeing on your screen. Second, I CAlled your original
Email message into XyWrite -- unaltered from what you sent and what I received
(the POP file I got from your ISP). A comparison of the characters at the byte
level indicates that they are identical. In other words, "á" -- a with acute
accent -- is E1 hex in your POP file msg. The character displayed in XyWrite
(Greek beta) is E1 hex.

So: CLIP is copying the characters with *absolute accuracy* -- E1 hex is E1
hex, and that's a fact beyond dispute. I wouldn't want it to work any other
way, and neither would you. What if you were CLIPping program code, and CLIP
had a mind to change the character codes? Unimaginable. Correct me if I am
wrong, but some of the characters above, e.g. A acute, don't even exist in the
the XyWrite character set, except when using Speedos above character number
255! That's why the U2 frame ANSI2XY can't convert them. How do you display A
acute in XyWrite?

You and I have been through this before, I think. We both know that XyWrite
uses the Ascii character set, whereas ANSI is ANSI -- and the twain only agree
100% in the area of 32-127 (hex 20-7F). You have to convert, I'm sorry to say.

> BUT, if I now copy the wrong characters from XyWrite to
> Notepad, the resulting characters are OK! Very curious.

Well, Notepad must be ANSI-based, I guess. That's the only explanation that
makes sense.

Anyway: this is why I'm not eager to write a XY2ANSI program! These are deep,
systemic incompatibilities, only resolvable with Speedos.

-----------------------------
Robert Holmgren
holmgren@xxxxxxxx
-----------------------------