[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Update of charset windows-1252, draft 2



> I see no problem in having cp1252 as an alias, esp. since
> there is already a cp936 alias for a related (same originator
> for same set of systems) encoding.

It may be too late to register cp1252 as a formal alias for
windows-1252. If I am not mistaken, MSIE does not support the cp1252
name, and MSIE is used quite widely.

> Why aren't all of the legacy MS defined encodings dealt
> with in a single batch?

Because I don't want to edit so many files every time we come up with
a single edit. That's why I'm starting with a single windows-*
charset: to figure out the final pattern for one of them, and then
apply that pattern to all of the others in one fell swoop (with minor
changes, if necessary).

> I really would like to see them being treated very similarly:
> * similar preferred names/aliases
> * similar set of other aliases
> * similar mapping references
> etc.
> Anything else would be needlessly arbitrary and surprising.

Unfortunately, we are not at the very beginning of the deployment of
implementations of these charset names. When making changes to network
protocols, you must take existing deployments into account, try not to
"break" anyone, and follow migration plans if necessary.

By the way, the windows-1255 charset has changed recently. See the
mapping for 0xCA in:

http://www.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1255.TXT
http://www.unicode.org/Public/MAPPINGS/VENDORS/MICSFT/WindowsBestFit/bestfit1255.txt

So we may want to update the out-of-date one (CP1255.TXT).

Also, windows-936 has already been registered, as an alias for gbk. So
we have to discuss whether we will add Mike Ksar's name to the
existing registration.

http://www.iana.org/assignments/charset-reg/GBK

> But I see no reason why that "best fit" file, or mapping APIs,
> would need any mention at all in a registration such as this.

It is merely being provided as "additional information". However, I
personally don't mind removing it if there is consensus to do so.

Erik