Re: TCHAR string?

Here's an example where Unicode is a sticking point that can't be worked
around on Win9x (even with MSLU):

Text in the Devanagari range of Unicode will be clobbered when calling
certain system functions that take a string (e.g. registry calls).

The reason this will never work on Win9x (even with MSLU) is that MSLU
depends on the presence of a code page to convert from "wide" to "narrow"
when going to such system calls, and there is no code page for Devanagari
Unicode (nor most of the Indic ranges of Unicode). So the internal Unicode
strings just get converted to "?"s and those are put into the registry.

So while I think MSLU is an good (or at least "interesting") idea, depending
on the range of Unicode your application is used for, it may or may not be
worthwhile.

Bob

"Pete Delgado" <Peter.Delgado@xxxxxxxxx> wrote in message
news:ObBhgwfFGHA.2648@xxxxxxxxxxxxxxxxxxxxxxx
>
> "Jonathan Wood" <jwood@xxxxxxxxxxxxxxxx> wrote in message
> news:OKI8vyVFGHA.1424@xxxxxxxxxxxxxxxxxxxxxxx
>> Pete,
>>
>>> I understand your point, however, but I think that if for some strange
>>> reason you looked at back porting an application to Windows 9.X, Unicode
>>> support would likely be the *least* of your problems!
>>
>> Not that it matters but my most recent program update no longer runs on
>> Win95/98/ME and Unicode is pretty much the only reason.
>
> Interesting. I have ported several large programs and while there were
> indeed technical challenges, it wasn't impossible, There have been others
> programs that I couldn't port because they relied on system services that
> weren't there in the earlier operating systems, but the point is that the
> sticking point was never Unicode support.
>
> I'd be curious to see exactly where your problems lie and whether the
> problems were insurmountable or whether the application could be ported
> with some modification. In my experience, if Unicode is your only issue,
> the likelihood is that the program can be ported. However, depending upon
> your design decisions, there are things that you could have done that
> would make a port very messy.
>
> -Pete
>



.

Relevant Pages

  • Re: GetGlyphOutline API
    ... libraries need recompiling in order to allow one to use MSLU. ... But it's dead easy in Unicode.] ... I'm doing such conversions all over the place, mostly with CStringW and ... which I find to be a very reasonable product for a reasonable price. ...
    (microsoft.public.vc.mfc)
  • Re: TCHAR string?
    ... > Here's an example where Unicode is a sticking point that can't be worked ... > around on Win9x (even with MSLU): ... The issues with Devanagari and Georgian in MSLU ... the Windows 9X operating systems had problems that went well ...
    (microsoft.public.vc.mfc)
  • Re: Any component for Unicode for window 98 and Win2k
    ... > G. Allen Casteran wrote: ... > are talking about MS Layer for Unicode, ... > regarding MSLU - MSDN library contains a complete list of what MSLU does ... Since we don't code for Win98 any longer it was not an issue I ...
    (borland.public.delphi.thirdpartytools.general)
  • Re: Setting character encoding
    ... MSDOS will never understand UNICODE. ... There were DBCS versions of MSDOS ... MSLU is for Win9x, ...
    (microsoft.public.vc.mfc)
  • Re: How to use 8bit character sets?
    ... I need to process German textfiles and generally work with strings which need to be processed using the local encoding. ... The only solution I managed to find was converting between unicode and latin-1 all the time. ... It seems that Python on Win9x doesn't really support unicode. ...
    (comp.lang.python)