[OLE #50] IRemUnknown Proxy/Stub Implementation

Vincent Béron vberon at mecano.gme.usherb.ca
Tue Jan 25 19:51:32 CST 2005

Le mar 25/01/2005 à 20:03, Bill Medland a écrit :
> On January 25, 2005 03:48 pm, Robert Shearman wrote:
> > Vincent B�on wrote:
> > >Le mar 25/01/2005 �11:51, Robert Shearman a �rit :
> > >>+ *  Copyright 2001 Ove Kåven, TransGaming Technologies
> > >
> > >Not sure we want UTF-8 in the source files... At least we don't have it
> > >(yet).
> >
> > I think it should be made policy that anything in the comments should be
> > UTF-8.
> Can one specify that a region of file has a certain encoding?

Not in a text file. Not even sure you can in a (x)html file (for the
different regions in a same file).

> Surely what we are saying is that the source codes are utf-8

Are? Only dlls/kernel/lcformat.c contain easily spotted UTF-8 sequences
in a source file. It (in Julio César Gázquez's name) comes from the old
ole/ole2nls.c. Alexandre moved some functions to their current location,
and while copying over the authors names, UTF-8'd them (they were in
Latin-9 before, also known as ISO-8859-1 (or 15)). I don't know if it
was willingly or a side effect of his editor (the change is dated May
22, 2003).

Should we keep it as is (and convert the other parts to UTF-8 at the
same time), or should we change it back?

Various .rc files also contain UTF-8 sequences, almost all in Nl.rc
files for e umlaut and i umlaut. I'm unable to make them print correctly
when output by Wine (ie, I see the individual UTF-8 bytes, not the
intended result). Hans, did you provide your translations in UTF-8
format by any chance? I think this should be corrected for that language
(and others if others have the same issue), as .rc files are encoded in
a language-specific pagecode, which is most likely not UTF-8.

> > I don't think the same can be done for code yet as gcc and other
> > compilers probably don't support this?
> Um! What part of the code would be other than the pure ascii portion of utf-8?  
> The only bits I can think of are comments and literal strings.  And certainly 
> on my system I can put utf-8 in them.

"Ascii" strings in UTF-8? Or Wide-char strings (or chars) in UTF-8?

The first looks like non-sense, while I don't think the second would
work out ok unless gcc understands UTF-8 (ie, is 'Ã¥' accepted by gcc as
a single char representing the same thing as 'å' if LANG=*.UTF-8, or is
it understood as "warning: multi-character character constant"? My
current testing points me to the warning.).

For the comments, it's mainly an editor thing (some do support UTF-8,
some don't). Of course, mail programs can also change the encoding while
attaching or inlining patches...


More information about the wine-devel mailing list