Mail Archives: djgpp/2003/06/27/18:32:02
"Eli Zaretskii" <eliz AT elta DOT co DOT il> wrote in message
news:2950-Fri27Jun2003205236+0300-eliz AT elta DOT co DOT il...
> > From: "deckerben" <bdeck AT lycos DOT co DOT uk>
> > Newsgroups: comp.os.msdos.djgpp
> > Date: Fri, 27 Jun 2003 11:59:14 +0200
> >
> > > > "Eli Zaretskii" <eliz AT elta DOT co DOT il> wrote in message
> > > > news:2593-Wed25Jun2003075056+0300-eliz AT elta DOT co DOT il...
> > > > > > From: "deckerben" <bdeck AT lycos DOT co DOT uk>
> > > > > > Newsgroups: comp.os.msdos.djgpp
> > > > > > Date: Tue, 24 Jun 2003 23:34:01 +0200
> Try a floppy written under DOS. If NT returns lower-case letters
> there, it simply lies to you, as all DOS directory entries are in
> upper case.
This isn't a relevant argument. Who is going to run PythonD from a floppy??
But just for the record:
A:\>touch HeLLo.txT
A:\>dir
Datenträger in Laufwerk A: ist W2PIFB1_EN
Datenträgernummer: 26CB-D36C
Verzeichnis von A:\
06/28/2003 12:20a 0 HeLLo.txT
Filecase looks pretty OK to me... I don't know what you are talking about.
> That's not the point. The point is that Windows has a
> case-insensitive filesystem, so making it case-sensitive (by setting
> FNCASE=y) is not recommended, since the side effects will generally
> surprise Windows users.
Not if it an NT user that's used to seeing filenames returned correctly. But
what DJGPP does *will*suprise them, believe me (I've seen it happen). I
think you are being *very* general here.
> That's because your DIR command goes to the shell which is a Windows
> program, as opposed to DJGPP programs which are DOS programs. You
> really should keep that in mind when you argue about these subtleties.
'ls.exe' is a DOS program, right? Note the output (without FNCASE set):
D:\>set FNCASE=n
D:\>ls -al
total 1617
drwxr-xr-x 17 bdeck root 608 Jun 17 01:13 .
drwxr-xr-x 17 bdeck root 608 Jun 17 01:13 ..
-rw-r--r-- 1 bdeck root 708836 Jun 16 01:10 Citizenship
docs1.zip
drwxr-xr-x 4 bdeck root 608 Mar 29 15:43 Download
drwxr-xr-x 29 bdeck root 1184 Mar 30 13:42 Project
dr-xr-xr-x 2 bdeck root 64 Sep 6 2001 Recycled
dr-xr-xr-x 7 bdeck root 160 Sep 7 2001 r
-rw-r--r-- 1 bdeck root 19968 Apr 14 21:00 sondervortrag.doc
drwxr-xr-x 2 bdeck root 32 Dec 19 2001 temp
This DOS program seems to be returning filecase *quite* accurately, accept
where all uppercase has been intentionaly coverted to lowercase as we
discussed before("temp" is actually "TEMP" and "r" is actually "R"). But
since my DJGPP programs anyway return filecase correctly under Windows2000,
I don't understand the need of even having the aforementioned conversion at
all.
I'm sorry, I guess I just don't see the point here. Maybe it's just beyond
me... no idea. But I think DJ's compromise might make everyone happy (I
hope).
"DJ Delorie" <dj AT delorie DOT com> wrote in message
news:200306271309 DOT h5RD9FWE002892 AT envy DOT delorie DOT com...
>
> > This is just wrong. NT returns filename case correctly, from both the
CMD
> > and COMMAND prompts. You are probably refering to Windoze32 DOS-based
> > thunking OSes. They are still DOS. NT is not DOS.
> You could put one patch in the python binaries to set that variable,
> instead of forcing your python users to cripple *all* their djgpp
> programs.
Maybe this is the best idea. Somehow, some here don't like the idea of
changing the NT environment (although I don't see why).
Naya, es ist mir alles Würst.
I will try to see what the best way to implement such a thing. AFAIK, it
only makes sense under NT systems anyway. I'll take another look at it all
next week.
Thank you all for your input!
Many regards,
Ben
- Raw text -