Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!ucbvax!decwrl!decvax!eagle_snax!geoff
From: geoff@eagle_snax.UUCP ( R.H. coast near the top)
Newsgroups: comp.sys.ibm.pc
Subject: Re: Funny Characters in filenames?!?!?!?!
Message-ID: <473@eagle_snax.UUCP>
Date: 16 Mar 89 22:33:58 GMT
References: <56@a.coe.wvu.wvnet.edu> <1748@csuna.csun.edu>
Reply-To: garnold@sun.com (Geoff Arnold @ Sun ECD - R.H. coast near the top)
Distribution: usa
Organization: Sun Microsystems - East Coast Division
Lines: 42

In article <1748@csuna.csun.edu> abcscnge@csuna.csun.edu (Scott Neugroschl) writes:
>In article <56@a.coe.wvu.wvnet.edu> manager@a.coe.wvu.wvnet.edu (Cris Fuhrman, Systems Manager) writes:
>#On one of the hard-disk machines here at school, we have some files and 
>#directories with funny-characters (smiley faces, hearts, diamonds, etc.) in 
>#the file names.  How do I delete these files?  How did they get there?  
...
>Change the names to clean up the funny characters and then delete them.
>Alternatively, you might be able to use alt-key sequences to enter
>the funny chars for deletion,but I wouldn't swear to it.
>-- 

Something interesting happened between DOS 3.x and DOS 4.0. In the
DOS 3.x documentation, IBM/Microsoft said "these are the characters
which are legal in file names"... and so by implication all others
are illegal. Since all of the legal characters were in the range
00H - 7FH, any programmer (including myself :-) could be forgiven for
treating any character with the top bit set as illegal.

Time went by, and then I started getting calls from Europe complaining
that our software wouldn't handle foreign-language symbols in file
names. Before I was able to shout "RTFM", my eye fell upon the
relevant page in the DOS 4.0 doc. In this, IBM enumerated those
characters which were ILLEGAL in file names - so by implication, all
others were legal. And this made all those characters from 80H to FFH
as legitimate as could be..... Obviously this change took place
during the latter days of DOS 3.x, but it only showed up as
a documented fact in DOS 4.0.

(And naturally we've fixed this now - no flames, please!)

DOS actually performs a fair bit of processing on these characters.
If you include the character "lower case e acute" (ALT-130)
in a file name, a DIR will show that DOS has mapped this into "E". On
the other hand "pi" (ALT-227) is untranslated. This game is
interesting for about 3 minutes..... pity the poor SQA soul who has
to wade through all of them, aided only by an ALT key :^)

-- 
Geoff Arnold,                              Internet: garnold@sun.com
Manager, PC-NFS Engineering                UUCP: ....!sun!garnold
PCDS Group, Sun Microsystems Inc.
*** I SPEAK ONLY FOR MYSELF *** (MY CHILDREN INSISTED THAT I SAY THAT) ***