Subject: Re: msword doc bug
From: Thomas B. Passin (tpassin@idsonline.com)
Date: Wed Jan 19 2000 - 22:44:58 CST
Arnold M.J. Hennig wrote:
> On Wed, 19 Jan 2000, Logan Hall wrote:
>
> > Leonard Rosenthol wrote:
>
> > > At 1:51 PM -0700 1/19/00, Logan Hall wrote:
> > > >sterwill@abisource.com wrote:
> > > >
> > > >> Perhaps very soon we'll have a Mac port,
> > > >> and I'm not sure what we'll do there. I believe it's common
> > > >> to preserve
> > > >> the extension in the file names when dealing with
Microsoft-tainted
> > > >> data on the Macintosh (is this true?).
> ...............................
> > > The main place where you see extensions on the Mac, is where
> > > there is a good possibility that the user will be taking to the file
> > > to a platform (Wintel or Unix) that requires file extensions for
> > > "typing" the files. This is certainly most evident with files
> > > destined for the web (.html, .jpg, .gif, etc.), though even standard
> > > office files (.doc, .xls) or graphic files being exchanged (.psd,
> > > .ai, etc.).
>
> > Ah yes.. i forgot about html and jpg and gif. also for doc files,
> > does word add the extention if you save it as a windows word file?
> ......................................
> > The point is that for the Mac port abi should just use the
> > type/creater codes that are already inplace to import and export.
> > That would make it easer to avoid this same problem of opening a
> > file that isn't what its extention says it is. Also could the
> > Linux version use the magic number to associate files? .......
>
> Some thoughts on this thread:
>
> 1. Please, let's not try to abandon extensions, they're very useful
> even when they're not required. Agreed, however, that using the
> internal typing codes, where they exist, ought to be used as a test
> for file format, even when the extension is present and appears to
> be interpreted correctly. Some Linux programs do this, others do not.
> However, incorrect format should not be cause to kill the program, as
> happens with some programs.
>
> 2. Windows 95 and 98 appear to the user not to use extensions,
> because they don't display them in the file listings, but in
> reality they are more dependent on extensions than previous versions
> of Windows and DOS. (A rather ridiculous paradoxical state of
> affairs, IMHO!)
It doesn't have to be so bad, but Windows uses default settings that I
change immediately. My computer displays extensions. Also, Word can be
told to load a file with any extension. When you do this, it tries to
figure out the file type by analyzing the file. It's not bad at doing this,
and it can be set to inform you what conversion it thinks it should use
(this behavior is version-dependent, though). Anyway, it seems clear that
on the Mac, you should use the info in the resource fork when it exists, but
use the extension and analyze the file to verify when it does not. You have
to keep the ability to read and write extensions if you want to make
interoperability with PCs easier.
>
> 3. The reality of the Mac side is not any simpler. Instead of
> extensions, they use a two-part file system, and again hide it from
> the user for the most part. Having never used a Mac myself, I have on
> several occasions had to explain to Mac users how to open a simple
> text file retrieved from our BBS interface, for which there was no
> resource fork, so that they could import our translations into their
> apps. (Same way we always did it with Dos and Unix programs - works
> every time, even on Macs ;-) )
>
> I do have it from a friend of mine who is a Mac based graphic artist,
> that the correct use of extensions simplifies the job of correctly
> importing files from other systems, and that current Windows users
> tend to be worse at this than Unix users.
>
Tom Passin
This archive was generated by hypermail 2b25 : Wed Jan 19 2000 - 22:39:13 CST