28 Aug
2006
28 Aug
'06
4:08 p.m.
Teo Zenios wrote:
----- Original Message -----
From: "Cameron Kaiser" <spectre at floodgap.com>
To: <cctalk at classiccmp.org>
Sent: Monday, August 28, 2006 6:30 PM
Subject: Re: "File types"
Yes, it's totally
transparent. This was very nice for taking 3.5" disks
between my IIsi and the DOS PCs at work back when I sneakernetted everything.
On a FAT volume the fork gets stripped away, so taking the file and putting
it back on the Mac without the resource fork the Mac OS will not know what
to do with it (ruins the file).
Well, there seems to be some disagreement here (?) :>
{I'll let you guys finesse the details...)
But, that's not material to my question. I really don't
care how Mac's handle different file systems, etc. My
interest in them was only as an example of a system that
did NOT fold "type information" into the namespace for
files.
Yet, it *seems* that even they are following the Lemmings (?)
I've still not heard any comment on *why* (hysterically)
this scheme was (apparently) replaced by one in which
names convey file type information (e.g. foo.sea.hqx).
Is it simply a "lowest common denominator" -- i.e. every
file system (on "every" OS) supports the concept of a file
NAME so that's where it *has* to go? Yet, why does it
*have* to go there at all? (i.e. why does file type
*need* to be part of the name -- is it just something
that users have grown accustomed to?)
>> However, when storing Mac files on, say,
a FAT volume, a hidden folder
is
made to hold the additional data in the resource fork.
This includes
type/creator information.
Ah, OK. So the OS already has provisions to handle it
and
hides those details from the user. I.e. it does the things
I was insinuating the user would have to do, manually.