File extensions, was: Re: [Coco] Portal-9 bug report

Theodore Evans (Alex) alxevans at concentric.net
Mon Oct 18 22:23:02 EDT 2004


On Oct 18, 2004, at 10:30 AM, Dennis Bathory-Kitsz wrote:

> And think about when it was implemented -- right after Windows 3.1, 
> which
> had 8.3 filenames. The implications of the long filename couldn't be 
> known
> yet, nor how people would use (or abuse) them.

There were many systems which supported relatively long filenames that 
weren't 8.3 significantly before Win 95 came out.

> Still, the suppressed extension made them significantly more 
> approachable
> for lay users. Indeed, programs, files and directories became programs,
> documents and folders. The presence of the icon distinguished what 
> sort of
> item it was, not the presence of a file extension. This was a good 
> idea in
> a kind & gentle world.

If MS didn't insist in the stupidity of identifying file types by 
extension (particularly considering that most modern, non-MS file 
format contain easy to recognize magic numbers you could name the file 
whatever you wanted and still identify the file type.  Incarnations of 
AmigaDOS from 3.0 onward had no problem with this.  Unix systems using 
both CLI and GUI interfaces have never depended much on file extensions 
to tell the user what type a file was either.

>
> I'm hardly in the habit of defending Microsoft, but it was a reasonable
> thing to have done (without the advantage of a crystal ball) to a lay
> user's advantage.

There were enough systems with better methods that predated Win 95 that 
it didn't take a crystal ball see that MS was putting forth a really 
poor solution.

-- 
Theodore (Alex) Evans | 2B v ~2B = ?




More information about the Coco mailing list