Ian Smith smithi at
Fri Aug 29 03:44:25 UTC 2008

On Thu, 28 Aug 2008 13:33:35 +0200 (CEST)
Wojciech Puchar <wojtek at> wrote:

 > CP/M was single-user and was used on floppies up to 360kB AFAIK, 

And MP/M was multi-user, using the same filesystem.  From memory, there 
was perhaps one byte that indicated which user owned a file :)

 > NTFS is a theft of OS/2 HPFS. they didn't even bothered to use other 
 > partition ID :), but they managed to f..k^H^H^H^Hextend it's 
 > functionality, so it's actually even slower than FAT, and too - does 
 > nothing to prevent fragmentation.

It wasn't (straight-up) theft; MS cut a deal with IBM to use HPFS and 
OS/2, more or less in exchange for letting IBM licence Windows 3.1 as 

When things went sour - google provides days of happy reading if you're 
interested - MS morphed it into NTFS for NT, cruelled the deal with IBM 
so OS/2 couldn't run NT/Win95 apps (signing OS/2's death warrant, though 
it took a long time to die) and stopped distributing OS/2 themselves.

 > This is normal, as Microsoft make a problems to be able to "fix" it 
 > (creating 3 times more others) in new releases, so idiots continue to buy 
 > new versions of windoze and new hardware, just to do as simple task as 
 > writing a few-paged document or view a webpage

Yeah, yeah :)  I'd be surprised if NTFS isn't as defrag-proof as HPFS, 
which as I recall had self-defragging garbage-collecting features built 
in; certainly I never felt the need to defrag any HPFS volumes, and I 
used it for quite a few years to run BBS and Fidonet stuff, not once 
losing any data .. HPFS was a very resiliant and reliable filesystem.

If you compare:
% find /usr/src -name "*hpfs*
% find /usr/src -name "*ntfs*

you'll go 'hmmm ..' and if you look through the sources you'll see whole 
large slabs of code that are shared between those two implementations, 
by the same author.

I've never tried writing to HPFS volumes, but I did recover many years 
of work and play from a number of HPFS disks and still hope to do some 
more someday, so I was glad to see the code is still there in 7.0 ..

cheers, Ian

More information about the freebsd-questions mailing list