Re: A simple opinion of SCO vs. IBM

From: Bill Vermillion (bv_at_wjv.comREMOVE)
Date: 11/15/03 

Date: Sat, 15 Nov 2003 15:05:01 GMT

In article <20031114175345.GA70255@alexis.mi.celestial.com>,
Bill Campbell <bill@celestial.com> wrote:
>On Fri, Nov 14, 2003, Wesley Parish wrote:
>>Jim Sullivan wrote:
>>>
>>> "John Collins" <jmc@nospam.xisl.com> wrote in message
>>> news:3fb25bb8_1@nnrp1.news.uk.psi.net...
>>>> Joe Dunning wrote:
>>>>
>>>> > Even his comment about IBM denying that AT&T developed Unix is wrong.
>>>> > The current definition of Unix is whatever the Open Group say it is. If
>>>> > the Open Group say that both Win NT and IBM's System/390 are "Unix"
>>>> > then clearly AT&T did not develop Unix: they developed one flavor of
>>>> > Unix.
>>>
>>> Well, until Novell gave away the trademark, there was only one UNIX and it
>>> came from the AT&T source code. To say that AT&T didn't develop UNIX is
>>> to insult the developers at AT&T who spent 25 years developing UNIX.
>>
>>Except there was the independent academic branch otherwise known as BSD ...
>>>
>....
>>Not forgetting of course, that IBM's Linux JFS - one of the matters of The
>>SCO Group's legal filings - is a port of the OS/2 JFS, which was an
>>independent rewriting of IBM's AIX JFS, which was a journaled file system
>>based on 4.2BSD's FFS; which coincidentally is also in Unix SVR4 as UFS.
>>SVR4's own S5 file system just isn't up to scratch.
>
>Then there's the little matter of the lawsuit AT&T brought against Berkeley
>in the early '90s which AT&T lost (possibly because it was found that AT&T
>had taken BSD code, stripped the BSD licensing text, and included it in
>SYS3/V themselves).

>Some of the most important usability and performance enhancements to *ix
>came out of BSD, not out of Bell Labs->AT&T (e.g. vi editor, file systems,
>TCP/IP, etc.).

BSD made the TCP/IP popular but as I recall it was developed at
BBN by Vint Cerf and another person - but Berkely was the first to
implement it widely.

I ran across the original vi sources recently and compiled it just
for fun. Small screen lines as many used it on Dec10
terminal/printers when it came out, and really limited compared to
what we have today, but anytime I'm on a system without it I feel
lost.

And the FFS spawned many similar and faster file systems. From
talking/reading I think the ext2 in Linux is essential the FFS
with the extents [fragments] removed.

But the spirit if BSD lives on and some of the original CSRG group
are doing the work on UFS2. That is interesting and it answers a
lot of questions/problems with today's larger file systems.

Inodes have gone from 128 bytes to 256 bytes. In addition to the
standard MAC [modify/access/change] times there is now a creation
time. [older systems often mis-documented ctime as creation time
not inode change time].

It can also 'snapshot' the FS so that you can run a snapshot [less
than a second typically] and then backup that snapshot on a running
file system.

The snapshot is also used on boot so that an fsck can be run in the
background while the system is in normal use. Considering how long
it could take to run fsck on the large drives and arrays we have
today you can that is important. Older limitations are gone
too. The only boot restriction is the boot files must reside
in the first 1.5TB of file space. Things keep changing. For so
long the SW limits were far ahead of the HW but in the past couple
of years that has changed drastically.

Bill 

-- 
Bill Vermillion - bv @ wjv . com

Relevant Pages

  • Re: Alternate installers for FreeBSD
    ... Bill> One big file system can reach up and bite/byte you when you ... Bill> are not looking. ... the most common problem with storage volumes is not the file system, ... Bill> a crashed system, you can remake the / filesystem, install ...
    (comp.unix.bsd.freebsd.misc)
  • Re: Court Ruling says M$ first OS was a ripoff!
    ... That would be a surprise for Bill to learn. ... He wasn't "the writer" of BSD. ... Joy designed and wrote the Berkeley ... Bill Joy used his own work, ...
    (comp.sys.mac.advocacy)
  • Re: Court Ruling says M$ first OS was a ripoff!
    ... That would be a surprise for Bill to learn. ... He wasn't "the writer" of BSD. ... And Bill Joy and Sun also used the work of others at UC Berkeley and the ... University of California, as the copyright notices on it attest. ...
    (comp.sys.mac.advocacy)
  • Re: Inode limit
    ... Thanks, Bill, but I mean in a number of sub directories inside the same directory ... or the inode table of the current file system (depends on the FS size ... La divulgation de ces informations est à ...
    (AIX-L)
  • Re: Slow Filesystem I/O
    ... > Bill Todd wrote: ... >>it onto the disk as compensation, ... Spiralog did not change *default* file system behavior in any ... when RMS specified that its internal buffer should be ...
    (comp.os.vms)