*BSD News Article 59613


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!nntp.coast.net!news.dacom.co.kr!news.kreonet.re.kr!usenet.seri.re.kr!news.imnet.ad.jp!wnoc-tyo-news!news.iij.ad.jp!news.CET.CO.JP!usenet
From: Michael Hancock <michaelh@cet.co.jp>
Newsgroups: comp.unix.bsd.netbsd.misc,comp.unix.bsd.bsdi.misc,comp.unix.solaris,comp.unix.aix
Subject: Re: ISP hardware/software choices (performance comparison)
Date: 14 Jan 1996 11:14:41 GMT
Organization: CATENA Enterprise Technologies
Lines: 76
Distribution: inet
Message-ID: <4daoj2$d5t@diablo.cet.co.jp>
References: <4cmopu$d35@vixen.cso.uiuc.edu> <4crnbe$8a@olympus.nwnet.net> <4cs2kn$kfg@cynic.portal.ca> <4cu7t0$mg5@engnews2.Eng.Sun.COM> <4cv8j1$59k@park.uvsc.edu> <4cvjpk$rpf@durban.vector.co.za> <4d4s3u$u3@news.voicenet.com> <4d70n5$6lp@diablo.cet.co.jp> <4d9i2k$qo9@park.uvsc.edu>
NNTP-Posting-Host: a04m.cet.co.jp
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 1.2N (Windows; I; 16bit)
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.netbsd.misc:1959 comp.unix.bsd.bsdi.misc:2103 comp.unix.solaris:57231 comp.unix.aix:68708

Terry Lambert <terry@lambert.org> wrote:
>Michael Hancock <michaelh@cet.co.jp> wrote:
>]
>] I guess there is no end to the bickering amoung the various Unix camps.  
>] Aren't the Unix wars supposed to be over?
>
>They were until Sun bought out their license for SVR4 sources
>so that their code could diverge again.
>
>BTW: it would be legal for Sun to put the SVR4 code they bought
>out up for anonymous FTP.
>
>] The positive side of friendly competition can described by looking at 
>] something like the file system.
>] 
>] SYSV has s5fs.
>] 
>] BSD makes ffs.
>] 
>] SYSV adopts ffs calls it ufs and implements it within its
>] vnode/vfs framework.  This facilitates the development of
>] new file systems such as vxfs journaling file system, and nfs.
>
>Uh, the Usenix VFS papers were all based on SunOS.
>
>As was Heidemann's Ficus work at UCLA.
>
>As was Rosenthal's work.

Sorry, I skipped some history.

>You are probably thinking of the "file system switch", not the
>VFS (which implies vnodes) interface.
>
>VXFS is a UFS derivitive.  The directory structure is identical,
>as is the flexname handling.
>

Didn't vnode/vfs facilitate it implementation?

>You also forgot "USL files suit against UCB and forces them to
>not distribute some portions of UFS".

I also forgot to mention Digital had gnodes.  We can go on and on.

>] The BSD community reimplements vnode/vfs and nfs; and adds
>] stackable vnodes, union mounts, and the portal file system.
>
>Sun invented VFS.  You could argue that it was a reimplementation
>of fsswitch[], until you looked at the location that the in core
>copy of the on disk inode data was stored in both models.
>
>Portals are old.  The Heidemann architecture for vnode stacking
>was work done at UCLA, not UCB.  Union mounts are a special case
>of volume spanning (like the Sun TFS and loopback mountin from
>4.x).
>
>] With WinNT amoung us the bickering needs to end and the Unix
>] vendors need  to find new ways to be cooperate and be
>] competive by delivering better quality systems and better
>] service.
>
>It's not going to happen.  There is no way to add value to
>non-commodity hardware, other than divergence.  There's no
>way for products to compete on commodity hardware without
>divergence.
>
>You can't pull a value proposition out of a hat.

Unless it were necessary for survival.

--
Mike Hancock
michaelh@cet.co.jp