Yes, this is a fair statement indeed.


> just to be fair, I have certainly seen Solaris accept out of spec behaviour in
> order to work with vendor XYZ's product.  until we see network trace data, I
> would hold off on assigning blame.  
> [as an example, HPUX TCP/NFS client sends data on the final ACK of the three
> way handshake; didn't bother Solaris, but left my company's box rather
> frazzled]
> 
> nathan
> 
> --- Matthew Jacob <mjacob@feral.com> wrote:
> > > 
> > > Well, I just dd-ed a > 4G file over NFS between a NetBSD 1.5 client and
> > > server, without problems. No special flags are needed, and never have
> > > been.
> > > 
> > > So it would seem that your problem is either in the application
> > > or the server.
> > 
> > Not quite that simple.  A plain dd (or lmdd with O_LARGEFILE on Solaris) has
> > no problem with this server. It's the *BSDs as clients to this box that have
> > a
> > problem.
> > 
> > Yes, it's probably a server 'problem', but from the point of view of me going
> > and and saying "Just use FreeBSD or NetBSD boxes in your test harnesses" and
> > having Solaris "just work" is a hard sell.
> > 
> > -matt
> > 
> > 
> 
> 
> __________________________________________________
> Do You Yahoo!?
> Yahoo! Auctions - Buy the things you want at great prices. 
> http://auctions.yahoo.com/
>