Fwd: Re: zlib-1.1.4 out - security fix

Bill Maltby LFS Related lfsbill at wlmlx1.wlmcs.com
Wed Mar 13 05:28:57 PST 2002


Don,

If memory serves, that old Univac was a "hoss" in its day. Never
worked on it, but I seem to remember American Express in Phx AZ
had one, not an IBM. And that AMEX loc was a _big_ high-volume
transaction processing site at the time.

And the room was impressive too. Of course, that was in the day
when all the big corps put "showcase" rooms in place with glass
windows so the visitors could be escourted by and suitably
impressed by the technology.

On Tue, 12 Mar 2002, Don Smith wrote:

> "Bill Maltby LFS Related" <lfsbill at wlmlx1.wlmcs.com> wrote in message
> news:Pine.LNX.4.10.10203121808240.3265-100000 at wlmlx1.wlmcs.com...
> > Ian,
> >
> > Just like the old Mainframes. 16 GPR, 15 is not IP, but is branch to
> > <snip>

> 
> About the time you  were using BALR, I was using LMJ's and LBJ's on
> UNIVACs. 128 registers of which 44 were directly modifiable in user
> mode. Just leave your return address in a register, no need for slow
> memory accesses.

You know, the above caused me to reflect on they way technology is being
influenced. It used to be that because of the cost, R&D expenditures
on this stuff was high and considered worthwhile, in light of potential
returns. Then Intel style stuff arrived. Now, it enabled the low-cost
we see now and that brought enough volume that we have (ultimately)
Linux and other *IXs on these boxes and _millions_ of us at home with
miltiple configs. A good thing. But now the dark side appears.

Zilog and many other competing architectures died. Mainframe development
slackened, but for a few "super-computer" orgs. Now there is no innovation
to speak of - all is biased toward x86 compatible in the micro area and
the biggies stick with old architecture - just scaling up. Yes, materials
research is still aggressive (.13u die and organic bases, etc), but
basically no real innovation in the architecture. Focus on more cache,
longer pipelines, extended pre-fetch/pre-code. Maybe predictive branch
is somewhat innovative. But the financials still dictate x86 compat.

It is sad.

> 
> Of course these things filled a very large and specially built room to
> get just 1 MIPS!

Remember the size and capacity of the disk drives. I seem to recall a
IBM 23?? that was about 20" inches across and a a whopping 7.2MB capacity.
I think seek times were in the 120ms range. Called DASD (Direct Access
Storage Device), it had special locking covers for insertion and removal
into the drives. And expensive. An operator dropped one once and he
blanched _totally_ _pale_ on the spot. Of couse that unit had 16 hours
annual inventory run results on it at the time.

><snip>
 
> 
> Or put Linux on them :-)

Thanks for showing me the light again :)

Guilt attack! I guess we ought to move this thread to the
Beer Joint down the street! We're probably boring the
youngsters and aggravating the LFSers.

My apologies to all.

Bill Maltby
billm at wlmcs.com

-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-security' in the subject header of the message



More information about the lfs-security mailing list