We lost a co-inventor of ENIAC, John Mauchly, on this day (that's 8th
January like my headers say) in 1980. ENIAC is claimed to be the "first
general purpose electronic digital computer", but I guess it comes down to
a matter of definition[*], as every culture likes to be the "first" (just
ask the Penguins, for example); for "computer" you could go all the way
back to the Mk-I Antikythera (Hellenic variation, from about the 100BC
production run)...
[*]
Analogue/digital/hybrid
Mechanical/electrical/electronic/hybrid
General/special
Wired/programmable/Turing
Prototype/production
Harvard/Neumann/quantum
Etc...
--
Dave Horsfall DTM (VK2KFU) "Those who don't understand security will suffer."
> From: Andy Kosela
> it appears this is a fundamental Intel bug that exists in all x86_64
> CPUs.
I'm highly amused by the irony. Intel throws bazillions of transistors at
these hyper-complex CPUs in an attempt to make them as fast as possible - and
(probably because of the complexity) missed a bug, the fix for which
involves... slowing things way down!
I wonder how many other bugs are lurking in these hyper-complex designs?
Didn't anyone at Intel stop to think that complexity is bad, in and of itself?
But I guess the market demands for faster and faster machines outweighed that
- until it bit them in the posterior. The real question is 'how many more times
will it have to happen before they get a clue'?
There's an interesting parallel between this, and uSloth's struggle with
security and bugs. For a long time, it seemed it was more important to the
market to add features (i.e. complexity), and security be damned - until poor
security really started to become an issue.
So now they're trying to catch up - but seemingly still haven't got there, in
terms of the fundamental architecture of the OS, as the never-ending stream of
bug patches attests.
The sad thing is that how to provide good security (not perfect, but much,
much better than what we have) was worked out a long time ago, and Intel hired
Roger Schell to add the necessary hardware underpinnings when they did the
386:
http://conservancy.umn.edu/bitstream/11299/133439/1/oh405rrs.pdf
Mutatis mutandis.
Noel
Clem Cole:
IIRC this is part of the argument Dykstra made with the THE paper years
ago, Parnas in his information hiding paper -- i.e. why microkernels and
proper layering are a good idea. Keep is simple and do one thing
well/protect yourself against other subsystems not being 100%.
=====
Indeed. Complexity creates needless RISC, er, risk.
Norman Wilson
Toronto ON
http://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/
Including here as it concerns Unix kernel and leaking memory from kernel
space to userland.
This is big -- it appears this is a fundamental Intel bug that exists in
all x86_64 CPUs.
It will be interesting to watch the ramifications and impact of this on the
industry as a whole.
--Andy
> A lingering gripe that explains my latent anti-Americanism goes back to
> when I had to support Uniplus 2.2/2.4 (sorta SysIII-ish) on the WICAT boxes
> in here in Australia. At installation time, we had to express the time
> offset as hours *west* of GMT; this left me with a lingering belief that
> Americans didn't want to be perceived as being backwards (yeah. it saved an
> entire keystroke out of the dozens that were otherwise required).
But east postive is an artifact of north up. I remember Australian
souvenir shops selling maps on "MacQuarrie's corrective projection",
in which south is up. In fact this orientation was not uncommon in
Europe between medieval maps that really were oriented, with east up,
and later convention that put north up and shoved Australia down under.
Surely an Aussie would prefer south up and west positive!
Doug
On Sun, Dec 31, 2017 at 7:15 PM, Dave Horsfall <dave(a)horsfall.org> wrote:
>
> A lingering gripe that explains my latent anti-Americanism goes back to
> when I had to support Uniplus 2.2/2.4 (sorta SysIII-ish) on the WICAT boxes
> in here in Australia. At installation time, we had to express the time
> offset as hours *west* of GMT; this left me with a lingering belief that
> Americans didn't want to be perceived as being backwards (yeah. it saved an
> entire keystroke out of the dozens that were otherwise required).
Dave I'm not so sure it's about being perceived as forward or backwards -
its just shallow, provincial and often lazy because the program did not
really knowing any better. The problem is too many American',
(particularly younger ones that experience our 'excellent' educational
system), have often never travelled that much and experiences other places,
cultures or social norms.
I admit this is extreme example, but about 8 years ago, my daughter had a
friend, who was approx 16 at the time, that we took to the big city
(Boston) to play in a orchestra concert at Symphony Hall when they both
were named 'All State' for the instruments. I don't remember why said
friends family did not/could not come - but it made sense and we said we
would take her with us. On the drive in-town, we were talking with her and
I discovered that she had never gone to Boston before ... ever -- she was
excited to see it (we live less than 1hr North mind you. Note quite the
boon-docks). She had not gone to a 'Bo Sox' game or anything. Never went
to the Science Museum, etc. She grew up in her town (mind you happy) and
using TV as her window to world.
Which brings me to >>my complaint<<. We, as American's, project so much
about 'us' via TV. The said truth is most Americans are not like what they
see on TV [e.g. Rice-A-Roni is made up!!, Benihana's is an American
invention, and "the big yellow school bus" is dirty/noisy and usually
without seat belts]. Sadly, many Americans do not know any better - queue
the famous quote about never under-estimating the taste of the American
public. But think about what folks outside the US see and think? Many of
my European friends in particular all want to visit NYC. [I tell them all,
visit Boston or Philadelphia first if you can. Those cities are much more
representative of America then LA, NYC or Dallas; if for no other reason
they are more 'European' in feel].
When I run into things like what you just described (and I seem to run into
then most often with MicroSoft based tools), I think to myself, it must
have been a cold day in Redmond, WA and some programmer did not want to
make an effort to do make her/his solution really general ;-)
Clem
FWIW: Not only did we take my kids all over the world as children, we
brought the world to them by sponsoring kids from all sorts of countries.
But I fear, my wife and I are less the norm then I would wish.
ᐧ
ᐧ
> From: Dave Horsfall
> What timezone did you say you were in again?
Ah; didn't realize you wanted the exact time, not just a date! :-)
Well, according to this:
http://www.rfc-editor.org/rfc/museum/ddn-news/ddn-news.n19.1
it was _planned_ to happen at "00:01 (EST) 1 Jan 1983". Whether it did happen
at that moment, or if in reality there was some variance, I don't know (IIRC,
I was off sailing in the Caribbean when it actually happened :-).
Noel
It was sort of a soft change over. All the did on January 1, 1983, is
disable link 0 on the IMPs. This essentially blocked NCP from working
anymore.
You could (and many of us were) run IP on the Arpanet for years before that.
In fact the weeks before we were getting ready for the change and we had our
TCP/IP version of the system up.
It crashed.
We rebooted it and set about debugging it.
It crashed again. Immediately my phone rang. It was Louis Mamakos, then
of the University of Maryland. He had tried to ping our system. I
called out the information to Mike Muuss who was across the desk from me.
We set to find the bug. It turned out that the ICMP echo handling in the
4.1c (I believe this was the version) that we'd cribbed our PDP-11/70 TCP
from had a bug where it used the incoming message to make the outgoing one,
AND it freed it, eventually causing a double free and crash. It was at
this point we realized that BSD didn't have a ping client program. Mike
set out to write one. It oddly became the one piece of software he was
most known for over the years.
The previous changeover was to long leaders on the Arpanet (Jan 1, 1981?).
This changed the IMP addressing space from eight bits (6 bits for imp
number, 2 for a host on imp) to 16 bits for imp number and 8 for a host on
imp. While long leaders were available on a port basis for years earlier,
they didn't force the change until this date. The one casualty we had a
PDP-11/40 playing terminal server running software out of the University of
Illinois called "ANTS." Amusingly the normal DEC purple and red logo
panels at the top of the rack were silkscreened with orange ANTS logos (with
little ants crawling on it). The ANTS wasn't maintained anymore and stuck
in short-leader mode. We used that option to replace that machine with a
UNIX (we grabbed one of our ubiquitous PDP-11/34's that were kicking
around). I kept the racks and the ANTS logo for the BRL Gateways. I
turned in the non-MM PDP-11/40. A year later I get a call from the
military supply people.
ME: Yes?
GUY: I need you to identify $65,000 of computer equipment you turned in.
ME: What $65,000 machine?
GUY: One PDP-11/40 and accessories.
ME: That computer is 12 years old... and I kept all the racks and
peripherals. Do you know how much a 12-year-old computer is worth?
The other major cut over was in October of 1984. This was the
Arpanet-Milnet split. I had begged the powers that be NOT to do the
change over on the Jan 1st as it always meant I had to be working the days
leading up to it. (Oct 1 was the beginning of the "fiscal" year). Our
site had problems. I made a quick call to Mike Brescia of BBN. This was
pre-EGP, and things were primarily static routed in those days. He'd
forgotten that we had routers at BRL now on the MILNET (all the others were
on the ARPANET) and the ARPANET-MILNET "mail bridge" routers had been
configured for gateways on the MILNET side.
> From: Dave Horsfall <dave(a)horsfall.org>
> the ARPAnet got converted from NCP to TCP/IP in 1983; ... have a more
> precise date?
No, it was Jan 1.
It wasn't so much a 'conversion', as that was the date on which, except for a
few sites which got special _temporary_ dispensation to finish their
preparations, support for NCP was turned off for most ARPANET hosts. Prior to
that date, most hosts on the ARPANET had been running both, and after that,
only TCP worked. (Non-ARPANET hosts on the then-nascent Internet had always
only been using TCP before that date, of course.)
Noel