Prev: No multicore grahics from Intel, Intel withdraws Larrabee!
Next: Several Ntoskrnl BSOD crashes over the last week, but different stopcodes
From: Robert Myers on 26 Jan 2010 17:47
On Jan 26, 7:32 am, Sebastian Kaliszewski
<s.bez_sp...(a)remove.this.informa.and.that.pl> wrote:

> And that's all because losses are bounded, and not having software at
> all would mean that lost gains would far surpass losses caused by that
> "bad" software we have today.
>
It isn't clear to me that having software of the kind you write and
sell is better than having no software at all.

The bound on computer risks is the collapse of civilization as we know
it.

Ask Google about which software is "mission critical." If their
answer didn't include IE6, I couldn't imagine why.

Robert.
From: Robert Redelmeier on 27 Jan 2010 08:48
In comp.sys.ibm.pc.hardware.chips Robert Myers <rbmyersusa(a)gmail.com> wrote in part:
> On Jan 26, 7:32�am, Sebastian Kaliszewski > wrote:
>> And that's all because losses are bounded, and not having
>> software at all would mean that lost gains would far
>> surpass losses caused by that "bad" software we have today.

> It isn't clear to me that having software of the kind you
> write and sell is better than having no software at all.

I presume you agree with the rest of Sebastian's long post
or at least found this small extract the most disagreeable.

Well, only you can speak for what you find clear or what you
consider "better". But many, many people have voted otherwise with
their dollars, euros, yen, yuan, etc. Repeatedly. The video game
industry is larger than movies. Commercial software enterprises
(incl MS, ORACLE, SAP, etc) are similarly successful.

Myself included, advocates of Free Software (and Open Source
if you prefer those nuances) should admit that Microsoft has
helped _tremendously_. And not just by setting expectations so
low :) Their eye candy sold _many_ machines and stole much RAM
and CPU cycles. It drove huge economies of scale and fueled
the performace race. Without Microsoft, WordStar, Lotus, and
WordPerfect, Linus probably never would have afforded his first 386.
The Internet as we know it would not exist


> The bound on computer risks is the collapse of civilization
> as we know it.

Oh, dear! Not the 95% annihiliation of life on earth?
This is possible, but highly unlikely.

As for the financial crisis of Sept 2008, computers had
almost nothing to do with it (unlike Aug 1987). The fault
lies in a series of incentives and decisions that are mostly
understandable and justifiable on microscale yet add up
to instability. This was predicted by many, perhaps even you.

> Ask Google about which software is "mission critical."
> If their answer didn't include IE6, I couldn't imagine why.

Yet you do not seem short of imagination!

AFAICS, Google's "mission" is to _quickly_ serve up
ranked WWW search results with attached ads. The indexing
spiders, dispatchers and database machines are absolutely
mission critical. All running Linux and in-house code AFAIK.
MS-IE6 might be in some offline testing suite.


-- Robert R

From: Robert Myers on 27 Jan 2010 12:15
On Jan 27, 8:48 am, Robert Redelmeier <red...(a)ev1.net.invalid> wrote:
> In comp.sys.ibm.pc.hardware.chips Robert Myers <rbmyers...(a)gmail.com> wrote in part:
>
> > On Jan 26, 7:32 am, Sebastian Kaliszewski >  wrote:
> >> And that's all because losses are bounded, and not having
> >> software at all would mean that lost gains would far
> >> surpass losses caused by that "bad" software we have today.
> > It isn't clear to me that having software of the kind you
> > write and sell is better than having no software at all.
>
> I presume you agree with the rest of Sebastian's long post
> or at least found this small extract the most disagreeable.
>
That's not a correct presumption. I only felt it necessary to rebut
his final claim in order to rebut his entire argument as it pertains
to computers.

It would be pointless for us to continue in mutual accusations of
misunderstanding if there is no common basis for understanding, as
clearly there is not.

> Well, only you can speak for what you find clear or what you
> consider "better".  But many, many people have voted otherwise with
> their dollars, euros, yen, yuan, etc.  Repeatedly.  The video game
> industry is larger than movies.  Commercial software enterprises
> (incl MS, ORACLE, SAP, etc) are similarly successful.
>
People do lots of stupid and even self-destructive things with money.

> Myself included, advocates of Free Software (and Open Source
> if you prefer those nuances) should admit that Microsoft has
> helped _tremendously_.  And not just by setting expectations so
> low :)   Their eye candy sold _many_ machines and stole much RAM
> and CPU cycles.  It drove huge economies of scale and fueled
> the performace race.  Without Microsoft, WordStar, Lotus, and
> WordPerfect, Linus probably never would have afforded his first 386.
> The Internet as we know it would not exist
>
Packet-switching is a very clever and useful idea. Whether the
Internet as it is currently constituted is a good idea is another
matter entirely.

> > The bound on computer risks is the collapse of civilization
> > as we know it.
>
> Oh, dear!  Not the 95% annihiliation of life on earth?
> This is possible, but highly unlikely.
>
The entire argument is about nearly limiting cases: the product of a
"negligible" probability and an intolerable cost.

NASA has gotten all the "ordinary" risks correct twice, and twice lost
human life and shuttles because of what I suppose must be
"extraordinary" risks--the second time *after* scathing criticism and
review of its risk-management techniques.

> As for the financial crisis of Sept 2008, computers had
> almost nothing to do with it (unlike Aug 1987).  The fault
> lies in a series of incentives and decisions that are mostly
> understandable and justifiable on microscale yet add up
> to instability.  This was predicted by many, perhaps even you.
>
Computers were a big contributor to this latest crisis. They led to
overconfidence about risk-management. Everyone was using the same
detailed computer model to put values on complicated securities.
Everyone made the same mistake of negligible probability over and over
and over again.

For all I know, they're all using random number generators that
correlate in some way that no single person knows about but that
contains the seeds of disaster. The pun was accidental, but I'll let
it stand.

> > Ask Google about which software is "mission critical."
> > If their answer didn't include IE6, I couldn't imagine why.
>
> Yet you do not seem short of imagination!
>
I have a very nice imagination for which I can claim no personal
credit.

> AFAICS,  Google's "mission" is to _quickly_ serve up
> ranked WWW search results with attached ads.  The indexing
> spiders, dispatchers and database machines are absolutely
> mission critical.  All running Linux and in-house code AFAIK.
> MS-IE6 might be in some offline testing suite.  
>
That is not Google's mission at all. Google's mission is to gather
valuable information and to use it to make money. They have barely
begun to tap the possibilities. If they can't protect the information
they gather, they will go out of business. Their servers were
compromised because of a flaw in IE6.

Are there copies of IE6 installed somewhere on Google's system? You
bet there are. They want to make sure the pages they create display
correctly in IE6, which are still widely used in enterprise. Are
those copies of IE6 better locked down now? The ones they know about
are, anyway.

Robert.


From: Robert Redelmeier on 28 Jan 2010 18:26
In comp.sys.ibm.pc.hardware.chips Robert Myers <rbmyersusa(a)gmail.com> wrote in part:
> That's not a correct presumption. I only felt it necessary
> to rebut his final claim in order to rebut his entire
> argument as it pertains to computers.

Those are your feelings, without logical validity.

> People do lots of stupid and even self-destructive things
> with money.

Oh dear, clearly more lessons in elementary logic are needed:
just because people do some bad things with money does not
mean everything done with money is bad.

> Packet-switching is a very clever and useful idea.

Hardly new -- the telegraph was packet-switched. Multiplexing
/circuit-switching was very clever but now is undone.

> Whether the Internet as it is currently constituted is a
> good idea is another matter entirely.

.... and how do you propose to convince people or even
communicate without logic? Shouting your feelings?
Yes, your feelings are heard. Why should anyone care?

> NASA has gotten all the "ordinary" risks correct twice, and
> twice lost human life and shuttles because of what I suppose
> must be "extraordinary" risks--the second time *after* scathing
> criticism and review of its risk-management techniques.

Many more. Apollo 1 and lots of "accidents". Unfortunate, but risk
management is not perfect -- it can only deal with identified risks.
Classifying risks as "extraordinary" is a cop out. Frankly,
I'm surprised there have not been more. I expect 10% loss.

> Computers were a big contributor to this latest crisis.
> They led to overconfidence about risk-management. Everyone
> was using the same detailed computer model to put values
> on complicated securities. Everyone made the same mistake
> of negligible probability over and over and over again.

Only an incompetant craftsman blames his tools. GIGO is a
_human_ failing, not a machine failing. Very complex because
the human failings were systemic. Herd instinct and a certain
blindness driven by short time horizons.


> Are there copies of IE6 installed somewhere on Google's system?
> You bet there are. They want to make sure the pages they
> create display correctly in IE6, which are still widely used
> in enterprise. Are those copies of IE6 better locked down now?
> The ones they know about are, anyway.

Should have been behind an adequate firewall. NIST has been
warning about MS-IE insecurity for 10+ years.


-- Robert R
>
>
>
From: Robert Myers on 28 Jan 2010 18:59
On Jan 28, 6:26 pm, Robert Redelmeier <red...(a)ev1.net.invalid> wrote:
> In comp.sys.ibm.pc.hardware.chips Robert Myers <rbmyers...(a)gmail.com> wrote in part:
>
> > That's not a correct presumption.  I only felt it necessary
> > to rebut his final claim in order to rebut his entire
> > argument as it pertains to computers.
>
> Those are your feelings, without logical validity.
>
I'm tired of you, and I thought you were tired of making a fool of
yourself. Apparently not.

You don't understand anything more about logic than the warsaw pact
dude does about probability and statistics.

I've caught you dead-to-rights showing off your "knowledge," and I'm
done playing games with you. Do you want to tell me again about the
Euler equations?

> > People do lots of stupid and even self-destructive things
> > with money.
>
> Oh dear, clearly more lessons in elementary logic are needed:
> just because people do some bad things with money does not
> mean everything done with money is bad.
>
No, a**hole, but it does show that your argument about what others do
with their money is at best a red herring. And stop trying to teach
me about logic. You have no clue how little you know.

> > Packet-switching is a very clever and useful idea.
>
> Hardly new -- the telegraph was packet-switched.  Multiplexing
> /circuit-switching was very clever but now is undone.
>
> > Whether the Internet as it is currently constituted is a
> > good idea is another matter entirely.
>
> ... and how do you propose to convince people or even
> communicate without logic?  Shouting your feelings?
> Yes, your feelings are heard.  Why should anyone care?
>
Your content-free nastiness is left in place for the world to admire.

> > NASA has gotten all the "ordinary" risks correct twice, and
> > twice lost human life and shuttles because of what I suppose
> > must be "extraordinary" risks--the second time *after* scathing
> > criticism and review of its risk-management techniques.
>
> Many more.  Apollo 1 and lots of "accidents".  Unfortunate, but risk
> management is not perfect -- it can only deal with identified risks.
> Classifying risks as "extraordinary" is a cop out.  Frankly,
> I'm surprised there have not been more.  I expect 10% loss.
>
"Ordinary" risks was one of the many nonsensical items in Sebastian's
post to which I did not bother to respond. Even NASA never would have
been so naive.

> > Computers were a big contributor to this latest crisis.
> > They led to overconfidence about risk-management.  Everyone
> > was using the same detailed computer model to put values
> > on complicated securities.  Everyone made the same mistake
> > of negligible probability over and over and over again.
>
> Only an incompetant craftsman blames his tools.  GIGO is a
> _human_ failing, not a machine failing.  Very complex because
> the human failings were systemic.  Herd instinct and a certain
> blindness driven by short time horizons.
>
No, thank you. The entire ponzi scheme is built on ersatz knowledge
that depends on superficial thinking for which people are apparently
due "Nobel" prizes. It was to such "wisdom" that Sebastian appealed
in defending his own recklessness. The key role that computers play
is that they allow vastly more and faster amplification of
incompetence. The valuation model that led to disaster was actually a
piece of competent work--as far as it went.

> > Are there copies of IE6 installed somewhere on Google's system?
> > You bet there are.  They want to make sure the pages they
> > create display correctly in IE6, which are still widely used
> > in enterprise.  Are those copies of IE6 better locked down now?
> > The ones they know about are, anyway.
>
> Should have been behind an adequate firewall.  NIST has been
> warning about MS-IE insecurity for 10+ years.
>
You should *definitely* be CTO or at least CISO at Google. Your
qualifications are painfully obvious.

Robert.
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5 6 7 8
Prev: No multicore grahics from Intel, Intel withdraws Larrabee!
Next: Several Ntoskrnl BSOD crashes over the last week, but different stopcodes