From: Steve Pope on
steveu <steveu(a)n_o_s_p_a_m.coppice.org> wrote:

>>glen herrmannsfeldt <gah(a)ugcs.caltech.edu> wrote:

>>>Muzaffer Kal <kal(a)dspia.com> wrote:

>>>(snip, someone wrote)

>>>>>Hmm. I can still remember old time arguments that it is impossible to
>>>>>communicate more then 2400 bits per second over the phone line.
>>>
>>>> I think one reason for that is over time what constitues a 'phone
>>>> line' has also changed. 2400 bps might as well be the limit for a end
>>>> to end analog line.

>>>Also on how much computational resources you put to the task.

>>>At the time of the 2400 bits/second claim it would likely take
>>>more processing power than was available at even the largest
>>>computers.

>>Another thing to remember... if it's worth remembering, is that
>>before turbo codes were developed, many people felt it was
>>impossible to communicate beyond the rate-distortion limit.

>What relevance does that have? Turbo codes played no part in taking 2 wire
>telephone line modems way beyond 2400bps.

No, but the rate distortion limit did add to the confusion
as to what data rates were possible.

Steve
From: Eric Jacobsen on
On 4/5/2010 10:00 PM, steveu wrote:
>>> This
>>> can, of course, work both ways. An investor who had been informed of
> the
>>> channel capacity theorem might have cut off all funding for MIMO work
> as
>>> bogus, because theory says you can't communicate that fast.
>>
>> MIMO doesn't violate any theories of which I'm aware. The basic
>> capacity formulas for single-use of memoryless channels have been
>> characterized as such from Shannon's earliest work. Applying formulas
>> with those restrictions to multiple use of multipath channels (with
>> memory) could be expected to lead one to incorrect conclusions, though.
>
> Of course MIMO doesn't violate any theories, or it
>
> - wouldn't work, or
>
> - the theory would have been disproven.
>
> You are not thinking like a novice in the field. If someone tells an
> investor the channel capacity expression, which it quite easy to remember,
> they might well (and quite reasonably) discount as bogus any claim that two
> boxes can communicate faster than that.
>
> Steve

Most VCs have some awareness of what they don't know or what they can't
verify on their own. The smart ones then get help in the technical due
diligence area. Sometimes they get incompetent help for that, but that
happens in any aspect of business.

Clearly there's a lot of "non-smart" money out there. My former
employer, despite being a global technology leader, had a reputation for
not being smart with capital investments. I saw first hand how that
process worked, and was not able to stop a few really bad investments
(although I did help ward off a few, and helped close some good ones as
well).



--
Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.abineau.com
From: Steve Pope on
Eric Jacobsen <eric.jacobsen(a)ieee.org> wrote:

>Most VCs have some awareness of what they don't know or what they can't
>verify on their own. The smart ones then get help in the technical due
>diligence area. Sometimes they get incompetent help for that, but that
>happens in any aspect of business.

>Clearly there's a lot of "non-smart" money out there. My former
>employer, despite being a global technology leader, had a reputation for
>not being smart with capital investments. I saw first hand how that
>process worked, and was not able to stop a few really bad investments
>(although I did help ward off a few, and helped close some good ones as
>well).

If they had VC level business intuition, they'd be VC's
rather than middle-managers at a huge corporation.

I once worked for a Dow 30 company. I won't detail the many
mistakes they made, but one troubling aspect is they could
conduct a huge venture that would be a massive failure, then
for the next huge new venture, they would put the same
individuals in charge.

Steve
From: Eric Jacobsen on
On 4/6/2010 9:18 AM, Steve Pope wrote:
> Eric Jacobsen<eric.jacobsen(a)ieee.org> wrote:
>
>> Most VCs have some awareness of what they don't know or what they can't
>> verify on their own. The smart ones then get help in the technical due
>> diligence area. Sometimes they get incompetent help for that, but that
>> happens in any aspect of business.
>
>> Clearly there's a lot of "non-smart" money out there. My former
>> employer, despite being a global technology leader, had a reputation for
>> not being smart with capital investments. I saw first hand how that
>> process worked, and was not able to stop a few really bad investments
>> (although I did help ward off a few, and helped close some good ones as
>> well).
>
> If they had VC level business intuition, they'd be VC's
> rather than middle-managers at a huge corporation.

The guys in the capital arm were dedicated capital guys, definitely
comparable to their equivalents in the other capital/VC firms we wound
up working with in deals. FWIW, the capital guys in that company were
WAY better than the A&M guys, who did several billion-dollar deals that
turned to dust.

> I once worked for a Dow 30 company. I won't detail the many
> mistakes they made, but one troubling aspect is they could
> conduct a huge venture that would be a massive failure, then
> for the next huge new venture, they would put the same
> individuals in charge.
>
> Steve

Ages ago when I worked at Honeywell that was such regular practice that
there was an acronym for it: FUMU = (Foul) Up, Move Up.

--
Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.abineau.com
From: Clay on
On Apr 6, 2:24 am, Muzaffer Kal <k...(a)dspia.com> wrote:
> On Tue, 06 Apr 2010 00:31:36 -0500, Vladimir Vassilevsky
>
> <nos...(a)nowhere.com> wrote:
>
> >steveu wrote:
>
> >> If someone tells an
> >> investor the channel capacity expression, which it quite easy to remember,
> >> they might well (and quite reasonably) discount as bogus any claim that two
> >> boxes can communicate faster than that.
>
> >Hmm. I can still remember old time arguments that it is impossible to
> >communicate more then 2400 bits per second over the phone line.
>
> >VLV
>
> I think one reason for that is over time what constitues a 'phone
> line' has also changed. 2400 bps might as well be the limit for a end
> to end analog line.
> --
> Muzaffer Kal
>
> DSPIA INC.
> ASIC/FPGA Design Services
>
> http://www.dspia.com

A plain old pots line worked at 32kbps - end to end. It was the 56k
revolution that exploited an assymmetry from the CO to the enduser and
that took advantage of looking at phone lines differently.