From: Joerg on
Jerry Avins wrote:
> On 7/10/2010 4:12 AM, Rune Allnor wrote:
>
> ...
>
>> For the past decade or so, a number of cable TV channels have
>> started producing shows on astronomical subjects. Stunning as the
>> computer graphics in these programs is, I find them totally boring.
>> There is nothing left for me to figure out or imagine. Those things
>> have degenerated to mere show-off pieces for computer graphics
>> artists. Or standard as the graphics seems to be - computer graphics
>> *engineers*.
>>
>> Nah, the real effort has to take place inside the student's mind.
>> As somebody once said "I can't teach you, but I can help you learn."
>
> I often, not always, prefer to read a book than see a movie made from
> it. Visualization is a powerful tool. Once, in a meeting, I gave an
> approximate (10%) numeric solution to transcendental equation that had
> just become germane. I simply visualized my slide rule. Visualizing a
> calculator doesn't work for me. :-)
>

When I did that in a company some guys looked over my shoulder, saw the
slide rule. "You know, we do have electric light around here".

But visualizing a slide rule can be very powerful. It is amazing what
can be done with paltry computing resources in a chip if you manage to
place a log-LUT somewhere.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
From: robert bristow-johnson on
On Jul 11, 4:20 pm, Joerg <inva...(a)invalid.invalid> wrote:
>
> But visualizing a slide rule can be very powerful. It is amazing what
> can be done with paltry computing resources in a chip if you manage to
> place a log-LUT somewhere.

or put in a simple 6th order polynomial to cover one octave of
log2(). if memory is tight but MIPS are in abundance, i might suggest
that. both LUT or a finite power series have some error, but can be
very small.

r b-j

From: Jerry Avins on
On 7/11/2010 11:26 PM, robert bristow-johnson wrote:
> On Jul 11, 4:20 pm, Joerg<inva...(a)invalid.invalid> wrote:
>>
>> But visualizing a slide rule can be very powerful. It is amazing what
>> can be done with paltry computing resources in a chip if you manage to
>> place a log-LUT somewhere.
>
> or put in a simple 6th order polynomial to cover one octave of
> log2(). if memory is tight but MIPS are in abundance, i might suggest
> that. both LUT or a finite power series have some error, but can be
> very small.

All very well. I was doing trig.

Jerry
--
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������