Re: Fwd: MuP21 programming manual

[Jeff Fox <Fox@xxxxxxxxxxxxxxxxxxx>]

Dear MISC readers:

> Jeff Fox wrote:
> [[[ ramblings deleted ]]]
> 
> You do realize, don't you, that instead of answering any of Greg's specific
> questions 

I have answered mass quantities of Greg's questions in email and
in detail with great patience and joked with him about all his 
absurd insults.  I have answered many of his questions here just to
show to anyone who happened in on one of his rants about how
I never tried to answer his questions or just lie to him how
absurd that is.  And yes, I did realize that I didn't even
try to answer any of Greg's specific questions in that post.
I hope everyone else also realized that.

I also wanted to express my own opinion, which is different than
Greg's, that the manual is not excrement, but pretty nice.  We
have a right to express a different point of view.  I think it
is nice.  He hates it.  Infer what you will.

> you've just rambled on and on about how great the MuP21 manual

I tried to give it a review as compared to many of the other
documents available.  I do type very fast, make many typos that
I don't see, and am long winded.  I guess you could shorten that
to I rambled on, but I thought it wasn't too rambling.

> is and how Greg must be an idiot for not seeing the answers 
> in the manual which, apparently, leap right out at you?

I did not say a thing about Greg being an idiot or a child
or any such thing.  But the inference is certainly there for
anyone who wished to pick it up.  But I wouldn't, and didn't
use the term idiot.  Those are your words.  I think something
else is infered much more clearly.
 
> This is _precisely_ the sort of thing Greg complained about. 

Of course it is.  And I make no appology.  I have a right
to my opinion that it is a good piece of documentation all
in all.  He has right to his opinion that it is all excrement
and that all the people involved are lying about every detail.

> Most of us
> on the list can't help Greg because we don't have a copy of the manual
> and/or haven't played with the chip. 

And those who do and have tried patiently to answer his questions
have been called names.  Of course those who can't answer his
questions can't help him.  

I just have to list myself as one of those who can't answer
his questions because I tried to and just got tired of not
getting anywhere and just getting insulted.  It is a little
like dealing with a three year old who asks a question
and recusively asks another question about every answer.
We have all played that game with a three year old before
and realized that the adult can't win at that game.  At least
three year olds don't usually curse and call names when
they throw their temper tantrums.

> But surely someone who sees the simple
> clarity of the manual and/or understands the chip well could quickly and
> easily answer some of his simple concrete questions such as:
> 
> > And in fact everything from 1E0000-1EFFFF may be the
> > conf. register.  Is this the case?

I wish it were true, but I am afraid that that is a question 
about the hardware implementation not about what is in or should
be in the programmer's manual.  The programmers manual SAYS that
the address of the conf. register is this address.  If you
access this address you will get this register.  It gives
you all the information that the programmer needs to deal
with the configuration register.  

Greg's question is outside of the range of a programmers manual
and if a programmer wants to run experiments to try to figure 
out how the wires are connected underneath by tring 
unassigned opcodes or unassigned addresses and see what happens
that is fine.  A programmers manual will not, and should not
answer questions about things like that.   A programmer has
to write a program, write to the official unassigned address
in question and see what happens to be able to answer
such a question.  

Have people done that? Yes. Is it in the programmers manual?
No.  And it shouldn't be.  That is a hardware question and
only someone who has seen under the hood, or been told
what is down there, or done the experiment themselves
can answer the question.

Not seeing an answer to that question in the manual does
not mean that Greg is an idiot.  That question is not 
answered in the manual.  If you are reading the programmers
manual to find out hardware information about the hidden
internal implementation details then maybe there might
be a better place to look.  The programmers manual doesn't answer
that kind question about hardware.  It does have pinouts and
electrical characterists and timing diagrams that engineers
need and extensive documentation of all the instructions and
things that programmers need to know.  It does not go
into details about how the transistors are connect up
internally to someone interested in that.  Ting has that
sort of stuff documented elsewhere.  That is not what Greg
asked for.  Greg asked for the engineering hardware
information to make it work and the software info to make
it work.  That is what is in that manual.

You will notice that I never did answer Greg's question
about what happens if you try those unassigned addresses.

> I don't intend to restart Greg's flame war, but a simple "yes" or "no"
> would be lots quicker and easier to type than a long rambling message
> about how anyone that doesn't understand the manual is obviously an
> idiot.

Agreed.  I will simply explain that if someone wishes to ask a
simple yes/no question about the hardware they will only get
an answer from me only if they do not precede it with a long rant
about how the manual they have is crap (even though their
question is outside the scope of the manual anyway) and the
author lied to him and I lied to him and I am name and
that name.

These are just my opinions anyway. Other people can form their
own opinions about whether a programmers manual's quality should
be judged by whether it answers all speculation about how
the hardware is wired up below the surface.

In fact that document does contain several of Chuck's talks
to fig about the designing of the chip.  He does talka about
how the hardware decodes work.  As I recall he does just
decode just the required upper bits to select each of the
address ranges and so it might very likely mean that
since that is the only control register on P21 that it
is visible over the whole range.  I may have even done
the experiment myself many years ago.  I can't remember.
That sort of detail of how a particular chip did a 
particular thing in hardware is not important from 
the point of view of a programmer.

Jeff Fox