Re: 8520/21

From: silverdr_at_wfmh.org.pl
Date: Sat, 07 Mar 2015 12:01:32 +0100
Message-ID: <54FADA8C.7080208@wfmh.org.pl>
On 2015-03-07 00:31, Jim Brain wrote:

>>>> Presumably they just wanted a binary counter for the Amiga. The
>>>> idea that they made the 8520 and then the 8521 was an
>>>> afterthought might be true, too :-)
>>>>
>>> I suppose they did this is crazy way just because they had the
>>> capabilities to do so. But, man, what a waste of precious
>>> resources.
>>
>> Have you ever worked in/with/for a corporation? Especially one
>> that's been relatively large and has undergone some M&A, Jim?
>>
>> My guess is not ;-)
>>
> Actually, I have worked at 4 publicly traded corporations in my life,
> 3 above $2b, 3 of which are the product of significant M&A.

Well, then Dilbert must have been on your reading list.

> It's not the waste per se.  It's the waste in light of the rationale
> and personalities of the earlier crop of designers.  The stories I
> read noted that time was always short, and designers were always
> looking for shortcuts to deliver working silicon.  The early designs,
> at least, were filled with elegant (and something inelegant) repeats
> and such to finish a design using a minimum of space and design time.
> Bob Yannes and his compromises he made to get the SID done, the
> tradeoffs in the various video ICs, etc.
>
> Against that backdrop, dropping just one feature doesn't make sense.

Please note that there's been some years between those chips. During
those years there were some changes in the company, including top
management too.

What I am trying to say that I haven't seen a corporation (I never
really worked as an employee of one but I consulted many years for
several) relatively big, especially after some M&A, where illogical
waste wouldn't be something that is just happening every day.

So, maybe there is some story behind the 6526/8520 or maybe that's just
one of the illogical things people in corporations do due to many
factors. Lack of information flow being one of the most prominent.
Budget spending requirements and cross-departmental wars or at least
competitions coming somewhere down the list among many other.

[addition to last night's one]

Please also note:

- AFAIR the C64 was supposed to be phased out about 1986.
- NMOS was to be phased out in any case
- Amiga didn't need the BCD based TOD

What I tend to speculatively believe is that the rational approach was 
something like: "We need to move HMOS so let's make the new CIA in HMOS 
but since our 8-bit line is passe, let's strip the new chip from 
unneeded baggage". That's how the 8520 came to the world. The fact is 
though that the C64 didn't want to die and continued to outsell all 
other machines despite the launch of both Amiga and the C128 (and the 
16/116/+4). The management was faced with a dilemma "to kill or not to 
kill the only hen that lays golden eggs". At some point then a decision 
was made to redesign the C64 (into C64C) rather than phase it out as 
previously planned. But at that point the 8520 didn't have what was 
absolutely needed to keep full compatibility with the previous C64 and 
all of its software (something they learned the hard way was important 
and possibly became - potentially unnecessarily - obsessed with). Hence 
an "afterthought" - "we need the BCD based TOD back NOW!". The 8521 came 
as the answer…

-- 
SD!

       Message was sent through the cbm-hackers mailing list
Received on 2015-03-07 12:00:05

Archive generated by hypermail 2.2.0.