Jump to content

Talk:Richards controller

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Introduction

[edit]
"The Richard’s Controller is a clever method (...)"

"Clever"... this does not appear to be neutral IMHO. --Edcolins 20:39, 17 February 2007 (UTC)[reply]

Well, it was clever, but it was also a pretty obvious use of msi. Others had seen this and created similar circuits.

Another problem is that by 1973, when this was published, the circuit was already obsolete. Field programmable read only memory (prom), with 32 8-bit words, primitive by todays standards, but a wonder at the time, had become available. This was much more versatile. Not only was the design easier, the machine could be modified by replacing the prom.

AJim 03:55, 19 July 2007 (UTC)[reply]

I believe that the article stresses the use of the technique in design, not implementation of state machines.

Dhatfield 14:26, 10 November 2007 (UTC)[reply]

I am talking about the use of this technique in design. In 1973 there were better ways to implement the designs that could be implemented using this technique. This controller, by the way, is a kind of state machine, as the first sentence of the article states. I thought it was clever at the time I was shown a version of this (pre 1973). and I used it for some designs, but when the prom came out (also pre 1973) we switched design approaches and never went back. Today a designer might use this technique in "virtual" hardware implemented in a field programmable gate array in order to save a few gates. The prom approach is not so different except that it is more flexible. There are sometimes still reasons to use a multiplexer and decoder, but it easier and more general to get the next state from the memory and use only a register, and not a counter, for holding the state.

I must say that there is one important concept here, and that is that the state is stored encoded as a binary number. Back in those days there was a competing design philosophy that said, "since flip flops are cheap, let's dedicate a flip flop to each state". This turns out to be a bad decision because the hardware has 2n possible states for n flip flops, and most of these states are invalid. There are too many ways for things to go wrong.

I believe there are parallels in biology. Stuart Kauffman analyzed the pattern of homeotic mutations (transdeterminations) in drosophila larvae and concluded that the determined state of a cell appeared to be encoded in binary.

AJim (talk) 19:55, 13 June 2008 (UTC)[reply]

Expert review

[edit]
The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.

The result was: There was no conclusive answer, I will reset the date counter. --B. Wolterding (talk) 11:33, 29 November 2007 (UTC)[reply]


This review has been announced on the talk page of WikiProject Computer science. Comments should be added on this page, below.

This review has been announced on the talk page of WikiProject Technology on Nov 08. Comments should be added on this page, below.

As part of the Notability wikiproject, I am trying to sort out whether this topic is notable enough to have its own article. Could you have a look at it? It's not clear to me whether this design is in fact widely used, or whether it is just something that was described in a single research article. The only hits I get on Google Books for the title refer to King Richard III.

Also, you might want to add the article to your project (I don't know why it is listed for WikiProject Philosophy).

Your opinions are welcome; please add your comments below. --B. Wolterding 14:10, 26 October 2007 (UTC)[reply]

Google? Google is your guideline for notability of controller design techniques? Although not an expert in this particular controller type, regarding a strict "notability test", mention in a 3rd edition 2001 text on Digital Design verifies notability as far as I'm concerned and the content is certainly good, clear and concise. Lets go and hunt stubs, rather than decent articles regarding (relatively) specialised fields.

Request to author: Please provide one or two additional references to this work from other secondary sources to provide a conclusive answer. Dhatfield 14:26, 10 November 2007 (UTC)[reply]

  • No, Google is not my guideline here. I'm not an expert in controller design. But if this concept is notable, one would think that it appears in typical textbooks on the subject. Since there seem to be no hits on Google books, that makes notability at least doubtable. So I asked for an expert review. --B. Wolterding 16:11, 10 November 2007 (UTC)[reply]
    • I will reset the date counter of the notability tag for now, to allow for more sources to be provided. But the sources given do not convinve me - Is the concept mentioned in passing in that book or is it really in-depth coverage? --B. Wolterding (talk) 11:33, 29 November 2007 (UTC)[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

modern information age

[edit]

The article currently says A few examples of devices that need a complex controller are CPUs, Graphics Cards, Network Interfaces, Hard Drives, Bus Controllers (USB, PCI, Firewire…) The modern information age would not be possible without complex controller designs like the Richard’s Controller. which is technically true, but misleading. It's like saying "the modern information age would not be possible without water, such as the water found on Europa (moon)".

Yes, all these devices are using "complex controllers". Yes, Richard's controller is a kind of complex controller. However, people making these kinds of designs are not using, and often have never have heard of, Richard's controller.

For a while I suspected that perhaps this idea of a "counter" in a "complex controller" may have originated with Richard, and people are using it without realizing who invented it -- like people today use electric lights without realizing electric lights were invented by Francis Hauksbee in 1705.

If anyone can point to any specific "modern information age" device that uses a Richard's controller (or a controller derived from it), please point it out in the article. I doubt any are -- my understanding is that "modern information age" devices use complex controllers derived from the 1951 Maurice Wilkes microprogramming concept.

I think all text about "control units in general" should move to the control unit article, the same way text about water in general should go in the "water" article (rather than leaving in the Europa article). --68.0.124.33 (talk) 17:18, 11 April 2008 (UTC)[reply]

Follow-up re: notability

[edit]

I'm finding several indications of notability in Google, including evidence of the Richards contoller in CS/EE courses at at UC Davis and Rensselaer, and its inclusion in the Google Book, "An Introduction to Electronics", which describes it as "a classic hardware implementation" of a finite state machine. Based on this and the support of the expert review from 2007 (see page history), I'm removing the notability tag. --DGaw (talk) 04:50, 11 December 2011 (UTC)[reply]