Talk:Counter-machine model
Add topic | This article is rated C-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||||||
| ||||||||||||||||||
Next work
[edit]Compare each model with the Counter machine:Reference model, and use the RefLib to show examples and comparing.
Complementar sugestion
[edit]At the 14:08, 24 October 2006 version of the article, there are a "The models in more detail" section and the 10 models.
Reading (over view) each model, we can see a "soon introduction" and "Details" parts. I think the "Details part" can be split in a "near standard" (and perhaps explicit) parts:
- Main diferencial characteristics: with the reference model (and/or predecessors).
- Main motivations.
- Typical instruction set: list of labels from RefLib. If necessary, a "translation table"... "John, in your original article, used
Jfor RefLib'sJZ,MULTforMUL, ... " - Notes: other notes, comparations, etc.
It may be to facilitate readers.
-- Krauss 26 octuber 2006
Fixing the description of Melzak's model
[edit]The description of Melzak's model used to contain the following sentence:
The phrases indefinitely large number of locations and finite number of counters here are important. This model is different than the Minsky model that allows for a finite number of locations with unbounded (effectively infinite) capacity for "markers".
I deleted it as both observations are false. There are no fundamental differences between this and a standard register machine, the author of that sentence just misunderstood what the quote actually said. Melzak's world has a potentially infinite number of locations, but this is just for the programmer's convenience. Each program in Melzak's world is a finite series of instructions, and therefore each program will actually use only a finite number of registers, just like a standard register machine. We just get to choose their names. And the "finite number of counters" is also just a misunderstanding. The "finite" in the quoted text is talking about specific configurations of the machine. Again, this is exactly the same as with a standard register machine: each register can store an arbitrarily large number but at any given moment of any calculation that number is finite.