Talk:Elliott ALGOL

Untitled
This page should be called "Elliott Algol" - but I don't know how to rename it!! (bil@beeb.net) — Preceding unsigned comment added by 195.248.107.182 (talk) 08:45, 5 October 2005 (UTC)

These two implementations are completely unrelated, except insofar as they are implementations of the language designed by John Backus and others and described in their Report on the algorithmic language ALGOL 60.

Elliott Algol was implemented by C.A.R.Hoare in the UK for Elliott Brothers to run on their Elliott 803 and subsequent Elliott 503 computers and first released in 1963. The first dialect of Burroughs Extended Algol was implemented by designers of the Burroughs B5000 (later B5500) computer led by Bob Barton in the US and released in 1961. The B5000, like its successor machines, the B6500/6700 and B7700/7800, was unusual in that it had no assembly language. The operating system was writted in an extended Algol dialect.

Owen Reddecliffe is an Australian who worked for Burroughs in Australia during the 1970s. Unless there is evidence to the contrary I don't believe he ever had anything to do with Elliott Algol. The only Elliott machines in Australasia were, I believe the 503s in Hobart and Wellington, in joint government-university computing centres. Both as it happens were superseded by Burroughs B6700 machines but the history of how that happened is not relevant here.

124.170.18.14 (talk) 10:05, 19 August 2008 (UTC)

which ALGOL
I think this page is confusing two different implementation of ALGOL. The one that ran on Elliott machines and the one that ran on Burroughs machines. Iccaldwell 10:54, 2 December 2007 (UTC)

The Elliott Algol I know about is the famous one associated supplied for the Elliott 803 computer in the early 1960's (although targeted at it's successor, the less-well-known but much faster 503). According to the original Elliott book I have in front of me, the company was known as National Elliott at the time, although the company changed names and owners regularly. On the back of the book it says to contact Elliott-Automation Computers Ltd, Elstree Way, Borehamwood (England) for scientific applications and NCR Electronics in Marylebone Road for commercial uses.

Elliott Algol was very definitely an implementation of Algol 60. Some of the brains behind it were Jeff Hillmore, Dan Simpson, Jill Pym and most notably, Prof. Tony Hoare (inventor of the quicksort algorithm, sometimes know as Charles Hoare but usually CAR Hoare).

I knew this compiler fairly well, but know nothing of any later connection with Burroughs. That's not to say that it didn't reappear on a different platform. I agree with Iccaldwell - I think there's confusion about implementations here. Can anyone add to this? I've since come to regard the Elliott Algol compiler as something of a milestone in computing. I didn't realise who good a language it was until I had to use some of the other's around at the time. Fjleonhardt (talk) 20:27, 4 May 2008 (UTC)

Merge proposal
It seems from the very limited amount of discussion and similarly limited page history that the language being described in this article is something to do with the Burroughs range and nothing to do with Elliott Brothers (computer company), so I'm proposing that the Burroughs article take anything that might be pertinent from here if it's not already in there, then I'll propose that this article be deleted unless someone can come up with anything notable about Elliot ALGOL, possibly from http://www.billp.org/ccs/A104/ (cited on the main ALGOL page).

--ClickRick (talk) 21:50, 18 May 2009 (UTC)

Paul Kimpel and I corresponded about this claimed connection between Elliott ALGOL and Burroughs ALGOL and having looked at both over the recent years, agree with Paul's analysis that they are unrelated, except for the possibility that both groups were involved with the general debate about the ALGOL language and may have discussed implementation strategies. Burroughs ALGOL was a single pass compiler, Elliott was multi-pass, Paul goes on to state: "... Elliot Algol had a number of significant limitations that B5000 Algol did not: no branching out of a procedure body to an external label, bounds of an own array must be constants, recursive procedures must not have scalar parameters called by name, no procedures as formal parameters, identifiers significant only in the first six characters, etc. All of this suggests the Burroughs compiler had a completely different design." This page should be kept however, as there is more to be added about Elliott ALGOL, particularly Bill Purvis's efforts, and hopefully soon more detail about the 503 ALGOL compiler differences (compared with the 803). --Nigwil (talk) 23:08, 3 March 2012 (UTC)