Talk:WATFIV

No New Compilers ???

See: Open Watcom. So there is more wrong than just "No new compilers"; need to link the continuing history. Why not merge, leaving only a stubb in Wikipedia? Rwwww 05:47, 16 August 2006 (UTC)

Open WATCOM FORTRAN-77

> "no new WATFOR-like FORTRAN compilers have been developed"

Open WATCOM FORTRAN-77 is the continued development of the WATCOM FORTRAN-77 compiler, a traditional optimizing compiler. No one is saying that development of this compiler is not continuing.

No new WATFOR-like compilers have been developed. WATFOR-like implies a number of things, chief of which is its ability to compile/link/execute on the fly. Turbo Pascal was a Pascal compiler that had "WATFOR-like" characteristics. But that was a Pascal compiler, not a FORTRAN compiler. If someone knows of the recent development of a new WATFOR-like FORTRAN compiler, please let us know.

Watcom's WATFOR-77 compiler has not been open-sourced.

Original research

"The source of this is personal experience." – Reporting personal experience is clearly original research (see WP:NOR). A large part of this article is a personal interpretation of using WATFIV. In particular, the assertions about WATFIV's uses in debugging should be sourced. John FitzGerald 20:49, 5 February 2007 (UTC)

In fact. there are many questionable, unsourced assertions in the article, such as "Non-experienced programmers could be taught programming at minimal cost in time and computing resources." John FitzGerald 20:58, 5 February 2007 (UTC)

I believe something like that is given in documentation as the reason for writing WATFOR, at least for the OS/360 version. Gah4 (talk) 09:08, 16 February 2019 (UTC)

The OS/360 compilers are designed for a separate linkage editor, and then go (IBM name) step. That is all fine for large programs, and especially for debugged ones, but is a lot of overhead. Gah4 (talk) 09:08, 16 February 2019 (UTC)

Since you're not supposed to have multiple templates on an article, I chose the NOR template as the one which best described the problems with the article. John FitzGerald 15:06, 12 February 2007 (UTC)

Since it has 14 references, it can't be all original research. Still, I am suspicious. Among others, the claim that WATFIV is WAT Fortran IV seems to be unsourced, and I don't remember that at all from the 1970's. Much is documented on the uwaterloo site. I am not ready to remove what looks like OR, but I am not against doing it. Gah4 (talk) 07:02, 18 June 2016 (UTC)

One reference, lots more available if you really want them

From: http://www.cs.uwaterloo.ca/40th/Chronology/1965.shtml

By November 1965, five Canadian schools, eleven American installations and one in Switzerland had requested a copy of WATFOR. By June of 1966, requests were received from an additional two Canadian locations, thirty-six American, and six international installations The WATFOR compiler was acknowledged to have increased the computing capacity at the university by 5 times, thereby saving the university dollars in hardware and software upgrades that most certainly would have been required to provide necessary computing service.

-- and --

The original WATFOR compiler for the FORTRAN IV computer language for the IBM 7040 computer had diagnostic capabilities superior to most of its contemporary counterparts so that users could find and correct errors. The program greatly expanded the potential for using computers in undergraduate instruction and it put the fledgling university on the map internationally.

 -- and from:  http://www.cs.uwaterloo.ca/40th/Chronology/1967.shtml --

WATFOR 360 was developed in 1967 and it greatly enhanced the speed and ease of use of the IBM 360 series of computers. The program was essentially WATFOR for the IBM 360 series, the latest in computing technology at the time. It was also the early student-friendly variant of FORTRAN IV. The group leaders of the team that created the compiler, Paul Dirksen and Paul Cress, won the Grace Murray Hopper Award in 1972 for their efforts. In 1973, WATFIV was augmented with structured programming constructs and character variables to create WATFIV/S.

When Cress and Dirksen completed their work on the WATFOR 360 compiler, the program increased the performance of the machine to the point where it could compile and execute a 45 card job in under 10 seconds. Comparable programs took some 13 minutes to execute the same job. The WATFOR 360 compiler became the basis for the world-famous WATFIV program

For full the article, please see: UW Special Collections. GA 133-57. Wes Graham Fonds. Series 1.2: Biographical: Articles. John Helliwell, PC 4, no. 7 (April 2, 1985), 204-209, ill.

 --

I invite you to read the WATFOR/WATFIV/WATFIV-S articles in the subsequent years' chronology. And the author of the books cited below, Prof Ken McLaughlin was the principal author of the chronology excerpted above (which was a commissioned online-only work created for the 40th Anniversary of the School of Computer Science).

For hard-copy verification, see:

Ken McLaughlin; Waterloo: The Unconventional Founding of an Unconventional University; UW Press 1997

Ken Mclaughlin; Out of the Shadow of Orthodoxy: Waterloo @ 50; UW press 2007

-- Trevor Grove, UW  —Preceding unsigned comment added by 129.97.176.165 (talk) 23:31, 23 November 2007 (UTC) 

Yes, we need to remove the opinions from this article and add citations to sources. It was a very important milestone in educational computing, so there are a few. I am working on putting the project in more context to a layman instead of just having a litany of acronyms and people who ever worked on the project (if otherwise not notable). Patience is appreciated. W Nowicki (talk) 20:47, 5 April 2011 (UTC)

Books

Why do we need to link to four different versions of the WATFOR/WATFIV books? Doesn't each supercede the previous? Is there important historical information that reading the older ones and then the newer ones will provide? --jpgordon^{::==( o )} 20:28, 5 April 2011 (UTC)

I do not know, have not seen the actual books, just the listings of them. Now that you mention it, the Cress series might indeed just be updates. Hard to tell since they are listed as different titles, not differet editions of the same title. I vote leave them in until somenoe can verify it. Is there some harm done by having them? This project is of very much historical interest, so probably all are relevant to its development over time. And it sounds like at least one might be a description of the compiler itself (for those maintaining it) vs. a textbook for people learing programming, very different. Hard to tell. Do you have copies? Glad to see someone is watching this neglected page, thanks. W Nowicki (talk) 20:47, 5 April 2011 (UTC)

Nah. I'm old enough that "WATFIV" is this new-fangled thing; WATFOR was good enough for me. I kept my old manuals for years, but decided one day I didn't really need my 1975 programming manuals. --jpgordon^{::==( o )} 00:32, 6 April 2011 (UTC)

Watfiv came out about when I was first learning Fortran (in 9th grade), and I remember

liking it much more than WATFOR. For one, I think there were some actual bugs (or deficiencies) but also I liked the WATFIV extensions. Gah4 (talk) 09:01, 16 February 2019 (UTC)

Why WATFIV?

The initial and latest implementations are known as WATFOR. Then why this article is named as WATFIV? I think, it should be renamed to WATFOR. --Vssun (talk) 17:32, 23 September 2012 (UTC)

Thanks for your interest! Nice to hear someone is working on these again. The guideline is generally the "most common" name, not necessarily the first or last. Sadly that is often subjective, which is why there are so many battle, which in my opinion may divert editors from actually improving the quality of the articles. Certainly I knew it as WATFOR (which might be dating myself) but my guess is whoever created the article knew it as WATFIV. I would not object to such a move. W Nowicki (talk) 15:07, 24 September 2012 (UTC)

Thanks for the comment. My intention was to translate this into Malayalam --Vssun (talk) 04:47, 27 September 2012 (UTC)

I assumed that there was also a WATFOR article. I do still wonder about the name. At the time, it was explained to me that it was WAT-FIVE, that is, four plus one. Also at the time, there was much discussion on when IBM would produce Fortran V. Some others would name their extended Fortran compilers Fortran V. But this article, and the UW references, seem to describe it as Fortran IV. But WATFIV is where the extensions that would becomre Fortran 77 were tested out. WATFOR mostly followed the IBM Fortran IV, but WATFIV extended it, such as with CHARACTER variables, expressions in DO statements, and expressions in I/O lists. I think keep this, but also add a WATFOR page. Gah4 (talk) 10:49, 24 October 2015 (UTC)

I am still interested in a reference on the naming, WATerloo Fortran IV vs. WATFIV (that is, for plus one equals fiv). Also, what, if any, is the connection between the compilers from the 1970's for IBM mainframes, and later compilers for IBM PCs? Gah4 (talk) 20:06, 27 October 2016 (UTC)

WATBOL was implemented in Z1

"Unlike previous compilers, a significant portion of WATFOR-77 was written in a portable systems language to ease the implementation of the compiler on other computer systems. Earlier WATFOR compilers were written entirely in machine-dependent assembly language."

The WATBOL compiler was written in Z1 (Zarnke's very efficient systems implementation language). This was not an assembly language (but it was built for the IBM System/360). Z1 was written in Comass (sp?), Zarnke's "Compiler Assembler", a kind of compiler-compiler. Zarnke's PhD disertation would probably cover this. DHR (talk) 06:54, 30 June 2016 (UTC)

Seems to me that WATBOL isn't a version of WATFOR or WATFOR-77 or WATFIV, so the statement is still true. Some say you can write Fortran in any language, but I think WATBOL is a little too far. Gah4 (talk) 20:46, 30 June 2016 (UTC)

References