home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!mcsun!uknet!acorn!eoe!ahaley
- From: ahaley@eoe.co.uk (Andrew Haley)
- Newsgroups: comp.lang.forth
- Subject: Re: Documenting
- Message-ID: <1492@eouk9.eoe.co.uk>
- Date: 7 Jan 93 10:27:04 GMT
- References: <C0EoE4.Go3@starnine.com>
- Organization: EO Europe Limited, Cambridge, UK
- Lines: 67
- X-Newsreader: TIN [version 1.1 PL6]
-
- Mike Haas (mikeh@starnine.com) says that "screens" (by which I presume
- he means blocks used for source code) are viewed by [users of other
- languages] as evidence that Forth will never be a modern language.
-
- Some people may believe this, but it flies in the face of historical
- fact; BLOCK was an innovation. Every computer OS around at the time
- (1970) supported text files, but Charles Moore invented a different
- method of accessing source text and other data, not because he was
- "old fashioned" but because he thought that he had a better system.
-
- I don't think that people who do not use Forth will be drawn to it if
- we change all our sources to use OS text files; they will carry on
- happily using C or whatever. The only difference will be that people
- who like to use blocks will no longer have their sources in that
- convenient form.
-
- There is a great deal of work to be done in getting the interactive
- development tools already used in Forth systems to work with text
- files. If Forth has any virtue, it is simplicity; rewriting existing
- tools to use text files would greatly complicate things.
-
- There is too much emphasis in this group on rewriting tools that
- already work very well in order to satisfy the expectations of people
- who are never going to use them.
-
- I've no idea where the notion that some of [Forth's] major features
- are architected to work best only in embedded controller applications
- comes from; that is contrary to what is known about Forth's history.
- Much early Forth use at Forth, Inc was in commercial/database
- systems. They found that blocks work extremely well in that
- environment.
-
- On some platforms, interaction with host operating system files is
- essential; fair enough. You might even run Forth as a task under
- another OS. I've done this in many applications, and sometimes it's a
- good solution, but Forth is not a programming language so much as an
- operating system. The Forth applications I've seen which impressed me
- the most used Forth as the OS. There are some wonderful multiuser
- databases which use the Forth operating system.
-
- It depends on what your goal is. If it's to get the maximum number of
- Forth users, maybe Mike Haas's route is the best. If your goal is to
- develop efficient applications on computers (and yes, efficiency is
- still important) then you might choose a different route.
-
- If the "mainstream" expects a standard set of functions which allow
- one to link with object files from other languages, then by all means
- let's supply one. But we shouldn't assume that they will use Forth if
- we do.
-
- Those who stay with source blocks don't do so because they don't care
- to "put the work in." It's because we find that blocks are a good
- solution to the problems we face when developing applications. If
- some others choose to use text files, that's fine. I don't want to
- convince others how great blocks are; I just want to explain why _I_
- use them.
-
- If others prefer text files, that's great, but I resent the
- implication that I haven't moved to text files because of laziness.
- It's because blocks work best for my uses. If others use text files
- that doesn't bother me at all.
-
- I am very disappointed to see the great "blocks versus files" flame
- war resurrected in this conference. I suppose this message will only
- prolong the latest outbreak; ho hum.
-
- Andrew.
-