Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!seismo!rutgers!topaz.rutgers.edu!brandx.rutgers.edu!webber From: webber@brandx.rutgers.edu (Webber) Newsgroups: news.admin,news.groups Subject: Making binary groups obsolete (was Re:Are binary groups necessary?) Message-ID: <303@brandx.rutgers.edu> Date: Thu, 23-Jul-87 15:49:58 EDT Article-I.D.: brandx.303 Posted: Thu Jul 23 15:49:58 1987 Date-Received: Sat, 25-Jul-87 07:52:48 EDT References: <266@brandx.rutgers.edu> <8225@utzoo.UUCP> <272@brandx.rutgers.edu> <153@hobbes.UUCP> Organization: Rutgers Univ., New Brunswick, N.J. Lines: 119 Summary: an offer Xref: mnetor news.admin:724 news.groups:1269 To: webber@aramis.rutgers.edu, uwvax!uwmacc!hobbes!root@seismo.css.gov In article <153@hobbes.UUCP>, root@hobbes.UUCP (John Plocher) writes: > [ > Instead of this META DISCUSSION why don't YOU use this front end and write one? > YOU, Bob Webber. Not someone else. > > Put up, shut up, or DO SOMETHING CONSTRUCTIVE about it! > Well, I never meta discussion I didn't like. And if I did everything that I said was easy, then I would have no time to do things that are actually difficult (of course, there is always time for the impossible). I maintain that writing a simple C compiler for a standard micro is easy and I will make you the following counter offer: Most people distribute public domain code from a wish to share with a community of likeminded people. I have no wish to share with a group of people too lazy to write a simple C compiler, so clearly you can't expect me to do it for this reason. You have elsewhere said that people on the net get things done the way they want because they have donated to the net software and administrative effort. I disagree and further maintain that if I did something for the net there would be no expectation that things would start working in a manner I think is more reasonable. Although everyone on the net maintains they are doing what is `right,' there is too much descrepancy between my notions of `right' and the behavior I have observed by the `backbone' and `moderators.' Thus, I am not willing to rely on their `good faith,' but I will enter into some variation on the following agreement if anyone is interested. The details to be negotiated in news.admin, but the basic thrust to be: 1) I will begin work on a public domain portable C compiler (to be described below). 2) A moderated group called comp.pd.compilers will be created for purposes of progress reports on this and similar future projects, bug reports, and compiler releases. [Moderator to be agreed upon; I mention moderation not because I think it is necessary to the project, but rather because I think it would be necessary in order to get the agreement mentioned below.] 3) Upon completion (to be specificed later, but to include self-compilation) of said compiler and port to some 808x machine (to be specified later, I assume what is wanted is some IBM PC), two other unmoderated groups will be created, comp.pd.discussion and comp.pd.sources for purposes of unmoderated discussion of public domain source-related issues and unmoderated posting of public domain sources. The complete compiler source will be placed in the public domain. 4) Enforcement: (On the net, the only thing that binds people is public opinion and a personal sense of responsibilty, so the following has been designed to maximally invoke such.) a) A posting of the general agreement and people agreeing to it will be made on news.announce.important (this is at least as important as other recent announcements there). b) A posting of the general agreement and list of people who originally agreed and a list of those who have since reneged will be posted on news.announce.newusers monthly. c) The initial list of people agreeing must include each member of the current quasi-net-government, i.e., the members of the backbone and moderators. d) Specifically, the backbone administrators are agreeing to carry the above mentioned comp.pd.* groups as long as they carry any moderated sources or binaries groups. e) Specifically, the moderators are agreeing that if they are moderating a group that carries sources or binaries and if the public domain C compiler gets ported to any of the machines of interest to their readers, then they will monthly announce the existance of the compiler and references on where more information can be obtained. f) New moderators and backbone sites will be expected to enter into this agreement before assuming said status. Failure to do so will automatically place them on the list of people reneging on the agreement. 5) Definitions: a) Moderation means any time when someone other than the poster must look at a message in order for it to be distributed across the usenet machines that carry the group to which it is being posted. b) A `simple portable C compiler' means the following. It compiles the basic C language as first described in Kernighan and Ritchie's The C Programming Language (Prentice Hall, 1978) with later clarifications in Harbison and Steele's C: A Reference Manual. Where Harbison and Steele describe alternate interpretations, I will be free to pick the one I like. Where Harbison and Steele describe extensions not present in the original 1978 description, I will be free to ignore these extensions (e.g., I wouldn't waste time on implementing enumerated types). Also, library routines are not part of the C language (quote pp 143 of K&R: ``Input and output facilities are not part of the C language''), so for the most part, I will leave libraries to people more intimately familiar with specific target machines and only supply with the compiler what I view as a minimal set. The general strategy will be to use an intermediate language similar to INTCODE and OCODE (described in BCPL: The Language and Its Compiler, Richards and Whitby-Strevens, Cambridge Press, 1982 and Richards, The portabilty of the BCPL compiler, Software Practice and Experience 1, 2 (1971)). The compiler will be written in two parts: 1) to translate C into the intermediate language and 2) an assembler/loader will be written to combine files in the intermediate language into executables for the target machine. So, the bottom line is I will trade you a compiler for a set of groups that will hopefully obsolete the binary groups. WHAT DO YOU SAY???? ------- BOB (webber@aramis.rutgers.edu ; rutgers!aramis.rutgers.edu!webber)