Xref: utzoo news.groups:28586 comp.sources.d:6621
Path: utzoo!news-server.csri.toronto.edu!rutgers!sun-barr!cs.utexas.edu!wuarchive!uunet!crdgw1!crdos1!davidsen
From: davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr)
Newsgroups: news.groups,comp.sources.d
Subject: Re: CALL FOR VOTES -- comp.sources.reviewed
Keywords: news groups source review
Message-ID: <3229@crdos1.crd.ge.COM>
Date: 7 Mar 91 17:18:56 GMT
References: <1991Feb28.160555.8446@sparky.IMD.Sterling.COM> <1991Mar1.150759.11832@rick.doc.ca> <1991Mar3.035928.28569@sparky.IMD.Sterling. <1991Mar06.151615.3212@scuzzy.in-berlin.de>
Reply-To: davidsen@crdos1.crd.ge.com (bill davidsen)
Followup-To: news.groups
Organization: GE Corp R&D Center, Schenectady NY
Lines: 36

In article <1991Mar06.151615.3212@scuzzy.in-berlin.de> src@scuzzy.in-berlin.de (Heiko Blume) writes:

| and _*DON't*_ forget revision numbers! doesn't tell me much if
| something is patchlevel 1, if i don't know of what revision.
| that's a pain in the ass for me all the time...
| perhaps using rcs from the beginning might be a win.

  Sounds as if this group is going to be so picky that alt.sources will
continue to be the main source of software. I fail to see why use of RCS
makes the software better or worse than software using SCCS or any of
the other programs which handle versions. Or even just a file called
version which holds the information entered by a human being.

  I don't see that making use of RCS or any other package has any
bearing on the quality of the software.

| >The sources will also have to be evaluated for portability, as much as
| >possible.  Did it compile on a variety of machines (at least those it
| >was intended for)?  This test for portability I see as one of the
| >hardest things to conduct, and I am not sure how to do it.  Obviously,
| >selecting reviewers that run on a variety of platforms will be a first
| >step.

  Let's not get carried away on portability. A clear statement of
required environment is fine, it doesn't have to run everywhere to be
useful. A submission should be tested for BSD, SysV, and in most cases
Xenix/386 and V.3.2/386 since they represent the largest number of users
of a single type of UNIX. Compilers needs should be tested on K&R, ANSI,
and gcc flavors of C.

  I would also support a requirement that all compile options be
documented by comments in the Makefile.
-- 
bill davidsen	(davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen)
        "Most of the VAX instructions are in microcode,
         but halt and no-op are in hardware for efficiency"