Newsgroups: comp.sources.d Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!decwrl!deccrl!bloom-beacon!bloom-picayune.mit.edu!athena.mit.edu!jik From: jik@athena.mit.edu (Jonathan I. Kamens) Subject: Re: v01INF1: Status - Status of comp.sources.reviewed Message-ID: <1991Apr15.171204.7235@athena.mit.edu> Sender: news@athena.mit.edu (News system) Organization: Massachusetts Institute of Technology References: <6787:Apr1420:38:3991@kramden.acf.nyu.edu> <1991Apr14.232818.15851@athena.mit.edu> <12409:Apr1513:14:4591@kramden.acf.nyu.edu> Date: Mon, 15 Apr 91 17:12:04 GMT Lines: 67 In article <12409:Apr1513:14:4591@kramden.acf.nyu.edu>, brnstnd@kramden.acf.nyu.edu (Dan Bernstein) writes: |> The gamma testing process found lots of problems that I did not |> anticipate, but that happened *before* I published the sources. First of all, your gamma testing process may not have found all of the problems. Second, the review process provided by c.s.r is, in my opinion, meant to be similar to the "gamma testing" process you mention. Many people don't have access to a lot of people on lots of different systems to ask them to test their software. They can send it to c.s.r and ask us to test and review it. If your gamma testing process was useful, and c.s.r provides a service to people which is similar to your gamma testing process, then it seems obvious to me that c.s.r is useful. |> I again invite you to explain what you would have done with the package, |> exactly as it went out on comp.sources.unix. You're saying that you |> ``may have found problems''; I say you should come down out of the |> clouds and try to find problems. The package is right there! Make an |> example of it! Have a blast! You persist in your belief that your one package, in your particular situation, can be used as a general proof of properties that would hold for all packages in all situations. I've already pointed out this gaping flaw in your argument once; you ignored me the first time. Are you going to ignore me again? |> [ in reference to some c.s.unix package ] |> > C.s.r would not, in my opinion, have allowed that mistake to |> > happen. |> |> Your guidelines give ``nlist'' as a prototypical package name, and |> you're saying the group is going to catch mistakes? Irrelevant and juvenile comments like these do not lend any credence to your point of view. I KNOW that the package I mentioned had mistakes that would have been caught by the c.s.r review process, because I KNOW that the package would not compile in many configurations as it was provided, and the first thing c.s.r reviewers would try to do would be to compile it. I also KNOW that they would have criticized the documentation for being sparse and vague, and asked the author to clean it up. I know all this because I would probably have been one of the reviewers for the package, and I would have done all of this. And I respect the other reviewers enough to believe that even if I had not been one of the reviewers, one of them would have caught the same mistakes that I did, especially since all I had to do to catch the mistakes was try to compile the program. You have made the generarl assertion that c.s.r cannot provide a useful service. As proof of your general assertion, you have offered a particular case for which it holds true. That is the first flaw in your logic. Despite the fact that it is unnecessary (in a strict debate sense) for me to refute your assertion since it is not logically valid, I did so by presenting a counter-example; one counter-example disproves a general assertion, as I am sure you are aware. Your attempt at refuting my counter-example consisted of the brilliant logical maneuver of making fun of the submission guidelines for the newsgroup. My admiration for your technique grows by leaps and bounds. -- Jonathan Kamens USnail: MIT Project Athena 11 Ashford Terrace jik@Athena.MIT.EDU Allston, MA 02134 Office: 617-253-8085 Home: 617-782-0710