From mboxrd@z Thu Jan  1 00:00:00 1970
X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me
X-Spam-Level: 
X-Spam-Status: No, score=-1.3 required=5.0 tests=BAYES_00,INVALID_MSGID
	autolearn=no autolearn_force=no version=3.4.4
X-Google-Language: ENGLISH,ASCII-7-bit
X-Google-Thread: 103376,1efdd369be089610
X-Google-Attributes: gid103376,public
From: dewar@merv.cs.nyu.edu (Robert Dewar)
Subject: Re: gnat-3.10
Date: 1997/06/17
Message-ID: <dewar.866559209@merv>#1/1
X-Deja-AN: 249094601
References: <m2soz7parq.fsf_-_@devo.ridgecrest.ca.us>
     <dewar.864929078@merv> <m2u3jksdnn.fsf@devo.ridgecrest.ca.us>
     <dewar.865690628@merv> <m2vi3kpuay.fsf@devo.ridgenet.net>
     <dewar.866126619@merv> <m2d8prd2ty.fsf@devo.ridgenet.net>
     <dewar.866213404@merv> <dewar.866214689@merv>
     <m2d8pms722.fsf@devo.ridgenet.net> <5o480g$tqk$3@mdnews.btv.ibm.com>
Organization: New York University
Newsgroups: comp.lang.ada
Date: 1997-06-17T00:00:00+00:00
List-Id: <comp.lang.ada>


Dale says

<<More importantly, ACT and Cygnus, among others are an absolutely
critical experiment in the software industry. They will answer the
question, "Is there any middleground between students/hobbyists
hacking away at GNU code and commercial software houses like Micro-
soft?" Only a little bit of GNU code has made it to the mainstream,
and it's still pretty well untrusted in the business world, and
it's all because of "support." Even though informal Usenet support
from the author/maintainer is usually better than any company,
business demands a business>>

The reason that business demands a business is that they need to be sure
they have support when they need it, and that they get to decide when the
need for support arises. It is interesting that more and more of our support
is not related to bugs in GNAT, but rather, as people build more complex
applications using GNAT, and using the more complex features of Ada 95,
with giving people assistance in the use of GNAT. Also a lot of our
business relates to special purpose needs.

Yes, sometimes informal Usenet support can be quite effective, but on the
other hand, it can sometimes be highly unreliable. In particular, people
giving free support can quite reasonably take the attitude that you should
spend a lot of your time creating an absolutely clear example, which can be
a very time consuing excercise.

We quite often deal with large programs which don't seem to work at some 
particular point, and we dig around. Sometimes the digging shows up a GNAT
bug, sometimes a bug in the customer code, sometimes just a misunderstanding.
This digging can of course be very time consuming.

Also, I think the informal model works best when you have a program that is
of a level that one person can maintain it as a spare time hobby. Free software
has faired quite well, even without formal support in this kind of situation.

However, this model does not work well for large complex pieces of software,
like GCC, or GNAT, where you need a team of people working full time. In such
cases, the Cygnus or ACT model seems more workable (a similar situation is
true with Linux/GNU where more than one company has sprung up to support
this system.