In general, the way for a problem to be fixed is for someone who knows
how to fix it right simply to do so. Any solution that involves
discussion should be avoided except as a last resort, because it is
inefficient.
Dale Worley posted it to a mailing list in '89, and I asked him where it came from:
Date: Mon, 18 Sep 89 21:55:13 EDT
From: drw@BOURBAKI.MIT.EDU
To: don@brillig.umd.edu
Subject: An interesting bit of philosophy from RMS
From: Don Hopkins <don@brillig.umd.edu>
From: rms@AI.MIT.EDU
In general, the way for a problem to be fixed is for someone
who knows how to fix it right simply to do so. Any solution
that involves discussion should be avoided except as a last
resort, because it is inefficient.
Interesting all right! When/where and in what context did he say
this?
It was on one of the Emacs newsgroups, where lots of reasonably
uninformed discussion had broken out on how to resolve the fact that
VM and Gnus (I think) interfered with each other, because they both
used the overlay-arrow mechanism.
Dale
Searching Google Groups for the first sentence found the unexpurgated version:
Date: 9/11/89
From: rms@ai.mit.edu
Newsgroup: gnu.emacs
Subject: No more on overlay-arrow-position
I don't think most Emacs users want to participate in a discussion
about how to solve a fairly obscure problem in how Emacs is implemented.
This kind of discussion doesn't help me solve the problem. I know you
mean well, but you aren't experts. Most of this discussion consists
of proposals that are completely wrong, or won't work, followed by
refutations and counterproposals. When I decide to fix this for
version 19, I won't read the discussion; I'll just fix it--it will
take less time.
In general, the way for a problem to be fixed is for someone who knows
how to fix it right simply to do so. Any solution that involves
discussion should be avoided except as a last resort, because it is
inefficient.
Remember, the purpose of info-gnu-emacs (and its repeater newsgroup,
gnu.emacs) is to carry the information that *every Emacs user will
want to know*. There's nothing wrong with non-experts discussing how
Emacs bugs might be fixed, but please don't do it here.
RMS also doesn't like people posting baby announcements to mailing lists that are clearly intended for making dinner arrangements (and the baby in question is 27 years old now):
In general, the way for a problem to be fixed is for someone who knows how to fix it right simply to do so. Any solution that involves discussion should be avoided except as a last resort, because it is inefficient.