Re: exchange with Richard Stallman

[Michael Alan Dorman <mdorman@lot49.med.miami.edu>]

In message <m0u0Zm2-00063cC@mongo.pixar.com>, Bruce Perens writes:
>The following is an exchange I've been having with Richard Stallman over
>yesterday and this morning. My conclusion is that what he is asking for
>is not stuff that it's reasonable for us to do, that that Debian can not
>work together with FSF.

Well, assuming that he is unwilling to bend, yes.

>1. We need to have packages with unstripped executables.

Um, no.  Simply unnecessary.  Silly and wasteful.

What I don't think RMS is considering is that Debian is not targetted
at (what I would say is) the FSF's traditional "market" of experienced
sysadmins looking for a higher quality tool, and/or those ethically
attached to the concept of free software.  Those people (the ones who
are used to the drill, or are specifically looking to help improve the
quality of tools by pounding on them) are already using the FSF stuff,
and they're compiling it themselves.

Debian, on the other hand, seems targett?ed at users with varying
levels of experience who are looking for a Unix-like system with good
tools and sophisticated packaging.  I doubt they are, in the main,
going to be recompiling packages (otherwise, why are we doing all this
work?), and are not, I think, generally going to be looking at this as
an opportunity to help refine and debug tools.

In light of this audience, Debian has an excellent (though maybe not
as carefully tracked as might be useful) bug-tracking system, so that
users can report bugs to someone who can act as "first line" support
to try and determine if its a configuration issue, etc.

_That_ person (who is already recompiling things to create packages in
the first place) is the person who is most likely to be building and
using debugging versions of packages as they try and figure out if
there are problems with the upstream sources, etc.

Given that model, it's wasteful of disk space and user's download time
to include debugging information in the main-line code.

>2. We need to encourage people to write documentation in Texinfo.
>[...]
>Can we agree on a compromise whereby HTML is used as a universal least
>common denominator format for delivering documentation, but we both
>encourage people to write in Texinfo?

I've reached a peace with info-mode and info, reminiscent of the peace
I've reached with vi---which is to say that I prefer Lynx.  But what
he says doesn't bother that.

As far as writing documents, I appreciate the fact that Texinfo allows
the production of printed documents, but I'd rather use LaTeX or HTML
and one of the various packages around for converting from one to the
other.  Texinfo gives me a headache.

But, if RMS can induce upstream authors to produce documentation in
texinfo, more power to him.

>3. The installation procedure needs to be further simplified.  This is
>actually not a single issue, but a collection of subissues.  For
>example, the X installation needs to be clearly documented.  The
>choice of time zones should be classified in a simple and consistent
>way.

And Bruce is working on it (and doing very well---thanks to the SCSI
bugs that've been cropping up in later kernels, I've been re-doing the
installation on our web server quite a bit), and it's much better.
There's still more to do, but it's much better.

>4. We need a complete separation between what we distribute
>and all non-free software.

I understand that this is something the FSF might require, but I think
the burden is also on them.  It would be polite of us (which is
probably to say, Ian) to create some way to mark packages and then
allow the FSF to kind of direct those, but if they're the ones
striving for ethical purity, it's got to be their burden to bear.

Mike.
--
"Don't let me make you unhappy by failing to be contrary enough...."