Re: backward compatability of tools

[Paul Eggert]

"Dr. David Kirkby" <address@hidden> writes:

> If the developers insist on dropping backward compatibility, would it
> not be better to at least do a test for old hardware/software

I wouldn't bother myself, because such a test would be a maintenance
hassle.  We don't have the resources to keep track of all the quirks
of ancient hardware.  We have enough trouble keeping track of the
oddities of live systems.

At some point, when a system is no longer active, it is better for all
concerned if we simply drop support for it.  The reality is that GNU
developers have limited resources, and efforts expended on inactive
systems draw needed resources away from more important areas.

> Should well written software not respond with informative messages, no
> matter how invalid the user input?

This is not a question of invalid user input.  This is a question of
bugs in the operating system.  Autoconf-generated scripts, no matter
how well-written, cannot survive arbitrary OS bugs.