Re: Text processor

[Alan Shutko <ats acm org>]

Robin Miller <roblimo@home.com> writes:

> The GNU/Linux problem, right now, is that when a mere user asks
> (begs) for a tool that fits a job, the developers instinctively
> respond with reasons why existing tools are adequate and/or with
> suggestions on how the user can change to fit the tools, rather than
> aking the users what tools they need.

A lot of the time, people haven't considered everything about existing
tools when they ask for something new.  Usually, the request is vague
enough that existing tools (or modifications to existing tools) sound
like they could meet the users needs.  Since writing something
completely new is much more work than modifying something else to fit,
it's natural that developers would like to see if it's absolutely
necessary that something new be written.

The best thing for a user to do in a case like this isn't to give up,
as you seem willing to do.  It's to ask for _specific_ things that you
need, and to be willing to accept a solution that would do everything
that you ask for.  For example, I could put all the tools I saw you
ask for into Emacs, but you've written that off as impractical for
day-to-day use.  Other people could write scripts for Nedit for you,
but you've written that off too.  Why are either of those solutions
somehow less than a completely new piece of software?  I can tell you
right now which would be done quicker.

-- 
Alan Shutko <ats@acm.org> - Looking for a job in Long Island!
Check http://rescomp.wustl.edu/~ats/ for a resume.
You will be married within a year.