[lug] WordPerfect for Linux is back --- more or less
efm at tummy.com
Fri Apr 30 11:29:56 MDT 2004
* On 2004-04-30 10:57 John Starkey <lug at breezedev.com> wrote:
> > > I'd like to see this succeed: I am, and have been, of the opinion that
> > > WP was the best word-processing program of all time. So I'd like to see
> > > this work, in spite of the sarcastic tone of this e-mail.
> > >
> > > I'm just not sure about paying someone to test their software for them.
> > > Shouldn't it be the other way around?
> $30 sounds pretty reasonable for something that you clearly think sucks
> less. If you're that psyched to see development continue, give them some
> compensation. Every software in the world could still be considered beta
> if "stable == perfect".
In product development, it's important to be solving the customer's
problems not just the developer's problems.
Open Source software has grown very rapidly and very effectively through
solving developer's problems. The big opportunities are in solving everyone
But, I see there is often a weak feedback loop between everyone else and
developers. Without feedback, developers will prefer to work on problems
they have a clearer understanding of, more praise for solving, and higher
reputation points for taking care of, than problems that don't feed into
those personal and community reward cycles.
In a market economy though, we have a mechanism for rewards that connects
people who have problems with people who can solve problems.
As someone who would like to solve a broader range of customer's problems,
it's extremely difficult to estimate the actual need of customers for a
proposed solution. It's very easy to get suggestions, but it's fairly
difficult to get actual statements of need. For example, we often get
suggestions for our Linux Distribution KRUD that are very specific and very
clear. But they are often too expensive for us to be able to accomodate,
because the problem they're solving has too few people interested in the
solution for us to be able to do it for them (assuming it's a problem that
we wouldn't be intersted in solving for ourselves).
I see asking for payment for beta software as a way of getting that crystal
clear reading of the willingness of people to receive a solution, in a way
which is both effective and honest. It may not be positively effective,
that is the process may not result in a shipping final product, because
demand was not really there for the solution. But it is effective, because
both the developer and the customer have made commitments to each other
which they are prepared to stand by. And it's honest, because the developer
isn't claiming to be shipping something that is more finished than it
actually is (vaporware is dishonest).
There is a need, for transparency in this process. If both develpers and
customers were able to see how much interest there was in solving this
particular problem, then they'd both be able to react more effectively. On
the developer side, knowing how many people are willing to pay $x for a
solution gives them feedback in choosing what projects to work on. On the
customer side, knowing that only 3 other people are willing to pay $10 for
a solution could encourage them to either rustle up more interest, or to be
more generous in the amount offered for a solution, or to realize that at
this point in time, the problem they need solving is just not valuable
enough to attract developer interest.
There have been several attempts at developing this sort of feedback loop.
The one that comes to mind is collab.net
I'm not sure why they haven't been more effective. If I recall correctly in
several years collab.net only had a couple of projects push through to
It's a real dilemma. Developers need to know what problems people have,
and people need to be able to attract developers attention to their
problems. But without the mechanism of real feedback, they won't be able to
Regards, tummy.com, ltd
Evelyn Mitchell Linux Consulting since 1995
efm at tummy.com Senior System and Network Administrators
More information about the LUG