>>>>> On Sat, 26 Aug 2000 23:38:42 -0500 (CDT), Dave Rolsky <[EMAIL PROTECTED]> said:
> So my question is how do I best promote this thing? I think its good for
> Perl because its got the potential to make database related development in
> Perl much much easier. How did previous major apps (mod_perl comes to
> mind) get initial visibility?
In a nutshell:
You need a good idea. You need to discuss the good idea publicly so
others can weigh in their views and help you get out the showstopper
flaws[1]. You need to provide the implementation that works at least a
little in the way the thing is intended to work some day. You need to
show continuously that you believe in the idea. You need to focus on
the idea and set the frame. You have to commit to supporting your idea
for a "long" time, i.e. at least several years. You need to know your
competitors and know why you cook up your own meal. The best bet on
this is when you can say that you desparately need the thing yourself
because it solves your job of "Doing XXX in an YYY way". You need to
do the initial advocacy yourself but not too early (no hype!). People
who want to "do XXX in an YYY way" will start joining you. Be happy
when they are more than 10. From that day you need not do advocacy
anymore, just be helpdesk and coder whenever the need arises, this is
what you have promised (or should have promised). Be ready to toss the
thing and rewrite from scratch if it turns out to have design flaws,
but try to provide backwards compatibility in every respect.
There are many reasons why this doesn't work out as in mod_perl: the
potential impact on the industry may be too low. The difference
between your product and another product may not be significant or
visible enough. Your expectations in terms of how much you get back
from the community may need some adjusting (Thou shalt not expect
anything). Your code may be hard to read, your documentation may be
hard to read or the combination of code and documentation may be
difficult to navigate for newcomers. None of these things are
showstoppers per se, thay may turn out as showstoppers, but they may
just need time to be solved.
And always keep in mind to release early, release often.
--
andreas
[1] Discussing without people is possible too: write a talk about it.
