Frank Wales
Sat, 29 Sep 2007 18:58:12 -0700
On 09/30/2007 01:55 AM, Howie Goodell wrote: > Why am I impressed by such an obvious process? Because I've seen too > many failures in the bad old corporate software development process > that missed most or all of the above! The process was driven by > widespread perception of customer needs. Nobody asked, "what will > this do to the sales of our existing products?" Everyone who cared to > could make suggestions and act on them. Believe me, these are huge > advances!
Actually, shhh! Don't tell anyone, but many pieces of software
that I've worked on came about this way:
+ somehow become aware of problem in need of a solution
+ experiment with readily available tools to explore
the problem space and candidate solutions
+ wave results in front of users and see if they bite
+ repeat until either everyone gets bored, or
a better solution appears from somewhere else,
or other higher-priority tasks claim your time
+ eventually, when it becomes apparent that you need to
get professional, make the software of adequate
quality to ship, or otherwise hand over
+ in the process of getting professional, do the
things that make QA, support and marketing people happy
A non-trivial amount of the software I use daily (Linux,
emacs, OpenOffice, ...) has probably come about using all kinds of
random variations of this approach, in addition to more tidy methods
everyone gets taught in school.
I just though that the process in the case I referred to would
be particularly good at giving some people conniptions:
+ no-one really clearly specified anything, leaving
the programmers to figure out what might be useful,
thus annoying anyone who thinks programmers shouldn't
be allowed to design things by themselves
+ test cases got written near the end of the process, which
could wind up Agile/XP people
+ specification got written last, and only to satisfy external
QA paperwork requirements, which might make numerous
software engineering advocates turn seven shades of purple
+ and, of course, this whole process would probably make
ISO-9000 advocates weep openly
Well, something has to.
--
Frank Wales [EMAIL PROTECTED]
----------------------------------------------------------------------
PPIG Discuss List (discuss@ppig.org)
Discuss admin: http://limitlessmail.net/mailman/listinfo/discuss
Announce admin: http://limitlessmail.net/mailman/listinfo/announce
PPIG Discuss archive: http://www.mail-archive.com/discuss%40ppig.org/