Your SDLC blog has a bad anchor tag originating in a repeated "http://"
_lifecycle.htm">Software Development Life Cycle (SDLC)
Aside from that, I don't know if I agree with Bruce Bahlmann's template for
software development, particularly if it endorses anything approaching the
waterfall approach to software development. Typically both customers and
developers learn too much during the course of development for them to be
satisfied with their initially agreed upon plans. This is particularly true
with new products with more up front uncertainties, both respect to
requirements and development. Frequent updates to established products can
get around this problem by giving both customers and developers belief that
their new ideas can be incorporated in the near future.
I also worry that Bahlmann's template seems to call for software developers
to estimate time to completion for projects before they understand the
problems or solutions well enough to make accurate estimates. In my
experience, developers are always encountering subtasks where the estimated
times to completion are bimodal; either your original hunch works or you
have to find a whole new strategy. There is always enormous pressure to use
the optimistic estimate as both managers and project management tools (e.g.
Gantt charts) have difficulty integrating subtasks with bimodal distributions.
I personally favor development environments that encourage early development
of prototypes. I favor continuous refactoring, often a hard sell to management because the strategy seems like an invitation break a working system without adding new functionality. Conceptually I like the idea of test driven development, but have never developed a serious project under that philosophy.
Respectfully and with love,
--Pop
