[Cryptography] I don't get it.

[dan at geer.org]

 | As for probable causes:
 | - underfunding;
 | - ever-changing requirements, as the users will imagine new
 | possibilities as the project matures;
 | - managers/customers thinking that software *development* is a sort of
 | conveyorbelt, just replace a programmer with the next and the work will
 | continue;


You know, yours is a point worth repeating; it affects software
projects as it affects armies: mission creep.

Would it be worthwhile to banter about various analogies here?
Clinical trials of new pharmaceuticals where efficacy and safety
get conjoined evaluation?  Jet engines where an "airworthiness"
certificate is not for the engine but for what we here would call
the build environment?  Probably been done so point me at something
to read, but a gauntlet that precludes mission creep might be a
useful construct.  I've certainly been involved in failed software
projects involving mission creep (especially when a weak-willed
and/or maniacal CEO listens to salesmen who say "If we only had X,
then I'd be able to double my quota!").

On the other hand, we're headed into a rat-hole...

--dan