What ever happened to Tandem and NonStop OS 1997
I have, and as good a book as it is, it isn't perfect. I like the one about, "if a woman...
Eric Chomko) writes: Nothing is - that's our point. Still, it has a lot of good things to say. That's one of my favourites, too - although Brooks expresses it as the converse (i.e. reality vs...
No, haven't found any tools beyond the standard stuff everyone uses, ($EDITOR, cvs, spec writing etc).
The process has come naturally to everyone that has extensive experience from working in software environments with high degrees of standards penetration. Programmers-analysts with experience from open source, Internet; but also banking, finance, military etc. I don't see the same mindset in Telecom or Process control, although it should have been present there as well.
It is important that this process can loop a few times before the specification is set in cement.
The first couple of loops is a pretty people-oriented process, getting feedback etc., and things naturally solidify if there is a clear emphasis on interfaces and prototype code to implement them.
As soon as a scaffolding framework is in place real development can begin. This then has the benefits that people really can work productively on the consbreastuent parts, and the scaffolding forms an early minimal test rig.
So, at some point we convene a meeting and take a vote on every interface down the spec. "Speak now or forever hold your peace". This is around a fourth to a third of the way down the timeline of the development.
LOL, no problem, AFAIK it is out of service. If it isn't they are welcome...
A separate exercise is to build "auditing encapsulations" for the various modules, to check that they really behave as they advertise.
In some environments you may want to have two separate implementations of every module in the system. When the system clears transactions for billions of USD daily such items are vital.
-- mrr
What ever happened to Tandem and NonStop OS 1998
Alt Folklore Computers from Newsgroups