Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!hplabs!hpfcso!hpfelg!beth
From: beth@hpfelg.HP.COM (Beth Crespo)
Newsgroups: comp.software-eng
Subject: Re: testing in alternative lifecycles
Message-ID: <3440006@hpfelg.HP.COM>
Date: 24 May 90 16:12:40 GMT
References: <1990May20.154723.15140@axion.bt.co.uk>
Organization: HP Elec. Design Div. -FtCollins
Lines: 26

Keith Richards writes:
>Alternatives to the `waterfall' lifecycle such as the `spiral' model,
>evolutionary prototyping, incremental development etc. are gaining
>in popularity.  Ideas for software testing however seem to be centred
>still around the `waterfall' model.
>
>Is this fair comment?  Is anyone working on how verification, validation
>and testing fit in with various alternative lifecycles?
>
>All information/ideas/pointers welcomed.  Thanks.
----------
As I see it there are two kinds of testing implied by the spiral model.
1) For each round you have a testing phase to ensure you have completed
   the current spiral. 
2) When you have completed a set of spirals there is an extra phase
   of testing to ensure the entire system works as expected.

If you view each round as a pass through the waterfall lifecycle, I think
you need to add one more 'confidence' testing phase to the last waterfall.
The sum total of test time is probably less than implementing and
testing as a 'big-bang' when coding is "done". 

Can anyone relate experiences using the spiral model on application products?

Beth Crespo, Electronic Design Division, Hewlett-Packard 
These opinions are mine, Hewlett-Packard may or may not agree with them.