Fog Creek Software
Discussion Board

Definitive software testing text

In my place of work they've recently (in civil service terms) created a QA department completely separate from the development teams (which I whole-heartedly applaud).  Unfortunately, it's been lumped in with the ISO and Project Management teams (which is a whole other rant altogether).

The problem is no one, from the assistant director to the manager to the testers, has any real, formal software testing background.  Test plans tend to include some vague ideas of functional testing and loading testing, but they generally waste a lot of time on stuff they shouldn't be doing, and miss a whole lot of other stuff completely.  It's not a good situation, and there's no way to change any of the players involved, so I have to work with what I've got.

My question is:

If there were a single comprehensive (but easy to read) book I could give to every single person in this division explaining the purpose of, fundamentals of and practical implementations of the software testing process, what would it be?

Dunno Wair
Thursday, November 21, 2002

Check out:


both by Cem Kaner, et al

Thursday, November 21, 2002


also, guides on good software development in general, such as McConnell's several books, if followed, will make it possible to get a lot more value out of the testing process than you otherwise would.

Done a bunch of testing -- it's what I do -- and those doing testing are almost entirely dependent upon the work products of everybody "upstream". So, for example, having a hierarchically structured, meaningfully detailed and unabmiguous set of requirements with full traceability back to operational needs (in gov't speak, you might see terms like COIC, OTMOPS, and MNS at that level), will add immeasurably to the value a testing effort can add to the situation.

Unfortunately, in too many cases, getting things to that wonderful state is not anything the QA / QC folks are empowered to meaningfully affect. Even having a separate department doesn't normally do it, in fact can make the problem worse because it adds a layer between the QA/QC folks and the sources of the info, meaning that QA/QC folks can be left out of the critical information loop during the majority of the early stages of a project.That reduces their ability to devise tests well, and in time.

So, while having all these wonderful things happen in the "upstream" part of the development cycle will not help you have good testing if the testing itself is crappy anyway, It can, as I said, add very greatly to the value of what otherwise would be good testing.

So, it'd be a good idea to see if you have any luck making headway in that arena as well. McConnell's books are just one example of good things such folks ought to be doing.

Good luck,

Thursday, November 21, 2002

Managing the Testing Process by Rex Black
Publisher: Microsoft Press

This book would well suit the person who is going to take over the management of software testing in a company that had, perhaps, not previously provided a proper test environment. 

If your company lacks an adequate testing system (infrastructure) then this book could provide the basis on which one could be developed.

one programmer's opinion
Friday, November 22, 2002

*  Recent Topics

*  Fog Creek Home