piątek, 30 maja 2014

From Here to Eternity: Pragmatic, Independent Testing

(You'll find more on this subject - the "presentation" mentioned below - on EuroSTAR blog in a few weeks)

Testing best in practice requires understanding the history of testing and of software engineering. If you do not, you will not be able to recognize recurring patterns, you will be led astray by what looks new and revolutionary but is not. Knowing the history of testing, you are able to see the mechanism of its development, its evolutionary logic, and be better able to see future trends. There can be no real context-driven testing unless you can see the whole context - including the historical context!

My presentation will describe both the history of testing as well as attempt to identify the mechanism of its evolution, and use it to predict future trends.

1. Ancient times 1940-1955. The era of geniuses. Software developers were exceptionally gifted people, many of them Nobel-prize-level scientists. This artisan period worked because of the exceptional skills of these developers, small-scale software and low business pressure. Testing was very much embedded in the development process in a natural way.

2. Medieval times 1955-1968. The era of emerging engineering. The volume of software produced could no longer be done by geniuses only - more ordinary programmers appeared. Software development was no longer programming only - requirements engineering, design and architecture, testing, integration and maintenance became more and more separate disciplines. IT was no longer a natural part of computer science, it became more and more interdisciplinary. Unless rules and processes were followed, IT projects failed.

3. Modern times 1968 - 1990. The size and importance of software projects grew, the failure rate increased, the need for organization, processes and specialization became obvious. Trial and error as a method of coping with uncertainty and complexity was gradually abandoned and predictive, systematic, bureaucratic ways preferred. Software testing became more and more accepted as separate discipline, formal standards appeared and were followed. Testing schools: analytical, standard-based and quality-oriented dominated.

4. The Great Schism 1990-2005. The context-driven school appears and gains ground, foundations of agile framework (as XP) are laid, but at the same time very formal ISEB / ISTQB approach starts to dominate the world.

5. 2005 - today. Modern times. The schism increases - exploratory and agile approaches become to some extent extreme and almost religious rather than rational, while ISTQB strangles its opposition, becomes political and joins the PMI/IEEE/PRINCE2/ITIL super-standards league. At the same time, common-sense, middle way approaches appear, like post-agilism... but nothing similar is yet visible for testing. Or is it? This presentation shows possible beginnings visible today and tries to predict future test trends from them.

Finally, some remarks on the rules that drive this development will be made: on social science and cultural anthropology: how new schools and new memes appear and change, and evolve.

Brak komentarzy:

Prześlij komentarz