 n the best software organisations, the relationship between developers and testers is an emotionally charged political minefield that transforms itself into a pitched war zone at the time of delivery or shipment. The reasons for the edgy relationship and overt antagonism are not far to seek. Psychologically, the skills needed for developing and testing are opposed to each other. Developers are believers in the workability of their code while even the worst testers are agnostics. Testing, by its very nature, is iconoclastic - questioning, investigating and probing the supposed operationality of code. Not surprisingly, the relationship is marrred and marked by an "us versus them" attitude where charges and countercharges are thickly traded, compounding deadline pressures and transforming the collaborative nature of the software development cycle into a competitive one.
In India, the coder/tester conflict acquires an added intensity, as testing is most often a second choice, a last resort for candidates and a refuge for "average" and "mediocre" engineers. The best of breed engineers opt for development as a career and believe testers are dispensable. At the lower levels in the development and testing teams, the antagonism acquires the face of a class conflict between haves and have nots. Consequently, developers accuse testers of harbouring an inferiority complex while most testers brand developers as arrogant. A software architect working with a networking company believes "six out of ten testers are pathetic" and a senior tester working in the same organisation dismissed "four out of ten developers as 'junk', who are merely fooling around.
Interestingly members from the two communities accuse each other of incompetence and lack of domain knowledge. "Testers in India lack domain knowledge. Unlike testers in the US, they do not file new and critical bugs. The same set of tools are used to test every product and testers often report non-existent bugs rather than critical ones, wasting crucial time on critical delivery days in the process," says Sanjay working with Financial Banking Software company. "He elaborates, "Even when one tells a tester that certain software embedded in ATM cards has to be tested for a future date, testers report back saying the software does not work on the current date. Days before a crucial release, we have spent hours on identifying the bug till one realises the tester has made a mistake, " he adds. "At times the product is tested on a platform for which it was never intended," expostulates Ramana from a Networking Company. Counters Surendra a tester with the same company: "Testers are accountable for bugs they file. Developers suffer from a frog in the well attitude. Their knowledge is limited to the component they code for, unlike the tester who knows the work of every developer in the team."
The ego tussle and low opinion between testers and developers is exacerbated by the nature of the software development cycle in many organisations. Ideally, most organisations swear by the V model of software development where the various stages of the development cycle: Requirements, Analysis, Design, Implementation, Testing, Acceptance, Handover and Success are well-integrated operations. A prerequisite of the V model of development is integration of testing from the requirement phase to the last phase. In the "get it done and get it out" mode every organisation however skips stages. In practice, the V model of development is often bypassed in favour of the spiral model (where project details are discussed in a back and forth manner that leads to communication gaps) or the waterfall model (where testing constitutes the bottom-end of the software development chain.) As the test team is not involved in the initial stages of the development process, knowledge gap between testers and developer on the specific function and nature of the project lead to projects being tested for non-existent functions. On the other hand an extensive project cover-up is launched if system testers identify an architectural flaw in a product at the last stage, says Sanjay Senior developer at Financial Banking Software company. Also, the lack of integration makes it difficult for the tester to identify the exact owner of the problem, with few developers ready to accept ownership. A relatively inexperienced tester is virtually made to run from pillar to post, especially if he finds a critical bug close to the deadline, admits a project manager. Experienced testers may play similar games with developers. " A tester can make life hell for a developer by continuously filing bugs," says Surendra
Organisational management lends encouragement to the adversarial relationship with a definite tilt towards development teams. Senior technical management by and large comes from the ranks of developers and believes testing is a second hand job. Technically, in a standoff situation between developers and testers, it is the later that has the power to give the go ahead but most often it's the development team that holds sway. It's a no-win situation for testers, branded bad guys when they find defects and blamed if the software goes live with problems. Most organisations pay lip service to the complementary role played by testers and developers but in reality do not believe that each team brings a perspective that the other may lack. Most organisations believe, testers are dipensable as developer can bug fix but a tester cannot write code. The fact that engineers from the testing community were harder hit than developers lends credence to the theory. Also salaries and incentives earned by developers and testers reveal a wide discrepancy. Developers earn twenty and thirty per cent higher salaries than testers.
So is there a solution. Ninety per cent of the developers interviewed testers should learn debugging skills while testers believe that developers need to get real and expand knowledge of the domain. May be developers and testers could rip off a lesson from Thomas Elliot's literary masterpiece The Wasteland that he dedicated to Ezra Pound, the mentor of modernist writers, for slashing its length to one-third the original size. Elliot called him the "master craftsman."
Email this article | Respond to this article
|
