Naming things

A recurring problem is thinking of appropriate names for new technologies. When you are the first company to come up with a new idea and implement it, you have to evangelise the concept as much as the product. Hopefully, once other companies see the potential of what you are offering then they will jump on your bandwagon and together you will create significant momentum and more or less agreed terminology will become commonplace. However, that’s for the future: thinking of a handy moniker to stick on your new technology is not always easy.

To take some recent examples, IBM has been successful with the concept of cognitive computing though I have not yet seen many companies jump onto that particular bandwagon. Conversely, I am not altogether happy with “self-service data preparation platforms” because a) it is a mouthful and b) there are/were pre-existing data preparation products from companies like Actian and Teradata, that not only don’t have the self-service capabilities of products like Paxata or Trifacta but don’t have their scope either. Somewhat more confusing, because its terminology overlaps with something completely different is Alation, which I recently wrote about here. In order to differentiate its offering from those of self-service data preparation platforms it refers to its product as a data accessibility platform. Bearing in mind the excellent work that my colleague Peter Abrahams does with respect to accessibility, I am not exactly enamoured of this nomenclature.

Of course, the alternative is not to try. As an analyst that makes my life difficult: it is much easier if I can put a label to something. I can define the label and then relate a product to the label. A sort of reuse, if you like. I currently have this particular problem with a product called Javelin, which Grid-Tools is about to release (it’s already available for existing users) as an extension to that company’s Agile Designer. Now, Javelin is a test automation framework. That is, it generates tests and runs them. That’s fine, there are a number of those in the marketplace, even if they may not be as sophisticated as Javelin (less automation, more scripting). However, that’s not the issue. What is the issue is a name for the sorts of test that Javelin addresses, which are things like user interface testing, flat file testing, testing of SOAP and REST calls, testing of message queues, and so on and so forth. In other words, testing of things that are not specific to a particular programming language.

There isn’t a name for this sort of testing that distinguishes it from testing that is language specific. So I’m going to give it one: language independent testing as opposed to language dependent testing. There, not so hard after all.