Has 'beta' lost its meaning?

And then there's Google, whose development model(s) have been described by its own practitioners as nebulous blobs of arbitrary nomenclature. Gmail, as many of its users will recall, was marked with a "beta" banner for the better part of five years. In the meantime, a number of folks, some of whom actually do test Google's Chrome browser rigorously, wrote in to remind us that Google maintains a development ("dev") track separately from its beta ("beta") track, and that we were confusing a dev build with a beta build.

One independent explanation put it this way: The beta builds are for public beta testing, whereas the public dev builds are for private beta testing.

What's astonishing here is that the explanations I've received from various sources at Google has contradicted our own readers, and at times defies any kind of logic I've been able to concoct. Suffice it to say that, for the most part, Google treats every product and service it offers, in whatever state it's in, as under development.

The relative level of stability determines whether a product gets the "beta" banner, and everybody should expect some Google software to develop glitches at some point. And every Google project has its own methodology for making that determination, if it has one at all. But if it says "beta," folks should take that to mean it's more glitch-free and debugged than something that says something other than beta, but less glitch-free than something that doesn't say anything at all.

Does beta mean "free?" That depends. Does beta mean "testing?" That depends. Does beta mean "feature complete?" That depends. Does beta mean "ready" or "not ready for prime time?" That depends. On what does it all depend? That depends.

There's a saying attributed to Sherlock Holmes which has, in more recent years, become more generally attributed to Mr. Spock: When you eliminate the impossible, whatever remains, however improbable, must be the truth. The problem with the modern meaning of "beta" is, if you trust everyone with direct responsibility or involvement in the matter to eliminate the impossible, nothing remains.

As we proceed, either headlong or tail-first, into an era where software is delivered in increments and hosted on the Web or in the cloud or through some nebulous recreation of the red tape we faced when surviving the public education system, there appears to be less and less incentive for any kind of software, in any shape or form, to ever be done, complete, finito. (Ironically, not even Finito.com is ready.) There's a big reason why this will be a problem, not just for Betanews but for the entire industry: If software publishers are incapable of determining for themselves just what your role is supposed to be, as consumers and/or testers and/or co-developers, then it may become impossible -- if not wholly improbable -- for them to develop a modern business model for software.

In other words, if you can't define the development stage of your market, then you can't define your market. If you don't know the difference between when something is under development and when something needs to be developed that hasn't been yet, then you should ask yourself whether you're really in business.

It's an indication that nobody really knows how they'll make money from this thing, that there may never be a discrete point in time when publishers will have the willingness or confidence to declare that even free software is "released," that it's worthy of being made commercial, that it's something of value. And if no one can stand up for the value of their products, then how can you as consumers or testers or whatever you are, be expected to attribute value to them?

Over three decades after I was first paid to test a software product, this industry appears to be suffering now more than ever from an acute case of immaturity.