In a recent article, Wired addresses the idea of “perpetual beta” cycles. It applies to all software, but browsers especially. A closer look follows.

More and more companies are releasing software and online services that seem to be mired in a perpetual beta-testing state, leaving users to wonder what it takes to get to a version 1.0. Many developers and users say the constant tweaking is a dangerous trend. By Daniel Terdiman. (read full article @ Wired.com

That always gets me with the folks who want to say “Well you can’t criticize Firefox, it’s only beta.” Well, yeah, beta for software that’s been developed for ages now.

Developing software like Firefox is tricky, though, because there’s little value (that I can see) in saying “This is a finished release.” The one thing they don’t want is someone who finds a version they like and stops upgrading. Things are changing too much to have folks left behind. I mean, it’s not like someone using Firefox 0.8 a year from now would be comparable to someone using Netscape 6, but folks who thought Netscape 6 was “ok” and then didn’t upgrade did cause a problem.

Opera’s beta test cycle seems to go smoother. They do a beta release (or several, in the case of 7.23 if I remember correctly), but it is fairly close to being ready for release, and when it gets stable, they release it and start to work on the next version.

The frustrating thing, of course, is when they hit a “final” like 7.5 and there are problems with it, new bugs that popped up (not very common) or old bugs that came back (rare, but happens) or bug that I wished had been fixed that weren’t (most common).

At some point you have to either say, “This is a stable, though imperfect, version that is the best we can do right now. It should do no harm, it has some new features and some bug fixes. And there will be another version in a few months.

The only real frustration I get is with programs like Office which get a huge release and then go several years before there is any real work that appears in public. Usually there are a few things that I find right away, and you know they aren’t going to get fixed for a long time and you have to live with them.

Somewhere between the perpetual beta and the “We’ll be back in 3 years” model is probably the ideal. Finding the tipping point is the real trick.