Good and Bad Betas
The meaning of "beta"--a term meant to describe a largely completed software release ready for the stress of real-world testing--has gotten awfully muddy in practice.
Some applications, such as Google's Gmail, can spend years in beta status but feel as polished and stable as any 1.0, 2.0 or 3.0 release. Others can be advertised as "finished" but still suffer defects that should have been found and fixed before release: Think of the complaints I've seen about Apple's Mac OS X Leopard, and the considerably larger number of gripes about Microsoft's Windows Vista.
You should exercise caution with anything explicitly labeled as unfinished. But the ubiquity of public beta releases makes it hard to ignore this category of software on principle; some betas justify the added risk. How do you tell? Consider the worst-case scenario--if this program malfunctions, how bad could things get?
One just-released beta, the Beta 2 release of the Mozilla Firefox 3 browser delivered this morning, seems worth the trouble. As long as you don't have any browser add-ons installed for older versions (which probably won't work in this release), you don't have much to lose--at worst, the browser crashes, you uninstall it and you revert to the current 2.0 release--and you do have a decent amount to gain.
The new browser makes it easier to find a page you just visited--just start typing a word that appeared in its title, and it will suggest recently visited pages that match. Its bookmarks-management screen offers additional ways to make sense of your Web favorites. The most valuable feature here, however, may be the promised performance improvements; anything that makes an application I use for hours a day feel more responsive is good news to me.
I might not add this program to my primary work desktop, but I certainly will put it on the work laptop.
And then there's the other beta announcement I received--a press release from Symantec, announcing the mid-January availability of a beta version of its Norton 360 security and maintenance suite. The tricky thing is that the first sentence of this announcement might make you think it's the actual, final version:
CUPERTINO, CA, Dec 18, 2007 -- Symantec Corp. (NASDAQ: SYMC) today announced that version 2.0 of its market leading all-in-one security solution, Norton 360, will be available by early-January 2008.
In fact, the real thing won't ship until "sometime in March," company publicist Spencer R. Parkinson wrote in response to my query.
Symantec is unwise to treat the release of a beta version of this program as something worthy of public attention.
I'm nervous enough about installing completed security software--it must operate at the deepest levels of Windows, where a malfunction can lead to catastrophic damage throughout the system. (Put another way: Would you rather have a fire break out on the deck of the ship, or in the engine room?) And now Symantec is going to suggest that people try out a security program that it's still working on?
No thanks. Let the company pay its own software testers to find the bugs.
Tell me about your recent beta experiences--what's the last time you regretted trying a beta release that needed more time in the oven? And when was it worthwhile to load a beta release that cured problems present in an older, "finished" version?
Posted by: BR | December 19, 2007 6:27 PM | Report abuse
Posted by: CPS | December 20, 2007 7:04 AM | Report abuse
Posted by: wiredog | December 20, 2007 9:22 AM | Report abuse
Posted by: Huizhe | December 20, 2007 8:37 PM | Report abuse
Posted by: Bush -- not related | December 20, 2007 10:40 PM | Report abuse
Posted by: Gosseyn | December 21, 2007 5:49 AM | Report abuse
Posted by: Internet2k4 | December 21, 2007 8:18 PM | Report abuse
Posted by: Jim | December 22, 2007 9:01 PM | Report abuse
The comments to this entry are closed.