For the past few days, the blogging world has been abuzz with that “dashboardadvisoryd” thing. Indeed, it seems Mac OS X v. 10.4.7 contacts Apple’s servers ever eight or so hours and ensures you did not inadvertently download a malicious widget. This, however, raises a great many grave questions.
Contrary to what one might think, I am not outraged by Apple’s decision to phone home. Do I like applications calling home without telling their users? Absolutely not. I can however understand the reasons that motivated this choice and, while I believe it to be a PR blunder of ignominious proportions, I’m sure it’ll be an easy fix for 10.4.8, much like that iTunes problem we had a few weeks ago was. Engineers come with good intentions and that is what matter. They scratched their heads to improve security and they deserve applause for that.
No, what worries me is that we have, basically, re-invented anti-virus systems: provide “signatures” for rogue applications, perform a check against a known database and raise a red flag when the check returns positive. That, essentially, means one thing: there is no way to prevent damage (or, at least, one has renounced) so one relies on the next best thing, that is checking after the damage is done. The virtual equivalent of the difference between a condom and emergency “day-after” medication.
Since its inception, Dashboard has been a security nightmare. For starters, its implementation was approximate at best, placing files in hidden folders and providing no easy un-installers. Of course, to us “geeks”, the files are easy to find but one has to keep in mind most users have no idea what the “Library” folders are for. Then, there were these questions about being allowed to override Apple widgets, the difficulty of knowing what, exactly was installed and the auto-install feature that surprised many a user. Sure, there were technical explanations for all these faults and sure they were corrected but still, one should not have unleashed a whole new layer of what, basically are full-featured applications, all the while implying they are fun and harmless. It was a bad decision from the start, if Apple has ever made one.
Funnily enough, I know of no people using Dashboard. Downloading and installing widgets like there is no tomorrow? Yes. The interfaces are lickable and many of them provide fun or marginally useful features that are tempting at first. Yet, the slow loading times, lack of integration with the rest of the operating system and awkward interface (see above) mean many people who praise Dashboard simply never use it. (And yes, I am sure millions of people around the world do use Dashboard for real on a daily basis too, I just don’t happen to know any of them.)
Is it time to retire Dashboard? My personal incline would be to say yes, at least in its current form. Obviously, Dashboard is here to stay and Apple will not pull one of the operating system’s major features a year after its release (look at Sherlock, it is still here*). Yet, Dashboard, despite its marginal added benefit, has been at the heart of countless issues, confusions and controversies. The idea is interesting but it has just been made into something it is not and that is the saddest part of it all.
As many users of the community have pointed out, Apple could hire a small (large? huge?) team and have a similar system in place for every application on the platform. .Mac Virex anyone?
* Sherlock, I love you.