Author Topic: Zone Alarm "Fans" - A warning  (Read 8512 times)

0 Members and 1 Guest are viewing this topic.

Offline OrangeCrate

  • Avast Evangelist
  • Advanced Poster
  • ***
  • Posts: 798
Re: Zone Alarm "Fans" - A warning
« Reply #15 on: June 24, 2006, 12:16:14 PM »
I do find it a good idea to look at the ZoneLabs forums when a new update is released to gauge the stability and compatability issues , as a rule of thumb.

Obviously only users with problems will post there , which may be a small percentage of overall users, as with most program updates of any software.

Abraxas - The comment you've made is exactly what people should do before installing new software, updates, etc. There would be fewer flawed products in the marketplace, if people did their homework.
_ _

I've referred to "statistical sampling" a couple of times in my posts in this thread, and as an aside, I thought people might be interested in how that works, and how it relates specifically to evaluating software in forums.

As a professor in one of my early statistics courses in college explained to me, statistical sampling works a lot like this...

Everytime someone complains about something (as is the case here, though you can substitute any parameter you want into the equation), think of it as the tip of an iceberg. Statistically there will be others who have the same problem, but have not commented in the forum. To assume that only one in multiple millions of users has that problem, is simply not mathematically correct.

When someone else complains about a specific issue with a product, evaluate whether it's an interpretation of a problem relating to an existing iceberg, or does it represent the tip of another iceberg. And, on and on, as you evaluate each complaint. (In ZA's case, vsmon problems is one iceberg, installation problems is another iceberg, uninstall problems is another iceberg, etc., etc., etc.)

Though, in real life, the equations can get pretty complex, as you analyze the specifics within each iceberg, statistical sampling, in a nutshell, involves nothing more than counting the number of icebergs in the pond. From a software evaluation standpoint, the number of icebergs will have a direct correlation to the quality of the product, no matter what percentage of users are represented in the forum. Just count the icebergs.

I certainly don't see many in avast!'s pond, but I see many in Zone Alarm's. Hence my comments, and this aside.
« Last Edit: June 24, 2006, 02:13:40 PM by OrangeCrate »