Credibility = f(omissions, errors)
Originally posted by yosoyjay
However, he does have a point about the standard for 802.11g not being finalized yet. It will not be standardized by IEEE until this summer after which I'm sure all stuff will be built to spec. Until that point however, stuff will be built upon what is most likely to be standard. All he is saying is be cautious about buying crap that is based on standards that have not been finalized yet and try to buy stuff of the same brand so you won't run into any problems. I see no problem with this argument.
The problem is its holes in conjunction with his obvious errors, and they hurt his overall credibility.
He may as well have done an experiment whre he ripped a beloved old 20M hard drive out of a 20MHz '386 and drop it into his 2.5GHz Pentium4, and then rip the P4 a new one, because he found out in his tests that his disk I/O speeds didn't drastically change.
On the actual issues he raises, he missed two big ones.
First, he misses the fact that sometimes Standards are finally signed-off only by people shipping real hardware.
Second, even if his "worst case" scenario happens, namely that this stuff doesn't meet some new 802.11g standard, that also requires a second failure that it can't be software-refitted. Under such a "sky is falling" scenario, these devices will still continue to work, but they will only be compatible with themselves. Furthermore, their backwards capability to the entire installed base of 802.11b is proven today, and it will not go away.
As such, the buyer's downside risk really isn't all that dire.
-hh