okaven
join:2001-12-02
New York, NY
| reply to Bobcat
Re: Let's get something straight...
In certain areas of product testing with do double check things like clock-speed and other possible areas of optimization, such as custom written drivers, etc. This mostly applies to newest graphics cards, CPU's, etc. However, there must be room for a little trust here, too. You cannot go out and say by default every vendor tries to cheat. In Network Infrastructure this is not too much of an issue anyway because vendors rarely know what tests we are running exactly. Besides, it would be fairly detrimental for a vendor if we caught them cheating and published this. Although times are bad economically, PC Magazine still reaches about 3-4 million readers out there (which for the vendors translates into customers). So, I don't think that vendors would like to take the risk of cheating and paying for the consequences. Besides that, to accomplish buying all the equipment we would probably have to sell the magazine for $300 an issue.  Especially when you start looking at stories such as the Ethernet switch story with products that are in the thousands of dollars.
Most importantly, most of the products we test are generally not available in stores by the time we start testing. Although most of the products are available when the magazine ships to subscribers, our testing is done way ahead of that date.
Oliver Kaven
----------------------------------------
Oliver Kaven
Project Leader, Network Infrastructure
PC Magazine Labs
oliver_kaven@ziffdavis.com
-------------------- |
|
Bobcat
Premium
join:2001-02-04
Bedminster, NJ
 ·Verizon Online DSL
| reply to okaven
said by okaven:
it takes great effort and alot of coordination to get over 20 vendors rounded up to participate.
So you get the products directly from the vendors instead of buying them yourselves? How do you know a vendor didn't crank-up the clock rate in the box they gave to you?
--
Without software, life itself would be impossible.
Optimum Online; $29.95 per month; average speed 7200/900 kbps |
|
okaven
join:2001-12-02
New York, NY
| reply to Daniel
Daniel:
The layer 3 speeds of all routers tested is generally sufficient if you look at common DLS speeds available. It becomes more critical when you have a relatively well working cable connection (with few people on your segment). When you start hitting > 1,000 Kbps download speeds your router will become your bottleneck. Granted, this does not apply to most of us (unfortunately) but how boring would our testing become when we just stop at some point and don't try to find the limits of a device.
-ok
----------------------------------------
Oliver Kaven
Project Leader, Network Infrastructure
PC Magazine Labs |
|
okaven
join:2001-12-02
New York, NY
| reply to Daniel
----------------------------------------
(Sorry..replied to the wrong message. This is meant as a reply to Mike):
There are a couple of companies that are unfortunately not in the review. However, it takes great effort and alot of coordination to get over 20 vendors rounded up to participate. Besides that, sometimes the vendors are not willing to play because their products are in-between revisions or because they fear that their product does not exactly fit the filter that we set for a given story.
Oliver Kaven
Project Leader, Network Infrastructure
PC Magazine Labs
oliver_kaven@ziffdavis.com
[text was edited by author 2002-02-05 19:34:36] |
|
Daniel
Premium,MVM
join:2000-06-26
Pleasanton, CA
clubs: 
|  Are these speed differences even noticeable to a broadband user?
I am wondering if the speeds of the Netgear are way above what a broadband connection can push through it. If that is the case, and since the LAN side is equal, then it is almost a moot point.
Well, it would have been moot if it wasn't for the damage to my pride as a Netgear user.
--
"While we are postponing, life speeds by." - Seneca (3BC - 65AD) |
|
