| Too many holes I see so many problems with their testing that it looks like swiss cheese.
Just to start did they ever hear of website throttling? I use it on my web server along with a connection limit which brings up the second hole.
What does the app do if it encounters a connection limit error? Does it keep trying to reconnect to the site till it makes it and than counts the whole time it took from getting a connection limit error to the time it finishes collecting the page data? Grant it if they are the most popular sites this may not be the case.
I could go on and on with this one but it is not worthy of any merit other than someone figured they had a better way to collect connection speeds.
Even one of their testers posted a comment on their article. They brought up issues with the test and it was actually a repost from the forum for the test and apparently it didn't get answered in the forum along with article discussion. |
|
ame101
join:2002-05-02
Southington, CT | »www.numion.com/YourSpeed3/ShowMe···,554,496 |
|
1 edit | reply to raptor1418
said by raptor1418:I see so many problems with their testing that it looks like swiss cheese. Just to start did they ever hear of website throttling? I use it on my web server along with a connection limit which brings up the second hole. See this for what they are measuring:
»Re: This is illogical...
They are measuring browsing speeds and NOT connection speeds.
--
--
Internet News
My BLOG
My Web Page |
|
| Yes but the article here and on PCmag's site is touting it as internet speed which would not just be browser speed.
PCMag title: State of The Union: Internet Speeds Across The United States
If they really want to use that as the title they need to include large file download speeds into the equation. They also stated that their R&D helped pick out what they called "outliers" from the data to accommodate for anomalies. I see a problem with that right there as they left data in where the ISP is "Don't Know" or Other.
I would consider that in accurate data to use at that point. |
|
|
|
Reviews:
 ·Comcast
| reply to raptor1418
said by raptor1418:I see so many problems with their testing that it looks like swiss cheese. Just to start did they ever hear of website throttling? I use it on my web server along with a connection limit which brings up the second hole. Unless you throttle by ISP then the numbers would be consistent from ISP to ISP.
If the test hit the same sites, and ISP A was consistently faster than ISP B, then it would seem to be result of the ISP, not throttling. |
|
