- Aug 25, 2001
- 56,570
- 10,204
- 126
Originally posted by: RadiclDreamer
Spidey will own your face in every facet of interweb. Its best not to argue with him, it never ends well.
I'm not arguing with spidey07, I respect his knowledge highly. That's why I don't get it when he said not to use "ping" to test latency - ping was specifically written for that, and it's on every windows box. Why not use it? Is there some defect that I don't know about? Or is he just obliquely referring to using something like PingPlotter, with fancy graphs and the like? He hasn't answered that.
I do disagree that consumer ISPs should expect to see 100ms swings in the latency reported to a specific internet site. Unless there is congestion, it should ping relative to the physical distance and number of router hops that it takes, and in the case of both my friend's and my connections, that variability is much, much lower than 100ms. (On the order of 20ms for my friend's cable connection, and on the order of 3ms variability for my DSL connection.)
I was NOT saying that there wasn't a difference of 100ms, when comparing pinging say, a site close to you, and then pinging say, a site in China. Of course there is going to be variability relative to distance. But given the same site, pings shouldn't vary 100ms. If they do, then there is someething very wrong with your connection to the internet.