rolf wroteLatency jumps in traceroute are most probably due to a slow link (satellite or big distance terrestrial...) or severe network congestion.
yh this can be a reason in latency jumps, i should have mentionned it...
rolf wroteBasically the aim is to give priority to more interactive (time-critical) applications at the expense of background traffic.
but that isn't necessarily how it is used :P
rolf wrotegive priority to VOIP calls or video conferencing in a company, or give priority to browsing over download, p2p or mail
Last time i checked VOIP and video conferencing are still shit because of the low speeds, so does it matter to prioritize them :P, anyway what you said is very true in general...
rolf wrotethe aim is to give priority to more interactive (time-critical) applications at the expense of background traffic.
I can argue which applications are more time critical than others, some websites needs you to load content from the server once. So having a bit high latency(1000-3000) on these doesn't matter, but others that use AJAX for example to provide new content loaded from the servers require blazing-speed latency to provide a smooth experience. Normal downloads do not require high speed latencies nor does file sharing but games are lantency hungry, so is shell control, and those are P2P applications.
So basically u can't say that browsing is more time-critical than P2P, nor the opposite way...
Well once you said that providers should give basically three types of packages(and sub-packages maybe) slow, medium and fast... That is a brilliant idea. But providers are too lazy to do that kind of work, they are just ignoring a group of users but what's they are blind is that this group is far too big :P, it just needs one provider to grow some balls and provide the right solution for them at an affordable price(<100$) before others are forced to do the same...( from my observations, IDM has been moving that way, let's see how long they can keep up :P, or they will drop the balls like WISE did :P
