Scott Brim <[email protected]> wrote: > On Fri, Mar 7, 2014 at 11:05 AM, John Leslie <[email protected]> wrote: >> Scott Brim <[email protected]> wrote: >>> >>> As I understand it, the ultimate complaint is that the encapsulation >>> allows traffic to be routed to a part of the Internet that might never >>> have seen such traffic before, by using UDP to get past blocking >>> points (NATs, for example). >> >> This actually isn't a separate issue. Indeed, encapsulation can cause >> that; but plenty of other stuff causes the same thing. > > A separate issue from what? Can cause what same thing?
I guess it's fair game: If I don't understand something you wrote, you're entitled to not understand something I wrote... ;^} I meant simply that many things cause traffic to appear on a portion of the Internet that hasn't seen it before. > Are you saying the essential issue is to avoid causing congestion in > a part of the network that has never seen it before, No. (I wasn't speaking to any "essential issue"; and I was largely still saying +1 to Matt.) > or that that is a subset of some larger issue, Yes. > or ... ? No. >>> I suspect the concern is that such areas might not be able to handle >>> even dropping a traffic load that is normal in other parts of the >>> Internet. >> >> I don't understand that statment. > > Let's try a Socratic approach: Why do you think it's a problem to > increase traffic in a part of the Internet, by injecting what is > considered ordinary traffic in other parts? I didn't realize I thought that... Nor, AFAICT, do I think that's a problem worth considering. Whatever traffic any part of the internet sees is pretty random. -- John Leslie <[email protected]>
