Hacker News Reader

Comment by sillysaurus3 | Hacker News Reader

Ah, you want contrarian! You've come to the right place.

As a paying customer, I'd love the ability for streaming sites not to impact my bandwidth quota. I get 1TB a month. That's it. I've run over that limit once, and you can guess where 90% of it went.

This is especially true on mobile, where the bandwidth cap is literally 100x lower. 10GB is not a lot, and you can burn it on one long train ride.

It's absurd that we have to deal with caps, but... The world has accepted them. So if they're here to stay, why not classify streaming sites into a different mental category? They're bandwidth hogs.

I'd rather keep NN in place, but that's one flipside.

View on HN

Suppose your town decides the mall is the most important destination in town, and dedicates this year's road budget to a really nice new highway between the mall and the suburb where most the mall shoppers live. What happens to the roads in the rest of the town?

I didn't say it was a good idea overall, just that there are some pros. And the poster was looking for pros, not just cons.

On the other hand, it's typically only specific streaming sites that get placed in a different bucket. Thus, it becomes more difficult for new (potentially better) services to enter the space. I don't think that's a net positive.

But with NN in place, we still have caps. So what the heck are you talking about?

Repeal NN => youtube/netflix no longer counts toward cap => cap no longer causes problems.

Some ISPs (verizon?) do similar things, where if you use their app then you're allowed to stream as much as you want. But I'm not sure whether that runs afoul of neutrality laws.

This is admittedly a weak argument, since companies would end up eating the cost, further entrenching the existing monopolies. But they were looking for arguments in favor of NN.

There are other contexts where non-neutrality might be useful: A local startup once tried to roll out a lower-latency service for gaming, based on intelligent routing. It was unlikely to succeed, but even if it were technically possible it would definitely violate NN to be prioritizing/rerouting certain classes of traffic.

> It was unlikely to succeed, but even if it were technically possible it would definitely violate NN to be prioritizing/rerouting certain classes of traffic.

This is not true at all.

Net neutrality rules allow for "Reasonable Network Management" which includes prioritizing classes of traffic. Here's a discussion on it https://www.reddit.com/r/networking/comments/6nddmx/does_net...

That doesn't seem to address the natural follow-up: what stops me from claiming all my traffic is VoIP so it gets processed faster? Is there some offsetting downside that disincentivizes someone from labeling it that way? (something like "any request for lower latency will also reduce your throughput, so don't call it VoIP if you really care about the overall transfer rate)

It's fairly straightforward: everyone gets high priority up to X mbits/s with no upstream oversubscription. Everyone gets best effort up to their last mile speed after that. If there's still available bandwidth, everyone gets idle-only traffic without a cap.

You could add more priorities in between, etc. as needed.

Then that just means everyone will call their first X mbit/s "high priority" and then relative priorities are unaffected and nothing can be privileged.

That makes no sense. The first X mbit/s will be privileged. If you waste your priority allotment on torrents instead of voip that's your choice, and that choice doesn't affect anybody else.

They mean the ISP can say "streaming from Netflix doesn't count towards your data cap!", either as a free bonus or at least as a cheap add-on option.

In a sparsely provisioned world, IMO usage (caps) is the most sensible way to bill.

(No one wants to pay for a fully provisioned world)

Reply on news.ycombinator.com