Live Blog @ the FCC (Part 2)

ø

4:39 – P. Clark: This isn’t a short term problem. The last mile to the home is always going to be expensive.

C. Adelstein: Where is the line between good and bad discrimination?

BT: I know it when I see it, and Comcast is bad.

P. Bennett: There wasn’t any data from Comcast. But, discrimination might be ok when the network is crowded, but it isn’t ok when it isn’t crowded.

P. Clark: If the network is spoofing packets that look they come from somewhere else that seems particularly troublesome.

Smyers: When there is industry consensus on a solution we don’t need to look any deeper. But when solutions don’t comply with industry standards then it needs a deeper look.

4:22 – C. Cobb: We don’t even have the questions, so how can we have the answers?

C. Martin: If we do not have enough information, should we not take any action? P. Clark: Some disclosure might be good and start a dialog with industry. We also do know enough about the current actions to decide if they violate the IPS or not.

C. Adelstein: Are our network management practices having an impact on innovation here? A: There is a lot going on in Asia now.

P. Clark: The networks today contain ways of regulating traffic. The Comcast response is a very nuanced response to traffic. The network today allows me to go really fast when nobody else is there (this is good), but then when lots of people are there I need to slow down. The question is thus how should we slow down when we have to b/c it isn’t feasible to build a network where we can all always go at full speed.

4:14 – P. Weitzner: If the IETF or other technologist can’t solve these problems I don’t know how else they would be solved. Could the FCC wave a wand and do it? Probably not.

P. Bennett: But standards bodies are very slow. Comcast should be out experimenting so that the standards bodies have data to look at.

Smyers (Sony): Companies do and must experiment so that solutions can be fully vetted. However, if one company finds a problem the solutions still need to be deployed broadly across the network.

C. Cobb: Do we have enough information to figure out what the right solutions are? Profs: No. Also, we don’t have very good data and the ISPs don’t want to share it with us.

P. Clark: It is very hard to do research here b/c we can’t get data, even under a NDA.

4:05 – P. Clark: I don’t like discrimination on the basis of the application. Rather I’d have discrimination that is chosen by the user (how do users do this?).

C. Martin: Does BT take advantage of the way that networks (cable?) are designed right now? BT: Protocol was designed to efficiently move files. Key is to not rely on a single server, wasn’t designed to fill a network.

3:57 – Back to questions…

C. Martin: Should we investigate usage caps over time? P. Clark: Yes, you need to be careful, but we should recognize that there are costs associated with usage. This allows ISPs to have a positive engagement with customers instead of a negative one. This is similar to the way things work in wireless.

P. Clark: The models don’t work very well. How do we quantify what is acceptable congestion? This is hard and Comcast has tried to say that it is what doesn’t interfere with others. How do we impose fairness among users? Nobody can send bits faster than anyone else at the same time (this is the historical approach). There has to be some way to deal with congestion.

3:48 – Video Comments

3:46 – Starting in on some more questions…

P. Reed: Back in the day (mid 90s?) people said that we shouldn’t pave the world with one-way streets. However, we ended up being pragmatic.

3:32 – Last panelist:

Scott Smyers, SVP, Sony – Is there competition in the video marketplace? Yes, but it is different that many people think. Internet video can be transformative and lots of other companies agree with us about this.

The internet drastically reduces the barriers of entry. These issues are complex. What is clear to Sony is that we need to think about how we use the internet to distribute video. Super bonus points for only being 20 seconds overtime.

C. McDowell: If there is congestion, the internet has ways or working around it? BT: Yes. C. McDowell: If we increase disclosure would that be good? BT: BT is run much like an open standard so we are already disclosing the limitations of our application.

3:14

Eric Klinker, Chief Technology Officer, BitTorrent – Why does BT exist? It was invented in 2001 to solve an emerging problem of efficiently moving large files. Since then online media has exploded.

Who is served by BT? Any budget conscious organization, BT reduces the cost of publication. (BT also makes screens go up and down… or something).

David P. Reed, Adjunct Professor, MIT – Disclosure is important for understanding what is going on. Comcast’s secret action suggests a need for stronger intervention when we have a duopoly or monopoly market.

He is doing some sort of demo with an envelope. Envelopes have addresses and content inside of them. (seriously, I can’t make this stuff up). Content is meaningful only to sending and receiving host. The internet is a best-efforts network. But when congestion becomes extreme you discard some envelopes. (Ok, I get this now).

Solutions for congestions only work when they are implemented internet wide. The IETF does things like this (aww yeah IETF shout-out!). Techniques that look inside of packets are bad (and this is what comcast does).

2:55

Daniel Weitzner, MIT – Doesn’t want to use the term net neutrality. We’ll give him that one for free but let’s see if he can manage that.

The web is already very much P2P. (Some might call that generative. P. Benkler mentioned layers early, but I haven’t heard anyone talk about generativity yet. I am hoping that we’ll see it here!)

Richard Bennett, Network Architect – An engineer (yay!). The internet needs some management in order to function, otherwise it would fail. Different kinds of traffic require different things of the network. It is reasonable to take low bw, low jitter, traffic and give it priority (i.e. VoIP).

If we want a non-discriminatory internet then we’d lose a lot of things. The implications from adopting rules against discrimination would be very bad. Let the market and engineers figure out how to deal with the problem of congestion.

Dr. David Clark, MIT – If lawyers will talk about technology, I’ll talk about pricing (and a wikipedia joke!). Something needs to give in this market… we should have pricing based on how much you use — the all you can eat model just isn’t sustainable.

Limiting is bad when the ISPs choose what is given what priority. But some discrimination is needed.

2:53 – C. McDowell (also know as the funny one) will have to leave early. Maybe.

Ok, getting started back up again. I’ll try to apply a bit more of a filter this time around — less transcription, more comments. Who knows if that’ll actually work.

2:51 – Totally spoke to soon. Some more blogs about what is going on are all listed here so I’m not going to bother even trying to duplicate the good job that they are doing.

Also, some photos are up here.

2:45 – Well that sure was a long 15 minutes. At least there is room to breathe in here again, and I managed to snag a real seat, no more (not so) soft wall for me!

Commissioners are starting to come back in so we might actually be getting started soon. My apologizes for the spelling in here, Firefox is refusing to play nice with WordPress and as a result I don’t have easy spellcheck!

(edit, to many acronyms… got that fixed)

Comments are closed.