3 ms·
This is such a superficial and terrible article, its like if you gave an intern an outline and they could only use wikipedia to complete it. While I appreciate
by bifrost 7y ago
This is such a superficial and terrible article, its like if you gave an intern an outline and they could only use wikipedia to complete it.
While I appreciate what they're trying to do, this comes off as ridiculous to those of us who actually have built or build parts of the internet.
The TLDR of this article is "With current technology you can move more bits with light and electrons than you can with RF", which is likely to stay true for the forseeable future.
What they left out is that most fiber service providers use PON, which doesn't give you the top end of what your physical fiber can do.
Here's a handy article about it:
https://www.electronicdesign.com/what-s-difference-between/what-s-difference-between-epon-and-gpon-optical-fiber-networks https://www.electronicdesign.com/what-s-difference-between/w...
- sp332 7y agoThe article was co-written by EFF's senior legislative counsel. It's intended to be a position paper to influence policy, not a technical deep dive.
- bifrost 7y agoYes, and it shows they don't understand the technology. This is almost as bad as "the internet is a series of tubes" to a network engineer...
- admax88q 7y agoHow do they not understand the technology? They clearly compared the theoretical max to each technology as options for last mile connectivity. If you have last mile fiber into your home, then you're capable if receiving 10Tbps. Yes a PON setup won't actually provide that, but they never state that someone would offer those speeds, only the theoretical last mile capacity. The point of the article is to counter lobbyist claims that we don't need fibre investment because cable and 5g are fast enough. This article points out that even in their best case that is not true.
- bifrost 7y agoThey don't understand how PON works, think of it like FDDI or even DOCSIS; You'd never be capable of receiving 10Tbps due to fiber topology... The 10Tbps number requires DWDM anyways, its like saying a Kia is a Lamborghini. The Cable/Wireless carriers will use this misdetermination to undermine what these guys are saying so in the end its unhelpful. They'd have a much better time pointing out that existing cableco's have mismanaged their DOCSIS networks and that Comcast is considered the WORST company in this country (even worse than Monsanto) because of how poorly they manage their networks. There's also so little adoption for "Wireless to the Home" that its pointless to consider that a reasonable option.
- didibus 7y agoI'm not sure I really understand your alternative. It's not like in practical terms, providers will achieve the theoretical max of RF or cable either. What are you suggesting instead?
- bifrost 7y agoWhat I'm saying is they're talking out of their behinds, which will ultimately hurt us all.
- aeternum 7y agoYou clearly know about this, but your comments are not very substantive. It might be more valuable to explain what the article is getting wrong. Many HN readers are fairly technical and interested in stuff like this.
- bifrost 7y agoBased on the downvoting, I'm unconvinced they're actually interested in this :) I can tell you this -> Any time anything network/internet infrastructure gets posted here on HN, its a mess. My friends in NANOG have stopped bothering.
- lisptw102019 7y agoMy facepalm is that a group like the EFF would likely want to point out the vulnerabilities in RF and 5G from a snooping standpoint. My experience with all three techs from an Outside Plant aspect, tells me that Fiber probably would be the toughest for an outside party to snoop on (i.e. biggest physical access to network requirements)
- bifrost 7y agoHaving done a reasonable amount of network debugging/tapping, IMHO its actually the hardest to tap copper networking equipment since its fairly signal sensitive. You can tap fiber passively by "simply" cutting some of the jacket off and bending the fiber or installing a fiber tap. From what I understand about 5G, there's some amount of encryption between the phone and tower, and tower and the PSTN so it could technically be safer from a surveillance POV. But in the end the 5G tower will be fed by fiber anyways. If someone is tapping the physical media instead of having CALEA, you've got a bigger problem than how its tapped...
- jkoberg 7y agoYou don't get that there's the equivalent of multiple entire RF spectrums in every fiber ? Wireless last-mile always seemed like a false argument against municipal fiber on behalf of very profitable wireless carriers
- bifrost 7y agoLol dude, I've built terabit DWDM systems and I'm a licensed RF engineer...