> As long as it terminates in WiFi or a standard Ethernet connector and passes IP traffic at the rate I paid for, I personally don't care what the physical infrastructure beyond the demarc is.
I get that attitude, but coax connections are fragile in a way that fiber is not. Water infiltration, bad crimping jobs, unterminated ends in the neighbor's house all cause signal degradation that is extremely time and labor intensive to identify and fix.
Cable is an excellent solution for distributing broadcast analog radio signals over a private line cheaply and efficiently, but two-way digital signals over the same medium aren't as robust as their fiber-based counterpart.
ISP outages are one thing, but if you ever experience service degradation—and not a full outage, but tear-your-hair-out crap like "my internet goes down for 15 seconds once every 16 hours with no discernible pattern"—and it ends up being some particularly coax-flavored problem, you'll understand what I'm talking about.
Cable is not perfect. Not by a long shot. It has a whole host of failure modes, some of which are very weird.
But at least in the minds of the general public, those failure modes are not new. They’ve been around as long as cable TV has been around — since about the 1970s. Cable installers know these failure modes well, and how to test for them, and how to remediate them.
Fiber, on the other hand, is still relatively new for most people. It has its own failure modes, some of which are very different from coax. And unless you’ve been locked in a high speed datacenter for the past couple of decades, those failure modes are likely to be much less well-known, and the solutions for them are likely to also be much less well-known.
For me, everything else being equal, I’ll take fiber over coax. But a well-managed coax cable plant is orders of magnitude better than a poorly managed fiber cable plant.
I just want shit that works. I don’t care much about how it works. And fiber to the neighborhood plus coax to the home, is a perfectly fine solution as far as I’m concerned.
If AT&T wants to litigate that, then they should explain why they haven’t already upgraded all their customers to fiber. Then they should be forced to shut the hell up until such time as that upgrade for all customers is complete.
> Cable installers know these failure modes well, and how to test for them, and how to remediate them.
> Fiber, on the other hand, is still relatively new for most people. It has its own failure modes, some of which are very different from coax.
That is true, but most of the common failure modes in the sort of fiber that competes with coax are things that coax is similarly vulnerable (mostly physical damage or bad connections) to and for which the concepts involved in diagnosing and repairing them are also similar even if the specific tools are different. Someone who understands coax is not going to have a hard time learning fiber unless they're simply unwilling to learn.
For what it's worth about a decade ago I had consistent recurring cable modem issues that resulted in multiple truck rolls with supposedly top tier service techs coming out, fiddling with wires at their distribution block, and then leaving without really changing anything. I eventually managed to find the modem's "password of the day" algorithm and wrote a script to log the SNR per channel, which let me determine that the issue was always with just a small set of adjacent channels which I found to align perfectly with a nearby AT&T tower. I pointed this out to the next tech to come out and they ended up replacing the entire line from the street to my house which finally resolved the issue. These were experienced techs working for a large regional cable provider (Armstrong) with whom I had previously had almost exclusively great experiences, but none of them were able to figure out what should have been a pretty simple coax-exclusive issue until I basically led them straight to it.
> But a well-managed coax cable plant is orders of magnitude better than a poorly managed fiber cable plant.
Outside of very small rural operators or campus networks that are literally one-man shows or very close to it I'm having a hard time envisioning a situation where an operator did a great job with coax but just fumbled everything with fiber. Especially because, as you even touch on, basically every coax operator outside of the absolute smallest are already running a hybrid fiber-coax system and thus are dealing with a lot of fiber already. Maybe the lowest level installers still only deal with coax but at that level we're not really talking about the deep knowledge or major decision making.
> If AT&T wants to litigate that, then they should explain why they haven’t already upgraded all their customers to fiber. Then they should be forced to shut the hell up until such time as that upgrade for all customers is complete.
IMO AT&T does need to have their feet held to the fire about a lot of things like what happened to all the federal funds they (and other major telcos) received to fund fiber deployment which got diverted in to cellular service and why they ever deployed FTTN in greenfield neighborhoods while they already had FTTP in other markets, but that's not really relevant to the validity of this complaint against Comcast's intentionally deceptive marketing. Them being right about this isn't related to what other things they've been wrong on.
my internet goes down for 15 seconds once every 16 hours with no discernible pattern
Actually that's exactly the type of complaints I've heard from those who upgraded to FTTH. My understanding is that fiber is a lot more fragile than copper, and installation and testing requires significantly more expensive equipment, although it is cheaper for longer distances and theoretically capable of higher bandwidth (but the transceivers on each end are also a significant part of the cost.)
Is it possible you need to disclose some conflicts of interest here?
Fiber doesn't have the noise or attenuation issues that coax does. Having worked in the industry and monitored plant, things like daily temps, proximity to radio transmitters, water in the soil can all cause signal to noise issues with coax.
Upgrading to fiber was the best thing I've ever done in terms of internet service. I've never heard of what you claimed. Myself and everyone I know, anecdotally, have had FAR better service reliability with fiber. Also, having massive upload capacity papers over other failings elsewhere in your tech stack, like a router that doesn't know how to properly queue. Well, throw a ton of bandwidth at it and there's no big queue backing up anyway.
The only time fiber has gone out for me, including power outages that impacted my home, was if a neighbor or another ISP was digging and cut my line. That said, this happened with coax, and would be a problem with anything except a WISP or Starlink.
Indeed, if I move and fiber isn't available but cable is, I'll probably just go Starlink. I'd rather deal with that variability than ghost-in-the-machine coax craziness.
The real life rates of these things happening to people is low. We humans remember these events precisely because they are not common. This is why each of you is adamant different solutions are better.
As for satellite I think I’d rather just not have internet than go that route. Maybe read more.
I used to work on a high-speed firewall some 25 years ago. Bending an optic cable too much caused a speed drop in our benchmarks (packet loss or corruption); if we got unexpected bad results we'd go check the cabling and re-run before debugging any software.
Fiber can snap very easily where's copper does not have this problem. Most of the time this is something only technicians could cause to begin with though.
Absolutely, anything like that has nothing to do with the medium and GP trying to say that's a fiber behavior instead of a badly configured ISP behavior is just nonsense.
Back when DSL was still a relevant thing I had this sort of nonsense all the time because ISPs would set the PPPoE session to reset or something on a multiple of 24 hours. That may have been fine when people were still dialing up or actually manually initiating PPPoE but in the world of always-on broadband shared through a NAT gateway it meant the client got recurring mid-day interruptions that would start any time they had a power or ISP outage and continue until someone was able to manually reboot the modem outside of business hours and get its cycle realigned to not be a problem.
24 hour or multiple of it long session on the PPP is still common unfortunately.
Still unrelated to anything at L1 except that DSL IIRC generally brought you ATM-styled information (being derived from a system for delivering digital voice line over POTS for ISDN last mile) while DOCSIS has its own weirdo framing and is IIRC the only one that does not do PPP by default
Kind of a big wat to this whole post. I've never, ever heard, or experienced myself or with clients, anyone across three continents having anything less then a rock solid experience with FTTH. Only trouble has always been getting it installed in the first place.
>My understanding is that fiber is a lot more fragile than copper
How did you reach that understanding? It's the opposite, fiber is a lot tougher particularly at high speeds. For outdoors I use stuff with a steel helix armor and then kevlar, you can direct bury it a few inches and drive a truck or tractor over it no problem. It's immune to electrical interference, you don't have to worry about surges to different ground potentials, water is no issue. People sometimes talk about bend radius but it's something like 10x diameter installed and 20x under tension (pulling) which is trivial given how thin fiber is. And high performance copper has a bend radius too, which isn't as high a multiple but simultaneous copper is a lot thicker.
>and installation and testing requires significantly more expensive equipment
On the installation side no, not sure what you'd be thinking of there. Pulling cable is pulling cable. On the testing side, sorta? A hobbyist who is just willing to redo things while they're learning can get away with cheap stuff, just like with ethernet. I guess it's true higher end testing equipment (like Fluke) ramps up faster than ethernet, higher floor, but the most serious kit is really expensive for both, and simultaneously is a fixed cost that is clearly meant for contracts that will pay it. Anyone spending $15k on a DSX ethernet kit is doing government or enterprise contracts or something where the cost will amortize out, same as with fiber.
The only area where there's a clear (but still one time) cost/irritation difference is if a cable is broken and needs to be spliced. But this too correlates with "if you're big enough to actually need to care the cost gets amortized".
>although it is cheaper for longer distances
Can't do long distances with high performance copper at all, setting aside cheaper (and yes copper has skyrocketed in price).
>and theoretically capable of higher bandwidth
Uh, "theoretically"? Fiber is much, much higher bandwidth, no theory about it.
>(but the transceivers on each end are also a significant part of the cost.)
A 20km class GPON (1G) transceiver new, no bulk pricing, starts around $40, and an XGPON (10G) around $50. SFP+ ones are less. Yes if you get into 100-800 Gbps class transceiver cost goes way up, but so does all the equipment one would use for that (100G is probably going to be at least $250/port last I checked) even before touching on whatever is actually making use of such a network. Transceiver cost would not be a significant part of overall project cost.
> It’s the cars which can’t have their head unit replaced with a simple stereo which worry me.
I agree. There's a lot of room for innovation, consumer choice, and competition that was previously guaranteed to be available in those one or two DIN units of space.
I haven't seen them present in any new car I've looked at in nearly ten years, but they were in every car I saw from the 50 years prior.
> People often forget how "good enough" the PS2 graphics were, and how big of a deal that was.
I remember going to an electronics store with my friend. I couldn't figure out why they had a broadcast football game playing on a 15 inch LCD display up front. I assumed it was a game but then watched it for a few cuts and was so certain it was live footage...
reply