Practically also with 50G PON already being standardized and 200G in the horizon it will take decades before the limitations will be relevant; with typical 1:32 split you get comfortably 1G service to subscribers. I do expect gigabit connectivity to be generously good for 99% of users for long time.
It is also noteworthy that while PON was originally standardized as asymmetric, it seems like ISPs have had a change of heart and are widely deploying symmetric PON (i.e. XGS-PON). I don't know what is driving that change (Twitch streamers and Youtubers? :D) but I'm happy about that.
You blame ITU for PON, but IEEE has been pushing EPON (ethernet-PON) for almost as long (GPON ratified 2003, EPON in 2004). Ultimately standards organizations are driven by industry, not the other way around. With the industry having some very big players in it, I have no doubt that PONs would have happened regardless of their standardization status.
While PON is shared medium which is conceptually yucky, in consumer world its impact is less because lines are massively oversubscribed anyways. It doesn't make much difference if you have PON or active fiber if the bottleneck is the uplink.
Everything is eventually a shared medium. You don't have your own fiber all the way to Facebook. So the question is just at which point do you share and that should be a decision made on throughput and cost.
That said, it is unlikely that major telcos will deploy DWDM to the home outside of niche markets. The savings in feeder fibres costs are nice, but the bigger concern is that there is a very real cost to hosting enough ethernet switches to provide an ethernet port per customer. Most of the GPON deployments around where I live use 1:32 splits, but 1:128 is viable for residential subscribers at shorter distances and when using XGSPON or 10G-EPON (although I stick to 1:32 in my own network). With 48 ports in 1U of space a carrier can serve up to 1536 to 6144 customers in 1U with PON. That would be racks worth of equipment using 1:1 ethernet. DWDM-only would drive up operating costs for space, power, HVAC and equipment maintenance by orders of magnitude.
The current situation is one where XGS-PON users can get 5Gbps subscriptions, whereas AON users are stuck at 1Gbps - seemingly because the telcos aren’t upgrading their point-of-presence hardware to support anything beyond 1Gbps.
I'm not sure how many home users order that, given the extra cost of 10g switches, NICs etc and then 90% of usage being via WiFi that only just makes it to 1 gbps. But it makes a lot of sense for businesses with multiple users sharing one connection.
poor souls, though can we care about the low-end first?
What is the low end? Austria has a similar problem. There are some quite old and unmaintained AON networks where people are stuck with 100MBit whereas even G.Fast copper eclipses that in some cities at this point.
from my pov: <100mbps
Sure, but it's pretty ironic if you are stuck on a 100MBit fiber connection and a few buildings down you get 300MBit over twisted pair. And the problem with AON losing support is that you often can't find an independent ISP that would actually give you service over that AON you have.
And when the poor souls on slow internet do get upgraded, AON vs GPON suddenly decides if they can get upgraded to the new higher speeds in the next ten years or not. 1gbps may be relatively slow in 10 years, but with a widely spread GPON you're not getting much more out of that, while with AON entire neighbourhoods can be upgraded by replacing a single rack in the local POP.
Except in a few places it has been exactly the other way round. AON networks in Austria for instance have been built a few years back, some random companies ended up owning that infrastructure and don't upgrade. On the other hand the incumbents have built fiber, have rolled out GPON and have in the meantime upgraded to XGS-PON whereas many on AON got stuck. It's slowly moving but very gradually.
I don’t think we will ever hit the limits of PON quite frankly and swapping out PONs for newer and better standards is rather trivial.
I strongly disagree. On a party line information flows along the copper cable to every connected endpoint bidirectionally. While it's true that incoming information flows to all subscribers, never does information that flows out and you only get scrambled data even on the incoming stream. So if you're trying to make a security argument: the system is also safe on a physical level.
> We can do better!
Depends on what "better" is. I was quite critical of PON in the past but I have come around. Practically at this point I think PON is a better way to run networks in most places. At one point you hit a bottleneck anyways and not having to run individual fibers makes for a more resilient and cheaper system.
25gbps being "short sighted" is a bit of a stretch imho (running with 100mbps dsl and not feeling disadvantaged yet)
I made a comment a few days ago about how I despair when I see anything modern datacenter related. I get the same sort of revulsion when I look at the list of all of the gpon hardware on that page and thing: how much duplicated and wasted effort has gone in to making dozens of different models of the exact same thing. A thing that's not really even needed if a halfway-competent ISP made an investment that's more than the absolute minimum required.
Nice directory democratizing some good reverse engineering, though!
</end soapbox>