and maybe even without internet, the approach being vulture carriers?
agreed, pigeons are too weak to carry modern storage media.
why not just put some 6to4 gateways specifically for those laggard customers and let the rest of the world progress peacefully?
should be rechecked because some time passed and my ipv6-preferring net link already gives better and better speeds with various services that i use.
Sure about that? Research claims the opposite.
so, this is how those multi-gigabit speeds were achieved in universities?
nothing else than some campus pigeons with microsd cards???
And also 4to6. Who will pay for that?
I’m not even mentioning that supporting such a transformation would require redesigning the entire protocol, and it would no longer be peer-to-peer. That would be a colossal expense.
I would rather suggest you host such a gateway for your IPv6-only node, this will not require protocol rework.
IPv6 usage by the customers is negligible in comparing with IPv4. Maybe sometime it would change, but not now.
The mandatory requirement to join the network is to have IPv4, you can also provide IPv6, our S3 gateways may use it (for the customers, who use our S3 Gateways).
The problem isn’t speed, but rather routing disruption caused by the greed of some internet exchange points. And they’re simply disrupting routing because they can.
While corporate traffic is working fine (for example, IPv6 works perfectly for our services), the problems are experienced by clients of regular ISPs. This includes both our clients and our node operators.
that would involve expenses like some vps with an ipv4 link, right?
well, my current old ipv4 link already costs me approximately the same as a vps would, so there will be no gesheft.
and this line of thought is probably what other node operators in my area adopted, thats why they stick with old slow links and not upgrading or switching to faster ones.
only that please dont misadvertise anymore multi-terabit transfer speeds that were achieved by some university campus pigeons carrying microsd cards if you really want a broader adoption without quick abandonment afterwards!
well, that routing disruption translates into speed as observed by naked eye experience.
when some direct route is disrupted, a route around develops (probably using several slower horizontal links) and introduces lags of course.
but they are working towards fixing it, and it seems that dudes like backblaze are already able to take full advantage of that!
specifically in our town one of my old acquaintances said in a podcast he achieves 8gbps on his 10gbps inet link.
and poor me who cannot saturate even my 500 mbps with storj at downloading!
btw, we are on the same isp and he is not with storj of course.
and he owns a local video hosting platform that storj wants so much as their clients.
only if storj could impress him with real measurements and not just some oncampus pigeons somewhere far away.
Paying customers aren’t pushing for IPv6. That has been true for years: every time this discussion comes up. And if the-people-with-the-money won’t pay for it… then SNOs won’t get paid for it either 
When Storj believes they’re losing substantial business because of it… that will change. If/when IPv6-only demand arrives: it will be easy for SNOs to supply it. No worries!
depends who qualifies as these.
i, for instance, as a client, am currently paying ~$50 monthly for storj services because precisely in this period i am using more than i provide as a sno.
and as a client i am dissatisfied, and as a sno i am trying to explain to you the underlying reasons.
so unlike you i know how the problem looks on both sides.
That’s fantastic: thank you for being a customer of Storj! 
You should open a support ticket with them, to let them know your usage is being limited because of their current level of support for IPv6.
That’s the type of feedback that leads to change. The more customers (or potential customers) that let them know… the better! They’ve been receiving comments like that for almost a decade now… but that doesn’t mean the tech can’t change. IPv6 is great!
You have no idea about my view on “the problem”, or how much I’m spending for Storj to hold my data. But I hear what you’re saying. Please open a ticket: as a customer your voice matters!
oh does it?
from what attitude i get in response it really doesnt that much…
A customer that cares enough about an issue to open a ticket… is something that a business cares about if they get enough of them. It adds up.
Some rando crying in a forum doesn’t mean that much.
Use the benefits of being a paying customer to have your voice heard. Open a ticket: even if the response from Storj isn’t what you’d like.
i am quite annoyed by those arrogant dudes that wonder why customers just leave without providing feedback.
in reality, if they’d really want our opinion, they would have looked everywhere and not just their support ticket system.
and nowadays on those support systems is usually just some ai bot that leads nowhere and is not responsible for anything it says.
and after all the client questions to support is about what a client could do.
and in this specific situation there is nothing that i as a client can do to fix it.
i was already been told to switch to the same isp where storj’s campus pigeons reside.
lol on that, because with the same success i can switch my cloud storage provider (and it won’t require the hassle of recabling, by the way)!
and for the moment i have nothing more to add, so have a nice weekend.
Our main concern is that customers will lose access to their data if they lose their IPv6 connectivity (due to a major route exchange or ISP, etc.) and if pieces of that data are stored on IPv6-only nodes more than necessary to reconstruct a file.
Right.
This is the point. While they are equal, there shouldn’t be any problem for you to host a node, if you decided to join the network.
This is a valid point.
In a best case. Unfortunately, mostly “no route to host” or “90% packets are lost”.
You’re confusing a corporate data center and its corporate connection to Internet exchange points with thousands of small private operators or customers who rely on their internal ISP, unwilling to overpay for IPv6 traffic to some greedy exchange points.
In our world, these IXPs are allowed to dictate their terms.
If we became a centralized data center, things would work exactly the same. But abandoning decentralization to support 5% of IPv6-only nodes (not even customers) doesn’t sound like a good idea.
yea.
those results that we achieved via our support chat with rclone (either rclone-mounted or tntdrive-mounted, same shit) turned to be only very temporary, for just a couple of minutes, then severe timeouts come in and the file never finishes downloading.
so, the verdict would be: storj is maybe something for someone somewhere far away, just not for me here and now.
pity.
mass adoption?
oh forget that, please!
mind you, this is the current ipv4-only storj implementation!
Haven’t said that, but you might be onto something. Storj over carrier pidgeons… that could open Storj to a new set of potential node operators.
More seriously speaking, I’m curious how do you measure the difference in throughput between your IPv4 and IPv6 connections?
by enabling one and disabling the other in adapter bindings on an experimental vm on my server.
dont spoil, because this is exactly what storj currently has!
only that pigeons in my particular area have limited speeds if they want to have static ipv4 for node operation.
as i explained a number of times (or at least tried) is that the old isps that offered static ipv4 in the past are too lazy to upgrade client links.
new fast links have gnated dynamic ipv4 and one ip shared to many clients, only ipv6 is static, probably because of ipv4 depletion problems.
they said they can give static ipv4 to me, only that it would cost me more than it costs to just keep the old slow link from the previous isp.
thats my story as a sno in short, and it applies to other node pigeons in my area of course.
and that why i have slow speed when i act as a client!
And? What numbers do you observe then?
The problem with this idea is that customers who lose access to their pidgeons would lose access to their data, so we kinda have the same problem as with ipv6 here.
with ipv4 it is usually quite stable but very low (20MB/s at most per connection) and ipv6 fluctuates but the average speed is about twice as much.
thats general speedtest.
with storj in particular (and only ipv4) the situation is as i described in the reply to alexey, see my reply to him above.
oh dear didnt we have the same bla bla when transited from x.25 to internet???
just for the record,
i worked at the first x.25 provider here which later became the first commercial internet provider here.
had banks as our x25 users and configured them pretty sophisticated setups with very expensive telebit modems with automatic dbu (dial backup) in case the leased line falled.
so im sort of a dinosaur in packet-switching networking.
but i still welcome improvements that give benefits.
and i am VERY sick of these repeated scenarios.
and also for the record,
i am currently a client to the first bank we had as a client, while that isp ceased to exist because failed to embrace new technologies after we the progressive people have left.
ha ha.
That speedtest connects to the same server?
Yep. That’s the whole problem, isn’t it?