All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
IPv6 judged by an experienced professional
I just came across an interesting article by Ole Trøan, an IPv6 specialist with extensive experience in network protocols and Internet standards for over 25 years who has been involved in the development of numerous IPv6-related RFCs.
Here's what I consider the heart of it, unedited quote:
128-bit addresses. The original proposal was for 64-bit addresses. Making the address space 2³² times larger than the IPv4 address space was clearly sufficient. Having an address that fits in a native type in computers make it a lot nicer to work with. The 128–bit was a compromise between those who wanted variable length addresses and those that wanted fixed length addresses. A bit like the 53 byte cell size in ATM then. The 128 bit address space has so far mainly led to waste (Google proposes to give every host (or container) 2⁶⁴ number of addresses. Or to semantic addressing, where the address bits themselves is used to carry information. E.g. in SRv6 the address bits are used to carry a segment identifiers, in other solutions IPv4 addresses and ports. Objoke: With variable length addresses combined with semantic addresses, you could just put all the headers and payload in the address itself. What should the IPv6 designers have done?
Stuck with 64-bit addresses.
(emphasis mine)
He discusses other points too which may or may not be of interest for you, but the quoted paragraph IMO is the heart of it and confirms my view that IPv6 should be rejected/avoided.
Please note that he puts it very politely, e.g. by saying "a lot nicer to work with" instead of clearly albeit bluntly stating "what every grown up engineer worth their salt would consider a binding guideline unless clear and tangible necessity requires ignoring it".
Link to the full article: https://ipv6.hanazo.no/posts/ipv6-missed-opportunities-1/
IPv6 morons, go back to the original proposal which was for 64-bit addresses!
(...and you'll finally quickly see better uptake)
Comments
I can get behind this
All IPv4 needed was 2 extra bits. Did Poettering supervise the project?
So, this idiot spammer opened another IPv6 related thread?
The previous one didn't go the way you wanted, @jsg?
As if we haven't gone through your traumas over and over again the past few years, refuting them one by one.
Would they?
Seems to me the biggest barrier to ipv6 is not the address size but rather lack of backwards compatibility. A 64-bit IPv6 could be just as incompatible.
I'm not even sure what backwards compatibility would look like, to be honest, but if IPv6 was backward compatible, adoption would be 100% by now.
Even so, it's not like ipv6 has stalled. I read that nearly 50% of global traffic is IPv6-addressed nowadays.
The percentage of global IPv6 traffic Google sees across all its properties from users did not cross the 1% threshold until 2013. Since then, it has risen dramatically, hitting around 48% at the end of 2024. Going by country, the United States is at 53%, while France, Germany, and India are at 78%, 76% and 72%, respectively. As of 2022, Akamai saw 52% of their US traffic as IPv6 and Facebook was seeing over 61% in the US.
Source
As much as I like the idea of IPv6, the very first thing I did every time when I grab a new server is to disable IPv6...
Not advocating for or against, but surely the 'waste' of the lower 64 bits ultimately became all about SLAAC, which relies on it being deliberately wasted and isn't mentioned here. Mentally block that out and you sort of have got 64 bit routable addressing these days.
Not entirely convinced that actually shortening the address scheme would make it any easier to use or improve uptake. People just don't like change.
Exactly, an anti-spoofing rule is needed in netfilter because the kernel somehow has no spoofing protection for IPv6:
ip6 daddr { ::ffff:127.0.0.1, ::1 } iifname != "lo" rejectWhich is to say, most installations are spoofable by their ISP out of the box. It's definitely intentional.
You may notice outlook on @yoursunny IPv6 list. My friend who use to work at Microsoft told me that his manager and other people at Microsoft hated IPv6. It caused more problems for them as soon as they enabled it on some service. Most notable was a huge increase of attacks (DDOS, brute force attacks, or anything they could try) as you know, Microsoft now have 2^128 more IPs they accept.
I think it's not an 'or' issue but an 'and' issue. But my point is mainly based on core practicalities and those are centered around the 128 bit idiocy.
I don't think so because there'd still be the 128 bit idiocy.
Kind reminder: Google also recommends handing out 64 bit subnets which means they do see the problem with the 128 bit idiocy (but are rich enough to play the IPv6 game) and such Google basically confirms that a 64-bit address space is more than enough, easily (because otherwise they'd not basically throw away 64 bits).
Moreover, the core question is not whether IPv6 has stalled. It rather is why IPv6 - after more than two decades! - still doesn't have an acceptance and uptake of nearly 100%.
I'm rooting for IPv69 ...
A Professionals Guide to IPv6, presented by an actual professional:
https://novalug.org/presentations/15-feb-2025-ipv6/
Luddites everywhere
it should be @yoursunny making the judgement.
I dream of a world where all my idlers a $1/month cheaper due to no IPv4
I'm deeply impressed. So, is it this Lee Howard "Lee Howard Senior Vice President IPv4.Global" (emphasis mine and intented)? Or is it that Lee Howard https://datatracker.ietf.org/person/[email protected]? I guess it's the latter, so let me quote about the "actual professional" ...
and this:
that is, the "actual professional" appears relevantly only in 2012 - which is about 15 years after the "not actual professional" Ole Trøan, about whom the IETF datatracker (https://datatracker.ietf.org/person/[email protected] ) does NOT say that he has no active role currently, but
that is the "not actual professional" Ole Trøan appears almost 10 years earlier as an RFC author or contributor than Lee Howard! Oh and btw. Ole Trøan also has more than double the number of entries. And he is also massively more cited ...
Let's be clear. As you seem to be in a vaguely academic institution you should - and highly likely do - know that "my expert is a real expert, unlike yours" "arguments" are inacceptable and usually avoided so as to not risk looking like a clueless idiot - but oh well, I give that to you, sometimes humans just can't resist to act stupidly, especially when standing on shaky grounds.
But at least don't use an irrelevant low level player when trying to hit at me or at a well established and recognized expert like Ole Trøan!
Try again in a decade, maybe by then you'll have a less poor uptake rate of IPv6 to show.
It seems he tried - and failed miserably and in a way that in academic circles is considered inacceptable and shameful ...
Anyway, the short answer is: NO, he should definitely NOT - and certainly not when a real heavyweight with dozens upon dozens of RFC citations has spoken.
Judgement is issued. This thread is over. Please ask mods to close. The IPV6 troll is right.
I know you have to be here.
It's the former.
He helps corporations deploy IPv6 so that some of their IPv4 assets can be sold through IPv4.global.
That would be a giant waste of time. It wouldn't be compatible with IPv4 and require another stack, but without all the other benefits.
Fucking people who don't know fuck all ranting about shit like they know shit about but don't.
IPv6 must have molested jsg was a kid. His refusal to understand the need for IPv6 is mind boggling.
You are telling the world you are incompetent and don't know how to use firewalls. Don't mention this to your boss if you are employed in the field.
I bet Lee knows how to properly use quotes instead of sounding like he has a mental illness.
Is there maybe an IPv6 fan here with verifiable, relevant and good facts and/or technical real arguments- or has the IPv6 fan crowd only zealotry and personal attack attempts on offer?
I'm asking because this topic concerns us all, yet so far I've seen only one single IPv6 fan, @raindog308, offering a halfway valid and relevant argument...
This topic has been discussed again and again. Even you have opened thread about it in the past. While most of us debated the technical aspects of IPv6, you repeatedly focused on so called "religious" (earth is flat and I am IPv6 victim...) aspects, completely clueless about the topic itself. You have no technical skills, no understanding of the subject, yet you keep bringing it up over and over.
It's been just a few months since some extensive, well elaborated posts about this same topic. Example:
https://lowendtalk.com/discussion/comment/4092615/#Comment_4092615
https://lowendtalk.com/discussion/comment/4092773/#Comment_4092773
...and some others before
https://lowendtalk.com/discussion/comment/3701339/#Comment_3701339
https://lowendtalk.com/discussion/comment/3702469/#Comment_3702469
So what more do you expect from us? To go through everything that has already been explained countless times?
And for what? Just to go in circles with your clueless ignorance over and over again?
How about no? Just f*** off, as annoying as you are.
I'm not a fan per se. I accept it exists. Wouldn't be the first poorly designed protocol we've had to live with.
Maybe "poorly designed" isn't the right phrase. Sub-optimal to be sure, though.
I don't administer IPv6 because I have no personal need for it:
So my practical experience with it is somewhat limited.
However, I don't think it's hard to come up with critiques of IPv6:
I'm sure a networking pro could come up with more.
As I said, there are all kinds of bad computer protocols. FTP, for example. POP3. WEP. Telnet. And whatever Windows does. At this point (actually quite some time ago), I think the IPv6 ship has sailed.
When IPv4 exhaustion was first emerging as a problem, a junior sysadmin I worked with said "just add a 1. to the front of 1.2.3.4 and that would fix it."
The young man was right, 40-bit addresses would have pushed the exhaustion problem 40 years into the future :)
It just pisses me off that whenever I learn something, I'm always just in time to see it being deprecated.
Yes, I also rolled my eyes a bit. "2 bits more" just doesn't make sense. "2 Bytes more" would make a whole lot more sense. After all, 34-bit addresses would on one hand hardly really add significantly more addresses (as in "plenty enough for at least another 100 years") while on the other hand still require processors that, at least back then, simply didn't exist (or seriously impact runtime, which in routers and such is really critical).
And IMO, in fact, 2 bytes more - which translates to about 65000 times the IP4 address space - would indeed have been a good solution, well kind of, because common processor architectures, i.e. the ones used on pretty much all servers, desktops, etc., tend to grow by the square and not just some factor, so the natural successor to 32-bit processors (which, again, were what we had back then) clearly is a 64-bit processor (which indeed is what's common today), so adding 2 bytes would be kind of wasteful.
A 64-bit address space or what I call(ed) "IP5" and what IPv6 originally - and very reasonably - aimed for (see OP link) was the right choice. But then e.g. the "var length addresses!" and other weirdos came along ...
You fool! 2 bits is even too much! Let's be rational (ℚ) for a moment: adding just 1.58496250072 bits to IPv4 would TRIPLICATE the existing address space, and it's barely just over a single extra bit. Basically 33.58496250072 bits! Who couldn't convene on this?
That may well be, but we're not discussing a philosophical or even technological question here. We are discussing an engineering question that is a question with practical premises and practical factors, some of which are at the very least wrt. actual feasibility non-conditional.
But I'm willing to ponder your proposal if you can show me a 33.585 bit processor *g
Besides and in addition to triple the available address space would be more of a crutch than a solution.
So, sorry, but there is one thing the IPv6 designers got right, well at least at the beginning: a true and good solution should (a) offer the largest reasonably feasible address space, and (b) be realistic and be (incl. cost) effective to implement - that's why I call their original approach, 64 bits, the probably best solution and meeting all reasonable conditions.