You didn't quote the rest of the sentence: "when most of the current base have no financial interest in converting". That applies to both IPv4/IPv6 and IPv4/modified-long-address-IPv4.

Backwards compatibility is difficult. A 32-bit IPv4 end system can reach 2^32 destinations. Those destinations must be shared between unmodified IPv4 end systems and the new expandable IPv4 stack. This makes having an expandable stack of questionable value until a big fraction of end systems have converted. So the success of the program depends on persuading businesses who have enough addresses to convert their end systems to the new system, even if it doesn't give them any short term gain. As well as the end systems, every backbone router on the internet between them needs to understand the extended protocol with bigger addresses. The global IPv4 BGP table needs to be extended. But that takes some serious consideration if backwards compatibility is a requirement. There will be routers with the current 32-bit routing table, and they will get a fraction of the new table with the 64-bit (or whatever) addresses excluded. Likewise DNS will need some kind of old/new distinction, and the old unmodified DNS servers have to be kept in the loop. Businesses, who have a system that is perfectly working already, have to convert their back office systems relating to IP address allocation, their DNS, their end systems, their routers, etc..

This sounds like the problems of IPv6, where progress depends on a lot of people with no particular motivation, as they already have enough addresses. But it also introduces an interesting wrinkle, which is that the new systems must fully interoperate with the old unmodified systems. There's an assumption that you can't rely on _everybody_ to move to the new system, so every single change you make to the new system has to work nicely with the installed base. That is quite different from dual stack, where the new system only has to interwork with itself.

The slow rollout of IPv6 demonstrates that _any_ new system has a hard time being established when most of the current base have no financial interest in converting. 32-bit/64-bit IPv4 (or whatever) has that obstacle _plus_ any exciting interworking issues breaking the 32-bit network.

Slightly longer than that, as IPv6 RFCs came out in 1995-1998. But the address space advantage of IPv6 only kicked in in the early 2010s when it became much more expensive to obtain IPv4 address space. In 2011 Microsoft piad Nortel $7.5m for IPv4 address space ($11.25 per address). It was only a matter of time before that sort of cost got passed on to customers by companies like AWS.

Yes, I did miss the point you were making. From this article it looks like 10 minutes of acceleration to get to 0.2c, which I think is an acceleration of 17000g. I have no idea if this is feasible. It should be possible to reach high speeds with gentler acceleration, though. For example, 120 days at 1g should achieve the same speed as 10min at 17000g.

Electrons can go a lot faster than that. If you plug 100GeV from the Large Electron-Positron Collider into this calculator you get more than 99.9% of the speed of light. Likewise for lead ions at >500TeV in the Large Hadron Collider.

That can't be the only thought in their mind, though. To quote Thomas Aquinas: "If the highest aim of a captain were to preserve his ship, he would keep it in port forever." They are asking for the MAX7 to be retrofitted at a later date, just like the MAX8 and MAX9 currently flying.

On the other hand, relying on people to remember to do things has a bad track record.

Not clear why parent post was modded "Flamebait"....

That also goes for the companies needing my two longest cancellation calls: Vonage and TiVo. The TiVo call was very impressive as I had no cable service at the time, so there really was no earthly reason to start paying for their service at the end of the initial contract.

There are probably some eternal beings logging our requests to make tech like it used to be. When we die, we'll be presented with Netscape 3.0, or Win95, or a phone with no maps available, or whatever else we've asked for.

Two weeks later we realize... "THIS is the bad place!"

There wasn't much AI text generation back in 2015. [ a previous comment by Spugglefink ]

There has been a terrorist action already, in 2015. Former high school classmates Anthony Sadler, Alek Skarlatos, and Spencer Stone were made Knights of the Legion of Honour for their part in subduing the attacker.