I don't have firsthand experience but I do know that IBM mainframes (Z-series) are still actively developed and lifecycled. Whether this is purely financial (as in: we keep the customer locked in to our ecosystem regardless if there are alternatives available which offer the same level of robustness etc.) or the platform really is better suited to the specific requirements set by banks, government agencies etc. remains to be seen..
Let's hope a decision on an alternative platform is made quickly (although I am sceptical due to the bureaucratic maze of NATO procurement). The current AWACS-platform (E-3A) still uses old engines which are loud, smokey and not as efficient. Neighbouring communities deserve a quieter platform as well as NATO needs a potent airborne early warning platform. European technology is not an issue (Thales, Leonardo etc. make great modern AESA based radars). Also interoperability of data between NATO members used in the C2 chain is not an issue.
Ow wow, worked ISP support in 1997 - 1999 and indeed recalled having to reinstall TCP/IP stack and tweaking setting to make stuff work. Next to that, gruesome installation of ISDN drivers (which technically were out of support for us) for a specific brand of ISDN modems the incumbent telco provided to customers.
The challenge of implementing this for internet routing is that you'll probably need a whole new protocol implementation as part of either BGP (currently the protocol responsible for Internet routing between networks) or something entirely new. Let alone that BGP is a path vector protocol and not a link-state protocol that uses Dijkstra (like OSPF and IS-IS).
It might optimize internal routing but getting this standardised across vendors etc. is not impossible, but probably takes a long time to standardise/govern etc.
The reason why we had something like the Delta works was due to a 12 Bft storm hitting our shores in 1953. This infrastructure is built to withstand these kind of storms and protect the land from flooding. The protections in place (movable doors in storm surge barriers etc.) are used a few times the last decades when storms did hit our shores. I don't know if this is useable in the Florida context. It's easy to say whenever a big hurricane hits the Florida shore: "Yeah, just ask the Dutch to fix this.." And I am sure some smart guys from our tech universities can pull it off, but you need money and political will. And it literally takes decades to built it.
Also, you need a place to build the dike. Look at a map of Miami, for example, and tell me where you want to build a dike. In front of Miami Beach? And how far does it go? All the way up the coast? There's 120 miles of continuous city on the Atlantic Coast. Also, the land is all very porous sand on top of porous coral. Even if you build a dike out of clay and concrete, water will still seep in from below. This is already happening at high tide.
Already happening at high ride? That is not a new phenomenon even though it is played up as as such.
Places like St Augustine, Fl or Alexandria, VA; and, although not a city, even Jamestown, Va all have records of regular flooding since their establishment centuries ago and well before the Industrial Revolution during strong king tides when you get a confluence of effects like the moon and the sun’s tidal forces amplifying each other, rains have swelled waterways and saturated ground, and the fact that they are situated and basically at water level. I’ve experienced it personally in a few places, ands considering that those places built a long time ago clearly have structures built to accommodate strong king tides is an indicator to me that they knew it happens every once and a while even before the Industrial Revolution.
Usually low-lying areas protected by dikes also deploy large pumps to pump out water that does get in. My guess is that it would be cost prohibitive to keep on pumping the water back out of Miami in 2075, but that is just a guess. There are probably people who know for sure, and it would be nice to read what they have to say.
It's bandwidth versus latency. Two different aspects. A live feed going one way works fine over a high('ish) bandwidth link as long as there's no need for interactive (back-and-forth) traffic. See it as a highway. The highway can be very wide with lots of lanes (=bandwidth) but to get from point A to B can cover a large distance (latency)
You probbaly need a specialised crew to do this and as such such fiber won't be installed in your own neighbourhood for your Fiber-to-the-Home connection anytime soon I guess. But, maybe in a few decades it will.
When such technology becomes practical for the large telco's it will be implemented soon as this saves on attenuation equipment.
The termination/splicing of HCF would likely occur at long haul endpoints, not in neighborhoods and last mile. There wouldn't be any meaningful upside to this. Crews are currently splicing fibers reliably in <5 minutes using gear from Amazon or AliExpress. We don't want to mess with something that is working this well.
If you want to compete you must also invest in associated PaaS and SaaS-services supported by a development framework to code and integrate it all. As mentioned, getting an IaaS product is not a problem now, it's about getting the developers onboard.
Isn’t it? Iaas is a lot more than just VPS and a few PaaS offerings, you need object storage, a command line and more. Last I looked the European cloud vendors were very lacking.
SAP has the BTP and all their different products, so IaaS is very much the issue for them and I doubt they have solved it without the help of Microsoft.
