More and more customers are looking for solutions that could keep their data as near and safe as possible. I view this move as a very logical economic opportunity.
Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
281–290 of 420 posts
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#282Earlier quoted context omitted.
At a personal|SME level? Who might want do DDOS John Smith or Pop's store? How much iron you need for such usage? That's without counting the various "datacenters horror stories" about how badly many providers are really without appearing. My homeserver is a simple NixOS, so I have to maintain just a config, easy to replicate anytime without manual setups or complex orchestration, it's a small celeron machine with 32…
I won't bore you with the details of my home setup, and I do host from home with both static IPv4 and IPv6 and everything (ok ok, two details: it's all freebsd on enterprise hardware) I will repeat: it is an option for some. and if you fall in the small group that has everything on green for a home hosted setup, it is the best option. But thinking this is for more then a very very small group of people is not confirm…
Well, while both option exists and obviously anyone is (almost) free to choose, I still fail to see convenience in living on someone else computer. Anyway, allow me a different scenario: you are Foo Bar, you have a bunch of documents and many photos/videos/music etc no computer skills beyond clicking around. You ask someone more knowledgeable, local or remote, but still a single human, if he/she can create your infra to own your data. he/she gives you a list of stuff to buy, instructions to assemble or came to you/send it assembled and ready to you, a usb stick with a live system to deploym the config (NixOS/Guix system) on it. You are now operational, your infra it's still a black box for you but you own the config, so you can give it to someone and in case of trouble or the need of changes you know who to contact. You pay a certain capex and small opex. Your infra evolve following you. Your contact disappear, another came and propose to rebuild anything, no data loss, he/she can use your iron, your data and knowing anything from the config. You are on again and the event loop keep running.
How different is from choosing let's say FileHoster inc who works well enough, albeit much less than your infra, and have no capex, than it experience a big issue (cfr. Gandi/FR two times few years ago) all your data are lost, you still have some here and there you get up again on another one, than it became too expensive, you switch from another, ...
In the two scenario:
- on one side you spent in capex more than in opex, so you spent at a specific point in time, in an inflationary economy, instead of being vampirized every month;
- you are tied to your infra reliability and consumer grade assembled iron it's pretty reliable for such usage, as most well know giants are for the same usage/point of view BUT on one side the reliability is in your infra, something you can tune, learn and check, on the other it's about third party decisions who can happen at every point in time without anything you can do.
You've certainly read countless of time about $BigName impromptu ban for instance. You've certainly experienced terms of use unilateral changes, sometimes ok for you sometimes not. Where is the balance?
If you live in a Korean goshiwon -alike you can't host while you might still have docs/photos etc you have no choice, but when you have it and honestly MOST people who need IT at a certain level do have the choice, at least formally, if he/she knows it exists, then where is the convenience? Trusting the market on one side, with little to no capex but much bigger opex and uncertainty or more capex, less opex and uncertainty? Think about schooling: what we do in most part of the world? A big initial capex (long school time) than profiting for life on the acquired knowledge or a jump in the wild than we will learn on the go?
In the mean people are honest, so trusting someone else IF you can verify or the exposure it's low it will generally end up well anyway, otherwise... I trust myself more than someone else shielded under a corp name, eventually in another country (so legal protection issues, geopolitical risks and so on)...
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#283Earlier quoted context omitted.
> EU-US relationship suddenly turning very hostile The economies are too integrated for this scenario. There are big R&D centers of many American corporations in Europe, doing big part of their tech. They won’t move, instead in a very hypothetical “Russian” scenario they will change owner and continue to operate. I’m sure EU regions of AWS or GCP won’t cease to exist, for example. There are some services that are ope…
At a personal level, it feels better to have something of your own to hold on to instead of someone else’s, so I think beaucrats will also respond accordingly. In a way, we’ll probably see more cloud fragmentation in the future, especially as other countries develop their IT sectors more and feel like they want more control over their own infrastructure, and whatever tertiary benefits can be extracted from that. Rela…
Let‘s not call it „cloud fragmentation“ please. It’s cloud competition. Cloud is an utility and probably should resemble energy market regarding the choice of suppliers and simplicity of the switch.
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#284Earlier quoted context omitted.
They produced a few fluffy documents in 2022 and then nothing happened. They repurposed the word milestone to mean agenda. It's just a list of events they're organizing. Because they have no actual milestones or goals.
I've joined some large EU efforts in the past, and it's always like this. Lots of different parties involved focused on producing tons of absurd documents, and nothing else. Some have good intentions, but it doesn't matter. There's a great thread on X now discussing the same topic: "25 years ago each major US company had a German and/or French equivalent. Today equivalents of US tech giants are in China and Europe is…
Here is what happened: fear. Fear of patriotism getting us a second Hitler. Fear of war.
This is IMO the root cause of why most public services are going down the drain in most Western Europe: people are there to work for themselves, not for their country. And the higher people are, the worst it is. Keep the status quo, embezzle if you can and shut your eyes to not see we're in economic and cultural wars against the rest of the world.
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#285Earlier quoted context omitted.
Hetzner and OVH are top of mind, Gandi is nice too. Not Amazon-scale, by far, but European companies hosting in Europe with decent service.
OVH is a joke (their data center burned because they had wooden roofs), Gandi is no more, and Scaleway gave up. There is no French host anymore. Only Hetzner is left in this business.
The fire occurred in their old datacenter, built in an era when OVH was aggressively cheap and experimental. It is in no way representative of today's OVH.
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#286I run a European cloud service, 80% of our customers are basically looking for a European alternative to the big clouds. The market is huge and in my opinion underserved. What makes it very exciting is that there not too much innovation required to compete
> there not too much innovation required to compete
Yeah, sure.
This is the kind of mentality which lead us to "Cloud" being developed by an online book seller and not hosting companies. Then those hosting companies needed 10 years for open source solutions to be available to close some of the gap. Next time a random company disrupts the hosting market, it won't come from hosting companies because "there not too much innovation required to compete" and those companies will wonder what is going on.
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#287Earlier quoted context omitted.
While most Western governments have gotten increasingly good at communicating with their citizens (i.e. making their web sites and forms accessible in human language, rather than bureaucrat language), and often even go so far as to offer versions in "Simple English" or local-language equivalents, the EU seems to be going the opposite way. I'd consider myself reasonably accustomed to and able to deal with bureaucracy…
Because the EU is not a national government. It issues no passports. It has no citizens. It levies no taxes. It has no army. It's an organisation that coordinates sovereign states. Often it doesn't even set the law directly but establishes a framework that allows it to specify some requirements that national legislative bodies then have to turn into actual legislation. Frameworks for how to talk about things is very…
It's amazing the EU has lasted this long really. The USA tried something somewhat similar back in 1781, and it was a complete failure: they organized a bunch of sovereign states (formerly colonies, but they became sovereign states after the Revolutionary War just before) into a confederation, where the central government had no real power at all. The resulting country couldn't even defend itself against pirates. They finally got sick of it in 1789 and threw out this form of government in favor of a constitutional republic with a much stronger federalized government. Over 235 years later, that form of government still persists, though it's really showing its warts and the Constitution really needs a rewrite IMO, but despite its enormous flaws in the modern age it's still a lot better than the decentralized mess that is the EU. If you want real economic power in the face of competing superpowers, you need centralized policy and authority, not a bunch of semi-sovereign states all squabbling with each other and no one able to make a decision.
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#288Earlier quoted context omitted.
The core feature of Gaia-X is that it makes it easy for small EU software companies to get up to €200k in "de minimis" tax gifts. It also funnels money into open source office software and "European data spaces", which is (very loosely) tools to replace US clouds with open source. I would say it's doing exactly what it's supposed to do, making it easier for EU companies to ditch US cloud services. And didn't exactly…
Need a third party citation on that. I'm highly skeptical by the press releases they bring out and seem to be taken over verbatim with very little due diligence. Given the fact they don't even allow public signups outside of DACH im highly skeptical of their claim they're doing billions of revenue in public cloud. It wouldn't surprise me if there is some interesting bookkeeping going on to boost the numbers. The curr…
Cloud&Heat receives GaiaX funding: https://www.cloudandheat.com/news-press/gaia-x-summit-2020/
Teams up with Schwarz group to create StackIT cloud: https://www.handelsblatt.com/technik/it-internet/schwarz-gru...
As for the revenue, I would trust a multi-billion international company that if they say billions in revenue, that'll be halfway accurate. Or else, that would be massive securities fraud. Here's their (independently audited) tax filling showing 1.2 billion € in revenue 2 years ago: https://www.unternehmensregister.de/ureg/result.html;jsessio...
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#289Earlier quoted context omitted.
The US government can still get at that data, because Amazon is still an American company. It doesn’t matter where the data actually lives.
There's published proof the US government actively will go to the extent of having submarines go to undersea fiber cables to tap into them. They have private lockers in almost all the datacenters in the world, etc. Doesn't matter if it's an American company or not.
Yes it does matter is the company is subject to US law or not. There is not 100% security but saying “it does not matter” is manichean and does not reflect reality. Yes the US have a very long arm but it’s all a cost/benefit even for three letter agencies.
Every time they make use of a zero day or a backdoor they run the risk of it being discovered. The harder it is to get a new one, the more they will think twice about using it for mass and low stakes surveillance. A non-US company will be less inclined/forced to cooperate with them, making it harder for them to siphon data out, hence lowering probability.
No one is 100% safe, agreed though. Probabilities and threat models is all we got.
Re: Lidl's Cloud Gambit: Europe's Shift to Sovereign Computing
#290Earlier quoted context omitted.
There's no way I could describe it without sounding conspiratorial, so I'm just gonna lean into it: imagine the EU-US relationship suddenly turning very hostile (like Russia-EU relationship already did), how long would it take before all of EU's tech suddenly stopped working? Not saying that's gonna happen or is even likely to happen, but I don't think investing money into bootstraping some sort of a backup is necess…
> EU-US relationship suddenly turning very hostile The economies are too integrated for this scenario. There are big R&D centers of many American corporations in Europe, doing big part of their tech. They won’t move, instead in a very hypothetical “Russian” scenario they will change owner and continue to operate. I’m sure EU regions of AWS or GCP won’t cease to exist, for example. There are some services that are ope…
> The economies are too integrated for this scenario.
This exact reason was put forward by political thinkers in the 1910s to "prove" that a major war in Europe was completely impossible. It was then put forward again in the 1930s to "prove" the same thing; for real that time.