Thread by @halvarflake

wired.co.uk/article/europe… - I should read up about the details of this, but it seems a typical DOA project. A thread:

The reason why Europe has no "proper" cloud providers is because Europe failed at producing an Internet giant. From the limited data sample we have, becoming a cloud provider requires two things: (1) Huge Capex and...

... (2) cloud-scale software to actually run. Amazon, Google, Microsoft, Tencent, Yandex all fit the bill. There is no cloud provider that did not satisfy both criteria.

Trying to build a cloud provider without (1) and (2) will fail, for multiple reasons:

A) The CapEx required in Cloud is beyond what any government can (and should) "subsidize": AWS and MS combined CapEx is ~$40bn, approximately Germany's defense budget. In order to compete effectively, a European cloud will need $10bn+ annually, for many many years.

B) Building a Cloud requires cloud-scale applications. The infrastructure layer and orchestration is only going to be "right" if you have quick feedback cycles; there is a reason why all of the relevant players in Cloud extensively eat their own dog food. If you don't have ...

... that, you will build a cloud that nobody wants to use.

I mean, I am one of the most pro-European folks you will find, and left-of-center to begin with, but I tell you: The only way to get a useful Euro-Cloud is to have European Internet Giants.

The way this project looks to me right now is the usual (sad) story of a bunch of incumbent Telco operators with a few universities siphoning off cash into white elephant projects.

There's also the issue that the actual solution to building a European Internet Giant is sufficiently radical that nobody is willing to discuss it; so better to light a few hundred million $ on fire and then claim things failed :-)