Hacker News Reader

Comment by tannhaeuser | Hacker News Reader

Does anybody have any technical details to share about Gaia-X? I mean service discovery, interchangeable services/interfaces, etc. has been around since the SOAP days (~ 2003) at least. Cloud can mean many things. Does there exist anything at all or is it just talk at the political level atm? Do they intend to fund new research into bringing these things, or will they pick existing tech?

View on HN

Well, if you know German there is this longwinded brochure:

https://www.bmwi.de/Redaktion/DE/Publikationen/Digitale-Welt...

Thanks, that's exactly what I feared and would call strategical/non-technical ;) I mean it mentions "open source technologies", and contains a vague general description of the concept of identity providers, and, of course, platitudes a la "industry 4.0" but nothing concrete. I want to hear about protocols, operating systems/POSIX/Linux, integration into TelCo provisioning platforms maybe if that's on the radar and they want to have 5G on board, payment standards, containers (chroot, VMs, Docker-like/Linux-only namespace containers, or whatever), or even IP networking. Also, going by the stock photos and the use case example, I'm a bit surprised this is targetting industry applications rather than public authorities, schools, medical etc.

This was published today: https://www.ovh.com/blog/gaia-x-catalogue-search-engine-unde...

Apparently more posts are coming.

That's at least something. Unfortunately, this reads more like a program to keep academic research institutes busy and rehashes phantasies of dynamically matching service providers to consumers and policies, something that was a hot topic in academia, and academia only, during the SOAP times, but wasn't used at all in practice (even today you're using lots of yaml to bind stuff on clouds). Moreover, it doesn't use any established logic programming paradigm and re-invents an ad-hoc service ontology/taxonomy and query language (this after the EU spend lots of money in the last decade to fund research on SemWeb ontologies) and they're just using Neo4j (which is fine as a graph DB but not a standard by any means). On top of that, they're using OpenAPI and JSON payloads; this isn't a good fit when the majority of established protocols one would like to use (and that have been created by lots of EU money) are based on WSDL and XML (payment protocols) or even EDIFACT (logistics). OpenAPI/SmartBear has certainly sold their primitive JSON Schema thing well I guess (that just exists on paper or in the form of broken Swagger Java apps). Didn't they want to push European tech? Then why are they promoting gitlab and k8s? I could go on and on, but this doesn't look promising. Tbh, making an undergrad project out of this is embarrassing for French academia who has produced some of the best work in the field.

GAIA-X: Technical Architecture: https://www.bmwi.de/Redaktion/EN/Publikationen/gaia-x-techni...

More on the project: https://www.bmwi.de/Redaktion/EN/Dossier/gaia-x.html

The "technical" architecture reads like an extensively annotated UML diagram about... generics. There are a lot of guiding principles, a lot of references to data sovereignty, but after reading it I have zero clue of what they are going on about.

That's because in such a big project, you have to start with the goals and the guiding principles.

More concrete specs and technologies are being built, but not ready to show.

One of the authors of the "technical architecture document" here. AMA, but I cannot share in-progress work.

Reply on news.ycombinator.com