I believe there’s no chance GCP will be abandoned or shut down. These fears are just HN’s usual exaggeration. Google is pretty ruthless with small projects, but

I believe there’s no chance GCP will be abandoned or shut down. These fears are just HN’s usual exaggeration. Google is pretty ruthless with small projects, but GCP is as significant for them as Google Workspace. Does anyone think they’ll shut down Workspace? De-risking is generally useful though. We’re using GCP as a high-quality cloud service (their offerings are usually well-designed and well-operated), but we abstract away or not use proprietary components. Also, it’s their expertise - they run their own mega-cloud anyway, it’d be stupid not to resell it. So: k8s, managed RDBMS, GCS - yes. Pub/sub: through Apache Camel. And so on. I’d do the same if we used AWS. Also we always got great support from GCP (actual engineers solving problems quickly). What they need to learn better is to listen to their users when it comes to new features I believe. For example I always feel that Googlers are not aware what the perception of their company is.