And yet they managed to make another complicated cloud service.

Would be nice if someone would cater for Startup owners that do not have time to study documentation and overlook multiple processes just to run a simple bot etc.

GCE is much simpler to use than EC2. Google came to the party late and they have seemed to take some time to not replicate AWSs annoyances.

The problem for me is that "much simpler" is not good enough. With DO for example when I go to their panel I never used it before it will take me tops 1 hour to find all the possible options, all perks and cons, everything. In just 1 hour.

I know, I know, AWS and Google Cloud is not DO and I agree. But would it hurt to introduce DO or Vultr style UX and make it easier for regular user to launch what he wants without all the expert language and need to jump in to documentation for hours?

Would button "Simple" and "Advanced" UI be really that hard to make?

I, as startup owner, dont really have time to focus on all the noise from large cloud providers, its easier to just launch 5 VPS with DO or Vultr and create my own infrastructure. And while I grow I doubt I will move my business to any of the cloud providers because I dont have time for it. Its easier to just build infrastructure on VPS's for $5 or just go for dedicated servers for backend.

Would you please care to share learning resources about GCE other than the official docs?

I've moved our infrastructure over to GCE and basically just clicked around in the interface until I had what I wanted. It's really quite easy (although I'm only using compute engine and bigtable right now).

I got 300$ in credits on every google account when I first vistited which allowed me to play around with their big data stuff and some public datasets, which was nice.

If only ...

If only there were cloud products and services that were dead simple and just did one little useful thing just the way you wanted it.

If only ...

That is nice in theory but nobody really wants that. In practice I do not want to run my app on 40 different services potentially owned by different companies. If you mean that inside one company is fine than I believe ec2, s3 etc seperately already provide what you say: if you would make them simpler it would get annoying as you would have to set up yet more services (within the same portal, which is a bit better) to get stuff done. I think the theory is nice to start off but I do not think there is a real market for 'dead simple with one little useful thing'. Then again: try it, who knows.

"Then again: try it, who knows."

You know what, I think I will!

What should I call it ?

What about when folks want lots of useful things?

Then they are outliers, and market themselves out of the 80/20 customer segment.

If the ads business goes to shit, the cloud business takes a beating, also.

Cloud is basically SaaS which is the antithesis of the advertisement model.

But okay, I guess you must know better.

I believe the person to which you replied meant that a lot of cloud platforms/hosting revenue depends on ad-funded customers. If the customers go out of business and aren't replaced, subscription revenue suffers.