So far, I have found CloudCraft from Datadog, but this only supports AWS and its automatically diagraming is still in beta (AFAIK).
I am considering building something custom for this - but judging from the lack of tools that support multi-cloud, or only support manual diagraming, I wonder if I am missing some technical limitation that prevent such tools form being possible.
I may be oversimplifying, but if I look at Web3 as the existing internet, but without central intermediaries; then most applications can be built this way today. If we want a Clubhouse, or Zoom clone, these could be built atop existing WebRTC infrastructure. A Whatsapp clone can be built atop libp2p messaging (again I oversimplify). From the internet infrastructure view, the part that was missing has always been mechanisms of identity, ownership, and payment (trust mechanics in general). Today, blockchains handle these in a variety of formats, such that a project can just pick the chain that fits best with the application architecture. In 'Web 2.0', if I want auth, payments, file storage, etc - I would simply stitch together one or more infrastructure providers such as AWS. This feels very natural today.
Is it possible that blockchain has already reached maturity for the value it brings to the decentralized web? Are there any popular applications today that could not be built peer-2-peer in this context?
The big gap I still see in this view is who pays for services that are free in Web 2.0 (such as Facebook). But even in this case, blockchains provide a variety of incentive mechanics where increasing usage could appreciate a digital token, which in turn gives incentive for people to run the underlying infrastructures.
I also feel crazy in that there seems to be consensus that blockchain is suppose to do everything, or should be doing more than it does. In all of this, I am viewing the problem space from the domain of infrastructure, agnostic of application. Is blockchain tech already everything we need it to be?