[documentation] Overhaul for 0.4

[magsol]

This is the checklist of documentation changes we'd like to make for 0.4.

Notes on web docs that don't make sense, aren't clear, or need expanding:

• QHub Architecture diagram (e.g., conda-store, rust-analyzer; some high-level suggestions) #1031
• Mention of CentOS should be struck from docs (as their images have been removed)
• Troubleshooting section (start with devs, mainly with upgrade paths) #1032
• [documentation] Documentation on conda environments failing to build on first try #1036
• Upgrade to 0.4.0 [documentation] #1048

Some existing documentation related tickets:

• [documentation] Keycloak docs #999
• [DOC] - Audit docs and adapt to Diataxis framework #975
• [documentation] Review QHub architecture diagram for Azure Provider #971
• Help Installing QHub on existing EKS (AWS) Cluster & Missing Local Testing Docs. #935
• Docs on how to better allocate resources for each cloud provider #796
• Documentation on kubernetes resources and functions for each provider to calculate resources #792
• Create Personas for QHub #634
• Create docs for destroying manually on GCP via console #606
• [documentation] Revise the tutorials section of the docs #487
• [documentation] - 101 Qhub #301

[iameskild]

I believe we can close #935, see docs here:
https://docs.qhub.dev/en/latest/source/installation/existing.html

[iameskild]

Given that CentOS images have been pulled out, we will need to remove their references in the docs:
https://docs.qhub.dev/en/latest/source/user_guide/experimental.html?highlight=centos#centos-jupyterlab-and-dask-worker-profiles

EDIT: Removed in #1003

[magsol]

@iameskild I'm realizing #935 seems less like a documentation issue; in fact it seems like the issue is still active, with the last post from the OP on Dec 17.

Has anyone reproduced this error since?

[iameskild]

@magsol I tried reproducing the error awhile back but I was having trouble doing so. Let me follow up with OP and we can go from there.

[iameskild]

With the release of v0.4.0 and issue #1076, I believe we can close this issue.