Backport #24413 by @earl-warren

Fixes: https://codeberg.org/forgejo/forgejo/issues/681

Co-authored-by: Earl Warren <[email protected]>

Backport #24428 by @KN4CK3R

Fixes #24418

Co-authored-by: KN4CK3R <[email protected]>

Many users report that 1.19 has a regression bug: the rootless image
can't start if the UID is not 1000.

#23632 (comment)


https://discourse.gitea.io/t/gitea-doesnt-start-after-update-to-1-19/6920/9


The problem is that the IsRunUserMatchCurrentUser logic is fragile, the
"SSH" config is not ready when it executes.

This PR is just a quick fix for 1.19. For 1.20, we need a clear and
stable solution.

Backport #22423 by @lunny

Now we have `All milestones`, `No milestones`, `Open milestones` and
`Closed milestones`.
Fix #11924
Fix #22411 

<img width="1166" alt="image"
src="https://user-images.githubusercontent.com/81045/212243375-95eea035-a972-44b8-8088-53db614cb07e.png">

Co-authored-by: Lunny Xiao <[email protected]>

Backport of #24419 to 1.19.

Backport #24444 by @silverwind

Remove the
[renderWhitespace](https://microsoft.github.io/monaco-editor/docs.html#interfaces/editor.IEditorOptions.html#renderWhitespace)
override, so the default value of `selection` takes over and makes
whitespace visible on selection.

<img width="128" alt="Screenshot 2023-04-30 at 19 09 41"
src="https://user-images.githubusercontent.com/115237/235366707-d598268d-03cc-4839-a195-c460bddae99b.png">

Co-authored-by: silverwind <[email protected]>

Backport #24487 by @fnetX

On the @forgejo instance of Codeberg, we discovered that forking a repo
which is already forked now returns a 500 Internal Server Error, which
is unexpected. This is an attempt at fixing this.

The error message in the log:
~~~
2023/05/02 08:36:30 .../api/v1/repo/fork.go:147:CreateFork() [E]
[6450cb8e-113] ForkRepository: repository is already forked by user
[uname: ...., repo path: .../..., fork path: .../...]
~~~

The service that is used for forking returns a custom error message
which is not checked against.

About the order of options:
The case that the fork already exists should be more common, followed by
the case that a repo with the same name already exists for other
reasons. The case that the global repo limit is hit is probably not the
likeliest.

---------

Co-authored-by: Otto Richter (fnetX) <[email protected]>

Co-authored-by: Yarden Shoham <[email protected]>
Co-authored-by: John Olheiser <[email protected]>