Windows CI failing

[philbucher]

Dear @KratosMultiphysics/all

You might have experienced failures in the Windows CI since yesterday evening. This is because Github no longer provides boost in their environment (see actions/runner-images#2667)

The @KratosMultiphysics/technical-committee is aware of this and working on a solution.
Unfortunately no PRs can be merged until this is fixed, sorry for the inconvenience

[loumalouomega]

At least we depend on boost as header only, it will be relatively easy to solve (but I assume that CI will take a little longer because of the time of downloading and decompressing boost)

[armingeiser]

At least we depend on boost as header only

Could we not provide boost headers the same way we do for pybind, eigen, etc?
Or is there a specific reason to not do so?

[philbucher]

At least we depend on boost as header only, it will be relatively easy to solve (but I assume that CI will take a little longer because of the time of downloading and decompressing boost)

thats essentially what we are trying to figure out right now

[philbucher]

At least we depend on boost as header only

Could we not provide boost headers the same way we do for pybind, eigen, etc?
Or is there a specific reason to not do so?

I think the main reason for this is bcs boost is quite big / has many files
Also we cannot only use ublas, there are many cross-references unfortunately

[philbucher]

The problem has been fixed

Please merge master in order to fix the Windows Build in the CI