Port of spdlog to v1.11.0

[Wmbat]

ci

[Wmbat]

There currently are errors in the CI builds at the Test phase

[Klaim]

Looks like formatter-bench.exe is waiting for specific input. Probably some testscript have to be updated following upstream changes.

[Wmbat]

The CI is still failing: https://ci.cppget.org/@57c6f5f0-49fe-4d3d-a739-727c1d2bae15
I'm not sure why it's not accepting the input of "all" that i'm passing through the testscript i added. Do you any idea of what i did wrong ?

@Klaim

[Klaim]

Looks like it should work... I'll have to check locally.

[Klaim]

There seems to be issues with the symlinks on windows, Ill fix these first.

[Klaim]

I can reproduce the issue now, will attempt a fix.

[Klaim]

Ok the issue was that the testscript was not recognized as such. testscript files need to either be named testscript or have the .testscript extension. Fixing that seems to run the benchmark as expected.

[Klaim]

I just launched a ci request: https://ci.stage.build2.org/@04f3e9c2-4263-4e3f-8fa8-01aa5585f70b

[Klaim]

Welp looks like something else is wrong...

[Wmbat]

we're getting configuration errors now:

error: unable to import target spdlog%lib{spdlog}

There is some problems when fetching the spdlog with git submodules

[Klaim]

Can you reproduce the issue locally? I can't.

I suspect there might be an issue with the test-packages relationships. welp it's not that

[Wmbat]

I can't reproduce it either

[Klaim]

I tried to simplify the fix I did on the symlinks but that's not it: https://ci.stage.build2.org/@31d93a0e-bd43-450c-9bef-76de48d69fed

[Wmbat]

I updated the package manifest file because i noticed that google-bench was no longer in "testing" but "stable".

CI: https://ci.cppget.org/@6cb41d52-e04c-403b-aa18-129ea371696b

[Wmbat]

@Klaim The CI results of my last changes seem pretty good. What do you think of it ?
I see it having trouble with mingw and emcc but otherwise, this looks quite good to me

[Klaim]

Looks like the tests are randomly failing or not? Macos/Gcc seems to segfault, that might suggests some memory usage error so maybe making sure the upstream implementation is supposed to be tested correctly might be necessary.

[Klaim]

IF the issues are from the original code, I guess we can publish the package or wait for a fixed version. If we publish it we can always revise or wait to publish the fixed version too.

[EmilRosenquist]

I accidentally made the "update" without seeing this pull request (#18). But as @Klaim pointed out this already existed.

I tested this and it works. I think unless we want to patch this release with the upstream pull-request that fixed the include (gabime/spdlog#2545). This is good go!

[Wmbat]

If we're all good to go with it. I'll merge the changes, add the tag and publish it to cppget. Doing another round check to make sure everyone is happy with it.

@Klaim
@EmilRosenquist

[Klaim]

LGTM

I think unless we want to patch this release with the upstream pull-request that fixed the include (gabime/spdlog#2545).

If we realize this is necessary we can always publish a revision for the same version.