v1.0.0 release: tarball creation/installation, plugin symlink management, release management

[valassi]

Hi @oliviermattelaer this is a WIP PR for tarball creation/installation and plugin symlink management

The part that is done is the usage of the MG5aMC_PLUGIN trick for the symlink management (nice trick, thanks!)

What I am still testing (also in the CI here) is the package creation

NB there is not yet any renaming of plugin directories or of repositories, this can come later

[oliviermattelaer]

Perfect nothing to comment on.

But please can we do a squash here?

Cheers,

Olivier

[valassi]

Hi @oliviermattelaer thanks!

But please can we do a squash here?

Sorry, no. Amongst other things, there are some test results that are only visible as individual commits (I then build all my summary tables for cnferences based on those). Eventually I/we need to find a better way, but so far, no.

And in any case I really prefer to see the full history, in case we need to go back and check some stuff. You might not believe it, but I am already squashing a lot (there are here around 100 commits visible, but in reality these were maybe 400, debugging the CI was really complex, and I want to keep track of that debugging).

Also for HEPInstaller by the way I would really prefer no squashing. Same thing, at least I can easily find back if ever I want to bypass info.dat (not that I think this will ever be needed, but one never knows).

Can I merge like this then? Thanks!
Andrea

[valassi]

PS Sorry I know I am difficult :-)

But this point easily gets very very very religious...!
https://dev.to/wesen/squash-commits-considered-harmful-ob1
Quote "A recurring conversation in developer circles is if you should use git --squash when merging or do explicit merge commits. The short answer: you shouldn't. People have strong opinions about this. The thing is that my opinion is the correct one. Squashing commits has no purpose other than losing information. It doesn't make for a cleaner history."

I am not saying "my opinion is the correct one", but definitely I do have a strong opinion that squashing is often bad. It will not surprise you that I prefer too much information than too little information...

Then I do squash sometimes, my own commits. With other people's commits I would even be more reluctant to squash, I want to be able to trace back what they did, if/when needed...

[valassi]

Thanks again @oliviermattelaer ! As discussed, merging now.

Then I will bump up the version in a later PR or commit

Andrea