Daniel Fahlke

> This PR is strictly dependant on https://github.com/OpenMage/magento-lts/pull/2138 because they both won't work if they're not both merged. why is it necessary? Not standing in a way, but I would...

we build a release package, which should replace the current .tgz download. It should be as runnable without needing to execute composer. It should support the "upload via FTP GUI...

first one does not work because it takes the config out of the composer.lock file, so the second one would probably be the best. I would do it analog to...

there would also be an option to use the magento Flag Table to store the Locks there

the main problem the locking solves are incomplete or non-existing flat tables. I have seen "table not found" issues in production environments, because the locking did not work for non-cron...

a config which makes the used locking configurable sounds good for me, then people can decide if they use "file", "none", and we can add other/better ones later, and still...

sadly our resources(people) were not able to fix the windows tests yet... therefore we are not able to validate if it properly works or breaks something :/

Technical, what you are removing in this Pull Request is the compatibility Layer. They do not contain any logic, no potential for any Bugs, just pseudo classes existing for keeping...

Hi, Sorry for the late response. getting such a big result for loadAcl definitely seems wrong, especially during login. Do you know where most of these are coming from? Do...

So what is the suggestion to get reproducible builds for gitpod? Ignoring the platform requirements might cause some compatibility issues, but at least is reproducable. Having no composer lock might...