Matt Davis
Matt Davis
Want to check the behavior of the newly vendored `pip==22.2.1` that is in main against this ticket because I saw something odd working on vistir where re-locking changed the markers...
I believe this is not an issue on the main branch as we force the usage of the vendor'd pip which eliminates the `using pip VERSION_X` with pipenv, as it...
@gitpushdashf Well technically it was vendoring of pip that fixed it and then we of course continue to vendor in new versions of pip as they are released. We recently...
@cclauss how would the `--use-local-pip` cause it to be less chatty? -- I think it would be more chatty because there would be more versions of pip in use for...
@gitpushdashf I've run `pipenv lock` on the pipenv project several times on the new `2022.8.31` and haven't seen the markers change yet -- wondering if you could check again?
@dqkqd are you able to reproduce it with `pipenv==2022.9.2`?
Awesome test project @dqkqd !!! On `2022.9.4` I am getting `same` each run: ``` matteius@matteius-VirtualBox:~/pipenv-triage/pipenv-4101$ docker run $(docker build --quiet https://github.com/dqkqd/pipenv-issues-4660.git) pipenv, version 2022.9.4 attemp 1: same attemp 2: same...
I got what you got too on `2022.9.2` -- so I think this is truly fixed now. My belief is that the sorting helped a lot, but there was still...
Perhaps too early to tell, but I maybe am onto something about this issue -- I removed the `lru_cache` decorators around the marker generation in requirementslib in this PR: https://github.com/pypa/pipenv/pull/5373...
@jfly I haven't experienced this issue before in the last months of running pipenv tests locally on almost daily basis. Could you re-check this behavior with `pipenv==2022.4.21`?