I’ve been burned by pipenv before on a large project where it was taking upwards of 20 minutes to lock dependencies. I think these days they use poetry instead, but I’ve heard the performance is still not very scalable
With that said, I think it can be a nice addition, but I think it comes down to Python packages not really taking dependency management as a top priority instead of favoring flexibility. This forces a package manager to download and execute the packages to get all the dependency information. Naturally, this is a time-consuming process if the number of packages is large.
On multiple instances I’ve seen projects abandon it for pip and a requirements.txt because it became unmanageable. It’s left a bad taste in my mouth. I don’t like solutions that claim to solve problems but introduce new ones.
I’ve been burned by
pipenvbefore on a large project where it was taking upwards of 20 minutes to lock dependencies. I think these days they usepoetryinstead, but I’ve heard the performance is still not very scalableWith that said, I think it can be a nice addition, but I think it comes down to Python packages not really taking dependency management as a top priority instead of favoring flexibility. This forces a package manager to download and execute the packages to get all the dependency information. Naturally, this is a time-consuming process if the number of packages is large.
On multiple instances I’ve seen projects abandon it for
pipand arequirements.txtbecause it became unmanageable. It’s left a bad taste in my mouth. I don’t like solutions that claim to solve problems but introduce new ones.