After having a strange time with vim, I found that my scoop shim for vim was pointing to a vim.exe from within the gow package. However, I also have the vim package installed. I've rectified the shim config and .ps1 files so that I get the vim I'm expecting, but I also expect that this could be destroyed by a subsequent update, so I'm wondering if there's any way to tell scoop that the shim from one package should take priority over others so that it doesn't get wiped out?
I would imagine that this problem isn't limited to my particular instance.
scoop reset vim
I would imagine that this perhaps resets to the last one that was implemented before now -- is there a way to _always_ pin a shim? What if I install 3 things in succession which provide the same utility? And updates after the fact? The onus is on the user to keep track of links -- which sounds like something the system could rather do.
If it's not a current feature, I'd like to propose it -- I can open a new issue or rename this one to make it clearer, if you like? Any package management system has to, at some point, deal with conflicts -- it would be really nice to have a deterministic solution to that (:
I'd love this too - it would be useful to keep multiple versions of say, PHP or Python up to date, whilst maintaining a default version in my shims so it doesn't change behaviour on my $PATH.
It's also a problem between sudo and psutils, and between python and python27, and between gow and coreutils and grep
Most helpful comment
I would imagine that this perhaps resets to the last one that was implemented before now -- is there a way to _always_ pin a shim? What if I install 3 things in succession which provide the same utility? And updates after the fact? The onus is on the user to keep track of links -- which sounds like something the system could rather do.
If it's not a current feature, I'd like to propose it -- I can open a new issue or rename this one to make it clearer, if you like? Any package management system has to, at some point, deal with conflicts -- it would be really nice to have a deterministic solution to that (: