I feel like that’s an instant turn off. I don’t want to manually manage all my stuff.
Or is there some middle ground where people just put a URL to the download in their package manager solution, whatever it is.
I feel like that’s an instant turn off. I don’t want to manually manage all my stuff.
Or is there some middle ground where people just put a URL to the download in their package manager solution, whatever it is.
I only really use npm packages for Electron; I usually have the libraries locally stored and don't do any soft preprocessing/packaging. Being able to "drop in" the source much like a C/C++ library ends up being comfortable for me for these smaller things. [1]
[1] https://github.com/danbolt/switch-compo/blob/820ce5fb4944e57...
Presumably you could do that here too, but I don't know how this host feels about hot-linking
<script src="https://cdn.jsdelivr.net/gh/bennettfeely/bennett@master/ztext/js/ztext.min.js"></script>
(You should probably link to a specific commit, though, like: <script src="https://cdn.jsdelivr.net/gh/bennettfeely/bennett@24f27b0f902821147ef726735b2717f27335c282/ztext/js/ztext.min.js"></script>
.)Although, this point is moot as soon as someone packages it for whatever package manager you use. No matter how hackily they do it. e.g. if there were a npm package called "auto-import-arbitrary-or-versioned-file-from-git-repo-or-http-url" that consolidated _all_ of your random single-file imports, that'd count IMO.
Of course, it's good to have the freedom to go both with and without a package manager and build tools, since this doesn't scale at all.