James Dyer has a short post on replacing some embark functionality with his own code. He does a nice job and you can check his post for the details but his post again raises the issue of why some users go to heroic lengths to avoid loading packages from ELPA.
Dyer’s stated rationale is that his system is air-gapped so of course he doesn’t want to be pulling down code from the Internet. I’m doubtlessly missing something essential here but that seems to me to present a chicken-egg problem. How do we bootstrap Emacs onto an air-gapped system? And whatever method we use, why can’t it also be used to install third party packages?
To be sure, there’s the problem of keeping things upgraded but that applies to Emacs too so, again, what’s the point of all this?
The last couple of times that I broached this subject [1, 2] some commenters schooled me on why, in certain circumstances it makes sense to avoid third party packages. Perhaps Dyer’s air-gapped system is one of those circumstances but I’m still having a hard time seeing how it solves any substantive problem.
What I do understand is the desire to try implementing these functions yourself. NIH has a bad rep, of course, but let’s face it, every developer can fall prey to it. Of course, Dyer isn’t saying that’s what he’s doing; he’s just trying to configure an air-gapped system.
Dyer is a regular poster and his posts have always made sense so I’m sure this one is no different. I just can’t figure it out.