[Feature request] Support early-init

[Nathan-Furnal]

Hi!

I'm not sure how much work that would entail but would it be possible to support early-init.el files as well?

Thanks a lot,

Nathan

[alphapapa]

Hi Nathan,

Probably not much work, but could you elaborate? What exactly would it do, and what would a use-case be?

Thanks.

[Nathan-Furnal]

Hey, it was mostly to mirror my own config as much as possible. I've mainly used the sandbox to explore one or a couple packages in isolation and get a better feel for them but it's nothing critical =)

[jeffbowman]

To add to this, I tried to use with-emacs.sh to test some crafted-emacs configurations, however, that project makes heavy use of the early-init.el file to bootstrap a packaging system and set some variables (among other things). Just running the init.el file breaks because some variables are not defined when init.el loads.

Using chemacs2 works, but has certain interesting side-effects, some of which have been difficult to sus-out when users report issues. I thought I'd try using with-emacs.sh as a different approach to testing a "default-like" configuration without the side-effects from chemacs2. Unfortunately, it doesn't work since the early-init.el file is not run.

---

Added PR #8 to resolve this issue.

[alphapapa]

To add to this, I tried to use with-emacs.sh to test some crafted-emacs configurations, however, that project makes heavy use of the early-init.el file to bootstrap a packaging system and set some variables (among other things). Just running the init.el file breaks because some variables are not defined when init.el loads.

Hm, well, with-emacs.sh does some initialization of the package system, too, so I don't know if we can be compatible with that. Although that can be disabled with the appropriate switches, so maybe we can.

Using chemacs2 works, but has certain interesting side-effects, some of which have been difficult to sus-out when users report issues. I thought I'd try using with-emacs.sh as a different approach to testing a "default-like" configuration without the side-effects from chemacs2.

Yes, that's the idea.

Added PR #8 to resolve this issue.

Thanks, I'll add a comment there.