ok krw@ fwiw. :-)
....Ken
On Wed, Sept 21, 2022, 8:05 a.m. Jeremie Courreges-Anglas <j...@wxcvbn.org>
wrote:
> On Wed, Sep 21 2022, Kenneth Westerback <kwesterb...@gmail.com> wrote:
> > Kenneth Westerback <kwesterb...@gmail.com> writes:
> >
> >> Kenneth Westerback <kwesterb...@gmail.com> writes:
> >>
> >>> Nam Nguyen <n...@berkeley.edu> writes:
> >>>
> >>>> Kenneth Westerback writes:
> >>>>
> >>>>> Alas, I can't make the emacs-28.2-gtk3 I downloaded from the link
> jca@
> >>>>> provided work. :-(
> >>>>>
> >>>>> If I do my normal 'emacs --daemon' from i3 I get two (?) running
> 'emacs
> >>>>> --daemon' processes and 'emacsclient -c' does not seem to do
> anything.
> >>>>> If I just do 'emacs' it flashes the startup screen briefly, goes
> black
> >>>>> and then doesn't do much else. At least I only get 1 emacs process.
> >>>>>
> >>>>> No idea if there is a newer package available to test, or if there
> is a
> >>>>> library dependency that I managed to break during my
> >>>>> g2k22/EuroBSDCon2022 travels.
> >>>>
> >>>> I cannot reproduce this with both building a package and jca@'s
> provided
> >>>> package. I see only the expected behavior with one `emacs --daemon'
> and
> >>>> multiple emacs clients being able to connect and see the same instance
> >>>> using multiple `emacsclient -c'.
> >>>>
> >>>> Here are some sanity checks to try:
> >>>>
> >>>> 1. make sure snapshot is updated
> >>>
> >>> Just installed the amd64 snapshot I built overnight from -current and
> the
> >>> unfortunate behaviour of 28.2 on my E595 did not change.
> >>>>
> >>>> 2. reinstall emacs-28.1. does `emacs --daemon' and `emacsclient -c'
> work
> >>>> there?
> >>>
> >>> Yep. Typing this in mu4e in doom emacs with 28.1, after emacs --daemon,
> >>> emacsclient -c.
> >>>
> >>>>
> >>>> 3. test in a different window manager like dwm
> >>>
> >>> I'll try whatever the default is (fvwm?) and see what happens.
> >>
> >> No behaviour change in fvwm. Except ... when I 'close' the black window
> >> I get a message box pop up and say it can't find something called
> >> 'cl-seq'. Which then quickly vanishes along with the emacs window.
> >>
> >> Reinstalling 28.1 and searching the whole disk did not disclose the
> >> presence of anything called cl-seq.
> >>
> >>>
> >>>>
> >>>> 4. use -q to ignore .emacs config
> >>>> emacs --daemon -q
> >>>> emacsclient -c
> >>>
> >>> I will add that to the test queue as the espresso machine heats up.
> >>>
> >>
> >> And that works!
> >>
> >>>>
> >>>> 5. maybe some alias issue? use full paths
> >>>> /usr/local/bin/{emacs,emacsclient} to be sure
> >>>
> >>> No aliases defined.
> >>>
> >>>>
> >>>> 6. try `emacs -q' by itself since you said even emacs flashes screen
> >>>> briefly and goes black
> >>>
> >>
> >> 'emacs -q' also works fine.
> >>
> >>> Will do.
> >>>
> >>>>
> >>>> When I try to launch emacs --daemon multiple times it errors out the
> >>>> second time as expected.
> >>>>
> >>>>>
> >>>>> This is on an amd64 Lenovo E595.
> >>>>>
> >>>>> Happy to provide any potentially useful debug info or test stuff.
> >>>>>
> >>>>> .... Ken
> >>
> >> So I will go see if I can debug what part of doom is getting upset
> >> enough to prevent 28.2 from working.
> >>
> >> 'cuz life without doom, magit and mu4e is not worth living. :-)
> >
> > Aha! I somehow skipped the required recompile of doom when running a
> > different version of emacs! 'doom sync' recompiled all the required
> > packages and I type this on an apparently fully functional 28.2!
>
> Thanks for pinpointing this. emacs -q/-Q is the most useful tool to
> debug an emacs problem.
>
> Since I have received no negative (nor positive...) feedback except
> yours, I'm tempted to go ahead and commit this update.
>
> --
> jca | PGP : 0x1524E7EE / 5135 92C1 AD36 5293 2BDF DDCC 0DFA 74AE 1524 E7EE
>
