Closed timlod closed 3 years ago
Actually, you see this with ps x
as well, although there it is OK since the definition is on one line, instead of systemctl status
's multiline output. Would it be possible to wrap that module inside another script that is called instead to prevent this?
This is a duplicate of #393.
Putting it inside another script is certainly possible but you have to be careful that it still works over tramp.
I'm not very experienced with elisp that touches outside processes - could you give me a pointer on how I could have that module external? Then I could submit a PR for that.
Edit: I have found how to wrap it in a script - I replaced
:args `("-c"
,anaconda-mode-server-command
with
:args `(,(concat (file-name-directory (locate-library "anaconda-mode")) "anaconda-mode.py")
where anaconda-mode.py
wraps that script. But as expected, it won't work over TRAMP, as I suppose the filename is expanded with the current TRAMP prefix...
Maybe a simple solution is to check (pythonic-remote-p)
and if that's true, send the content of anaconda-mode.py
instead of passing the filename.
Other option would be to copy the file to the remote host first?!
I made a pull request fixing this - #407. It's my first change of this kind, so it may not be exactly what you're looking for - let me know!
Thanks. Had a quick look and lgtm. I'm pretty busy atm, so in case @CeleritasCelery isn't faster than me it'll take some time for a proper review.
Fixed in #407
We have noticed that whenever
anaconda-mode
is running, the entirety ofanaconda-mode-server-command
shows up in thesystemctl status
under theipython
process. We are on a multi-user server with several instances ofanaconda-mode
, and thesystemctl status
output is thus quite polluted.The log will look something like this:
Can we somehow prevent this from happening? Running emacs 27.1 as a daemon.