-
-
Notifications
You must be signed in to change notification settings - Fork 402
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add FreeBSD service #2333
base: master
Are you sure you want to change the base?
Add FreeBSD service #2333
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for the PR!
It's not clear who the target is for the readme, but I think answering that question might let it be more concise and less duplicative of the contents of the script.
contrib/FreeBSD/sopel
Outdated
profile="$1"; shift | ||
|
||
run_rc_command stop "${profile}" $@ && | ||
sleep 1 && |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Necessary?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The sleep 1
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah - will stop
ever return before sopel has finished stopping? If not, is there a reason for the sleep?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, some IRC servers may block an IP with two or more simultaneous connections to prevent flooding. The message Session limit exceeded
should be printed in the IRC client. I will use sleep 1
because daemon(8)
will run sopel in the background without waiting. I know sopel has a flag called --fork, but the daemon is convenient in context because the output is redirected to /dev/null (sopel already prints the output to /var/log/sopel
, so there is no need to have two outputs).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fair, but I think that belongs in the start function - and it looks like there already is one there, making this one unnecessary, unless I'm misunderstanding run_rc_command
or something.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Due to a slow connection or a network problem that delays a packet before disconnecting it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Normally Sopel sends a QUIT
message and waits before exiting. Unless this stop
doesn't wait for sopel to exit, or you know of routine cases I don't, you should be able to assume that if the process exits you're all set.
Line 635 in 71380df
def quit(self, message: Optional[str] = None) -> None: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
As I mentioned in another comment, the service uses daemon(8) which does not wait for the program to run.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Do you agree with the change I mentioned?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Starting two:
echo;sleep;fork;echo;sleep;fork;echo;sleep;fork
Restarting two:
stop;wait_for_pids;sleep;echo;sleep;fork;stop;wait_for_pids;sleep;echo;sleep;fork
So for each, to start it simply waits 1 second and then forks, for a total delay between instances of one second.
To stop, it's sending a signal, waiting for Sopel to exit, waiting one second, then in start
it's waiting another second, then forking the new one. The total delay here is 2 seconds between each instance.
I still don't see why there needs to be twice the delay when restarting vs starting.
It is for FreeBSD users. The README explains how to set variables for sopel using the FreeBSD tools and explains how to get more variables to configure sopel. My intention is to show how to configure sopel as another FreeBSD service through |
* Delete parse variables in the docs * Reduce long lines to keep between 80-120 chars
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If the readme is intended for end users (rather than packagers), I think even fewer of the options need to be specifically pointed out. If someone has unusual needs (in particular, changing sopel_prefix or sopel_interpreter) they probably understand how to do that just by the sopel_profiles instruction and the list of defaults.
The average end user will be reading documentation with one goal in mind, making it work. They're looking for "pip install sopel, then copy this file here, that file there, and run these two commands. For these common changes do [...], or for less common changes apply that pattern to one of [variable list]", and the doc should be structured and written to support them in that endeavor.
Co-authored-by: mal <[email protected]>
Co-authored-by: mal <[email protected]>
The `sopel_prefix` is intended to packagers, not end-users
…explicit about the scripting language used.
@half-duplex When this is satisfactory, are you up to walking OP through the process to squash & amend commit message to meet project standards? |
I apologize, I will comply with rule 5 in CONTRIBUTING.md. |
@DtxdF Don't worry about it just yet. We allow time to squash/tweak the history after PR approval. 🙂 |
Did this one fall by the wayside? I would defer to @half-duplex's expertise here, so just looking for the 👍 that it's ready and a squash of the intermediate commits + messages meeting guidelines. |
Description
Service support for FreeBSD.
Checklist
make qa
(runsmake quality
andmake test
)