Yet another hook request (and question about ssh) about canto-next HOT 8 CLOSED

That's an interesting idea, but I'm sort of knee deep in the command infrastructure at the moment (and still trying to find time as always) so I'm going to have to TODO that until after all of the serious changes.

As for canto over network, I haven't looked at it in a while, and obviously the network adds some latency vs. a local socket, but last time I used it over LAN it wasn't very noticeable. Your commands look just fine. Totaling the amount of daemon socket traffic I see on startup (which should be the highest load), with maybe 1000 items moving over, I get about 200k of traffic over thirty seconds which isn't even close to enough to choke a LAN connection...

SSH won't speed anything up, but you can use port forwarding to encrypt your traffic (see http://codezen.org/canto-ng/manual/#Remote-Networking) if you'd ever want to connect from outside your LAN.

from canto-next.

Thanks for the reply, maybe it's my bad with my local network. Currently it's not connecting at all. At least now I know that it should work without noticeable latency :) Maybe it's just my RPI issues.

About the hook upon reading news - when I would have some time for it I'll try to make it myself (thx for the great example) and perhaps send a pull-request.

from canto-next.

Well, it seems like it all works over ssh (port forwarding), but nothing works over lan (I get connection refused error). So for now I'm going to use ssh-way.

from canto-next.

I have created a pull request about these hooks
themoken/canto-curses#14

About networking problems - I have fixed them, it was my LAN connection what wasn't working. Although, canto-daemon shuts down after some time working, I don't know why. I wrote a script which will restart canto-daemon if it's not running so it's not problem anymore.

Also, although these hooks and plugin should work, I have tested them on 0.8.4-2 version from Arch Linux AUR canto-curses version (not git). So check it out just in case.

from canto-next.

By the way, currently I don't need these hooks anymore and new version of canto perfectly works both over LAN and SSH, so I guess this issue could be closed unless you want to implement such hooks some day.

from canto-next.

Could the info on the remote networking made be available again?
It got removed some time ago and I'm currently have problems with this.
Wanted to check, if I did something wrong with the configuration.

The problem I notice is, the daemon with -a "" and -p 11000, the daemon doesn't seem to fetch anything, unless a canto-curses session is running.

from canto-next.

Sorry, I moved to inoreader and reading news on the go from my mobile phone now.

from canto-next.

No worries. I think I found the cause of my problem.

Edit: tl/dr;
User units are alive as long as the user is connected. Therefor no wondering, if I cannot connect via LAN if no session is running, and all is working well, after I connected via ssh first :/ Know your tools or so

All is working well, as long an user who owns the canto user units is active on the pi (ssh or whatever)
And after reading those bits a few minutes ago:

Aug 25 01:59:25 Clio systemd[778]: Stopped Timer for forced updates of custom canto-daemon service - 1Hour

(A timer + service file which does every hour an forced update, as it seemed to help in the past, when the LED on the pi was having fun, and ssh into the pi reveaveled via htop, that the canto-daemon hogged quite some resources and tried to do something)

And the actual service for the daemon:

 $ usystemctl status cantod             
● cantod.service - Canto RSS Daemon
   Loaded: loaded (/home/omnius/.config/systemd/user/cantod.service; enabled; vendor preset: enabled)
   Active: active (running) since Do 2016-08-25 12:24:32 CEST; 3min 32s ago
 Main PID: 1118 (canto-daemon)
   CGroup: /user.slice/user-1001.slice/[email protected]/cantod.service
           └─1118 /usr/bin/python /usr/bin/canto-daemon -a -p 11000

Aug 25 12:24:32 Clio systemd[1108]: Started Canto RSS Daemon.

I closed the last ssh session at 1:59 and opened the new at 12:24.
So this whole thing is handled by systemd the fact (which I kinda forgot) that the systemd --user instance is a per-user process, and not per-session.
Additionally as a feature which came with 230
( https://github.com/systemd/systemd/blob/master/NEWS#L274 ), which would also be a candidate but as far I can tell is disabled on Arch and ALARM.
So I need to read about how to keep that stuff alive.
The nohup equivalent

from canto-next.