Announcement: HADashboard v2 Beta3!

i tried it with pushbullet like this:

{"message":"even testen of een % teken gestuurd kan worden"}

no error, the message gets send without any problem.
lannouncer has no problem with it also.

so i am curious what message you try to send and how, because json has no problem with it.

Tried with your exact same message. Same story: with the % nothing happens, without it the message gets sent without any problem, and with the % escaped Hass throws the json error.
So I think we can conclude that the bug resides in the Join component

no.
if that was true, you should be able to send it to GSM or chrome without problem.
you say you tried to GSM or joined, so my guess would be the GSM component has a problem with it.
i tried lannouncer and pushbullet together and there is no problem there.

but i think you can better open a new tread to ask if someone else has the same problem with the GSM or chrome component, because this has nothing to do with HAdashboard V2 :wink:

Run with the -D DEBUG on the command line and you will get more info.

Yes.
Convinced it IS true, because gsm / chrome are not components, but are simply different incarnations (think “targets”) of my Join notifiers (Join is the only notify component I’m using…).
But I agree this turned out to not being an Appdaemon, let alone HAdashboard, problem - as my conclusion already infered. Thx for helping in pinpointing the guilty one !

2 Likes

This is what I got running with DEBUG flag;

2017-05-16 09:48:28.831400 WARNING ------------------------------------------------------------
2017-05-16 09:48:33.835726 DEBUG Refreshing HA state
2017-05-16 09:48:33.835949 DEBUG get_ha_state: entity is None
2017-05-16 09:48:33.847217 INFO Got initial state
2017-05-16 09:48:35.849750 WARNING Disconnected from Home Assistant, retrying in 5 seconds
2017-05-16 09:48:35.849953 WARNING ------------------------------------------------------------
2017-05-16 09:48:35.850044 WARNING Unexpected error:
2017-05-16 09:48:35.850116 WARNING ------------------------------------------------------------
2017-05-16 09:48:35.850318 WARNING Traceback (most recent call last):
  File "/usr/local/lib/python3.5/dist-packages/appdaemon/appdaemon.py", line 1208, in appdaemon_loop
    read_apps(True)
  File "/usr/local/lib/python3.5/dist-packages/appdaemon/appdaemon.py", line 987, in read_apps
    if root[-11:] != "__pycache__":
TypeError: 'NoneType' object is not subscriptable

2017-05-16 09:48:35.850410 WARNING ------------------------------------------------------------

@biacz - Could you share your code for your weather widget with the two “subtitles” (Wolken and Windchill)? Thanks!

When it comes to running this in a docker container, how are you setting up your dash_url? The actual ip of the container is going to be dynamic on the docker network, right?

Nevermind: I’m listening on everything and it’s working:

2017-05-16 10:51:17.327826 INFO Listening on ('0.0.0.0', 5050)

Back to basics.
I’ve been running dashboard for a while, but now I’m about to rebuild my Pi as some things are getting a bit shaky with HA and other things I’ve been working on.
So, I notice there are three different sets of instructions for installing HADashboard with some very different steps:

  1. From the HA documentation at:
    (https://home-assistant.io/docs/ecosystem/hadashboard/installation/)
    This one clones into ‘hadashboard’ and relies on Dashing and I assume is superseded.
  2. From appdaemon on git at:
    (https://github.com/home-assistant/appdaemon/blob/hadashboard_beta/DASHBOARD.md)
    This one clones into ‘appdaemon’ and relies on appdaemon install.
  3. From the start of the thread here, and points to instructions that clone into ‘appdaemon_dashboard’ and don’t rely on the appdaemon install instructions.
    Any recommendations on which one to use?
    I will be using hassbian and not Docker.
    Thanks

Use 3 - when the beta is over I will tidy up the conflicts.

Thanks @aimc, all of your work and support are greatly appreciated!

1 Like

Is there already a kickoff date for the next release?

1 Like

When it’s ready - no firm date yet due to extreme work related busyness.

2 Likes

okay, so my 0.0.0.0 config sort of bit me. The links in the default page use this setting to link to other dashboards. So am I able to configure this to work behind a reverse proxy (nginx) rather than configuring the links to go to http://ip address:5050?

okay, another challenge using my config…

I can’t get any values elements in my dashboard, looking at the javascript console, i’m seeing why…

VM1222:1 GET http://0.0.0.0:5050/state/sensor.multisensor_temp net::ERR_CONNECTION_REFUSED

So how are you guys configuring this inside a docker container? Obviously dash_url=0.0.0.0:5050 is not the right method. And is anyone doing this behind nginx?

I think @ReneTode is. And I believe @aimc runs nginx as well.

i run HA behind nginx, but i still have no clue how to put dashboard behind it.
i actually dont care at this moment because i can reach HA from outside, dont need to create dashboards as well.
outside my network is only for emergency

HA behind nginx works great. I’m trying to get dashboard to work in a docker container, behind nginx.

i dont use docker :wink: