Yeah, I posted the way to do it in the post you replied to
o really? I thought that was your effort and that had failed too. It does here adding that to the action block as condition âŠ
There is a list of over 1100 issues, all wanting to be fixed. join the queue
Iâm here just hoping a new update will actually fix my HA installed 115 broke it restart now takes forever 6+ minutes up from 90 seconds on 114.4
Core.restore_states isnât being saved on restart so states are wrong once restart complete.
Media player is missing play pause button now was there on 114.4
Iâm questioning this mode of thinking. Integrations and features are being added with every release while we have a backlog of 1100 issues? Surely itâs better to get the current ones working before we add even more stuff. I love this project and on the most part it is very stable but when things break and we get someone telling us to âjoin the queueâ it doesnât give anyone a good impression. Priority should be fixing breaking changes not adding features to a broken system, nobody wants to walk into a broken house because HA is falling apart.
There are only a limited amount of people to fix issues, you also have to remember they need a dev who is familiar or own the hardware of the integration. You cant expect those developers who do it in there free time to fix issues they either have no interest in or capability. Also remember a lot of the integrations and features are added by developers outside of the core nabu casa ones, they cant be forced to fix issues on stuff they didnt implement
I already knew about the limitations in terms of open source development and I know there are many people like me who âwould jump in and help if they knew codingâ, doesnât change a thing I know, but this is the point Iâm making: if nobody can, the integration shouldnât be available. If this integration cannot or wonât be fixed for whatever reason HA team just has to remove it from the catalog, problem solved, say that itâs no longer available on the release notes so we can remove it and move on, but to just sit there with a broken integration hoping it will be fixed when it never will or it will take an unreasonable amount of time is not good for anybody.
Since version 0.115, I cannot restart my HA instance from the GUI.
If I call homeassistant.restart from Developer Tools > SERVICES nothing happens.
If I go into Configuration > Server controls and under server management press RESTART, nothing happens. But if I press STOP, then HA stops.
This is the same on Windows with chrome browser, Macbook with Safari, and iOS home assistant companion.
I have to login to the console and initiate a restart.
Anyone else having this issue?
But its working for some people, myself included why should it be removed for those its working for. i use HA container and google tts is working fine
from looking at the issue in the github, it only seems to be on certain architecture arch | armv7
Mine is |arch|x86_64|
| â | â |
|dev|false|
|docker|true|
|hassio|false|
|installation_type|Home Assistant Container|
The github issue has been open for 9 days. Thatâs hardly âan unreasonable amount of timeâ.
Also, per your original post, itâs poor etiquette to directly tag developers to get a response.
Please see #16 in âHow to help us help youâ
ânever will or will take an unreasonable amount of timeâ. Key word is âwillâ. The future. Sure 9 days isnât, but if this drags on for 3 months then thatâs when my statement applies. Apologies for tagging but it doesnât seem like this integration is tied to any maintainer!?
It makes sense that youâre not affected if youâre not running on ARM. When I do a config check this shows up:
Platform error tts.google_cloud - Error relocating /usr/local/lib/python3.8/site-packages/grpc/_cython/cygrpc.cpython-38-arm-linux-gnueabihf.so: backtrace: symbol not found
I researched, it seems that the Cython/PyZMQ has to be updated but even doing a docker exec -it homeassistant bash and then âsuâ on the shell to become root, I cannot do apt update / apt-get update, the apt binary isnât present, so I cannot even try.
This is what Iâm saying, it appears the issue is simple to fix so it shouldnât take a crazy amount of time. I stand by my comments, if the implementation is not going to be supported for all architectures this should be clearly noted on the documentation. Iâm not the maintainer so I wonât touch the documentation, does anyone know how to update the packages inside the container?
It has maintainers, they just havent been assigned this issue yet (and that doesnt mean their not aware of it or looking into it, just not formally assigned on github).
Per CODEOWNERS, the TTS integration is âownedâ by Pascal (pvizeli) and google_cloud is âownedâ by ĐŃĐ±ĐŸĐČĐžĐș (lufton). It also seems that Pascal may be on vacation this week, FYI.
It looks like individual people (whom have permissions on the repo) go through and tag the specific integration the github issue is related to, then the bot goes ahead and adds the maintainer to the assignees based on the codeowners list.
The issue in question just hasnt been tagged to a specific integration yet, but there are definitely maintainers associated with both google_cloud and TTS.
Thank you for explaining that makes a lot of sense. Hope heâs enjoying his vacation.
problems upgrading to 0.115.3
Canât install homeassistant/raspberrypi3-homeassistant:0.115.3 -> 404 Client Error: Not Found (âno such image: homeassistant/raspberrypi3-homeassistant:0.115.3: No such image: homeassistant/raspberrypi3-homeassistant:0.115.3â).
Do you have the required disk space to perform the upgrade?
i see this in the log
20-09-25 15:46:41 INFO (SyncWorker_5) [supervisor.docker.interface] This error is often caused by not having enough disk space available. Available space in /data is: 13.7 GiB
hassio also reports 47.9% disk use percentage
the first time hassio noted in the logs it crached and rolled back
after that i got the message above. I already rebooted my pi but got the same errors so posted it here on this forum.
Also my zigbee2mqtt was broken: Unknown options availability_blacklist
but got it fixed with reset do defaults and changing my settings again.
Search your automation file for minutes: where the value is ââ, it needs to be â0â instead.
Thanks for the feedback but I donât understand what I exactly have to fix in the automation file.
I found something like ;
minutes: â/30â
looks the only entry for minutes.
Also my pacakges donât contain âminutesâ
try deleting the / as it wants it to be a number between 0 and 59. BTW I am mostly just guessing as I just fixed my system with a similar error, which was a ââ for the hours instead of â0â.