@ -18,45 +18,23 @@ To address this issue, we have implemented 'lxml' variants for each Kometa branc
To transition to the lxml branch, simply prepend 'lxml-' to the branch name you are currently utilizing. For instance, `kometateam/kometa:lxml-master`, `kometateam/kometa:lxml-develop` and `kometateam/kometa:lxml-nightly`.
This should cover it:
## LinuxServer
The first image offered in unRAID for Kometa is the Linuxserver.io image [`linuxserver/kometa`]
This image is different to the official image [kometateam/kometa] in a few ways that cause a variety of problems particularly for new users.
One typical error is something like:
```
Path does not exist: /run/s6/services/kometa/config/SOMETHING
```
The result is that the stock config file and a lot of the examples found in the wiki and config repo don't work.
This image is different to the official image [`kometateam/kometa`] in a couple ways that could cause issues for users who follow our installation guides.
If you use the LSIO image you should be aware of the following.
The LSIO image:
1. Requires absolute paths in the config. /config/Movies.yml, not config/Movies.yml. Because of this most of the examples in the wiki and config repo don't work as-is with lsio. Config files that work outside of docker often fail because of this with an error referring to `/run/s6/services/...`.
2. Only has `latest` version, no `develop` or `nightly`. If you want to switch to `develop` or `nightly` to try a new feature, they aren't available with LSIO image.
3. Doesn't support [runtime flags](../environmental.md), only ENV vars. This means that a command like:
4. Doesn't do manual runs correctly; they loop over and over. That command in the previous bullet point [which uses `KOMETA_RUN` to run it right now] will run over and over until you manually kill the container. The same thing using the official image will run once and quit, as expected.
1. Advises and provides examples of absolute paths in the config, for example `/config/Movies.yml` and not `config/Movies.yml`, however, relative paths should still work as expected.
5. Resets ownership of entire config dir every run. In tests, the ownership of the config dir and its contents was set to 911:911 with each run.
2. Resets ownership of entire `/config` dir on startup based on the values of the PUID/PGID environment variables. If not set, the ownership of the `/config` dir and its contents is set to 911:911, which can cause permissions issues.
Generally speaking, we suggest you use the official image instead of lsio.
Generally speaking, we suggest you use the official image instead of LSIO. We do not provide support for issues which relate to using third-party images [LSIO or other].
## Others
@ -64,4 +42,4 @@ There are 8 other images listed at dockerhub aside from the official image and l
Unless you have some compelling reason to use one of those, you shouldn't. You should stick with the official image.
If you choose to use another image than the default, you should be prepared for the possibility that published examples fail in various ways.
If you choose to use another image than the default, you should be prepared for the possibility that published examples fail in various ways.