I'm not sure why I added the attribute previously, in
0afd511b9cbc30e9586d388c32c72d4d25ddddaf, but it's not necessary, and
has caused some confusion. Given that, this commit's removing it.
This commit adds two Antora Playbook files, so that the docs can be
built, stand-alone, from the remote repository and from the local
working copy. It also documents how to build the docs locally.
It's come up discussion around the migration to Antora, that the client
docs should be able to be built standalone, and not just as a part of
the larger documentation set. By adding an Antora Playbook file
(site.yml), the client docs can be built both standalone and as part of
the larger ownCloud documentation site.
See https://github.com/owncloud/client/pull/6785#issuecomment-425391756
After re-reading the relevant section of the Antora documentation, I
believe that my previous understanding was incorrect as to how
navigation files are built. Consequently I used too much indentation and
included an unnecessary header, that was duplicating the name draw from
the antora.yml (the component descriptor file).
While not strictly necessary, the rename makes it that much clearer for
contributors to know what the file is for and whether it's the correct
one to edit. It also better organised the "advanced usage" files, so
that the file structure is that much more meaningful as well. So, a
little bit of work now, to save time later.
It was deleted when the original rST content was migrated to AsciiDoc,
and the .gitignore file had a broad exclusion, which it was caught by.
The exclusion in .gitignore has been fixed as well, in this commit.
This reintroduces it, after converting the content to AsciiDoc.
The names, ordering and indentation of the content was incorrect in the
original version of the navigation file. This commit updates it to
reflect the now deleted index.adoc file, which contained the original
ordering from the sphinx-doc version of the documentation.
They may not be, strictly, required at the moment, but as they're part
of how Antora organises content, it's good to have them tracked as early
as possible. See https://docs.antora.org/antora/1.1/component-structure/
for further details.
Antora implements a specific file and directory structure for
components, which you can find out more about at
https://docs.antora.org/antora/1.1/component-structure/. This commit
reorganises the existing images to follow that structure, as well as
optimises them, so that they're as small as possible, without losing any
quality.
