See #175 for the discussion about this feature.

These have happened in the wild, and the validator should alert people
to these issues, as they are always unintentional.

The "id" type was simply missing due to an oversight, and the
"mediatype" type will likely exist in a future release of the
specification, so let's support it early.

Having those could lead to circular references between components so
that they can not be freed properly anymore, or even worse may crash
AppStream in an infinite recursion when building the data pool.
Since we don't fully control the data we are processing, it's good to
protect against these issues as much as possible, even though they
should not ever happen on a system that receives well processed
metadata.

This adds a few checks and cleanups for the release metadata, most
importantly validating the release dates for correctness.
See issue #245 for more information.

This resolves #245

This was already made clear in the spec part of the documentation, but
the quickstart page was not up to date.

The answer is: Pretty much use .metainfo.xml all the time, unless you
have a desktop-app in which case you may choose .appdata.xml (which will
likely be supported forever for legacy compatibility).

We don't use deprecated API anymore and are slightly faster.
In theory we can get rid of a few more allocations here, but let's see
if that is actually needed for better performance.

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 100.0% (197 of 197 strings)

This is nice for automated tools that need to read this data for further
processing.

Previously, AppStream would read and releases from metainfo files, then
ensure that they were sorted and make a sorted list of releases
available to the user.
Now, we only sort before data is serialized, not when it is loaded.
Additionally, the validator will emit a warning (failing validation)
when releases in a metainfo files are not sorted correctly.

A latest-to-oldest release info is now be strongly advised ("should")
for metainfo files, while being mandatory for collection metadata
("must").
We might enforce this ordering for all metainfo files soon, if there are
no issues found with this change.

Currently translated at 100.0% (197 of 197 strings)

By default, `appstreamcli validate*` commands will now display a terse
list of found issues. You can pass `--explain` to the command in order
to get a detailed explanation on the individual issues.

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 79.7% (157 of 197 strings)

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 100.0% (197 of 197 strings)

This is rudimentary for now and will look much better soon.
But it allows to test the new validator more easily in the master
branch.

With this new validation system each issue gets a long explanation as
well as a unique short tag which can be used for display it better in
UIs, ignore it or just quickly infer what the problem is.
This is a semi-breaking change as it will change the validator output,
so this will receive further updates in the next few days to make it
work better and display the output better.

Currently translated at 100.0% (197 of 197 strings)

Currently translated at 79.2% (156 of 197 strings)