It's a saner default. Much of the time it won't just work but it'll work *significantly better* than `download_full_article=true`.
Additionally it should reduce the chance of being seen as abuse, cf. <https://github.com/koreader/koreader/issues/12953#issuecomment-2565973079>.
Also see #3172.
Since both the UI and the default config include the value, impact may be surprisingly limited.
* implements a Provider singleton, to be used by thirdparty plugins
* exporter: support for thirdparty providers
* splits plugin loading into two steps: discovery and load
1. get a list of all candidate plugins to load for the different paths
2. sort providers before on the rest of them and try to load them
Remove unused UIManager:clearRenderStack().
Previously it was used by CoverBrowser plugin when extending the FileManager popup file dialog with aditional buttons.
Now the buttons are inserted through the api added in https://github.com/koreader/koreader/pull/12845.
This allows to set days to zero. As a security fall back setting days and hours both to zero will yield 5m autosuspend time. (to avoid shutdown loops).
fixes#12831
Before this change, on hold articles were deleted if `Remotely delete
finished articles` was enabled. To make this configurable new option
`Remotely delete on hold articles` has been added.
Fixes#12698.
Wallabag can be really slow to generate epub for large articles. For
example, some Wikipedia articles can take 20+ seconds on app.wallabag.it
instance that I use.
Tweak `mergeTitleBarIntoLayout` to keep the current selection (adjusted for the added rows) instead of resetting the focus.
It only actually called `moveFocusTo` on `hasDPad` devices, which explains why this was missed during testing.
Fix https://github.com/koreader/koreader/pull/12583#issuecomment-2423554770
Keep things least likely to be terrible at being interactive shells
first ;).
We penalize ZSH a bit, because RPROMPTS aren't really handled by the
plugin, which makes things confusing.
Most older NTX boards will horribly implode when attempting to suspend
with the Wi-Fi chip powered on.
And the others may do very dumb things re: power management anyway.
(This falls outside of the existing machinery that deals with this case,
as it happens *before* sending the Suspend Event).
Fix#12614