lazy_load_full is meant for siturations that you don't need
the constraisn of passing a single container parent and adding
a sigle widget to it.
Reimplemnted lazy_load on top of lazy_load_full.
If you clicked unsub, then undo and then the same show widget you
would navigated to the previous ShowWidget and the unsub button
would still be insensitive.
Use .then() combinator to override the result and return
Ok(()) even if the task fails. That allows us to use join_all
instead of the custom written collect_futures function.
Follow http 301 permanent redirects by using a future::loop_fn.
It's kinda funcky, match_status still returns status_codes as erros
and a new DataError Variant had to be added to distiguise when we
should Loop::Continue. This could be cleaned up a lot.
While ideally we want episode widget to determine their states
themselves and avoid refreshing the whole view, currently there
is no infrastructure for that which resulted in views not being
updated their EpisodeWidgets stuck in the InProggress state.
Previously we would refresh all the views when download/cancel
button was clicked. This was done mainly to avoid zombie widget bugs
that would arise from shared state.
Now we still refresh all the background views but not the visible one.
Instead the widget has the reponsibility of recalculating it's state.
make the data: Vec<T> mutable, then reverse the vector
so it can be used as a stack, and then use the ::pop()
method to retrieve the item.
This also avoid the constrain for Clone on T.
They are invalid in `pango` markup so theres no reason they
should are not needed. Also add some paranoid .trim() calls.
It returnes a &str slice so it's cheap.
The performance boost is not good enough to justify the
code complexity it add and the memory overhead of
yeat another threadpool.
We will start refactoring the whole pipeline implemantation
and might transition to either rayon-futures or tokio-runtime.
