Force re-frame to stop warning about subscriptions in non-reactive contexts
while executing subscription tests (i.e. unit tests using the macro
test-helpers.unit/deftest-sub).
The net result? No more hundreds of useless warnings in the output of make test.
re-frame: Subscribe was called outside of a reactive context.
See: https://day8.github.io/re-frame/FAQs/UseASubscriptionInAJsEvent/
The warning is actually useful in production code, but in a subscription test we
already know we're never inside a reactive context.
This PR resolves the issue where Composer becomes detached from the footer during the chat history loading phase, specifically when the skeleton screen is displayed. Fixes#17237.
fixes#17103
### Follow up from #16865
This commit focuses on enhancing the user experience by replacing our legacy skeleton loaders with the latest version. Additionally, it introduces a refined approach for deprecating outdated components and functions to maintain a cleaner codebase.
### Key Changes
Replace existing skeleton loaders with the latest skeleton-list component for improved performance and usability.
Implement a standardized deprecation method, using either naming conventions or metadata annotations, for phasing out old components and functions.
Unshadows all remaining vars in status-mobile, including non
cljs.core/clojure.core ones. The only exceptions are cljs.core/type and
cljs.core/name (which happen quite often, so I'm not sure if it's worth
unshadowing them).
The bug was caused, again, by the fact that the component react-native/image
source was a string and it was not wrapped inside the map {:uri "..."}. The
bug was likely introduced by commit 255a3b917265d381fed57e26422277b4cc8772e4
I decided to change react-native.image for good, similarly to what we do in
react-native/fast-image
(ebd38295c6/src/react_native/fast_image.cljs (L21))
so we can prevent this type of bug in the future.
Fixes https://github.com/status-im/status-mobile/issues/17157
Fixes: #16123
This commit changes the behavior when joining communities:
1) If the user hasn't joined a community, they will see the
non-collapsed header.
2) If the user has already joined the community, they will see a
collapsed version of the header, in order to have quicker access to
the chats of the community without having to scroll a lot.
One edge case that is taken care of is the following:
1) User visits a community that has not joined, requests to join.
2) The user is on the community page when the state changes from
not-joined -> joined.
In this case the user won't see the collapsed header until they leave
the community view (go back to home say and back to communities).
This is done in order to avoid some janky transition between the non
collapsed & collapsed version of the header.
This is slightly different from as mentioned in the issue "visiting the
page for the 2nd time", as that will require larger changes but it
should be an acceptable behavior, since most of the times they will have
to visit the overview to request access.
