## Ticket
Part of 1368 ("if user overrides the category homepage language with a different language setting + use 'search this language' option, use that instead. It's not typical this will happen, but probably expected.")
## Change
- Refactored to move more logic into `resolveLangForClaimSearch`
- Replaced the ternary version to make it more readable (hopefully).
- Inverted the language filter precedence by honoring the user's setting first.
- Note: for both User and Category, URLSearchParams will always take precedence.
- I got the previous logic completely wrong.
- Wild West's release time now comes from the homepage definition, but since the definition currently does not support "start of week", continue to hardcode the customization here.
- This logic also fixes the release_time for Tags.
Re-used lots of Anthony's code + made fixes to a few areas including the publish page.
a
new videojs
video.js@7.18.1 + http-streaming@2.14.2
remove console log
Mainly only useful for Wild West, which doesn't use `channel_ids` (otherwise, the homepage maintainer could just exclude the id from `channel_ids` directly).
It was previously done in `GetLinkData`, but accidentally removed in d854a992.
It is DRY-er to do it here rather than at the client side.
Also needed to filter in the homepage-sort dialog, so move the definition to a constant array.
## Ticket
1074
In mobile, the second livestream tile in Following gets missing because of the Ad being in 'replace' mode.
## Change
Switch from 'replace' to 'insert' mode in Mobile layout.
If missing (replaced) tiles is not a desired behavior, the true fix would be to always use 'insert' mode, and either live with the uneven tiles or wait for our design master's solution.
## Issue
1345
## Change
The refactor in 04c5ac46 broke the logic, because `!hasCost` would incorrectly include an undefined `costInfo` (haven't fetched) as free content.
## Ticket
1307 newly added categories aren't adding to existing customized homepage
## Change
In Customize, new categories were appended at the bottom, but forgot to repeat the behavior on the actual homepage itself.
Ticket: 1180
I believe the original intention was "if a claim was previously resolved, but without authentication (no `is_my_output` data), then resolve again with `include_is_my_output`"
Update to exclude that logic for Incognito.
## Aside
The check for `!canonicalUrl` is interesting:
- Are there claims that really don't have `canonical_url`? If yes, would that end up in an infinite `resolve` loop?
## Issue
- Adding `content__non-clickable` everywhere probably isn't the right way when we move forward to separating `all.scss`.
- Not intuitive -- easier to use if it's an attribute to `ClaimPreview`.
## Change
- Consolidate to a re-usable `non-clickable` class.
- Made Wallet Send search preview non-clickable as a temporary band aid until the channel-finder is consolidated.
- `listBlocked`: something is overriding the pointer, so non-clickable isn't working. Removed for now -- doesn't really matter.
## Issue
Maybe related to https://github.com/OdyseeTeam/odysee-frontend/issues/99#issuecomment-1042384649
## Notes
`fileInfo` does not contain `metadata` anymore -- not sure what happened. With zero duration, `videoPlayedEnoughToResetPosition` was never true.
But the claim usually contains the duration already, so we can use that. There are some audio/video claims that don't, but those are a minority.
Expanded homepage pins to support 2 types of input (if both are passed in, pinnedUrls take precedence):
(1) pinnedUrls --> uses doResolveUris (resolve)
(2) pinnedClaimIds --> uses doResolveClaimIds (claim_search)
Instead of injecting the pinned URLs directly, we inject from `resolvedPinUris`, which will be blank until the uris are resolved, thus preventing it from being resolved individually from the tiles.
There's additional complexity with the `claim_search` method, as the rest of the code deals with uris instead of IDs. Fortunately, it's all contained with `useResolvePins`, so removal would be easier when the batch `resolve` issue is fixed.
Ticket: 1228
## Code changes
- Pass through `getTimeAgoStr` so that the value gets localized.
- Pass the simpler `number` around instead of the `moment` object for better memoization.
## Notable differences
Due to how `getTimeAgoStr` is written, we now get to see the time actually counting down, vs "in a few seconds" currently in production. I think the counting-down behavior was the original intentional, since a 1s timer was used (otherwise, a 1-minute timer could be used) ... or maybe not since streams may not start on the dot.
