1.4.13

Docker Tags

  • ghcr.io/asimons04/tesseract:1.4.13
  • ghcr.io/asimons04/tesseract:v1.4.13
  • ghcr.io/asimons04/tesseract:latest

Bugfixes

  • [0ed97991] Don’t badge-ify hashtags if the hashtag linkify option is disabled
  • [ce5a310a] Make markdown text reactive (and re-run pre-processing step) to changes to the enable/disable hashtag setting.
  • [704e2d78] When clicking the user/community links in the modal, close the modal when navigating to the target (previously modal stayed open)

Enhancements

  • [3b8ef653] Update banner when viewing a post on a remote instance to use more concise verbiage and add an extra button to link to the post’s canonical instance.
  • [ec8c4bfe] User and Community links are now a elements rather than button. Allows middle-clicking or right-click-> open in new tab to immediately open the profile/community in a new tab without having to click through the modal. Regular clicks still open the modal.

New Feature: Synthetic View Modes

Under the hood, there are still only two main types of view: Card and Compact

However, Compact view is affected by several options which can be combined to create different view styles.

The “compact/card” switcher button has been replaced with a selector menu to select from one of 6 views:

  1. Card: The flagship Tesseract post view. Posts are shown as cards with all the media embed bells and whistles. Post body preview is set to 240 characters.

  2. Compact: The classic “compact” view as of 1.4.12. Feed margins are present, post images are thumbnails, the post body preview is 240 characters, and no media is embedded unless the post is expanded into card view.

  3. Wide Compact: Same as “compact” but without the feed margins. Posts span the full width of the display. Clicking the thumbnail image or the “expand” button in the post action bar will expand the post into card view.

  4. More Compact: Same as “wide compact” except the body preview length is set to zero and can be expanded.

  5. Ultra Compact: Same as “wide compact” except the thumbnails and the post body are hidden. Post body cannot be expanded, and you will need to click into the post to see it. Post can be expanded to card view only with the “expand” button in the post action bar. Expanding into card view does not reveal the post body, only the thumbnail image (if present). If the post is a media post, the embed will be available (either in full or click-to-play depending on settings)

  6. Reader: Same as “wide compact” except the entire post body is shown (up to 10,000 characters).

More views may be added along the way, but for now, those should cover more use cases without having to manually fiddle with the various options.

I’m really liking the “Reader” mode, myself.

  • Admiral PatrickOPMEnglish
    2•3 months ago

    Good issue report. I’m able to replicate that using my Lemmy World account. The gist of it is that Lemmy’s search is kinda awful in some (cough many cough) ways.

    The autocomplete on the post form is used to prevent errors when creating posts by fetching the community ID that’s needed for the post is submitted. It uses the search API with the community name as the query string and the search type set to “Communities” . The search API doesn’t recognize the name @ instance format and just seems to search by name. It’s also limited to 50 results (which is what I’m using), and I don’t have pagination logic in those autocomplete components.

    In this case, I guess ‘[email protected]’ isn’t in first 50 results. That’s not something that would have come up in my tests, unfortunately (small instance problems lol)

    I have an infinite scroll component that I believe I can shim into the autocomplete dropdowns which should take care of search pagination fairly seamlessly. I could also allow manual input (e.g. community @ instance) and lookup the community ID if not found by the auto complete.

    Thanks for reporting that. I rarely use the Create -> Post menu, lol. Usually I use the sidebar or the community modal.

    Will make this a priority for the next release.

    • @gedaliyahEnglish
      2•3 months ago

      I usually navigate to the community before posting as well, so I never came across it. Thanks!

      • Admiral PatrickOPMEnglish
        1•2 months ago

        Ok, I think I have this working a lot better now.

        1. Looks for an exact match for the query (local communities resolve with or without the @instance and federated communities with the @instance.
        2. If an exact match is found, it will be the first item in the results
        3. Query is passed to the search as before
        4. Any results from that search are appended to the results after the exact match (if any)
        5. The autocomplete dropdown now does infinite scroll pagination for the search results.

        I’d love to be able to make it more granular (e.g. news@lemmy. would return news@lemmy.world, news@lemmy.xyz, and news@lemmy.abc) but the search API leaves much to be desired.

        I’ve got this all roughed in and working for the community autocomplete component. Just need to tie up a few loose ends, refactor it a bit, and run through some more tests. I also need to apply the same logic to the person autocomplete component since it works the same way and has the same 50 result limitation.

        Hope to have this deployed in 1.4.14 and released sometime this week.