cross-posted from: https://lemmy.world/post/12315554

Post is displaying URL image in Thunder but not Voyager

Hi,

I’m comparing Lemmy Android app clients and I’m between Thunder and Voyager.

I noticed something weird with Voyager. There was this post it displays an image in the timeline using Thunder but it doesn’t in Voyager.

Is it some configuration that I need to tweak in Voyager?

Also, is there some config to display short text about the post if has image? Like some description below if available

Thanks

  • @darklightxiM
    3•10 months ago

    What should I share other than the screenshot of the app showing the image of the link and the setting for Scrap Missing Link off?

    I would try to share as much detail as possible for us to be able to reproduce the issue. If we can’t reproduce it, then it’s very unlikely that we’ll be able to find the cause of it (and therefore fix it). The best way to do this is to start from a clean install of Thunder and perform the minimal steps to get the issue to show up.

    Aside from steps to reproduce, I would also suggest putting the version of Thunder you’re using, and any other relevant information! (Android/iOS version)

    Also, do you mind commenting about privacy and security concern about this feature on?

    Of course! I’ll try to give the best summary based on what I remember, and all the source code is available on GitHub if you’re interested to take a look.

    If you have the setting turned on, Thunder will attempt to fetch missing thumbnails for posts that don’t already have one (in particular, posts that contain an external link). If the instance already contains a thumbnail, then it will prioritize that one and won’t try to find thumbnails even with the setting turned on.

    If there is no thumbnail (which is generally pretty rare now), then Thunder will make a request to the external URL (linked from the post) and attempt to fetch any metadata information from it. This metadata sometimes contains a link to a thumbnail which Thunder proceeds to use.

    This process would be no different than if you were to tap on the link yourself to view the site. This does mean that Thunder will connect to the site in order to retrieve the metadata from it. The only difference here is that Thunder changes the “user agent” when performing the request to fetch information from the site to a more generic one. The user agent often contains information about the application and version being used to contact the external URL, so this alleviates some (but definitely not all) privacy concerns associated with this.

    This is the extent of what I can remember, and as always, feel free to suggest any feedback that you see fit (whether it be privacy/security concerns or otherwise)! I’d be happy to take a look at them and see if it’s feasible to do. Thunder is still mainly maintained by a only a few people, and we all do it on a voluntary basis so there’s definitely always room for improvement.

    I hope this answers your questions!