• Off Topic
  • What feature/app do you use that the majority never talk about?

Murcielago
i use KeePassDX for barcode stuff. One less app and more secure.
https://github.com/Kunzisoft/KeePassDX

mythodical
YTDLnis has more functionality and seems to be more stable.
https://github.com/deniscerri/ytdlnis

mumbo6868
MJ PDF is more polished.
https://gitlab.com/mudlej_android/mj_pdf_reader

jet_silver
I tried it but I like Musicolet better as it has a more intuitive layout and just as many options.
https://play.google.com/store/apps/details?id=in.krosbits.musicolet

N1b
I prefer local music so Seeker is a great option.
https://github.com/jackBonadies/SeekerAndroid

I find Fossify Voice Recorder more easy to use. It also has an inconspicuous widget for quick start.
https://github.com/FossifyOrg/Voice-Recorder

    Macrodroid for device automation. Like Tasker but free and (imo) much more intuitive.

    Shizuku for safer root-ish privileges, like silent app updates and on-device ADB shell.

    Lockscreen Widgets for, well, lockscreen widgets!

    Croak3114

    I briefly tested Sendvisee. The most relevant advantages of the respective systems (for me) are

    Wormhole

    • file size (5GB vs 2GB)
    • online viewer (you can play images, audio, and videos without downloading)
    • Possibility to delete files before the timer expires

    Sendvisee

    • completely open source
    • Possibility to create password-protected download links

    Thanks for the tip, I will definitely use Sendvisee more often in the future.

      MarsTrue

      Catima is used, for example, to store barcodes from bonus cards, which you can then present at the checkout and have scanned. I also use Keepass and try to use as few apps as possible, too. Hence my question: How do you do this with the presentable barcodes in KeepassDX?

        Murcielago file size (5GB vs 2GB)

        1. Wormhole does up to 10GB.
        2. The 2GB for Send only applies to the official vis.ee instance. Other instances can go up to 30GB. There's also the option of running your own instance and having whatever size limit you want.

        Also, Wormhole seems to just be a fork of Send, or am I crazy? The UI for them are the exact same (the only difference is "Wormhole lets you share" vs "Send lets you share" in the description paragraph).

          Murcielago you can add one or more photos to a keepass entry and tap on them to go fullscreen. I have one entry on the main screen for all my barcodes.

          Dumdum

          OK, I didn't know that. Thanks for the correction.

          fid02 Futo works much better. it also integrates well in the vanadium search widget and open board
          https://voiceinput.futo.org

          they also make Grayjay, arguably the best YouTube/Odysse/Rumble/twitch/SoundCloud app
          https://grayjay.app/

          I would also like to mention two sites that offer hundreds of thousands of free, high resolution images for your home/lock screen.
          https://www.pexels.com/
          https://unsplash.com/

            MarsTrue Futo works much better.

            In what way does it work "much better"?

              fid02 more accurate speech recognition, be sure to select the high quality settings. better integration as well. try them both and see for yourself.

                AppVerifier is a tool I can use to check the signing certificate hash of an app against the signing certificate hash I get from a trusted source and/or its internal database so I don't have to trust GitHub (or other sites) as much not to tamper with APKs I download from them. If I recall correctly, it's made by a GrapheneOS dev too, so I'm surprised it doesn't get more attention.

                  MarsTrue Hi, I'm the creator of Tranacribro. Version 0.2.0 comes with a bigger model (74M) than 0.1.0 which significantly improved quality, but note that FUTO Voice Input allows you to choose an even bigger model (244M). Transcribro doesn't yet have the option to use a different model, but it is planned and it might be implemented in 0.3.0 in the form of being able to import a downloaded model file which can be one that the app links you to or a custom one. It would mean you can make your own decision on the amount of quality vs speed you want to tradeoff instead of being limited to a few models!

                  Can you elaborate on "better integration"? It could help with improving Transcribro so feedback would be appreciated!

                    concrete it's made by a GrapheneOS dev too

                    Community member, not a dev (the user above in fact). It also has gotten a decent amount of attention (at least, recently) so not sure about that comment.

                    soupslurpr The main issue I have with Transcribro is a usability one. On the browser search widget it doesn't show the voice input icon. I've tested it with multiple keyboards and browsers. this means that every query now requires two aditional taps to change the keyboard which is very inconvenient.

                    The voice input key does seem to work on the keyboard when not typing in a search box, after activating it in settings.

                    Additionally there's no audio feedback when using it which can be a a bit confusing.

                    I've tested them both again, on 74. futo is faster, more accurate and it doesn't suffer from the above mentioned UX issues.

                    Hope this helps.

                      Seek which works without an account, offline, and without location permission to ID plants, bugs, and fungus fairly well. You will want to go into settings and toggle on the "automatically take photo when the identification reaches speces". Once IDed the website www.pfaf.org has the best info you cant find any where on the edibility and medical propeties of the plant. Happy foraging!

                      MarsTrue ah, the browser search widget. I'll see about adding it.

                      Okay, audio feedback, that makes sense. Will work on that.

                      FUTO and Transcribro are actually around the same speed with FUTO being a bit faster because of their acft fine-tuned whisper models (https://github.com/futo-org/whisper-acft) which means they can use a lower audio ctx than Transcribro can. FUTO also may seem faster because it pastes the words as it is processing instead of processing then pasting the transcription. FUTO may indeed be slightly more accurate because whisper.cpp's beam search is too slow to be used right now. Once that changes, Transcribro will also use beam search and they should be around the same accuracy.

                      Thanks for the feedback!