Skip to content

the overdue update drop - #282

Merged
JovannMC merged 171 commits into
mainfrom
maya/conversion-things
Oct 4, 2026
Merged

JovannMC merged 171 commits into
mainfrom
maya/conversion-things

Conversation

@JovannMC

@JovannMC JovannMC commented Sep 11, 2026 •

Copy link
Copy Markdown
Member

this PR is here to make it easier for others to track the status of the update drop (should've done this earlier). this includes many new things, but below are most of the main things. you may have been directed here if your PR was closed.

  • Completely local video conversion with Mediabunny
    • by default, videos will now be converted locally with Mediabunny
    • a toast will warn you if a conversion needs vertd (manually selected, fallback, automatically due to unsupported format)
  • per-file conversion settings for: images (imagemagick), audio (ffmpeg), & videos (mediabunny/vertd)
    • ..with a LOT more conversion settings to mess with
  • choosing a specific converter
    • mediabunny/vertd is the first multi-conversion category, will allow us to add more possible converters to stuff like images <3
    • welcome back libvips maybe?
  • vertd-related stuff
    • vertd servers can now limit file size
      • this is mainly for instances who want to host to the internet
      • official instances will continue keeping no file size limit
        forever or time out :3
  • language updates
    • added norwegian, czech, and polish languages
    • updated turkish and spanish translations
    • re-organization of translation keys, migrating existing languages
  • many, MANY bug fixes
    • better handling for vertd connections
      • if you suddenly drop connection to vertd, the UI will actually immediately cancel conversion instead of being stuck |
    • handle mislabeled file names (e.g. png named a jpeg)
    • ping servers instead of relying on privacy-invasive ipapi.co
    • many others i don't remember honestly, there's a lot

TheFrostBunny and others added 30 commits January 29, 2026 15:08
translated all missing sections that were introduced over the last few months to ensure full localization coverage.
pressing "download all" with multiple files of same name gets overwritten (e.g. when pasting screenshots/from clipboard)
adds support for animated webp to video by using middleware: webp -> gif -> video (vertd)
god kill me i fought with paraglide for a while
doing this on the way to a national robotics competition lmao
this took a bit lol, restructured the conversion workers a little bit - i prob broke vertd and ffmpeg audio for now
translated all missing sections that were introduced over the last few months to ensure full localization coverage.
im sleepy this is probably a terrible implementation rn
removed the settings auto detection (since it made UX a bit worse), added conversion speed slider - need to implement in vertd
+ fix re-showing vertd size limit toast on navigation
…s, image settings updates

i forgot my yubikey at home while coding at uni so uh, this is all in one commit!
format dropdown, settings modal, get (better) animations. tbh ai did the animateHeight stuff cause i dont know how the hell this works

other minor ui stuff fixed
improve memory usage for multiple files, reduce progress report overhead
+ only partially read files for file detection, allow .png with apng data to convert
ensure errors are properly caught, gracefully handle failed files in zip archives (generating a text report) instead of failing entirely
tbh had ai do this one quite a bit, but looks good to me. will test ofc
- make per-file settings use file ids instead of filenames
- clear sw cache, localStorage, and sessionStorage independently (so one failure doesn't skip the rest)
- reject invalid audio channels + warning toast, fix mediabunny metadata setting
- remove rtf as input (it was rejected converter side anyways?)
@JovannMC
JovannMC merged commit 3e95d75 into main Oct 4, 2026
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants