Skip to content

Latest commit

 

History

History
65 lines (55 loc) · 2.71 KB

TODO.md

File metadata and controls

65 lines (55 loc) · 2.71 KB
Raw

TODO

aka a lot

P2

Planned for later.

  • Soft errors and retry logic
    • Big caveat: When the server still responds, like 50x errors which would be recoverable, we actually have no way of knowing it did in respond in such a way. See P4 - Handle Errors remarks.
  • Delete files (well, as far as the browser allows)
  • Inter-addon API (basic)
    • Add downloads
  • Chrome support
  • vtable perf: cache column widths
  • Download options
    • This is a bit more limited, as we cannot modify options of downloads that have been started (and paused) or that are done.

P3

Nice-to-haves.

  • Landing Page v4
  • Drag-n-drop ordering
  • Drag-n-drop files
  • Inter-addon API (hooks)
    • Manipulate downloads (e.g. rewrite URLs)
  • Native context menus?
    • Would require massive reworks incl the need for new icon formats, but potentially feasible.
  • Import/Export
  • Download priorities (manual scheduling overrides)
  • Dark Theme support
    • os/browser define be default
    • overwritable
  • Get and cache system icons (because Firefox doesn't allow moz-icon: for WE, but makes them kinda accessible through the downloads API anyway, essentially copying them via a canvas on a privileged hidden page into a data URL... ikr)
  • Remove any types as possible, and generally improve typescript (new language to me)

P4

Stuff that probably cannot be implemented due to WeberEension limitations.

  • Avoid downloads going "missing" after a browser restart.
    • Firefox helpfully keeps different lists of downloads. One for newly added downloads, and other ones for "previous" downloads. Turns out the WebExtension API only ever queries the "new" list.
  • Segmented downloads
    • Cannot be done with WebExtensions - downloads API has no support and manually downloading, storing in temporary add-on storage and reassmbling the downloaded parts later is not only efficient but does not reliabliy work due to storage limitations.
  • Handle errors, 404 and such
    • The Firefox download manager is too stupid and webRequest does not see Downloads, so cannot be done right now.
  • Conflicts: ask when a file exists
    • Not supported by Firefox
  • Speed limiter
    • Cannot be done with the WebExtensions downloads API
  • contenthandling aka video sniffing, request manipulation?
    • PITA and/or infeasible - Essentially cannot be done for a large part and the other prt is extemely inefficient
  • Checksums/Hashes?
    • Cannot be done with WebExtensions - cannot actually read the downloaded data
  • Mirrors?
    • Cannot be done with WebExtensions - no low level APIs, see segmented downloads
  • Metalink?
    • Currently infeasible, as we cannot look into download data streams. Headers-only might be an option. But then again, metalink's features are mirrors and checksums, and we cannot do those.