0.1.0 Roadmap

[aaronleopold]

Roadmap

This issue will serve as a roadmap outlining the major features planned for the first release candidate of Stump, v0.1.0, as well as some notes regarding potential post-release goals. This should be treated as a mostly static document, as the features listed are what I consider to be the minimum set to reach a viable release. The section dedicated to post-release features may receive more updates, though. Also note that a few features listed will be more-so nice to haves, rather than hard requirements.

Pre-Release Items

Below are all of the essential features that will need to be completed with at least minimal functionality before the first v0.1.0 release candidate is slated:

• [ ] Miscellaneous maintenance and code smell cleaning
  • [ ] Reorganize and consolidate some of the client code, its very messy
    • [x] Split api into separate package to house all the axios functions
    • [x] Rework gross hooks + callback types
      • [x] Fix base query hooks, e.g. custom useQuery
      • [x] Use corrected base queries for all existing hooks
  • [x] Revert the whole prelude crate, organize in other crates where it makes sense
  • [ ] Audit dependencies, prune where needed
  • [ ] Fix all the peer dependency warnings (where needed, F you storybook lol)
  • [ ] cargo report future-incompatibilities --id 4 --package rustc-serialize@0.3.24
  • [ ] local-ip-address check if release yoinked comes back okay
  • [x] Optimize [moonrepo](https://moonrepo.dev/docs) setup
  • [ ] Tighten logging noise in core
  • [ ] Revisit rework of StumpConfig for more efficient environment parsing
• [ ] General server upgrades
  • [x] Migrate to Axum v0.6
  • [x] Look into sqlite session store at some point to replace the in memory store.
  • [x] Consider refactoring queries to use more current select/include patterns
    • https://prisma.brendonovich.dev/reading-data/select-include#many-relation-options
  • [x] Change HTTP endpoints to be unpaged by default (except opds)
• [ ] Add proper testing
  • This will not hold up a release, it is a fairly large and complicated undertaking. I feel that initially, while not ideal, it would be OK to release without this. Then, to to help ensure general code safety, testing should be prioritized alongside new features and bug fixes.
  • [ ] Rework integration_tests
  • [ ] Add unit testing throughout core
  • [ ] Add unit testing throughout server
  • [ ] Consider testing UI elements
• [x] Refine Stump color palette (gray colors are :barf:)
• [x] Start building out, but don't use quite yet, custom UI components in new package components
  • [x] https://github.com/shadcn/ui
• [ ] Rewrite job system
  • Note that these are super complicated with Rust, and they all might not fit into the first release. At a minimum, jobs will be cancellable.
  • [ ] Introduce more stateful-ness with job variants
  • [x] Support cancelling jobs
  • [ ] Support pausing jobs
  • [ ] Save job queue / active workers state during a graceful shutdown
  • [ ] Restore job queue / active workers state after graceful shutdown
• [ ] Complete major UI scenes to at least MVP stage:
  • [x] HomeScene
    • [x] Continue Reading, Recently Added, Etc.
    • [x] Statistics
      • Note: I'd like more, but I think this is enough for MVP
    • [x] TopBar
      • Currently unsure what should go here
  • [ ] book/epub
    • [x] BookOverviewScene
      • [x] Show progress (if any)
      • [x] Show next in series (using a new cursor page param?)
      • [x] Render book description
        • [x] Plain text
        • [x] Support markdown
      • [x] File information and tags
    • [ ] readers:
      • [ ] ImageBasedReader
        • [x] ToolBar
          • [x] Fix overly eager image loading
        • [ ] Support two-page layout
          • This may be post-release
      • [ ] AnimatedImageBasedReader
        • This will not hold up a release, as it is mostly “extra” functionality on top of non-animated variant
        • [ ] Correctly implement desired page swipe animation.
      • [ ] EpubLazyReader
        • [x] syncs epubcfi and progress_percentage tracking updates with server as reading
        • [x] Restores progress accordingly in UI (i.e. go to correct page based on epubcfi)
        • [ ] https://github.com/aaronleopold/stump/issues/114
  • [x] series
    • [x] SeriesOverviewScene
      • [x] Add series information on 0th page
      • [x] Support filtering and ordering of grids/lists
  • [ ] library
    • [x] LibraryOverviewScene
      • [x] Add library information on 0th page
      • [x] Support filtering and ordering of grids/lists
    • [x] LibraryExplorerScene
      • This will not hold up a release
      • [x] Support basic filesystem navigation
      • [x] Double-clicking media should route to appropriate scene
    • [ ] CreateLibraryScene
      • [x] Replaces all functionalities from the previous modal implementation
      • [x] Support configuring scanner options
      • [ ] Tag management
    • [ ] UpdateLibraryScene
      • [x] Replaces all functionalities from the previous modal implementation
      • [x] Support updating scanner options
      • [ ] Tag management
  • [ ] settings
    • [x] GeneralSettingsScene
      • [x] Users can update their username and password
      • [x] Users can update their locale preferences
      • [x] Users can update their avatar image URL
      • [x] Probably one or two more things (this page is rather bland ATM)
    • [ ] UserManagementScene
      • [x] Admins can create users
      • [x] Admins can delete users
        • [x] soft delete
        • [x] hard delete
      • [x] Admins can view users
      • [ ] Admins can update certain user attributes
        • [ ] e.g. remove offensive or otherwise unwanted avatars
      • [x] JobsAndConfigurationScene
        • [x] TODO
      • [ ] SeverConfigurationScene
        • [ ] TODO
  • [x] Remove all those Loading... placeholders 🤢
• [x] Store redirects for unauthenticated access via HTTP calls
  • I.e. if I visit /some-page/1, and get redirected to /auth, should redirect instead with query params to restore state: /auth?redirect=/some-page/1
    • [x] All client side routing and API calls
    • [x] Swagger UI
• [ ] Scanner
  • [ ] Create new option for series dedicated scan
  • [ ] More content-aware scan resolutions
    • This is very complex, and might be pushed to post-release.
    • [ ] Support restoring missing library if newly scanned library “matches” missing one
    • [ ] Support restoring missing series if newly scanned series “matches” missing one
      • Note that there is a VERY naive implementation of these, but it simply won't cut it. So I am not counting them as done.
  • [x] Restrict concurrency to avoid performance degradation
  • [ ] Configuration
    • [x] Per-library basis
    • [ ] Support .stumpignore files to alter what the scanner will look at
    • [x] Support generating thumbnails https://github.com/aaronleopold/stump/issues/44
      • [x] Generic, downscaled webp generation
      • [x] Specific, user-configured, scaling options:
        • [x] Custom scaled height and width (e.g. 0.5 x 0.75)
        • [x] Custom scaled (e.g. 0.5, equal to 0.5 x 0.5 above)
        • [x] Custom size in pixels (e.g. 331.5 x 512)
      • [x] Specific, user-configured, output formats:
        • [x] png
        • [x] jpg
        • [x] webp
• [ ] Minimal RBAC / ABAC https://github.com/aaronleopold/stump/issues/41
• [ ] Basic reading list support https://github.com/aaronleopold/stump/issues/37
  • At the very least, I'd like to be able to have users create a reading list and edit/delete them.
• [x] Format support
  • Each format should generally support the following:
    • Basic metadata collection (pages, file size, generation of checksum hash, etc)
    • Page extraction (excluding EPUB until https://github.com/aaronleopold/stump/issues/62)
    • Thumbnail extraction
    • Progress tracking while reading with Stump readers
  • [x] ZIP
    • Maybe support https://github.com/iComics/comicbookinfo for pre-release.
  • [x] PDF
    • This may be a post-release item (I hate PDF)
  • [x] RAR
    • Currently excluding Docker, as there is a nasty bug that renders RAR unsupported in Docker
    • [x] Support RAR-to-ZIP converting
    • [x] If 0.5 unrar release lands:
      • https://github.com/muja/unrar.rs/issues/26
      • [x] Move off of custom read_bytes implementation
      • [x] Try and resolve https://github.com/aaronleopold/stump/issues/38!
  • [x] EPUB
    • [x] epubcfi and progress_percentage for progress tracking
    • [x] Generate TOC from meta
• [ ] Finish CI workflows https://github.com/aaronleopold/stump/issues/93
  • [x] Setup self hosted runner
• [ ] Add some automated system for versioning/releases
• [ ] Documentation
  • [ ] Finish writing it lol
  • [ ] Ensure all configuration options are documented
• [ ] Thorough user testing!!
  • [ ] I’d like a solid 2-week period of at least a handful of testers using Stump trying to break it wherever possible

Post-Release Items

First and foremost, the top priority after the initial release will be bug fixes and maintenance. If you've read the entire Pre-Release Items section, you'll know that v0.1.0 is intended to be a sort of MVP, and not a fully featured option yet. That being said, the following are things I'd like to explore afterwards, also not necessarily in priority order:

• [ ] Desktop App
  • Note: this should be very little effort, and will only really require:
    • [ ] GitHub actions to run the builds/uploads
    • [ ] Discord presence would be cool https://github.com/aaronleopold/stump/issues/60
• [ ] Tachiyomi Support https://github.com/aaronleopold/stump/issues/48
  • [ ] While one day I would like to create an in-house mobile application, since I can re-use some of the code for web/desktop, priority will be here
• [ ] Look into better validation for some HTTP inputs via https://github.com/Keats/validator
• [x] Rethink HTTP endpoints allow better range of filter options
  • [x] Basic structure implemented, just a matter of prioritizing it time permitting
• [x] Consider paying for small hosted Stump instance for demo purposes
  • [x] Support a sort of demo mode with restricted functionalities
• [ ] Start work towards benchmarks crate

[aaronleopold]

Closing this as to not maintain multiple lists. I'll just be using projects