codedread / kthoom Goto Github PK

Especially the menu button

As per the original thoughts in Issue #49, use the ComicVine API to try and search and auto-populate or suggest metadata for books.

There should be an option for kthoom to remember which files were opened previously and automatically restore (load) them from the network upon restart. The option and which books are open can be saved as a cookie.

Hi,

bitjs is not synchronized with the bitjs repo

It currently creates a small bug when deploying the app.

Regards

With Service Workers, we can make this work even if the user is offline.

Probably need to introduce a node module with a test script... maybe add it as a GitHub action?

... that's embarassing

The old way of Google sign-in is deprecated and will stop working in mid-2023:

https://developers.googleblog.com/2021/08/gsi-jsweb-deprecation.html?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+GDBcode+%28Google+Developers+Blog%29

Once loaded (from any source), it should be possible to save the file locally to the hard drive (via downloading from the browser)

Things we can remove:

• The need for traceur
• Use ES2015 method definitions on objects

When retrieving from Google Drive, sometimes it can take a longer than expected time for the first bits to come over the wire. While we are waiting for this to stream, we should show some indication that things are happening in the progress bar.

If you have a file opened, you should have the option to add it to the Interplanetary File System

Good example: https://www.gutenberg.org/ebooks/27805

There is no visible progress when fetching a book from IPFS. Need to determine if the js-ipfs library gives us any indication of the download progress.

Paginate eBooks using something like https://github.com/pagedjs/pagedjs

kthoom should let a user load a reading list (some file in an agreed-upon format) that is a list of comic books to load, in-order, into the Reading Stack.

1. Load an EPUB, displays fine.
2. Try to load a second EPUB. It loads but nothing displays (JS errors)

You have to refresh the web page and load again

Kthoom should let the users enter true full-screen using this API: https://developer.mozilla.org/en-US/docs/Web/API/Fullscreen_API

Can also remove hiding the UI manually after implementing this.

Markup for the kthoom menus take up about 125 lines of HTML, about 35% of the total payload. Since the menus are not visible upon first load, we can reduce initial HTML payload and further componentize the menu and make it more reusable in other contexts.

To reproduce:

• Go to the last page of a book
• Press the right arrow quickly several times

Because it takes some time for the next book to load, the book only has 1 page when the next arrow keypress is detected, bringing to the user to the next book.

Basically when the user presses 'next page' within a small enough time increment, and the book is still loading, ignore the 'next page' request.

In Issue #23, a Service Worker was introduced to cache files and make kthoom work offline. However, in looking at the code and observing behavior, I think the SW prefers the cached version over the network version and its' hard to "convince" kthoom to get the latest updates.

Need to investigate strategies and apply a fix: https://jakearchibald.com/2014/offline-cookbook/#network-falling-back-to-cache

Kthoom does a pretty good job of fetching, unzipping and displaying pages. This should be extended beyond comic books to support other e-book formats, starting with EPUB.

https://en.wikipedia.org/wiki/EPUB

... or equivalent. Research to do:

https://web.dev/generic-sensor/
https://w3c.github.io/gyroscope/#gyroscope-interface
https://developer.mozilla.org/en-US/docs/Web/API/AbsoluteOrientationSensor#permissions_example
https://caniuse.com/?search=sensor%20api

Add support for SVG images inside CBZ/CBR files

For large comics or ebooks, the entire file may not be needed to be downloaded at once. For example, as the user opens the book, they may only need the first few pages. Kthoom could fetch enough data for the first five pages and then pause the fetching until the user gets closer to the "edge" of what has been fetched to fetch more.

This would save on bandwidth from client and server perspective.

A new API is available, currently only accessible via Chrome flag: https://wicg.github.io/native-file-system/

This could allow kthoom to do some interesting things:

• access directory contents within the app and let the user browse books
• update and save metadata to local books
• save books locally to disk

Let's say a user is at kthoom/?readingListUri=foo.jrl

They go to the next book. That should update the URL to be kthoom/?readingListUri=foo.jrl&bookUri=baz.cbz so that if the browser tab reloads, baz.cbz should be loaded by default.

If a comic book archive file has a corrupt image (JPG/PNG), kthoom should still be able to load in the whole book and then render the bad page to let the user know.

kthoom should support 2-page viewing

https://github.com/naptha/tesseract.js looks cool and could help with automating metadata search (issue #54 )

Now that all browsers support the EventTarget constructor, we could remove EventEmitter.

This would imply that we need to remove all usages of the subscribeToAllEvents() method.

After issue #18 and issue #42 are implemented, we can then do two major things:

• Use the ComicVine API to find metadata for the comic book and update the archive.
• Support re-archiving the comic book to the file system directly (requires zipping)

please support PDF files

This actually seems to work well already on my iPhone. Just need a slightly better UI for it.

I may attempt this myself as I haven’t done much web front end dev

Users should be able to change the order of Books in the Reading Stack. Most convenient for mouse users is being able to drag-drop books.

Think carefully about designing this feature so that when accessibility is added (see Issue #15)

Support &page=11 when the bookUri query param is present.

• A View sub-panel that has icons for 1-page/2-page, rotate, fit-mode.
• An Open sub-panel that has icons for Files, URL, etc?

Toolbar visibility should be toggle-able, but default to on.
Remember user's toolbar visibility in saveSettings_()

Find some SVG icons for curved next / prev page arrows. Rotate them 90 degrees to account for page orientation.

Not really an issue, but it would be nice - especially on mobile devices - that comic books can be read panel by panel. Something similar as what the Comic Trim app does or the French comic book site Izneo (On the bottom of the page, click on the second icon from the right to trigger picture by picture mode)

I figured out there is already some code on github, I just have no idea how to integrate it into Kthoom by myself (I'm still no coder... )

And thanks for all the great work, the 'optimized for streaming' code you wrote a while ago is simply amazing on my site.

It would be good to have a tool that can be run that "fixes" comic books archives that ensures the files in the archive are stored in reading order and includes metadata to indicate whether it is optimized for streaming

If you select multiple files to load, but one of them happens to be an invalid comic book file, then only the first file loads.

I realize this is probably not the best place to post a beginners question, but I couldn’t find a better way to contact the developers of this great script - my apologies!

I'm trying to integrate Kthoom in my comic book related website. Works fine, only in the menu is the ‘open from Google drive’ field missing. Pressing 'G' also doesn't do anything.

Obviously I’m missing something - I guess I need some api thing from Google - but I can not find more information on the net.

So if someone would be so friendly to point me in the right direction? I have to confess that my knowledge of coding is almost zero (but I’m an expert in copy/paste)

Thanks a lot,
Alex

Loading files from the same web server that kthoom is hosted on should be possible (and easy).

This issue is spawned from the following: https://github.com/comicweb/kthoom/issues/3#issuecomment-363669519

Maybe a window for comicbook.xml files inside the archive.

Found some discussion here: Tenma-Server/Tenma#32

I tried using kthoom on a few cbz files, and loading worked inconsistently. For some files, the console displays a message about 0 pages loaded, and the gui seems stuck in a weird state. Other files seem to work fine. Any pointers what I should do to try to fix this?

I noticed the decompression supports forward only streams (I think)

Could the loading of images also be progressive in that i can view the first page ASAP instead of waiting for everything to be loaded?

Today it does nothing. We should instead:

• automatically load the next book in the reading list (if there is one)
• otherwise load the previous book (if there is one)
• otherwise properly handle returning to the zero state (no book loaded)