Consider testing EPUB with the titles from epubtest.org

[GeorgeKerscher]

We at the DAISY Consortium test EPUB readers. We publish the results and also maintain a Round up of reading systems on inclusivepublishing.org.

I installed Paperback on my Windows 11 machine and did a quick test using the Fundamental Basic test book, and the Fundamental Non-visual test book. I tested with NVDA.

The App looks promising! It was failing some of the test, such as go to page, reading alt text, presentation of MathML, language shifting tests.

Happy to help. I am George Kerscher, kerscher@montana.com

[trypsynth]

Hi @GeorgeKerscher and thanks a ton for your interest in Paperback!
I'll definitely give these a look, I'm taking today off from Paperback dev for the most part to deal with some life commitments but I want to make the most robust epub parser that exists, and this will definitely help with that. Thanks again, and I'll let you know if I need anything from you/you can help!

[GeorgeKerscher]

You know about Thorium from EDR Labs and the Readium project. Open source for many of the parts.

Thorium is the benchmark I use to compare EPUB reading systems. I use NVDA, Jaws some, and am starting to test with Narrator. Microsoft is putting a lot into Narrator at this time. It is still a little rough around the edges with Thorium.

[amirsol81]

You know about Thorium from EDR Labs and the Readium project. Open source for many of the parts.

Thorium is the benchmark I use to compare EPUB reading systems. I use NVDA, Jaws some, and am starting to test with Narrator. Microsoft is putting a lot into Narrator at this time. It is still a little rough around the edges with Thorium.

@trypsynth @GeorgeKerscher With all respect to Thorium and its developers, I personally find Thorium quite heavy and complex. It is an Electron-based app that essentially embeds a full browser inside itself. Even on my high-end Windows 11 machines, it does not run very smoothly.
Its use of hotkeys and the way it moves the user between different windows often makes it hard to stay focused on the actual text. For example, unlike most Windows reading apps, Thorium presents the Table of Contents as an HTML page instead of a tree view, which makes navigation between items unnecessarily difficult.
I truly hope Paperback does not follow the same path. Thorium also displays only one section of an EPUB at a time, rather than giving access to the entire book as a continuous document. Many readers, myself included, have not been able to connect well with its user interface.
If we are comparing reading experiences on Windows, IMO QRead and Bookworm still provide a much more fluid, text-centered, and accessible reading environment than Thorium.

[GeorgeKerscher]

I agree that having the chapter ending and forcing the user to move to the next file is a problem. It would be much better if it were seamless.

Regarding the TOC view, I like that I can navigate by h1 or h2 etc. to get to the various parts of the book.

It is a sophisticated application and the features may be too much for many users.

I look forward to see what you can do. In the non-visual reading title, there are tests with tables, images and alt text, MathML, and language shifting tests.

Let me know if we should add Paperback to EPUBtest.org. I think you may want to work on some items so the scores look better; I would think we do not want to publish with low scores.

[a-singer]

I agree with this, I want an app which will get out of my way and let me read the book. Sadly, if I were ranking readers on Windows, the furthest from that goal is Thorium, not because of its features but because of its sluggishness, while the closest would be a tie between paperback and QRead, with the edge to PB because of its speed of opening/navigation. I may say my primary interest is not the program, it's the material. If the material is made less easy to use (open, close, navigate, select, bookmark, etc), then the program is not of use no matter how much it can do. At best I will use it to deal with things only it can do and return, with relief, to the faster and smoother application. I really hope PB doesn't go down the route of bloat, if a feature would slow things down, that's one reason for not adding it even if it would be great otherwise.

[GeorgeKerscher]

Hi again. We at the DAISY Consortium produced extensive user requirements for reading systems. I think it will be officially adopted by the Consortium in the next day or two.

I think this would be a good document to review as you work on Paperback.
DAISY User reading system requirements

[trypsynth]

Hi @GeorgeKerscher I was looking over this document, and it seems Paperback currently falls short in these areas:

Read Aloud: Text-to-speech functionality provides an audio option for text-based content, helping users with visual impairments, dyslexia, or reading challenges. The app must support seamless playback with proper language handling, visual emphasis, and playback controls.

Easy enough to add, just use Microsoft SAPI on Windows, AVSpeech on Mac, etc.

Embedded Audio: This covers pre-recorded audio content (human-narrated or TTS) that comes with the publication, either as audio-only or synchronized with text. Users must be able to control playback, see synchronized text highlighting, and customize the listening experience.

Now this one worries me a bit. I have no problem with making Paperback play audio embedded in books, but it is not an audiobook player. The focus is on text navigation front and center, not audio, which would require a whole other UI. That's sensible for embedded audio, but not for entire documents. That said, if I support embedded audio and just not the full playback of audiobooks, maybe it won't work against me too hard.

Visual Adjustments: Flexible display options allow users to customize text appearance including font type, size, spacing, colors, and contrast to improve readability. These adaptations reduce visual stress and cognitive load for users with low vision or dyslexia.

Yeah, can do in time. I'm completely blind myself, so testing these features for me as an individual will be difficult, but I 100% see the value and intend to add them.

Highlighting: Users must be able to mark important text portions with customizable colors or categories for quick reference and review. Highlights should be distinguishable, hideable, and automatically saved across sessions.

I think I could fulfill this requirement by making bookmarks and notes show up visually.

Entering Answers: Learners must be able to input responses directly into embedded fields within publications, supporting various input types from text to multiple choice. This functionality helps users with print disabilities complete exercises without switching to separate documents.

This one also worries me slightly, although in the long run it could actually be incredibly cool if done right. The only format I know of that lets you do this really is PDF, and if we turned Paperback into a way to fill out PDF forms, that would be sick. That said I think it may be a bit out of scope.

The misc requirement also mentions password-protected documents, but that's in progress and will be coming very soon.

Thanks, and please let me know any thoughts!

[a-singer]

You can, of course, tell me to go to hell if you like. This is your project, and the excellence is do to your work, I make my comments keeping that in the forefront. Having said that, I figure I want to mention this anyhow. If any of these things would make PB more unwieldy, far bigger, etc I, for one, would hesitate to put them in no matter how bad PB may look on a scoring system because the program doesn't have them. Part of this, of course, is because I'm blind myself, and I'm scared of having features which will make using PB like the walk through molasses that is current Bookworm, Thorium, or Adobe's various products. Allot of this stuff sounds reasonable in the abstract, but may well be a bit of a disaster when implemented or, to put it more accurately, I've seen most of this and have always seen it in a thing which I hate using. Whether I hate using it because the developers don't understand how to put all this stuff in or whether it's because nobody can actually put the stuff in without making a horrible mess, I don't know. I say that despite the fact that some of the stuff would be lovely, particularly the PDF form filling. I put that up in issue 55, and you explained the difficulty, which I get. Having said that, it would be great to have if, again, it doesn't destroy the current excellence and speed. I'm sure, at some point in the past, MS Word loaded quickly and let you write well. Look where it is now. I write this knowing that you understand this sort of thing just as much, or more, than I do, I just want to be sure it doesn't fall through the cracks because of these ideas about what PB should be able to do.

[GeorgeKerscher]

I suggust you focus on only the fundamental basic and the non-visual reading titles. The visual adjustments and Read aloud titles are for low vision and dyslixic folks.

I get that you want a fast light weight reading system and focusing on screen reader users sounds great.

IMO the App should be able to present all the information, so images with the alt textables, MathML are essential. MathCAT is freely available for the math, but I think you only need to do is expose the MathML and MathCAT in the screen readers will do the rest. , t

[aryanchoudharypro]

I suggust you focus on only the fundamental basic and the non-visual reading titles. The visual adjustments and Read aloud titles are for low vision and dyslixic folks.

I get that you want a fast light weight reading system and focusing on screen reader users sounds great.

IMO the App should be able to present all the information, so images with the alt textables, MathML are essential. MathCAT is freely available for the math, but I think you only need to do is expose the MathML and MathCAT in the screen readers will do the rest. , t

hi, is their a way to do mathml without actually makeing it use a webview?
I meen we are eventually going to use webview to implement table navigation, but any ideas here would be great, especially because i don't have the full knowledge to integrate math. Thanks!

[GeorgeKerscher]

I do not know, it it is out of my area of expertise. I do believe this comes with the web view implementation. If that is in your plans anyway, I suggest you wait and see if it comes for free.