Improve UX of TTS Message option

[pilot51]

A Google Play review mentioned that the TTS Message option is a little convoluted. I agree it could be better.

Currently my idea, as detailed below, is to use tags instead of plain text for each message component.

• Improve descriptions of the components.
• Basic and advanced modes:
  • Basic: Checkbox list of components that can be reordered either by drag-and-drop or arrow buttons. They are separated by newline in the TTS text. (Thanks @Robio!)
  • Advanced: Text field as it is now but with tags as described below.
    • Short text in the tag identifies which component it is in a more self-explanatory way.
    • Each tag is listed and clicking on a tag in the list adds it to the text field at the cursor position.
    • Drag-and-drop to move tags around in the text field or to add to a specific position from the list.
    • A tag is removed when it is either deleted by the keyboard or pulled out of the text field.
    • The tag's text code (e.g. #M) is converted to the tag on the fly. The text codes should still be shown in the description or tag list for users who prefer to type.
• Allow saving a format to a list of presets. This would be especially useful now that per-app settings exists.

This may change as I give it more thought and find out how difficult it is to implement.

[Robio]

Hi @pilot51, I'm the reviewer who inspired this ticket. :) This sounds pretty good, although I'm not entirely clear on your description. But here's a thought: How about a list of checkboxes to select kinds of content to include, with grippers to drag those items into the order the user prefers (which is what I assume rearranging the #tags does in the current version):

≡ ▣ App Name
≡ ▣ Ticker
≡ ▣ Subtext
≡ ▣ Content title
≡ ▣ Content message
≡ ▣ Content Info
≡ ▣ Big content title
≡ ▣ Big content summary
≡ ▣ Big content text
≡ ▣ Text lines

BTW, I'm entirely unclear on what most of these descriptions mean. App Title (I've changed to App Name for clarity) I understand. Content Title and Content Message I figured out from using the Test feature in Settings. But I think they all need either a description or names that are more self-explanatory to end-users, like...

• App Name (currently #A)
• Notification headline (currently #C)
• Notification message (currently #M)
• Time stamp (currently #T — I think)

As for the rest, I don't have any idea what they mean. :)

[pilot51]

Aside from being very easy for me to implement, the main reason it's a text field instead of just an order is because punctuation causes TTS to sound different and I don't want to force anything on the users.

As you can see by the current and old defaults shown in the description, I couldn't settle on how to separate the components to prevent TTS from making it all one sentence. Originally #A: #T, then #A. #C. #M., and now newlines instead of periods.

This also allows users to do things such as Notification from #A with title #C which, with the test notification, would become "Notification from Voice Notify with title Content Title".

I like the idea of essentially making features "open source" for regular users to change however they want even at an advanced level, especially if it reduces my workload as a developer. Though now that I think of it, that makes a case for basic and advanced modes for the setting, where basic would be something like your suggestion.

As for the descriptions, I agree they aren't very clear and could use improvement. Some of them I don't totally understand myself. I carried them over mostly verbatim from the names of Android's notification APIs, and the documentation doesn't help much. Some are shown differently between versions of Android, so I can't necessarily rely on my experience to form an accurate detailed description.

#T is currently the ticker message, which used to be shown briefly in the status bar the moment a notification was posted, but on recent versions of Android it isn't shown anywhere.
I've considered adding a format option for timestamp, but that opens a whole other can of worms with how to format the timestamp itself and there isn't really any demand to justify the effort.

[Robio]

Just getting back to this. Having just recently figured out how much difference can be made for TTS by using punctuation and line breaks, I see where you're coming from regarding flexibility.

This would probably be a lot more work to implement, but what comes to mind two tiers of settings for the spoken content:

1. User-friendly: My suggestion above (checkboxes and drag handles), which would automatically include a short pause at the end of each statement — [App name] ... [content title] ... [content message]...
2. Advanced: The current functionality — #A, #C, #M

User-friendly could be the default, with an "Advanced" button at the bottom. When clicked the "Advanced" screen could have a little explainer, like "More control over how text is ready, using commas, periods and line breaks" then the chart of tags.

/just spitballing :)