After logging in, you can create projects, which would ideally correspond to a single text you are reading. When creating a project, you specify title, description, source (original) language and target language. Filling out the description properly will allow the grammatical analysis, translations and chat to optimize responses better (e.g. targeting the responses at the specific type of work or genre).
After you create a project, you can add segments to it, i.e. pieces of text. You can type or paste the text when you add the segment, or leave the field empty and read the text off a photograph once the segment opens — either with the “Extract from Image” button or by pasting a screenshot straight into the original text field.
In the segment editing panel, you can modify the text of the segment, input text of the segment via an image of text, analyse grammar of the text, translate it into the target language, and chat about it. Every panel here saves itself as you edit — the line under each one tells you where its text stands — and the arrows beside the segment title walk through the text without going back to the project.
Translation of the text can be performed either by Google Translate API or by an administrator-enabled large language model. You can change the model in the segment editing view when AI models are available.
Ideally, it is best to generate grammatical analysis of the text before generating translation. If you generate translation through a large language model, the model has access to grammatical analysis and project description and can optimize for them.
If you have special instructions on how to perform grammatical analysis or translation for the given project, please input them in the project description. The language model will have access to it and can optimize the analyses and translations based on these instructions.
Chat has access to the project description and segment information (grammatical analysis, original text, and translation), but not to other segments in the project. I plan to introduce the ability of the chat to refer to other segments within the project in one of the future versions.
You can add notes to the project. It is handy if e.g. you have access to published translation and want to store it along with your translation of the segment.
You can add annotations to the original text, translation, grammatical analysis and notes. Annotations are displayed at the bottom of the page. You can also colour-code annotations for different categories.
Reader Two has basic collaboration functionality. You can invite other users of the application to 'friends' list by inputting the email they used to log in to the application. You can also share projects with them.
When sharing a project you grant one of two roles, and the share form states what each one means: an Annotator can read everything and add annotations, but cannot change the text; an Editor can change the text, translations, analyses and notes. The same two words are used everywhere the role appears.