komanton/dicty

Multi-Language Dictation Tool (speech-to-text) with supporting: Skype, Slack, Word and much more.

• 👋 Hi, I’m Anton Komarov
• 👀 I’m interested in TypeScript, NodeJs, Angular, Lambda/Serverless
• 🌱 I’m currently learning ✨Rust✨, Event-Driven Development, Domain-Driven Desing
• 💞️ I’m looking to collaborate on a project which can help to save the environment or help to grow small enterprises
• 📫 How to reach me [email protected] or StackOverflow

dicty's Issues

Register domain name

See #9

it's need to register domain name for the homepage: dicty.me
https://ru.godaddy.com/domains 12$

1. register domain name
2. clean old links
3. and make release of desktop dicty
4. Bind git hub pages to godaddy
5. https://jinnabalu.medium.com/godaddy-domain-with-github-pages-62aed906d4ef
6. https://docs.github.com/en/pages/configuring-a-custom-domain-for-your-github-pages-site/about-custom-domains-and-github-pages
7. https://hackernoon.com/how-to-set-up-godaddy-domain-with-github-pages-a9300366c7b

Register homepage into Google

1. ad keywords to metadata the site
2. register homepage in Gogle search
3. create an acount for dicty
4. Add email to main page

Fix non-english-language in Linux

this is a bug inside robot js

Create Dicty homepage

The idea not just create a home page, but merge web part and home page into single website.
It should help to monetize the project by ading some advertisement on the homepage.
Elso it should help to make some web-analitics for the project.
the homepage should help to promote the dicty into web search

Confirmed: that dicty works perfectly with github pages.

Need to migrate web-part from dicty repo to dicty-web repo

Check list:

1. make homepage from spech.html Use github pages: https://komanton.github.io/dicty-web/speech.html
2. Make up of speech.html: content from readme, download, github repo, styles, remove 'web speech api demo', https://templatemo.com/tm-563-seo-dream
3. Publish it.
4. Clean dicty repo from: web static server, url to local host, move folder speech,(let's support bothvariants) make release with new URL to public site
5. Make new release for desktop-part

Short cuts

https://www.electronjs.org/docs/api/global-shortcut
Control+space

Prepare Build for macOS

1. Take the distribution build of electron for macOS
2. App files from src to the build for macOS
3. Rebuild robot package, see distribution instruction in readme
4. Test app
5. Zip all files, like for windows.
6. Publish it into releases

Another word, we need the same, like here https://github.com/komanton/dicty#distribution but for macOS

Desktop buttons design (concept)

Language switcher v1

1. Send message switch language to web
2. Web: add text box to filter languages just use navigator.languages to set up current languge on event handling also check this value in settings window on desctop (probably it grab current windows language)

Jira/Confluence formating plugin

Plugin which supports formatting for Jira and Confluence, like Bold, Italic, Strikethrough etc.

Can't switch languages in case of several languages in Chrome

If you add into chrome:

English
Russian
Danish
English(United States)

'Do not close this page' label

Migrate to electron v15

1. Fix the enableRemoteModule: true,
2. Fix robot js built script

[Desktop] Switch language on button press

Note: Due to Transparency activated, the button can be clicked only if it has a special Style. See previous buttom

Idempotency for start-stop mesage

it's nead to have start-stop mesage to synchronize part desktop web part. another word it's no ned to switch two oposite volume 4 start-stop mesage

Distribution (Windows)

1. Create archive with electron https://www.electronjs.org/docs/tutorial/application-distribution#manual-distribution
2. Add js files from current repo
3. Add instruction for installation
4. Publish to current repo artifacts

Auto connect to desktop-part

1. conect to the desktop part nope only on load Elsa after each second because distal part may be loaded after the page is already loaded. this is why mis it loading phase

Rename electron.exe to dicty.exe

and fix instruction

Try Kickstarter

Let's Kickstarter answer should we continue the project

1. Demo on YouTube for Kickstarter
2. Create a description and road map
3. Calculates: current MVP + feature development+ UX + design
4. List of features
5. Inline check list, smart select/rewrite, devmode, punctuation, support platforms, hot commands and commands batching

The strange space at first place

each new dictation begins with strange space

Support for 'no speech detection' timeout

If no speech so client moved to the state 'No speech was detected. You may need to adjust your'
Need to hanlde it gracefully.

Just send back message from web part to desktop part and handle it in toolbar component to disable button.

Language switcher v2

1. React on switching language
2. Move language drop-down from web to desktop part in settings window
3. Convert drop-down into multi-select
4. When language selected send message to host and store it in global state.
5. When language selected add language into separate select control with single select option
6. React on hot keys, and send message to web part and settings
7. In settings handle message about switch language and show label

Option list formating plugin

Create a plugin which will be format text like option list. When user say 'Option 1 buy apply Option 2 but tea', the plugin should add a new lines with numbers:

1. Buy apple
2. Buy tea
3. ...

Add installation instruction for users

In readme and in txt for distribution

Add more languages

https://cloud.google.com/speech-to-text/docs/languages

Integration with ChatGPT (plugin)

Like here
https://www.npmjs.com/package/@abhagsain/ai-cli

Open link by voice command

Create a command which responsible to open some link, like: 'Open: Pull Requests', 'Open Jira Board' and etc.

Leave only last sentence on preview screen

no ned to remember the history of the translation.
ned to remember only last sentence due to security reason

Make release of Desktop-part with new URL

1. Clean dicty repo from: web static server, url to local host, move folder speech,(let's support bothvariants)
2. make release with new URL to public site

Supports 'domain' dictionaries

The idea is about to support dictionaries from diferent busines domains, like IT, Healthcare, Finances, Legals and etc; slangs words, abriviations.

Probably, we can achieve this by use ofline voice recognition for instance with help of TensorFlow and sugest results from tensorflow like an alternative variants of voice recognition.

It can be a plug-in feature.

User can create its own voice-DSL

It can be intermediate step to suport development mode (elso like a plug-in)

Emoji plugin

https://cloud.google.com/speech-to-text/docs/spoken-emoji

Support punctuation for different languages

1. we can recognize punctuation after recognition and
2. find symbol in dictionary of symbols 4 current language,
3. Afters that, we can replace the text symbol with symbol itself
   https://cloud.google.com/speech-to-text/docs/spoken-punctuation

Make button more clickable

increase the size of the clickable area near the button
decrease the size of the cros icon

Integration with devhints.io (plugin)

It would be great to translate some keywords into real commands and snippets by searching them into descriptions in devhints.io. For instance, user say 'start node app' -> 'npm start'

Try open collective

Waiting account approval... https://opencollective.com/dicty-me#sponsors

Smart detection the first leter

[Blocker] Enable https

Need to enable HTTPs otherwise microphone permissions

Add web server for static file to fix issue with microphone permissions

1. Use this server to server speech.html https://github.com/sindresorhus/electron-serve
2. Add alert after start with url to speech.html (to open it in Chrome by click on link in alert)

Add use case about making notes during a Skype cal

Abbriviation command

should understand the command: 'abbriviation cnn' -> 'CNN'

probably wil require tensorflow to implement this feature