kokushkin/peoplevocabularies

Created with CodeSandbox

peoplevocabularies's Issues

Translation doesn't work (in mobile versions specifically) when the word starts from upper case letter

This is screenshot from PC version but initially it's mobile problem, cause there typing starts from upper case by default.

markup issue

Some "tricky" words aren't translated

Microsoft doesn't translate "tricky" words. So, consider to add some more translators.

Problem with number 1950

It should not be here. It's number.

Have uploaded text and can't find word in the vocabulary

So.. after text has been uploaded it must be in the vocabulary!

Network error

Get when sign in. But, If I go through another network (for ex. mobile internet) all good. Very strange
In half hour it disappears.

couple of 's , 't endings

Tracking users by queries

buy real domain

Strange case with the word "super"

Add words listening

Buy domain

buy real domain

Improve coding infrastructure in VS code

• jump on create react app
• setup linter and code formatting
• setup typescript

Add spins for long time operations

Todo list

• restore deploy
• make footer
• make landing page
• integrate landing page with registration and auth
• rebuild component architecture in application (decrease duplication)
• make words list scroll (only for desktop)
• implement algorithm for special learning (multistage learning)
• make delete words from vocabulary passable (check again)
• fix colors for header and footer
• make search to delete desirable word
• fix bug of " in the word. For ex. I can't delete word with "
• fix bug if we delete all words from vocabulary (I guess it doesn't work correctly)
• make translation of words in returning list
• re-implement(ask Julia) special algorithm for (ing, ed and other cases) (look for elastic help) - I've used the clouds in fact
• sort words according to how often they encountered (2500 words)
• unregular words (verbs)
• make translator like another part of the vocabulary
• publish without domain
• add recapture
• check in real situations
• buy domain

next iteration ------------------

• handle situation in translator, when word not in the context

• fix bug when you analyze again the same text it doesn't work (doesn't run effect cause text the same)

• refactor list in different components

• go to styled components

• go to typescript (?)

• thinking about typos, pron nouns and so on

• include 2500 words and add new words + information how often they are encountered

• what about using google translator to form basis vocabulary - you take any text that big enough and then sub struct from it last searches in google translate

• or how about integration with google translator and then just have a suggestions what do delete (cons and pros because in suggestion will be words what I maybe just wanted to check)

sometimes it's not enough words in translation

Some podcasts to big to analyse

Try,
https://softwareengineeringdaily.com/wp-content/uploads/2020/01/SED980-Josh-Wills-Slack-Search.pdf
This is limited with 30 sec timeout. Limitation of the Amazon. It can't be changed. But maybe we can optimize something (watch logs).
So, it works but very unstable on big podcasts.

Move list down in the tablet version

Make footer unfixed, because it hides button (issue in mobile and tablet)

Get error xml when reload any page (when served from Amazon, not in local IDE)

Make passable translation of the word without context

Sometimes it's very convenient.

Make video about the site (user part)

Make video about the site and place references on the site

add another irregular forms for adjectives and so on. Also what about have-has? Looks like we should keep first irregular form also.

Take styles under control

Reimport bootstrap properly and change base colors.

Consider using of some component library and/or styled-components. But this rather for next time.
https://create-react-app.dev/docs/adding-bootstrap/
https://react-bootstrap.netlify.app/components/accordion/
https://www.youtube.com/watch?v=MpGLUVbqoYQ&t=2337s

Some words which Google Analyze Syntax thinks they are proper nouns

Some words which Google Analyze Syntax thinks they are proper nouns, not just nouns. That leads to

stale translation results. It's better to say - you are trying to get translation for proper nouns or smth like this.

Problem with some names

Seems like google doesn't know who is Feynman.

Some artifacts still here.

Add captcha

fix markup (look at the left margin in the list of translations)