LogoLogo
  • audit and issues
  • useful references
  • positioning idea
  • custom theme
  • Brainstorming
LogoLogo

audit and issues

things to improve

  • fix - if a domain is already in use in our system, we should throw an error for that during adding the domain. Maybe we are doing it for a different organization / different site -- vercel does throw an error in inngest. We should make sure its properly handled

  • fix - when a user leaves / closes the window, we are stuck with the caret

  • migrate web service to aws, vercel is super expensive in our case because of electric and long polling!

  • ship icon (svg) with the tree itself for better loading initially

  • BUG - code blocks with html

  • BUG - mobile navigation for spaces changing not working in renderer

  • BUG - if I refetch the json / yml for openapi spec and add it. It gets added in the UI but doesn't get changed when I publish again

  • BUG - because of our backend constraint of not able to add same api spec, backend sends an error which fails the frontend later

  • fix - while deleting a space, if its attached to a published site we should not let them delete and ask them to first remove the space from the published site, republish it and only then let them delete it

  • BUG - RangeError: Unknown node type: emoji

  • fix - Ability to drag and drop custom html components

  • fix - UI for theme

  • fix - UI for custom component

  • BUG - when I rename the space, the slug for it is not correctly renamed

  • fix - add new doc button / spec spec / group (we get name when hover)

  • fix - we need kotlin, swift, dart as a language (currently we have typescript as default), people might want something different (maybe we need ability to customize or give like all the major languages compared to what we have currently given to start with

  • Custom icons for things like google, etc. We need to have the ability to provide an svg somehow

  • Based on tags, we should create new groups and add them there. That would be much easier to do. Else manually have to put them into categories is hard. We should give the option on the UI to add it to the parent or create sub folders based on tags

  • BUG - Custom HTML component going back to the default after setting it up

  • BUG - Navigating on api links, rerenders the DOM for search api

  • BUG - on click, the scroll for sidebar goes to the top: can not resolve for now as we are using nodejs runtime

features for mvp

  • better ui for docs-renderer

    • favicon

    • logo

    • structure

    • header / specific links if needed

    • footer with social links

    • TOC

    • performance optimization for ssr

    • full text search

    • Optional - somehow ability for custom css / classes

    • Optional - inject javascript?

    • custom code block where we can add fancy components like cards and others

  • / for adding new elements

  • support for videos / html embeds like loom, youtube videos

  • oauth (signin with google)

  • improvements - ability to unpublish a site, how do I remove it / delete it completely? remove it from google where it has been indexed

  • SEO related settings so that the page gets listed on google correctly

  • API documentation - yaml / json format support

    • with playground to actually make the request and try out

  • posthog integration for analytics

  • ability to add internal links

  • ready for opensource development

    • better readme.md

    • better contributions.md

    • license

    • proper CI/CD pipline and github actions

    • removing migrations from vercel build command!

  • hide a single page from deployments. each page also has an option to hide so that it / all its chidren do not get published when we publish a site!

  • cmd + k with generate optimized static images to be served

    • or we do it automatically when user uploads an image somehow while we are generating the static site

  • Custom theme

  • Neighbour link at the buttom.

  • url instead of json for openapi spec with a 1 day refetching and redeployment

roadmap

  • versions of a particular document

  • todo - to test revert changes for a deployment

  • agent which connect with github repo

  • collapsable side navigation

  • currently we only support CNAME i.e. docs.yourdomain.com but we should also start supporting things like yourdomain.com/docs which some people might want for SEO purposes

  • AI agents on the docs-renderer

  • Github connect on web and using agent to edit / create documentation something like cursor

  • permissions

  • owner, admin, member to have different roles like edit, view etc. We should change it to owner , editor and viewer that would make it much better

  • docs-renderer layouts

    • help center

    • api docs

  • enhanced code block

  • ux improvements - after domain being added, if the docs has not been published, we should show a banner at the top that please publish the changes as domain has been changed

  • publish a doc for single link sharing

  • find relevant lucid icons automaticlly using llm

  • ability to scroll during drag and drop of individual element inside the editor

  • analytics for my sites / docs which I have created.

    • imagine I am using docs.radhe.space as my api documentation, total number of visitors, time spent etc.

  • On a new version of the UI, we should show a banner somewhere that a new version is available, please refresh

  • better search and replace inside the editor

  • custom code block (language as docufy ) where people can write their own custom code like mintlify so that we can have extreme customization

  • custom docs slugs inside a site

  • when I delete a doc, we should also delete the associated index db with it

  • floating bar for changes in tiptap editor

  • Custom icons (svgs only)

  • multi lingual support (en, es, fr...) [basically, for the same content, we can have different versions][the blob would be different, in the thing we would have a new route maybe something like es/spaceId/slug -- api reference might not support this, we will have to think about that how make that that in the product

  • comments inside editor

  • proper markdown/html copy when I copy something from the editor. SO, that one can paste it

    • it seem to work correctly if I copy from docufy to slack / notion but doesn't work when I paste on openai chat input white_check_mark emoji

  • Publish markdown docs for llm

  • cmd + k with search within the docs AND navigation

  • In the documentation tree, we need to have the ability to multiselect and move to: if I want to move 10 docs to a folder, it is currently a nightmare to do that. Potentially later we will add cmd+k and if something is selected, we will show the option of move to in that as well

  • ui improvements - while moving a doc, its image

null
  • Was this page helpful? with thumbs up / down and analytics in the app

Done
  • fix - ability to click on a folder in the UI of documents

  • fix - doc editor code block

    • remove auto detect language

    • color formatting not working

  • ui improvements - editor immediate improvements

    • collapse block

    • todo blocks

    • table blocks

    • emojis with :

    • slash command

  • fix - slugs

    • org slug

    • site slug

  • fix - copy to clipboard in sites page

  • BUG - Deleting a parent doesn't delete all its children

  • BUG - link sharing with subheading doesn't automatically navigate

  • BUG - order of pages is incorrect after publish

  • BUG - Need to add a constraint that type = api can not be at root level (would create an issue with publish)

  • ui improvements - sonner / toast in the application instead of console.error / console.log

  • fix - mobile nav missing in api-reference page

  • BUG - When i delete a page and I am on the page, we get stuck on Loading...

  • fix - UI for rename

Next
useful references
Last updated 3 months ago

On this page

  1. things to improve
  2. features for mvp
  3. roadmap