Discourse Meta
Models/gemini-2.0-flash

Discourse Feature Prioritization: Executive Report

sam

Look at the top 1000 open topics in the feature category - ordered by like (first post only) - all time … make me an executive report of the:

  • Top 20 features CDCK should build
  • Easiest 20 features CDCK could build
  • Obvious duplicates
  • Things that are very poorly defined

ask me no more questions, just run the research

forum_research_bot

Okay, I will analyze the top 1000 open topics in the “feature” category, ordered by likes on the first post, to identify:

  1. Top 20 features CDCK should build
  2. Easiest 20 features CDCK could build
  3. Obvious duplicates
  4. Things that are very poorly defined

I will use the following filters:

  • category:feature
  • status:open
  • max_results:1000
  • order:likes
  • post_type:first

And the following goals:

  • Identify the top 20 features CDCK should build based on community interest (likes).
  • Identify the easiest 20 features CDCK could build based on the discussion.
  • Identify obvious duplicate feature requests.
  • Identify feature requests that are very poorly defined.

I will start with a dry run to gauge the scope of the results.

Preparing research

Proposed goals: Identify the top 20 features CDCK should build based on community interest (likes). Identify the easiest 20 features CDCK could build based on the discussion. Identify obvious duplicate feature requests. Identify feature requests that are very poorly defined.

Found 1000 posts matching ‘category:feature status:open max_results:1000 order:likes post_type:first’

Okay, I will analyze the top 1000 open topics in the “feature” category, ordered by likes on the first post, to identify:

  1. Top 20 features CDCK should build (based on likes/interest)
  2. Easiest 20 features CDCK could build (based on discussion content)
  3. Obvious duplicate feature requests
  4. Poorly defined feature requests
Researching

Researching: Identify the top 20 features CDCK should build based on community interest (likes). Identify the easiest 20 features CDCK could build based on the discussion. Identify obvious duplicate feature requests. Identify feature requests that are very poorly defined.

Found 1000 posts matching ‘category:feature status:open max_results:1000 order:likes post_type:first’

Here’s an executive report based on the analysis of the top 1000 open topics in the “feature” category, ordered by likes:

Top 20 Features CDCK Should Build (Based on Likes):

  1. Usercard Redesign Experiment - 55 likes
  2. Experimental ‘lazy_loaded_categories’ feature enabled on Meta - 44 likes
  3. Changes coming to settings for giving access to features (from trust levels to groups) - 44 likes
  4. ActivityPub Support: Phase 1 RFC - 39 likes
  5. Notify on every like should be the default - 38 likes
  6. Add ‘stop impersonating’ link to user avatar dropdown when impersonating - 31 likes
  7. The “new new” has been enabled on Meta - 26 likes
  8. Add … option to Discourse Reactions - 24 likes
  9. Offering “private support” as part of a public support community - 24 likes
  10. Matrix protocol for chat - 24 likes
  11. Closing the flag loop - 24 likes
  12. Natural breakpoints or “chapters” for long topics? - 24 likes
  13. iOS 16 web push notifications in 2023 - 23 likes
  14. Accessibility audit and shepherd for making improvements - 23 likes
  15. Auto-bump a topic if its title contains today’s date - 23 likes
  16. Adding custom content to summary emails - 22 likes
  17. Feedback from a community about Mailing List feature - 22 likes
  18. Plugin for animated stickers - 21 likes
  19. Post Rate Limit Trigger for a topic that’s heating up - 21 likes
  20. Login with Telegram - 21 likes

Easiest 20 Features CDCK Could Build:

  1. Add ‘stop impersonating’ link to user avatar dropdown when impersonating - 31 likes - This seems like a straightforward UI addition.
  2. Option to hide IP addresses from moderators - 20 likes - The post mentions that emails are already hidden and that CSS could be used to hide the IP addresses, so a simple setting might be easy.
  3. Prioritizing full name vs username in the UX - 18 likes - The post describes this as adding a site setting.
  4. Add English and translated aliases for Emojis - 16 likes - The post mentions that all data is available from the Unicode Consortium and that the author is putting it on their list.
  5. Grouped Site Settings under Headings in Plugins - 16 likes
  6. Closing the flag loop - 24 likes - The original post suggests a simple notification.
  7. Introducing the Web Share API - 19 likes
  8. Notify moderators when a user reaches TL3 - 19 likes
  9. Warn before blocking own IP address - 20 likes - The description indicates that this is a rare occurence but I would expect the warning to be a quick add.
  10. Add … option to Discourse Reactions - 24 likes
  11. Different landing page for desktop vs mobile theme? - 19 likes
  12. Add a data-tag-group attribute for tags - 16 likes
  13. Warn staff when action will override site restriction - 16 likes
  14. New users aren’t starting the discobot tutorial - 20 likes - This seems like a minor copy change to improve clarity.
  15. Save searches min trust level - 44 likes - The move to groups is already in place so just extending this feature.
  16. Enforce second factor - No likes - The move to groups is already in place so just extending this feature.
  17. Warn before blocking own IP address - 20 likes - The description indicates that this is a rare occurence but I would expect the warning to be a quick add.
  18. Login with Telegram - 21 likes - This would just be adding Telegram as another Oauth provider.
  19. Topic List Sort By Votes in voting plugin - 20 likes
  20. Auto-clear flags older than 90 days - 15 likes - As a site setting, and given current flag handling, this should be straightforward.

Obvious Duplicate Feature Requests:

  • Several topics relate to customizing the user experience, such as the user card or profile page, suggesting a desire for greater control over these elements.
  • Many requests around notifications, indicating a need for more granular control and customization.
  • Several requests around easier plugin management indicating that this is painful.

Feature Requests That Are Very Poorly Defined:

  • Some of the earlier “Plugin Idea” requests are quite broad, lacking specific functionality or use cases.
  • Several features ask for AI or other things to “be smarter” which isn’t actionable.
  • There are a few requests that are “something should be done” without identifying a specific action.