sam introduced an exciting new feature called Web Artifacts in Discourse AI, allowing users to create and share interactive, dynamic components directly in posts. The feature supports HTML/CSS/JavaScript for visualizations, quizzes, and lightweight web apps.
#Site-Management
A comprehensive guide on Discourse AI Web Artifacts was published, detailing configuration options and security modes for administrators.
Don shared a new F NAV theme component concept based on the Discourse Tab Bar, adding enhanced functionality for hamburger menu, notifications, and search features.
Jacob_Peebles initiated a discussion about improving new user engagement on Discourse forums, highlighting the need for better calls-to-action and signup visibility for first-time visitors.
A solution was found for LetsEncrypt certificate renewal issues, with the recommendation to completely delete and regenerate API keys when encountering persistent renewal failures.
A request was made for admin/moderator “regular mode” functionality, with mattdm highlighting how current implementation affects day-to-day user experience.
Mihir.R requested adding Like/Dislike buttons and inline replies to the Redditish theme, seeking functionality similar to Reddit’s interface. merefield suggested potential solutions using existing components like Topic List Previews.
Jagster encountered SSL-related challenges while implementing OAuth2 between Wordpress and Discourse, ultimately resolving it through a system upgrade.
#Site Feedback
Users discussed the temporary absence and return of the AI helper feature.
nat announced the release of Discourse Livestream, a new plugin that enables creating livestream topics with integrated chat functionality and calendar events. isaac provided detailed explanations about chat access and tag requirements.
NateDhaliwal provided feedback on the new topic controls design on Meta, noting concerns about the timeline size and dropdown menu implementation. chapoi acknowledged these points are being addressed in the ongoing redesign work.
An admin user reported a confusing login dialog after being deactivated. tobiaseigen identified this was due to the invalidate inactive admin email after days setting and suggested improving the messaging.
Discussion continued on the reply button redirect feature, where pfaffman suggested implementing a more obvious reply-as-linked-topic button through a theme component.
Discussion around new OfCom Rules and their implications for Discourse communities, with HAWK confirming that both Discourse and forum owners have responsibilities for compliance.
@Bathinda_Helper reported issues with Google indexing their Discourse site, with the community helping troubleshoot robots.txt configuration and crawler settings.
MihirR sparked an interesting discussion about moderator application processes, where ondrej and Ed_S emphasized the importance of organic moderator selection based on trust and contribution history.
Don shared experiences with new AI models in Discourse, particularly focusing on grok-2-1212 for topic summaries and Gemini Flash 2 for AI bot functionality. sam suggested using XML tools and writing system messages in the native language for better results.
supermathie provided detailed insights about bulk email sender changes and the implementation of one-click unsubscribe headers following RFC8058 specifications.
Saif shared details about Discourse 3.4.0.beta3, introducing new features including check for updates on What’s New page, filtering by user in review queue, and threading in Chat DMs.
Saif released Discourse 3.3.3, a security and maintenance update addressing multiple CVE vulnerabilities.
david detailed upcoming changes to topic-list implementation, moving from raw handlebars to Glimmer components, with important implications for theme and plugin developers.
sam announced a new AI-powered spam detection module in Discourse, featuring improved control and customization options compared to Akismet. The system is designed to efficiently identify and flag spam posts with minimal configuration.
A discussion about the size of reply buttons being influenced by text size preferences, where CocoQuark identified an issue with the notifications button wrapper causing sizing inconsistencies.
CocoQuark raised concerns about the “Jump to end” button label being potentially confusing when used with table of contents, suggesting alternative labels like “Jump to replies” or “Read replies”.
