A discussion about adding Condorcet Voting to Discourse Polls gained new interest, with suggestions to adopt Debian’s tested version of the voting system. merefield noted they had already contributed Ranked Choice voting as an alternative read more
david provided detailed explanations about new privacy features for admin access to topics and PMs, including information about a new site setting to suppress content visibility.
Multiple users encountered Docker rate limit errors during launcher rebuilds, with md-misko discovering it was related to IPv6 on DigitalOcean droplets.
A detailed discussion about the New User of the Month algorithm emerged, with RGJ suggesting mathematical improvements to the scoring system using SQRT and LN functions to better balance post quantity versus quality. Ed_S proposed considering distinct liking accounts to prevent voting rings.
wenqin reported color changes after a forum update, particularly noting differences in like button colors between devices. elmuerte identified this as stemming from recent changes to the discourse-simple-theme.
In the Discourse Chatbot thread, Sidi_Codet inquired about supporting additional AI models beyond OpenAI, with merefield explaining the focus on OpenAI helps manage complexity.
Discussion around Enter to newline functionality in chat with multiple users debating the standard behavior. joffreyjaffeux confirmed the team is working on implementing configurable enter behavior for chat submissions.
In the FKB Pro - Social theme discussion, users discussed potential modifications to the comment icon behavior following recent Glimmer implementation changes.
tobiaseigen participated in the trust level discussion, suggesting solutions and seeking clarification about use cases for profile visibility restrictions.
lindsey provided UX input on the chat functionality discussion, emphasizing the desire to avoid additional settings while maintaining consistent behavior across devices.
Thanks for reading, and I’ll see you again tomorrow!
martinmodrak reported an issue with hCaptcha preventing user signups when the plugin was enabled but not configured. tobiaseigen confirmed this isn’t enabled by default and submitted a PR to improve guidance in the settings.
A discussion about restricting profile visibility for low trust level users led to sam proposing conditional disabling of the feature for certain forum configurations.
anurag encountered difficulties with AI-powered spam detection configuration, leading to sam explaining the experimental feature rollout process and how to enable it through the “What’s New” feed.
Ninad sought help with rebuilding their forum on an outdated Ubuntu 16.04 server, with recommendations to upgrade or transfer to a new server to maintain functionality.
Saif introduced Smart dates, a new AI-powered feature for Helper that converts text containing time and date information into Discourse-friendly timezone-supported dates.
A discussion was started about hashing usernames in profile URLs, though thoka questioned the necessity given usernames are already customizable and visible throughout Discourse.
Don reported an issue with the glimmer post menu being selectable by long touch, affecting discourse reactions. saquetim later shared a PR adding user-select: none to resolve the problem.
A detailed guide was shared about Smart dates functionality in the Helper feature of Discourse AI plugin, explaining how to enable and configure timezone-friendly date conversion.
Saif initiated a discussion about using Discord alongside Discourse, seeking insights on community experiences and announcing an AI-powered Q&A Discord bot experiment.
one1 raised an interesting security concern about advertising user’s security practices through badges, suggesting it could potentially help hackers identify vulnerable accounts. sam agreed this could be dangerous and suggested considering private badges instead.
A locale issue was reported regarding untranslated graph titles in the community health dashboard. sam flagged this with the team and invited community contribution through a PR.
pierrox discovered an emoji rendering issue where inline emojis were being converted to text, which Don pointed out could be resolved through a site setting.
An insightful discussion about category usage and organization where noahl shared how they simplified their forum structure from 16 categories to just 3, resulting in improved engagement. Bas highlighted Wix’s approach to radical simplification.
A discussion about hiding user profiles revealed some gaps in the current privacy features, with bloomexperiment demonstrating that users remain visible in certain areas.
A critical login/logout bug was discovered and resolved where Cloudflare cache settings were conflicting with server management panel caching, causing users to be incorrectly logged in/out. Full discussion
User experiencing issues with installing themes online in the website backend, particularly with GitHub access from China
Request to add functionality to turn digest emails back on for suppressed users, particularly useful for communities with occasional important announcements
A forum owner sought help with breached forum rules where users were sending unauthorized links through PMs. supermathie clarified that Discourse staff cannot intervene in disputes between forums.
Discussion around AI bot capabilities where users expressed concerns about the AI’s ability to process detailed content in summaries and newsletters.
char reported issues with embeds on mobile clients, experiencing random problems with images, webms and YouTube links in their Discourse 3.4.0.beta3-dev installation.
Fehler in der Deutschen Übersetzung thread, with tobiaseigen and Moin collaborating on improvements