First off, we are still looking for feedback on the Photon & Alexandrite alternative Lemmy UI trials. Please comment in the linked posts.
Technical server updates
Server migration was completed last month and it is generally working well on this new 8 threads, 16GB RAM machine. There is a problem left with occasional stalling of the database writes for which the source has been identified, but it will likely require some hardware replacement and thus a small down-time later this month. But having everything on a dedicated machine allows me in theory to give additional server admins access, but we first need to identify people willing and able to help out with that (more on that below)
Due to active ongoing upstream development the CSAM scanning tool was not yet enabled, but preparation work is complete and it should be easy to start soon.
XMPP chat integration
Still a work in progress, but the auth-script against the Lemmy database is functional and deployed here, so you can already use your regular SLRPNK Lemmy account (username@slrpnk.net) as an XMPP account since a few days. The Movim web-client is also up and running on https://movim.slrpnk.net, but it needs some further testing and a few known issues make it not ideal to use in combination with other XMPP clients until the next Ejabberd server release (Edit: updated, lets see if that fixes this problem). In case you want to use Movim on your Android or iOS device, use the install to homescreen option in your mobile browser’s menu (Movim is a PWA with Webpush). End-to-end encryption is disabled by default, but can be enabled in the user settings and works reasonably well. However this feature is still somewhat experimental in Movim and you are probably better off using other XMPP clients for this. Audio and video calls are currently not fully functional yet, and will require a bit more work on our server setup.
Community wiki
Based on our previous discussion, work has started on a community wiki linked to our Lemmy communities. It is already available on https://wiki.slrpnk.net but the account integration is not yet functional. The plan is to create an individual namespace for each Lemmy community to have their own community specific wiki that community moderators can edit. The wiki will likely also be automatically synced with a git repository.
Fedi admin guild
One of the goals of the F-hub.org organisation hosting SLRPNK is to enable better community management of FOSS projects. To this end a new service was recently added that allows creating coordination forums. As a trial, I created one such group with the goal to invite active Fediverse admins and moderators. The general idea is to have a better managed space for exchange and mutual-aid in server administration as well as to make the decision making process for topics like instance defederation more transparent. If you are interested in this please DM me here and I will send you an invite email. I hope that the moderators of the most active communities here might be also interested in this, as it is not only meant for sysadmins.
Additional site mods
I would also like to ask if any of the already a bit longer active community moderators here would be interested in becoming site moderators for SLRPNK. Additional responsibility would be to help with account approvals (not many lately) and reacting to user reports for communities that have not very active moderation. Ideally it would be also nice to have some sort of vote by the other currently active community mods on who they would trust as a site mod, as it comes with quite a lot of power in regards to moderation decisions in all communities.
Open discussion
If you have any other topics related to our community here on slrpnk.net please comment below. Feel free to also highlight some new communities here on SLRPNK you have recently created.
As mentioned above, we will have a scheduled maintenance this weekend due to the identified hardware issue with the SSD drive controller. If you have wondered why the page stalls sometimes since 2-3 weeks, this is why.
I have an idea how to fix this without throwing away good hardware by moving this SSD that is part of the main storage pool of the database to another drive controller and use the stalling one for less high performance purposes which should not overload it like that.
I expect the down-time to be around 1-2 hours as I plan to do some other off-line maintenance as well (full database vaccuum etc.). Lets see and hope nothing unexpected happens 😅
Maintenance was completed and everything seems to be back online. Lets see if the database stalling has been solved by this, but it looks good so far.
Edit: I think it got a lot better, but it still rarely happens. My guess is that I can further improve this by switching the Pict-rs database to Postgres with the next version update.