Item — HN Reader

> That is like saying ice cream makers should focus on the 'core' tastes, claiming that that is banana and vanilla because you like those.

From my point of view, I'm suggesting the ice cream makers should stop trying to make paninis to go with the ice cream and instead focus on making sure the ice cream cones don't have holes in the bottom.

I nearly lost all my files because server-side encryption went wonky. That's core functionality that should, imho, be fixed before going off into ActivityPub.

More generally, I have trouble being enthusiastic about "We added activitypub" focus because of what it seemed to do to MediaGoblin. That project seemed to be chugging along and making steady progress until it seemed they focused a lot of their energy into ActivityPub. Since then, MediaGoblin hasn't really had any of its functionality updated, as far as I can tell. They certainly don't seem to have released a new version in over 2 years.

1 comments

All for avoiding Panini's, but we're a Content Collaboration Platform - and Nextcloud Social fits certainly well in that.

WRT server-side encryption, it gets work all the time, though it is mostly an enterprise feature that gets love when enterprises pay for that. Very few volunteers work on it, if any. Help is of course welcome, in the end - either somebody pays for it, or somebody puts in free time.

This is true for every open source project, of course.

WRT MediaGoblin, no worries, it isn't like Nextcloud Social has all our attention. I guestimate it is at most 5% of our engineering time, if that. That is enough to make it work and improve it release over release, and we'll put in more if there is customer interest or lots of community contributions, but we always start such new things small. Talk started as a night-long-hacking-to-prove-it-could-be-done and now has 3-4 engineers on it full-time, because customers want to pay for it. Which also means it doesn't take away from other things - if we didn't do it, we simply couldn't afford these engineers in the first place. It isn't like they would work on server-side encryption ;-)

Okay, that makes sense and makes me feel better about it. Thanks for responding and clarifying.