I think, due to an ill-advised attempt (or perhaps more accurately ill-timed - late at night right before going to bed not really being ideal) to update my reverse proxy, I might have somehow caused it to block posts from all instances. Either that, or everyone's been really quiet for the last day. I'm trying to get it fixed now.

The intent was to switch from blocking all traffic from snac instances to blocking only their POST requests. I'm not sure yet what, if anything, I messed up.

Snac has a default setting to send all posts to all known inboxes, to improve reach for small instances; what I read on my #microblogpub instance is essentially the whole instance inbox. I do not find the result acceptable. I'm going to have to get more creative if I ever want to follow someone on a snac instance, but for now I thought I'd try to make it so that snac users could maybe make their instances retrieve my posts, even if they couldn't talk to me without someone else boosting them over the wall. Unfortunately, something seems to have gone awry...

I got quiet for a while, huh? Here's the latest...

I noticed #microblogpub rendered the posts I was replying to with u-comment where I thought it should be u-in-reply-to, so I changed it. Then I realized the reason it was like that to begin with is because the same block is used for rendering replies to me as comments. So I thought about that for a while, and concluded that since the most likely consumer of that markup is something checking to see what I'd said because of a webmention I sent, and therefore it is way more likely that said consumer will see things I am replying to than things that are replying to me, it made sense to leave my change in place until I could do better.

Then I went to commit the change (since I'd made it in production) and while I was at it, I figured out how to make it depend on whether it was part of the thread before or after the post the permalink is for, so I did that instead. So that's what's new around here.