This is mostly for Dave but I welcome eye-rolling discussion from all.
I’ve been doing a thing for a little while now, but only Dave has asked about it. And so, in our cozy, little internet for two, I’m just going to talk to him about it now.
Dave, hello, old friend. Thank you for paying attention to my RSS feed and this long, slow ongoing conversation about blogging that makes other people’s eyes glaze over sooooo quickly.
You asked:
@lloyddavis.bsky.social please can you show me how you do your micro.blog > WordPress daily thing. I think that’s probably the best way for me to achieve this: da.vebrig.gs/2024/10/30/%…— Dave Briggs (@da.vebrig.gs) December 10, 2024 at 5:52 PM
Well. So. Yeah. Hang on a sec.
In the summer, I was having cognitive capacity issues, keeping track of who I was talking to where about what. So I looked at the cross-posting capabilities of micro.blog – I was attracted to it first by seeing it being used by my early blog heroes Adam Tinworth and Robert Brook and I’m a big fan of how Manton Reece eats his own dogfood. Each of these people do their own thing with it and none of them are to blame for the stupid things I might choose to do when I pick up the same tools. I just mention them because I don’t want to look like I invented all this or that I think I’m cleverer than anyone else.
Where was I? Yeah so one of the cool things Manton has kept on top of is cross-posting to these different platforms. It was the most straightforward way of writing a post once and having software post it in several places (btw they also go to my ancient tumblr – and could go to medium, linkedin and nostr if I wanted to go bananas). And recently, I’ve been able to see my BlueSky and Mastodon replies there too, but that’s another complication and we’ve got enough complications here already, I think. Focus, Lloyd, focus!
Cross-posting deals with the copy/paste problem, but I was still feeling a gap. Ever since I got on Twitter, I wanted to have a place where I could find all my stuff instead of it being in different places. I couldn’t get my act together to go full POSSE, so maybe it’s cousin – PESOS might work. PESOS means publishing elsewhere, syndicating to own site. It feels inferior, and a bit wrong, but it kind of works for now.
I feel the need to just step back and acknowledge that I don’t know what I’m doing here. I’m just bumbling along having a life and now and then saying stuff on the internet. I don’t have Information Strategy meetings with myself to work out the ideal infrastructure and architectural approaches. I’m just the same bumbling blog-hacker, driven by novelty, that I’ve always been.
So I decided that what I’d do now, since I’d got all the chunks being made on micro.blog, it might be nice to make a daily digest post on here (perfectpath) of all the chunks I’d made that day. (turns out, it’s hard to get *all* the chunks, I’m human, sometimes I forget and just post something on one platform directly – also replies – we’ll come back to that).
I was also interested in how to use ChatGPT for helping me remember how to write code. So I asked it to write a script using the WordPress XML-RPC API (because that’s all I could remember there was) to take my RSS feed from micro.blog (because that’s all I could remember there was) and automatically post it here at about midnight every day. After some iterations I got to this version of the script. I then made a cron job on my home desktop computer to run at 11:55 each night, redirecting the output to a log file.
55 23 * * * /usr/bin/python3 "/Users/lloyddavis/blog digest/micro_digest.py" >> ~/microblog.log 2>&1
It’s all a little bit shit, but it’s my little bit of shit. Before we start with the “why don’t you just…” here’s where I can see it needs improvement:
- It only handles my original posts, no replies (but see json stuff below)
- I see that WP has lots of APIs – I’m tempted to believe that my first choice of XML-RPC is not the best (given it was basically down to ignorance of the others)
- I also realise that micro.blog will give me feeds in json format which might well be more elegant to handle (and less error-prone?) than trying to parse the RSS into a custom data structure and then remember what I called the variables. There’s a whole API for doing more complicated stuff (including replies!)
- The formatting is horribly basic (but then so am I).
- If I want it to run more reliably, perhaps I should put it on someone else’s computer rather than my mac mini which is prone to attack via the feline keyboard marauder and my own stupidity.







