I took a look at the way I'd designed it, and came to the sad conclusion that while it could support a range of inputs and outputs, it really was a bit of a hack, and what it wouldn't do was support multiple inputs going to a single output - i.e. Delicious+Pinterest -> Dreamwidth.
At which point I started a redesign, put the whole thing together again from scratch, with proper data design, etc.
And I got married.
And Julie got cancer.
And work's been quite busy.
And frankly, I spent a lot of time being avoidant, and not having the energy/enthusiasm to work on it very much.
However! As of ten minutes ago, I got it working. Not _nice_ - it's still pretty ugly, the forms don't have validation, and there are around twenty different issues that need to be fixed before I let other people use it properly.
But it's working well enough that I've switched over to using it myself. And at midday tomorrow, with any luck, a post will appear on my journal from the new poster, rather than the old one.
All of which I've written _purely_ to jinx myself. Because what are the odds of it all working first time? You can look forward to garbled, upside-down, text appearing on my journal tomorrow, and me scrambling to fix it and get a normal link post out. Probably :->
Original post on Dreamwidth - there are comments there.