Maybe I did! It seemed the posting account was a "Denise", and recent changelogs on the project website only attributed documentation changes to her.

Her approach to "hacking" sure differs from mine ... from using "#!./tbenv/bin/python" (rather than "/usr/bin/env python" which will find the virtualenv link) making the code unusable for most people out of the box, to requiring a bunch of third-party libs to complete a simple task, to copy-pasta even preserving tutorial docstrings, you have to disconnect from the IRC server to reload the application logic ... it just seemed very sloppy to me.

I'm not hard to find, if you are interested: http://www.linkedin.com/in/jessamynsmith

I didn't know there was a more elegant way to do the shebang. I've fixed that now. It hadn't occurred to me that there was a particular advantage to being able to reload without restarting, but I am considering that now.

Personally, I prefer using libraries to reinventing the wheel. YMMV.

For me, that depends on the size of the library and how much I am using it. I really don't need a library to handle the logic of connecting to a server and sending things as simple as IRC messages. This is not something that I really expect to use outside the context of a bot. However, if there were a good library for creating a flexible bot (which would thus include the small low-level IRC stuff) then I would be interested. But investigating what was out there for creating IRC bots in Python, I really wasn't impressed. Most every bot framework is designed for sitting in one channel and listening for specific commands or command prefixes. Nothing wrong with that, but that only covers one sort of IRC bot.

Er, while trying to research Jessamyn, I think you instead tracked down the leader of the Dreamwidth project, Denise Paolucci. Yes, she updates the Dreamwidth documentation, among other things...

This is all perfectly appropriate for a quick hack. She's not yet hanging her career on the polish displayed in her Python bot.

Have you considered submitting a pull request? I'm sure it would be welcomed.

I wouldn't submit a pull request because I'm not interested in ever using this project. It accomplishes one narrow goal and would need to be rewritten to do much else -- and the twisted requirement is onerous.

I guess it's a special effort for people to write good, simple software? Whether a quick hack or not, I think most of us would come up with something a lot nicer.

Anyway I guess I wasn't supposed to look at the code even though it's open source ...