Which is why I am having issues with turning AutoPostBot into a finished general service for everyone. Ok, so the survey shows that AuroPostBot will not tear apart the fabric of LJ blogging, but will just be a mixed blessing, trending to the manageable side, for most people.4 So I am looking at the code for this little service and I know exactly what I need to do to make it a robust general service that should be able to easily scale to at least 1000 users, and to do it semi-right. And I think, ugh, coding. Write, debug, write, debug, write, debug, whatever.
I'll also have hosting issues. I'd find it more interesting if the project would be a way to explore Google or Amazon's new platforms, but AutoPostBot is just a continuous (blocking) loop waiting for input either by polling an RSS stream or waiting for an IM from Twitter to come in. So far it just polls a specific Twitter RSS stream because Twitter can't get IMs to work reliably from my POV, even though having Twitter instant-message AutoPostBot would be far preferable to me over having it check a website every minute to either find out what somebody just Tweeted or to get a private message from a new user to ask to be followed. Still, it ain't a webservice, so Google Application hosting is not interesting, and Amazon will just be expensive and overblown since AutoPostBot needs to always run. Plus there is the very thorny issue that AutoPostBot would have to store people's LJ passwords, and that squicks me. Responsability, ouch!
I could easily just have it running on my print server as it does now. But there's nothing in it for me but kudos, and whiny scorn from people who want more features. You can't monetize an ancillary service to a free service really, especially one as ephemeral as Twitter. So, part of me wants to do this just to see an idea through to the end because I hate dangling projects, and to have my name out there, and part of me has just no motivation to do this. I finally have the time, tools, and experience, but the geek fire is out, people.
Well, except my desire to keep
1Oh ok, alright, if a gig needs to call me Information Architect I'll do that too, but having gotten my minor in UI design before 1995 means I have an instinctual dislike for the term because, to my thinking, either an IA is a UI Designer who has been asked to wear one technical hat too many, which is problematic, or an IA is not a UI designer so then what the hell is that person doing designing user interfaces? I know, I know, it is now an established term so it will be used and people have to call themselves that to short-hand what they do, but I do not have to like it on an aesthetic level any more than I have to like the word 'hottie'. I loathe 'hottie' as a word.
2Add the word 'Mobile' in front of any as required.
3Pervasive right-click context-dependent menus in Swing 1.4, mofos, and the programmers working with me could add their actions for their custom views without fuss by implementing just one interface, and the system would pick the actions up automatically and show them in the right place. And that was just one of the things I made JAVA do for me.
4Around 250 followers but only around 40 people found the time to check a box and click a button? You passed up a chance here to shape the web, people. Or I am off more default filters than I want to consider.