Now I have to figure out what to work on.
I've got a list of things I want to do to sightseer to improve it. Add more map services (live, yahoo, etc) as well as getting street view working. Adding area specialists, so people more knowledgable than I about local geography can approve submissions from the areas they know about. Mobile support. Various things. I've got a list somewhere. BUT, having rewritten it, I'm kinda wanting to work on something else for a while. Yeah, making it better might make it more profitable, but the money isn't really the driving force.
There's lunch-o-clock, which I thought up just over a year ago, and which has been sitting around since then. Done right, I think it could be a pretty damned useful site. Design it with google ads in mind, so it too can support itself and pay its own way. Design it with mobile browsers in mind, as well as ensuring the hooks are in place to tie it into twitter and jott and whatever else seems like a good idea. Doing it in ruby/rails will make it relatively easy - sightseer was a pain even in RoR, because I was recreating an existing system as exactly as I could, this would be all new and thus MUCH easier. I could probably have the basic structures in place in under a week, REALLY basic functionality in a fortnight, and bring it to beta-ready within a month. Probably. Maybe.
Or, there's the con management system that I've been drawing up plans for since 1998. I need to go charge up Dactyl.. the old Newton's got my original notes on the project, which I started designing at my very first real con ever, and continued to add features to until at least as late as OSCon 2000. With the advent of SteamCon, it's come back to my forebrain that this is something I really REALLY wanted to do. On the other hand, I'm not sure I could have a system that I'd be comfortable putting in a real production setting with real people (and real money) depending on it in time to be useful. It's also complicated by the fact that it really needs some expertise beyond my own. I'm almost exclusively a web and cli programmer, and only a saintly level of charity would call me merely 'rusty' at local GUIs. And particularly for the registration system, that's really what you want.* So, I dunno. I need to go find my notes, and post about this specifically and see what people think.
There's also taskomat, which still needs to have Hobo ripped out of its innards. It's just a question of cleaning out the gunk and replacing what's lost as a result. I think it'd improve performance by a couple orders of magnitude. And there are a few improvements I've thought up along the way. It's probably got a fortnight of serious work needed.
* Heh, I just had the most wonderful and horrible idea. It's a steampunk con. Screw GUIs. Self-checkin using steampunked and/or brazil-like terminals. The software itself would just be text-mode, green-on-black (maybe amber?) and you can look up your prereg, or reg new, there'd be a magstripe reader so you can pay by credit card, etc. And you can use the same system for the volunteer-managed registration. There's no reason to put a GUI on top of a system that just fills out a form.