If I were to make a skeleton using Mr. Bones that was built on Ramaze, what would people like to see in it? It's for my use first, but the community's use a close second if I can generalize it enough. I was thinking along the lines of openid support, having a decent home page and logged-in-user gateway, appropriate nav-bars, haml+sass, jquery+ui, etc....thoughts? What do you find yourself doing at the beginning of every website project that you'd like to not have to repeat every time you start a project?
Well, another day, another quest for game ideas. An online game would be fun to make, but my problem is always I think too damn big. My last idea was a Utopia clone, and I got so far as to create a scripting language that compiled into Ruby, but I just never got around to finishing it. What would be nice is a game that's fun in its simplicity -- simple to play, simple to program, just something you spend half an hour or an hour on a day, tops. There'd be a way to interact with other people easily; there'd be chat rooms and forums, and discernible "hot spots" where a lot of people were chatting or posting messages. But the most important thing is core gameplay: unique, rewarding, and fun. I'll get right on that...
I think it'd be great if there was a hosted repository of web services that you could hit up for free.
If I were to make one...
It'd be hosted and allow a certain number of requests/hour, in the ballpark of 100. It'd be hosted by Merb, be heavily cached, and have certain plugins available like hpricot for parsing HTML and XML. Finally, it'd be able to serve up resutls in HTML, XML, JSON, or YAML.
Each service made would have to be reviewed by me before getting public access. Anyone would be able to contribute to the site, though that might be somewhat complicated. Additionally, heavy users could be charged minor sums in order to get more use out of the site. A small premium could even be charged for the benefit of the person who put together the site (me) and possibly the person who wrote the service. Later on, there could be more computationally intensive services, like converting a .doc or .txt into a .pdf.
So I always have a lot of ideas bouncing around in my head. I think I'm going to try to write about them here before attempting to implement them, partly so I can just walk through the idea and get it down in writing.
On that note, one of my recent ideas was this idea of a "distributed webserver". Okay, great, turns out it's actually been done and/or researched before. But I hadn't quite found the idea I was thinking about. My idea was to have a [email protected] the same way you have [email protected] or [email protected] or BOINC. The goal would be to have a distributed webserver that ran on untrusted clients with full site rendering and database capaibilities, serving pages to users' browsers. Sounds like sort of a cool idea, right? Distributed stuff is always cool, plus you might actually get better end-user performance if you whip some geocoding into the mix (i.e. servers are picked based on proximity to user). To install on your server, you could download what would essentially be VirtualBox plus an image, and then run a bit of config and be good to go.
My inspiration for the idea is that it would be cool to be able to get a VPS for myself and sell the excess cycles/bandwidth/storage to web sites I like.