There are several motivations for participating in Spooky Distance, and one of them is the desire to be part of our own company. (I won't say 'be your own boss', because no matter what your business is, the customer is boss.) We've occasionally bandied alternative business ideas about, so here's a place to record them.
;On-line Game AI Services: This would be an interesting sub-field of on-line gaming on which to focus. Example: We would contract with on-line gaming companies to provide a certain number of AI player/hours per week. They provide us with protocol specs (and perhaps libraries) that let us interface with their game the same way that their clients do. We develop AI software and run it on our own servers. We closely monitor the software (easy since we run it ourselves) and improve it over time.
Istvan It has come to my attention that game developers frequently "farm out" segments of their coding to other development shops as "piece work". I had no idea. Constitution of the coding team as a formal company able to take on contract programming would be a requirement.
;Database product ideas: Write a better database API, an alternative to SQL, which is an absolutely crappy relational expression language. (A .NET library, perhaps?) Initially, the API interfaces with various existing database systems, usually by generating SQL. Eventually, we provide a better DBMS as a 'native' engine for the API. This would be an expansion of the work I've been doing at Creative. There is room for better products in this genre. It would be interesting to see if we could buy the source code I've developed from Creative; they shouldn't feel threatened by such a proposal, since they're not in the software tool business. See RRDBAPI for more detail.
JDH Create a simple n-tier database service like [Quickbase] that clients use on a pay-per-use model. J2EE provides a great deal of horsepower for managing n-tier database based systems - that horsepower is very easily harnessed. JBoss makes a great J2EE app server at a great price (free). Rather than explain the idea in too much detail - check out our main competitor in this field: Intuit's [Quickbase]. Intuit charge $295 (and up) a month for this service! This sounds to me like money for old rope. By building a solution on top of J2EE (mainly using the container managed persistance (CMP) and JSP aspects of that spec) and using JBoss as our middleware and something like Postgres as the database I think we could offer something for far less and I think we could offer a richer set of functionality. Furthermore I expect that rather than selling the hosting, administration and the software all as one package we could seperate it out and offer a much more flexible model than Intuit's (in fact we may want to start by not offering hosting at all - or it might make more business sense to look for a partnership with an existing hosting service). I think this group has all the skills required to get this going fairly easily and seeing at how much the press seems to like Quickbase it might be an easy money spinner...
JDH Ack! I was looking at the page for "corporate pricing" (read "rape and pillage") - the small business option is far, far cheaper. It doesn't mean we couldn't do it better though - but it might not be the easy money spinner I was dreaming of - oh well ...
; Investing Tool: DWM Pop fly to left field: For years I've been mismanaging my investments. (Don't tell my wife that I admitted that.) I've got a handful of good 'how-to' books, most of which I've mostly read, but lack sufficient interest and discipline to put the knowledge to use. Seems to me it should be easier to gather and analyze information leading to stock and mutal fund trading recommendations. Then again, maybe the products that are already available (which tend to be pricy) already do a good job.
JDH My brother-in-law is now into day-trading (actually he calls it something else but it's definitely short term trading). He has some very slick looking software that has a wealth of features and he sits in front of 3 19-inch monitors working out how to loose the most money (or something like that). But the main thing that this software is good for is real-time analysis of info - and so I believe it's sold on a monthly subscription basis because you need to have a stock feed (which necessitates a monthly agreement). If we were to go into that market place I think we'd find it very competetive (because there's a high level of BS around the whole day trading thing). However if we went into the longer term investing market (which is what I imagine that you were referring to) then we might have some more luck - our competitors there would more likely be people like Motley Fool and I also think some of the online brokers provide some analysis tools (I quote these examples because I'd imagine we'd be trying to prodcuce something for the layman rather than the hardcore investor (who probably has one of the more expensive software packages you mentioned)). My impression is that these tools aren't as aggressively marketed as the "real-time" ones are. I think we could do something relatively easily and distribute it on a shareware basis or try and persuade an online brokerage to buy it. But I don't know if we could generate much development enthusiasm around such a product.
FWIW: I give this: Achievability 'A', Marketability 'B', Fame&Fortune 'C', Interest* 'C-'
- I suppose at a push we could use this system as a way to generate stock market info for our game and that would increase the interest level. But I think that's a stretch.
DWM Yeah, I didn't expect a lot of interest in this. I'm just trying to find ways to get myself interested in taking care of this chore. :)
Istvan No real argument with John's closing points, and I'll add that the reason I'd be concerned about building such a product would be potential liability from frivolous lawsuits from people who use the product and lose money anyway. We'd need bulletproof legal defenses. I actually think such a product might be riskier than a game.