Well bear in mind that we need to do market research anyway, so that we would only be starting out with one auction at a time.
So maybe we can start with just manually holding such auctions, to see if there is even enough interest to make it seem worthwhile to even just make a few shell scripts or python scripts or whatever to ease the manual work the auctioneer actually turns out to need to do?
The usual free open source route in other words: you have someone who does a certain thing (in this case holding auctions) and gradually over time he or she hacks up some little scripts or figures out some database fields and records to use in mysqladmin or maybe more likely simply using a command-line mysql client, and once that has evolved into a useful final set of fields and records and scripts and so on they gradually tie all the parts together more until eventually they have a full collection of database tables and scripts showing what to do to those tables when various events happen, at which point anyone who feels there is need or demand for a GUI or a CGI or whatever has all the specifications, in working code, for all the moving parts other than the user-interface, which initially is the command-line or maybe the commandline for the scripts and mysqladmin for editing the database if for some reason mysqladmin is found to be more convenient for some reason than the commandline mysql client...
Another idea, totally a change the shape or orientation of the box kind of idea, is fix the price of the things to be "auctioned" and auction off the points they are priced in.
Like hey this car costs 1000 points, how much am I bid for one point today?
-MarkM-
as far as mysqladmin or whatever it is because kids coming out of college or high school don't know what a command line is. (tongue in cheek)
With all the follow on parts you were talking about. Do you think that anyone would do that once the bounty was paid? Someone would have to be really into the idea to follow up for that amount of time and that amount of data collection and work to create those scripts and stored procedures after the site was launched and the bounty was paid, unless there was a continuing policy of adding bounties for more automation, but then really who is to say that the work done is in line with the best interest of devcoin or the community with the one installation of the auction site (or would people just take the open source auction site and make it their own, adding thier own scripts and what not to work it out if there was no bounty or even in lieu of the bounty.)
There would not be any bounty.
Someone would get a share a round to be our initial market-research auctioneer, run an auction. Or maybe it'd need four one share a round people so we get forty hours out of it at minimum. (Remember the hours per week or month to get shares is a minimum, you don't get more by doing more hours.)
The whole idea of using a bounty for this is massively premature. We don't even know yet for sure what needs doing so are not in any position to start enumerating all the little things needed let alone assign to each step or component a reasonable bounty.
We could for example start with a normal auction site and auction off our "bids". Or plan sell them for a fixed buy now price.
Then, using that same normal type of auction, auction off, for bids, the price a thing will be sold for. Basically the total number of bids placed will, at close of that auction, tell us how many pennies or devcoins or whatever the item is to be sold for.
Then we sell the item for that price to the winner of that auction.
For this initial market research we could even just let the auctioneer auction something themselves and keep the profit themselves.
If the model is in fact profitable presumably the auctioneer's pay can easily come out of what they manage to auction stuff for.
If not maybe they are not a qualified auctioneer and we need one who is better able to excite people into bidding.
-MarkM-