(1) A busy programmer would like to extend his/her toolkit/website in some particular way, knows it can be done and that others have probably done it or something similar, but he/she has never done anything like it before, and there's not enough time to search/learn/experiment and write from scratch or adapt someone else's code. It's worth a few bucks given to a good cause to plead for a "turn-key solution", and it all works out quickly. Everybody wins.

(2) A PHB or deadbeat student (or a clueless volunteer a DirtPoorCommunityService.org) hears about a web site where people write Perl code for you -- all you have to do is donate a few bucks up front to some obscure "public-domain software foundation", and then tell'em what you want. Do they know the difference between a snippet, a subroutine and a whole app? Are their needs intrinsically bound to the contents of their HTML/Excel/Access/Word/whatever data files? What do they know about modules? Do they even have Perl installed, or are they looking for a precompiled executable? And how well can they articulate all this? Ultimately, nothing works out and everybody loses (but at least YAS gets the donation).

The first group gets a great deal, because they have the right expectations, they know how to communicate what they want, and they know what to do with what they get; often, the successful answers just point to stuff that has already been done -- but it's still well worth the cost of the donation to have someone else do the search and assessment for you. (Of course, writing a spec for someone else to code can be no small feat, especially in an unfamiliar domain.)

The latter group engenders mutual confusion, disappointment, frustration, and/or disgust. You see a lot of dialog cycles just trying to figure out what the seeker is really looking for; and/or just as many cycles trying to clarify how to use a proposed solution (or which of several solutions ought to be used, or why the request is nonsensical or impractical), and these cases all end (at varying temperatures) with "I give up -- this is really not helping".

Naturally most cases would fall somewhere between the extremes, but no matter what the "median" performance really is, the failures will tend to consume (i.e. waste) the larger portion of resources, or will certainly be perceived as such. Frankly, that scares me a bit. (And God help us when folks start to add voting/XP issues into the fray...)

It might work somewhere else; if it were here, I would probably wish it were somewhere else.

I'd vote that we continue to encourage the first group to use SoPW, and continue to ask generally for donations as is done now -- maybe make the donation results a little more visible: more people should be shown (in some not-too-obnoxious manner) that donations do get made, and they might feel more inclined to mimic that behavior (or be shown that donations aren't being made, and feel a stronger sense of obligation to do so).