I partially agree. The Alcugs list is clogged with useless Ages, as most likely 98% of them will never see the light of day. This is almost like squatting :p
But how high can the list go anyway? 65536 I guess? If so it is tempting to think that we can register as much as we want.. And even by registering a lot we are not likely to use all the available numbers.. hmm..
Statistics time!
(
I obviously can't know the numbers for the ULM list, although I guess they are much lower. So I can only talk with the Alcugs list as a reference..)
Right now there are exactly 262 numbers registered on the Alcugs list. I count a bit over 30 numbers that are actively used (ie: released Ages, or test Ages belonging to active forum members). Probably more are 'actively' used, but even if it is around 50 that's less than 20%; and half of those are test Ages. So we have around 90% of uselessly registered numbers. Ouch. But again, thats 'only' 262 numbers out of 65536.. Less than 0.4% registered over two years and a half. If numbers keeps being registered at this speed we will run out of numbers around the year 2630.
(yes, I actually calculated that
) Also over 40% of those numbers were registered just during the first two months following the release of PyPRP (October/December 2005). That was the initial 'gold rush' and logically things have been slowing down noticably after that.
I agree this Alcugs list is a mess, but I think the real problem comes from the fact that we can *see* it, and it looks like a mess; but given the very high number of avaible Prefixes there isn't a real danger of running out of those any soon.. (unless somehow the number of active Writers suddenly goes through the roof; which would actually be quite cool) I know you want to prepare the future, but there is no real danger IMO. (Although I certainly don't mean to support the curent anarchy.)
What if we reserved a set of prefixes (around 10 of them) to be used only for testing Ages?
That's an interesting idea, but I like the possibily of registering numbers for a test Age (for the record I have only 1 and I plan to keep it that way). And I think other people do too. Even if it's just a test Age it's something that can be seen as 'personnal' and people may want to have their own number for it.. (Even though, as you mention, most people will never actually release it.). I stick with a single test Age because that's all I really need, and I really think of this Age as a 'real' Age, something I want to keep and that I may release one day (if only just for fun). So I need a unique ID for it.
(Matter of fact, I'm surprised that not more people have registered a test Age.. how do you guys test your stuff?)
IMO, it would be much better to hold off on reserving a "valid" prefix until you are actually distributing the Age via ULM, and then ULM can generate one for you.
I agree with that. Although I fear it will be difficult to enforce that policy.. If only for one thing: people who work on several Ages at the same time (and I wouldn't be surprised if many of us do) need to register a separate ID for each Age if they want to be able to have them all in Uru. That's probably where your idea of having a fixed public set of prefixes would work (we might need more than 10 just to be safe). This sandbox lot would allow people to test and export freely their Ages before final release. Your idea makes sense..
(In the end however unless we put in some technicall block (through the ULM site) we can't control what people want. If they want to register 28 numbers just because 'why the heck not' there's not much we can do.. We can only update the Wiki(wikis?) and make sure to tell people each time they ask..)
