[pmwiki-users] recipe usage statistics
pbowers at pobox.com
Sat Aug 30 07:39:34 CDT 2008
Someone was recently concerned about the "alpha" status on a recipe I
had posted. In looking at some other recipes (both those I've
authored myself and those that others have authored) I see that there
are an awful lot of recipes with a status that is a lot more
conservative than the recipe probably deserves. I.e., this might be a
recipe that is regularly used by dozens of sites without any problems
but it still has a "beta" or even "alpha" designation.
One possible reason for this is that developers have no way of knowing
how widespread the use of the recipe is. If only 2 sites are using it
then the "alpha" designation is appropriate, but if 28 sites use it
extensively on a daily basis and have been doing so for 6 months then
probably it can be justifiably upgraded to "stable". And how would
anyone know whether it is on the 2-site scale or on the 28-site scale?
My question is, is there any reasonable way to collect this
information? Probably asking someone to "register" every time they
try out a new recipe is both unreasonable (makes it too hard) and
ineffective (i.e., someone tests it and it doesn't work but they are
still registered so the recipe looks better even though it should look
But would it be possible to do this in an alternate way? Perhaps in
the script that checks for up-to-date recipes could it anonymously
collect the information of which recipes are installed? Or some other
Not only would this help developers, but administrators would also
have a way of seeing which recipes were most often used (and thus most
likely to be helpful, although the one doesn't necessarily imply the
other). I know as a pmwiki site administrator I would be very
interested in a list of "top 10 recipes" that other, more-experienced
administrators are using...
Any thoughts on either the value of the idea or on how it could be
implemented? Maybe it's already done and I don't know about it...?
More information about the pmwiki-users