On Apr 14, 2004, at 16:12, Erik Moeller wrote: I'd really appreciate it if these straw polls were *labeled* as such, instead of being given the appearance of official binding referenda. -- brion vibber (brion @ pobox.com)

On Apr 14, 2004, at 19:02, Erik Moeller wrote: Please reconcile the above with "The point of these polls is to get a reasonably representative *sample* of Wikimedia opinions." -- brion vibber (brion @ pobox.com)

On Apr 14, 2004, at 22:24, Arvind Narayanan wrote: The difference is authority of the outcome, and that difference is relevant to how the process is presented and handled. It's the difference between an informative survey, which gives developers some data points to work into our decisions on how to implement the software, and a formal election which determines what the developers must do (unless there are "profound reasons"). Erik's votes are shams. <strawman>Can I just unilaterally *call a vote* and declare the winner to be president of [insert country here]? You don't think the winner should be president? Well, you'd better have a *profound reason* why he's not good enough!</strawman> A survey, an informative straw poll to gauge preference and relative interest is fine, but *treat it as such*. Don't call it a *vote* with a *deadline* yadda yadda. For goodness' sakes, Erik sent the announcement of this little poll to the Wikimedia Foundation list -- the list that has replaced wikilegal-l. By not being honest about the real nature of these surveys (that they are just straw polls which will either be ignored completely or will just be a footnote to the implementation), a lot of people get worked up that they "missed the vote" or that the offered choices weren't fair or that yadda yadda. Self-selected poll responses are not going to give accurate samples, either... Software development is not driven by votes; real people have to do real work. Erik, if you really do mean these polls to be informal surveys to gather ideas and get a general sense of what's well-liked, please do something about the way you present them because they *look* like official votes which will formally bind the entire development team under the sole authority of your having declared there to be a vote. -- brion vibber (brion @ pobox.com)

On Apr 15, 2004, at 02:11, Erik Moeller wrote: On the wiki? Probably not. ;) Anyway, I apologize for venting, but I would continue to argue that the way these polls are presented makes them seem more official, binding, or important than they really are (or, I think, than you intend them to be), and that creates needless friction over the legitimacy of votes that are really just informal opinion surveys that are informative, not binding. -- brion vibber (brion @ pobox.com)

Brion- Please reconcile the fact that only 50% of the American population voted for either George Bush or Al Gore with the numbnut that is currently in office. Every vote only draws from interested individuals. In our case, that will hopefully make for better results, as the people not interested in highly technical and insubstantial matters are also the ones who are unlikely to have an informed opinion. Furthermore, a less significant vote needs a lower minimum turnout. We don't need to ask 1000 people when we get a reasonably clear picture from asking 100 people. What reasonably clear is depends on the context. For the comparatively insignificant matter of whether we should type <math>2 + 2 = 4</math> or, say, >>>>(XXXX>>>|2+2=4|<<<XXXX)<<<<, it is arguable that we already wasted far too much time getting an unreasonably *large* sample of community opinion, especially with one option clearly dominating all others (20 votes for my proposal, 1 vote for the 2nd place proposal, all others negative). The only reason I started a poll at all is that people kept bringing up all kinds of wacky ideas and we needed to come to a solution. Who decides what is an important matter that deserves a Wikimedia-wide poll with announcements on the Main Pages and a personal message on each user talk page, and what can be done simply by announcing it to the mailing list and Wikimedia News, or maybe just a talk page? Ultimately, whoever sets up the poll, but also those who watch them do so. If you had felt the need to bring this to the attention of a larger group of people, you had ample opportunity to do so. The difference between such a poll and a straw poll is that a straw poll is entirely unannounced, has no deadline, no prior discussion phase, no defined outcome and a very low, if any, minimum turnout. Straw polls answer the question: Are we anywhere close to consensus? They do nothing else. The Extension Syntax vote was *not* a straw poll by any reasonable definition. Of course, if people just listened to me we wouldn't need any stupid polls. Regards, Erik