Proposal: Programmable logic and FPGA design
My understanding is that the motivation for the reputation requirement during the commit phase is that the site is more likely to succeed if it has:
- more users that are familiar with the SE concept; and
- more users who have demonstrated that they are active on the SE network of sites.
(see additional discussion here: http://meta.stackoverflow.com/questions/65779/area-51-commitment-critical-mass)
I think that "familiarity with the concept" is insignificant -- to the credit of its creators, the SE concept isn't difficult to grasp, particularly for a technical audience. Also, every site has editors that make sure things are conforming to the style.
There's probably plenty of evidence that active SE users do contribute to new sites' success, but, overall, I think the commit phase requirement is detrimental to the expansion of the Stack Exchange network as a whole. This is because it effectively discourages the creation of sites that will bring fresh participants by requiring a (too) significant "internal" support. Additionally, there's no guarantee that even high-SE-reputation committers with an interest in the topic of the new site proposal are actually expert contributors on the topic, which is really what a new site should attract. Those experts may well come from the "outside", but have no way of significantly contributing to the "commit".
Two additional subtle observations:
- The three commitment limit could unjustly be hampering the progress of good proposals. Since sites can be months in the commit phase, an opening is quite infrequent (if all three slots are used, which is perhaps typical for high reputation users). I wonder what the "uncommit" rate is (in order for people to prioritize) -- my feeling is that once people commit to a proposal, they rarely uncommit.
- New (and probably many experienced) SE users are not familiar with the formula used for the "percent complete" calculation (it's not on the FAQ). If someone comes in and sees "50 committers" and "10% complete" they immediately recognize that 500 committers are required, which may deter them from going through the process of understanding the SE way of doing things, registering, and committing. (The argument that says "well, then they shouldn't commit anyway" is somewhat valid here, though shouldn't we at least try to achieve a more intuitive/simpler inclusion process?)
My suggestions are the following:
- remove the reputation requirement completely during the commit phase;
- keep the 200 committers requirement, or even increase it slightly as compensation... it would be far simpler to convey the progress this way;
- be more strict during the beta phase, the point at which the community is actually being put to the "SE test";
- consider allowing more than 3 commits -- is there any evidence that a limit of 3 is effective?