Re: On PEAR quality issues and natural selection

Hi! Lukas Smith wrote:

This has been our policy since I remember. I get tons of pear account requests to propose new packages. 50% of them are for packages which already exist. I send them a link to the existing package and tell them that if their package has a different approach I will open the account (note I am not evaluating this, I leave the decision to them). If its not then I suggest to contact the author to see if any thing in the proposers package might be added to the relevant existing PEAR package.

In other words: right now the competitive packages are effectively turned down before they even reach the proposal stage. Q.E.D. That's the difference in our position: you --- the member of Ye Inner Circle --- see these requests and I --- the member of unwashed masses --- *don't*. Thus I'd like them to be made public, I am quite sure that 95% of them will be turned down, but I don't know whether you also dump that other 5%.

Interestingly very few people seem to be going the route of contacting the original developers to merge in their improvements. Atleast it looks to me that way since I would expect to see new features being added to the existing package. So maybe we are naiv to be hoping for cooperation. However under these obervations the idea of every developer being allowed to modify any package seems even more naiv.

Lukas, you can't really *force* people to cooperate. But: you can add additional incentives for cooperation. Heavy breathing of competition on the back of one's neck can be such an incentive, don't you think?