Re: PEAR a way forward

--On October 4, 2006 11:02:36 PM +0200 Lukas Kahwe Smith <lsmith@php.net> wrote: >> I believe we have a race condition: How do we make decisions, resolve >> issues and move forward, when we clearly have trouble making decisions, >> resolving issues and moving forward? > > this is why i proposed at the very start of these discussions that people > who know eachother get together in smaller groups and work on proposed > structures in a semi open but highly iterative way. many people have > opinions, but few people have the dedication to put it all together. > however we need something put together at some point. > > based on these proposals, we can have feedback and hopefully in the end > each of the groups behind the proposals will be able to consolidate the > number of proposals to a sensible number (1-2 with only a few variations > each) based on the general feedback. > > these proposals do not need to decide on everything just yet, but they > should for one define the general structure and the common goal > underneath. maybe a few technical aspects, but that can hopefully follow > once we have the structure for decision in place along with a common > vision. I agree with what you're saying, but I'll try to make the same point I had hoped to convey in my previous message: recap: The problem PEAR faces is difficulty moving. We seem to be unable to make large decisions in a clear fashion. We disagree about how to solve the problem. Lukas has mentioned (above and earlier) that we need to have semi open meetings to develop proposals, and then proceed to decide on them. This is basically a "working group" tasked with assessing a situation and making a proposal to the community at large. (This is a very good way to reach a non-trivial goal.) But, we have no fair and open way to create or manage such work groups. Nor even any way to track the *goals* we think PEAR should strive for. Without goals, work groups are doomed to failure. (Rhetorical question: If I asked what the most important goal is -- the goal we set for the first work group -- how many different answers would I get?) We need a system (rules, procedures and software) which breathes life into PEAR as a thing separate from any one person or work group. Into the system we put our principles. We put in *all* the goals for PEAR. We put in the details like coding standards and naming conventions. We plan ahead for dealing with things like forming work groups, moving work group proposals through revision and acceptance, for guiding proposals through peer review, for freshening the PEAR group (I'm just tossing out things that come to mind -- I like the "stability" PEAR group provides to PEAR), or whatever else anyone wants to contribute. A parallel example... There is *one* reason why the entire Internet continues to exist and function: The RFC system. When some humble grad students drafted RFC 1, they were not thinking about the Internet as we know it. Instead, their ideas about *how* to organize their work on one project, how to present their work, and how to invite collaboration resulted in what we see today. The RFCs (and the subsequent STD docs, etc) lay a groundwork upon which everything is built. Back to concrete things... So that noone confuses my efforts: I do have an RFC in proposal about pearweb. But it's only meant to attract discussion about pearweb. In that RFC I'm only trying to capture the energy everyone is putting into discussing pearweb -- hoping that I'm making it easier for people to break off workable chunks in the near future. Unrelated to that RFC, I have specific ideas about how to implement the system/structure I'm describing above. (Ideas which definitely need discussion.) But I'm not going to pour effort into laying them out, unless there's at least a little interest. After all, if I'm the one person heading for left field, I'm not helping *at all*. But if this idea seems useful, then I'll be happy to start hammering out details. In the immediate future, the web gang is working on immediate, pressing issues. These issues are urgent (and therefore very worthy of the excellent work being applied), but pearweb itself is not important compared to the overarching issues PEAR faces. ---- PEAR web/CVS: cconstantine cconstantine@php.net / craig.blurfl.com