Re: [POLICY] Lazy account creation
I think we have two cases here that we need to treat differently: public and private code. In the case of publicly developed code, we can easily check the public source repository and mailing lists and say hey, these people did contribute to the project. I the case of privately developed code, we can not make this determination and lazy account creation seems like a good way to determine who is actually contributing. -dain On Mar 19, 2006, at 6:23 PM, Noel J. Bergman wrote: robert burrell donkin asked: Geir Magnusson Jr wrote: What we'll probably do is run it like we're running Harmony. The list of committers on the proposal are the people we expect to show up, but we won't be creating accounts by default - we'll need to have each person say yes, I'm ready and will be contributing immediately before making the account for them. this seems a useful tradition: is it too early to codified this into policy? I'd say so, yes, it is too early. Could probably use more discussion. I'm not quite sure what to make of it. What is it trying to accomplish? If someone comes up and says that they are ready to commit, what is this doing? Fast tracking? Geir already responded with a concern of his own, and suggested modifications if the rule. So I'll ask a more basic question: do we need a rule for this, or can the human beings responsible for managing the project make a human decision based upon the situation? Keep in mind something that I've said a lot lately: generally, I prefer to use SHOULD, rather than SHALL, leaving decisions in the hands of people instead of pre-packaged rules. SHOULD provides guidelines -- hopefully best practices -- while still providing leeway. I believe that we've already established a policy that recommends a relatively low bar for Committer status. At least thus far, what that means has been up to the PPMC for each podling. With respect to the message subject, I certainly do not have a problem with not creating accounts until someone is ready to actually do something, but keep in mind the lag time involved in the account creation workflow, including the CLA gathering phase. Beyond this, what else would be helpful to recommend, and is the recommendation Incubator specific? --- Noel - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [POLICY] Lazy account creation [WAS Re: [VOTE] Accept OpenJPA as an Incubator Podling]
+1 to the general idea. Yoav On 3/19/06, robert burrell donkin [EMAIL PROTECTED] wrote: On 3/19/06, Jacek Laskowski [EMAIL PROTECTED] wrote: 2006/3/19, robert burrell donkin [EMAIL PROTECTED]: On 3/19/06, Geir Magnusson Jr [EMAIL PROTECTED] wrote: What we'll probably do is run it like we're running Harmony. The list of committers on the proposal are the people we expect to show up, but we won't be creating accounts by default - we'll need to have each person say yes, I'm ready and will be contributing immediately before making the account for them. this seems a useful tradition: is it too early to codified this into policy? Before it goes to the policy, could someone explain me who's going to accept those who's claimed 'yes'? Possibly mentors? Are they 'we' in Geir's email? Will they - mentors - commit incoming patches? There must be someone who will do it. I liked the idea at the first sight, but then lots of questions came in. i'm assuming that the initial mentors consistute the initial ppmc and vote people from the list as committers when they show up. seems a good way to get used to the usual apache process. could be a useful way to get to know folks as well if people feel like posting potted bios or blog urls. but hopefully gier or leo will jump in and explain how it worked in harmony... - robert - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] -- Yoav Shapira Nimalex LLC 1 Mifflin Place, Suite 310 Cambridge, MA, USA [EMAIL PROTECTED] / www.yoavshapira.com - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [POLICY] Lazy account creation [WAS Re: [VOTE] Accept OpenJPA as an Incubator Podling]
robert burrell donkin wrote: On 3/19/06, Geir Magnusson Jr [EMAIL PROTECTED] wrote: What we'll probably do is run it like we're running Harmony. The list of committers on the proposal are the people we expect to show up, but we won't be creating accounts by default - we'll need to have each person say yes, I'm ready and will be contributing immediately before making the account for them. this seems a useful tradition: is it too early to codified this into policy? I think so. One problem is that it places special status on some people that others don't have, once you get rolling. So n months into the project, if one of the listed people pops up and gets working, fast-tracking commit might make others that have been working in the project (and don't have commit) unhappy. Maybe the solution is a well-codified policy with some time limit, so that it doesn't seem arbitrary and capricious. geir - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [POLICY] Lazy account creation
On Sun, Mar 19, 2006 at 09:51:40PM +, robert burrell donkin wrote: On 3/19/06, Geir Magnusson Jr [EMAIL PROTECTED] wrote: What we'll probably do is run it like we're running Harmony. The list of committers on the proposal are the people we expect to show up, but we won't be creating accounts by default - we'll need to have each person say yes, I'm ready and will be contributing immediately before making the account for them. this seems a useful tradition: is it too early to codified this into policy? Yeah, very much so. If Geir hadn't been running around like crazy for the first few months to handle the incoming patch stream it would not have worked. I think Harmony has 7 or so mentors who are all ASF members, which is quite a luxury. We also had an excuse of sorts to not set up accounts (eg it wasn't a completely concious decision from a community perspective) -- we wanted to sort out some of the legal stuff first. This made things easier to explain. We were also (deliberately) starting from scratch, whereas most projects around here start from a single existing codebase, and that is also a big difference. You can't just go and bring a bunch of people from an existing project in and remove their commit karma. I haven't committed anything to excalibur in months (I think) but I would be somewhat unhappy if a move to the FooBar organisation would mean I'd lose commit access (I use the code and sometimes I find a problem and then I want to just commit the fix to trunk). I think its sometimes a useful tradition to just go and do what seems right rather than what people wrote down as policy as long as you comply with the spirit and intent of the policy and don't mess up any of the tricky bits (like all the legal ones). But that is not so easily codified and perhaps as a written policy can lead to problems. LSD - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
RE: [POLICY] Lazy account creation
robert burrell donkin asked: Geir Magnusson Jr wrote: What we'll probably do is run it like we're running Harmony. The list of committers on the proposal are the people we expect to show up, but we won't be creating accounts by default - we'll need to have each person say yes, I'm ready and will be contributing immediately before making the account for them. this seems a useful tradition: is it too early to codified this into policy? I'd say so, yes, it is too early. Could probably use more discussion. I'm not quite sure what to make of it. What is it trying to accomplish? If someone comes up and says that they are ready to commit, what is this doing? Fast tracking? Geir already responded with a concern of his own, and suggested modifications if the rule. So I'll ask a more basic question: do we need a rule for this, or can the human beings responsible for managing the project make a human decision based upon the situation? Keep in mind something that I've said a lot lately: generally, I prefer to use SHOULD, rather than SHALL, leaving decisions in the hands of people instead of pre-packaged rules. SHOULD provides guidelines -- hopefully best practices -- while still providing leeway. I believe that we've already established a policy that recommends a relatively low bar for Committer status. At least thus far, what that means has been up to the PPMC for each podling. With respect to the message subject, I certainly do not have a problem with not creating accounts until someone is ready to actually do something, but keep in mind the lag time involved in the account creation workflow, including the CLA gathering phase. Beyond this, what else would be helpful to recommend, and is the recommendation Incubator specific? --- Noel - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
