Re: [Zope3-dev] Re: Pluggable authentication id management
Philipp von Weitershausen wrote: [snip] I'm sorely tempted to do this for 3.3. -1 -1 too Regards, Martijn ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com
Re: [Zope3-dev] Re: Pluggable authentication id management
On 7/31/06, Philipp von Weitershausen <[EMAIL PROTECTED]> wrote: If you ask me, the slipping of the Zope 3 June release is mostly due to the lack of a release manager. If we had a release manager who'd kick people's asses, we might have more betas since the first one and we might even have a final release. Jup. Also, I think that just before christmas and just before vacations may not be good places. I'd like to suggest august/september and february/march for release dates. These may be equally bad for other reasons, though. Let's have a bug day this Friday. +100 -- Lennart Regebro, Nuxeo http://www.nuxeo.com/ CPS Content Management http://www.cps-project.org/ ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com
[Zope3-dev] Re: Pluggable authentication id management
Jim Fulton wrote: On Jul 31, 2006, at 2:02 PM, Philipp von Weitershausen wrote: ... If you ask me, the slipping of the Zope 3 June release is mostly due to the lack of a release manager. If we had a release manager who'd kick people's asses, we might have more betas since the first one and we might even have a final release. I don't agree. Except for a few people, there hasn't been much interest in fixing bugs. Releasing betas when bugs reported earlier haven't been addressed is counter productive. Why should people bother testing if they don't think the bugs they report are going to be fixed. Disclaimer: I'm not involved in the Zope 3 release process, so what I say below is completely general and not specific to any events having gone on in the Zope 3 community. I'll note from my observations in the Plone community that not having someone bitching is a pretty good way to ensure things don't get done. Everyone has a lot of projects and personal lives that get in the way. I have several bugs assigned to me in the Plone tracker that I never seem to get around to doing. But when Alec or Limi or someone else tells me, we need this or else we can't ship, that makes them jump up the priority list. Having someone vocal with authority to instill a sense of urgency in people is a fairly important part of managing open source projects. Martin ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com
Re: [Zope3-dev] Re: Pluggable authentication id management
On Jul 31, 2006, at 2:02 PM, Philipp von Weitershausen wrote: ... If you ask me, the slipping of the Zope 3 June release is mostly due to the lack of a release manager. If we had a release manager who'd kick people's asses, we might have more betas since the first one and we might even have a final release. I don't agree. Except for a few people, there hasn't been much interest in fixing bugs. Releasing betas when bugs reported earlier haven't been addressed is counter productive. Why should people bother testing if they don't think the bugs they report are going to be fixed. Jim -- Jim Fulton mailto:[EMAIL PROTECTED]Python Powered! CTO (540) 361-1714 http://www.python.org Zope Corporationhttp://www.zope.com http://www.zope.org ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com
[Zope3-dev] Re: Pluggable authentication id management
On Jul 31, 2006, at 2:06 PM, Philipp von Weitershausen wrote: ... So I imagine that when I add an authenticator plugin to a PAU, I'll have to specify a fully qualified prefix, not just its prefix within the PAU. Right? Right. JIm -- Jim Fulton mailto:[EMAIL PROTECTED]Python Powered! CTO (540) 361-1714 http://www.python.org Zope Corporationhttp://www.zope.com http://www.zope.org ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com
[Zope3-dev] Re: Pluggable authentication id management
Jim Fulton wrote: > When I did the initial design for the pluggable-authentication utility > (PAU), I came up with a strategy for managing principal ids, in > retrospect, is overly complicated. This suspicion is supported by the > fact that I originally got the implementation of this wrong. Right. > An authenticator plugin, among other things, manages principal ids. > Principal ids need to be unique system wide. In a misguided attempt to > make life easier for plugin authors, I decided that that the PAU should > have a prefix that it adds to principal ids. > This means that plugins that manage principal ids can't get at principal > ids without accessing their PAU, which further means that a plugin can > only be used with a single PAU. > > I'd like to get rid of the PAU prefix and simply require that > authenticator plugins provide system-wide unique ids. This can be done > by providing suitable prefixes on each plugin. > > I suggest that, for 3.4, we get rid of the PAU prefix option and provide > a generation evolution script that, for PAUs with non-empty prefixes, > just prepends their prefixes to their plugin prefixes and clears their > prefixes. +1 So I imagine that when I add an authenticator plugin to a PAU, I'll have to specify a fully qualified prefix, not just its prefix within the PAU. Right? > I'm sorely tempted to do this for 3.3. -1 No more surprises for Zope 3.3, please. Zope 3.4 is only 3 months away. For all I care, the only feature it can have is to get rid of the PAU annoyance... But Zope 3.3 needs to stay frozen. We'd be making fools of ourselves (if we aren't doing that already) otherwise. Philipp ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com
[Zope3-dev] Re: Pluggable authentication id management
Benji York wrote: > Jim Fulton wrote: >> I suggest that, for 3.4, we get rid of the PAU prefix option and >> provide a generation evolution script that, for PAUs with non-empty >> prefixes, just prepends their prefixes to their plugin prefixes and >> clears their prefixes. I'm sorely tempted to do this for 3.3. >> >> Thoughts? > > If we cancel the June (hah!) release and requeue 3.3 for the November > release, then 3.3 would be back open for features (hopefully only a few > though). -1 We can't just postpone a release every time we're slipping the deadline. Sure, we're slipping badly, but that doesn't mean we shouldn't make an effort to close this chapter now. AFAIK there are only three critical Zope 3.3 bugs. What we need is a bug day and everyone (incl. me) getting off their butts and fix those bugs. If you ask me, the slipping of the Zope 3 June release is mostly due to the lack of a release manager. If we had a release manager who'd kick people's asses, we might have more betas since the first one and we might even have a final release. Let's have a bug day this Friday. Philipp ___ Zope3-dev mailing list Zope3-dev@zope.org Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com