On 22 May 2010 21:15, Mike Orr <[email protected]> wrote: > On Sat, May 22, 2010 at 12:12 PM, Rufus Pollock <[email protected]> > wrote: [...] >> Below I've given a list of some examples of the sorts of things that >> could be "plugins". What else should be on that list? Is there >> interest in having plugins? >> >> Regards, >> >> Rufus Pollock > > Pylons is intentionally small, so that things that would be plugins in > other frameworks can usually be implemented on top of Pylons. So for > instance, Repoze.who/what is a plugin system for auth^2, and there are
Right, we have used repoze.who extensively :) While this is pretty pluggable, if you want, for example, user/password/email confirmation you have to plug this together with the db, controllers etc yourself (sure repoze.who gives you the form but you heed to plug it together with db, emailing etc).[^1] Now I'm well aware that there will *never* be a one-size fits all solution. However, I think there is enough commonality across enough apps that a small plugin/example solution could be valuable to the community -- and yes I know there are examples on the wiki (which were very useful) but they aren't the same as some running code bundled up as a small package (for one thing it is much easier to keep such material up to date over time) [^1]: Looking around for this sort of thing already being done there is (AFACIT): http://blog.sadphaeton.com/node/39, http://pypi.python.org/pypi/BlastOff (but does much more) and we ourselves implemented this as part of http://www.isitopendata.org/ (though not factored out as yet) > Pylons application templates on PyPI that preconfigure a > Pylons/who/what application for you. Eventually the best of these > templates might get official status and be featured in the docs. Right, that would be great. It seems like templates would be getting pretty close to plugins if they included, say, sqlalchemy model code. > Including them in the Pylons distribution would be more difficult -- > we'd be committing to supporting them forever, and the state of the > art in auth, CRUD, etc changes so frequently. Sorry if this was not clear in my email: IMO *plugins* should *never* be in pylons core (as that should be kept lean and mean). Plugins are extensions of specific functionality However, you might want to have a set of "semi-official" plugins that had good support from the core community. > There has been some work to base Pylons 2 on zope.component, which > would allow Pylons components themselves (routing, action calling) to > be pluggable and replaceable. An application could use the same > component system to configure its auth plugins, etc. Another goal of > Pylons 2 is to make applications more nestable, so that they could be > used as inner apps similar to Django applications. So this may help > for plugins, but first we have to get Pylons 1 finished. This sounds a little more orthogonal as its about making "components" of pylons pluggable rather than about plugins extending pylons core. [...] >> ### 4. Commenting >> >> * Commenting is a common functionality across many applications >> * A basic plugin that providing commenting functionality either >> directly or pased on an existing pluggable service (e.g. disqus) would >> be extremely useful > > If you look on PyPI you'll see the existing Pylons applicaition > templates. You might get together with those authors and set up a > central list of these, and find some way to distingush the best ones > and get rid of the worst ones so that users can see, "This is the best > recommended template for doing X." Ok this sounds promising. First step I guess is to start a wiki page listing wants and existing work and then go from there. Regards, Rufus -- You received this message because you are subscribed to the Google Groups "pylons-discuss" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/pylons-discuss?hl=en.
