This is a classic example of why the term "open source" was poorly chosen. The forces that power what's usually called the "open source software movement" aren't tied to the creation of software. There's definitely strong ties between the blogging phenomenon and our software movement, but the term "open source" hides them. Lawrence Lessig would prefer that we call blogging an example of the power of "Free culture." If you think that's too grandiose, I'd offer "Free journalism."
That's a losing argument. Open source is pretty big; there's innovation out there somewhere. I'd cite Python's identation-based syntax, Firefox's Live Bookmarks and Evolution's Virtual Folders (and older feature, but it was innovative when it came out I believe). There's plenty more, but I'm trying to restrain myself to the popular projects, as you mentioned.
Support for more advanced forms would be great. Apparently Novell is working with Mozilla to implement XForms. That's nice, but XForms doesn't degrade will face a chicken/egg situation with sites/users. The WHATWG has a Web Forms 2.0 specification mostly fleshed out, and it degrades well in current browsers, meaning that sites can start implementing it today without pissing off their users. There's nothing I'd like more than Gecko support for Web Forms 2.0.
I'm pretty well versed in Zope 2, and am worried that Zope 3 seems too serious. One of the more common complaints about Zope has been its learning curve; I see no reason that those complaints won't continue with Zope 3. In Zope 2, if you wanted to create a new custom feature, you could create a product. I'm still wrapping my mind around the new concepts, but in Zope 3, it appears that you have to create a component, and maybe an adapter, and maybe a view. And you'll be strongly encouraged to write unit tests for all of it.
Also, Zope has 2 templating systems: DTML and ZPT. ZPT is supposed to be the more recent, more correct system, but as far as I can tell, most people, myself included, are continuing to use DTML (I actually think DTML is conceptually superior, since it offers more precise error handling). Both systems are in Zope 3, but it appears that the new DTML has had features actually removed from it, because they theoretically encourage bad code. Zope is written in Python, and one of Python's philosophical guidelines has always been "We're all consenting adults here." I'm worried that Zope 3 in general is strongly pushing developers to create formalized, overdesigned solutions instead of the conceptually simple ones that Zope, and Python in general, currently encourage. I'm worried that Zope is becoming more like J2EE.
Slashdot Mirror
This is a classic example of why the term "open source" was poorly chosen. The forces that power what's usually called the "open source software movement" aren't tied to the creation of software. There's definitely strong ties between the blogging phenomenon and our software movement, but the term "open source" hides them. Lawrence Lessig would prefer that we call blogging an example of the power of "Free culture." If you think that's too grandiose, I'd offer "Free journalism."
That's a losing argument. Open source is pretty big; there's innovation out there somewhere. I'd cite Python's identation-based syntax, Firefox's Live Bookmarks and Evolution's Virtual Folders (and older feature, but it was innovative when it came out I believe). There's plenty more, but I'm trying to restrain myself to the popular projects, as you mentioned.
Support for more advanced forms would be great. Apparently Novell is working with Mozilla to implement XForms. That's nice, but XForms doesn't degrade will face a chicken/egg situation with sites/users. The WHATWG has a Web Forms 2.0 specification mostly fleshed out, and it degrades well in current browsers, meaning that sites can start implementing it today without pissing off their users. There's nothing I'd like more than Gecko support for Web Forms 2.0.
I'm pretty well versed in Zope 2, and am worried that Zope 3 seems too serious. One of the more common complaints about Zope has been its learning curve; I see no reason that those complaints won't continue with Zope 3. In Zope 2, if you wanted to create a new custom feature, you could create a product. I'm still wrapping my mind around the new concepts, but in Zope 3, it appears that you have to create a component, and maybe an adapter, and maybe a view. And you'll be strongly encouraged to write unit tests for all of it. Also, Zope has 2 templating systems: DTML and ZPT. ZPT is supposed to be the more recent, more correct system, but as far as I can tell, most people, myself included, are continuing to use DTML (I actually think DTML is conceptually superior, since it offers more precise error handling). Both systems are in Zope 3, but it appears that the new DTML has had features actually removed from it, because they theoretically encourage bad code. Zope is written in Python, and one of Python's philosophical guidelines has always been "We're all consenting adults here." I'm worried that Zope 3 in general is strongly pushing developers to create formalized, overdesigned solutions instead of the conceptually simple ones that Zope, and Python in general, currently encourage. I'm worried that Zope is becoming more like J2EE.