I wanted to post something here last night about these conversations
I've been having in email with Ben Bangert, whose has the weblog
http://www.groovie.org/ and has written tools like Routes
http://routes.groovie.org/ - a tool for people who don't have the
benefit of nice zope.publisher style URLs to make (and regenerate)
nice URLs :)
These are some comments he made specifically about finding Zope 3
information. They may seem a bit crude, but he seems genuinely
frustrated. Now - I'll add that I'm impressed with tools like apidoc,
the books, and all of that, but if you compare the Zope 3 wiki front
page with that of http://www.rubyonrails.com/ and
http://www.turbogears.org/ you'll see that there's a big difference.
And you know that I personally have a big distaste for Wikis. Someone
(I can't remember who) said recently that a good problem with Wikis is
that it's hard to grok the relevance of a certain page - does it still
apply to current thinking? There are a lot of historical artifacts in
the Zope 3 development wiki. And I'm not saying that we should get rid
of them, but it's just hard to know that PageA is a hypothetical
dreamland item from three years ago and PageB right next to it is
valid information and documentation for Zope 3.1. Anyways, I'll stop
grand-standing on that and share Ben's thoughts. Well, after also
saying that I know that we're all busy developers with real jobs,
consultant gigs, research work, education, sprints, and so on, so I'm
not volunteering myself nor expecting anyone else to take up the lead.
But it would be great if someone did ;), and Bottlerocket may be able
to help... when our current rush of jobs settles... (sigh).
Alright. Ben's statements:
The documentation on the Zope site....ugh. That alone has driven me
nuts more than I can remember. The docs are horrible, frequently
outdated, only occasionally actually work, scattered around the site
with little organization. The site navigation is absolutely horrible,
things rarely indicate where you are, how you got there, whats next,
etc. While I can see the little nav thing at the top, I'm referring
more to the left sidebar that never indicates you are in that section.
- A good example of why Zope3.org should be its own site.
Zope 3 hardly looks ready for anyone to use if you actually go to the
website. It's slotted in under the Zope Corp page, and has the
appearance of a science school project. It really deserves its own
site devoted to it without all the extra navigation and confusing
headers leading all over. Zope 3.1 needs a colorful, enticing,
approaching site with excellent documentation that actually works and
- We have doctests that actually work and are maintained, lets get
some of that online! Lets get other doctest style documents online!
- Zope 3.1 is really really really a great release guys. I'm very
impressed with the simplification of the component architecture, with
the deprecation system, and with the generations system. It shows that
Zope 3.1 is a serious release. I think it's a great candidate for
getting up on the rooftops and shouting about.
I'd really like to give Zope 3 a try, and I keep trying to. The docs
are just nauseating. They might look good or fine to someone who's
used Zope for years, but to someone new they're horrid. As I
mentioned, the site is laid out horribly for someone who wants to
learn Zope 3. Why is the left bar saturated with links when I just
want Zope 3?
It's incredibly frustrating and disappointing to hear about all the
cool stuff you can do in Zope 3, and not see anywhere that shows it
actually being done with descriptions on how it works, etc. Where are
the examples? Where are the recipies to "do cool thing X"?
The developers I see talk about Zope are all in companies that use
it, that have teams that use it, that have tons of actual knowledge
that doesn't exist on the website. I really really want to give Zope
3 a spin, I have a few fairly complex projects I'd like to try out
with it. How do I get started?
Ben is an intelligent Python programmer who is a big fan of Myghty,
but might be just the kind of target audience we want for Zope 3 right
now - educated, enthusiastic, interested developers with a history of
web development. We all fit that bill here, but I imagine many of us
have been using Zope for years, some of us going all the way back to
Bobo and Principia. And I know that for me - Zope 3 is goddamn
exciting. But how does that message carry over?
Personally, I really like the Z3ECM project site - z3lab.org. It has a
combination of blogs, papers, demos (animations) and documentation.
That's still a project in its infancy, but it's a good looking site
with a variety of sources feeding into it. If there were to be a model
for a Zope 3 web site, that would be it. The development wiki should
still be inside of it, of course. But weblog entries like mine, like
Martijn's, like Benji's, should all go into it as well (not as a
"planet zope" type stream, but more dedicated so it's known that
relevant information shows up).
I think that Zope 3 has its own momentum away from Zope 2. Five is
what is going to be appealing to Zope 2 developers learning about Zope
3. There should be something that appeals to the general Python crowd
that I still believe to be the main target for Zope 3 as it stands
right now - a shorter path between your Python objects and publishing
them on the web. Or a shorter path between writing some plain old
Python code and seeing a result.
I need to read more of the other thread about this. I just know that
there's some debate of whether Zope 3 should be a part of the Zope.org
site. I'm for having it be separate. There are a lot of Python
developers out there who have bad memories of Zope 2 but just may be
the kind of developers who'd like or love Zope 3 if they could really
see and understand what makes Zope 3 so great.
Zope3-dev mailing list