For the big stack frameworks, it's quite major. As he said, without the sheer market share that Rails has for Ruby…

That's true for big full stack frameworks, where most of the contenders disappeared. Real competition seems to happen mostly in the mini framework area, due to people being forced to ditch more and more parts of their full stacks and wondering why they couldn't instead build upon a small platform. Flask is gaining traction here.

And then there's app engine, where things run a bit differently, and forcing django et al on top of it can feel a bit unnatural. Tipfy seems to be en vogue there.

In Python country, you always had quite a number of projects that ran outside the mainstream, though. Tornado, cherrypy, web.py, web2py etc.

I have used Pylons extensively, but not Django. I really prefer Pylons's packaging and methodology. Django seems more like a "classical" web framework where everything is included and much has been reinvented. Since my exploration in programming is often a matter of personal taste (i.e., not forced by employer to learn something very often), I have not really spent much time with Django as that style of framework does not mesh with my preferences.

That said, I have recently been tinkering with a Satchmo cart. I am under the impression that Django apps that utilize the Django backends can be converted into plugins for other Django apps with trivial or near-trivial effort, and that gives Django-based applications a lot of power, because if both Satchmo and some Django-based forum use the Django admin and user database, I can plug in or replace forums really easily without creating a bunch of extra overhead for my users. That is very appealing to me, and the first reasonable use of a "big" framework (like Django, Rails, or Cake) as far as I am concerned. It has encouraged me to explore Django/Pinax more thoroughly.

I find Django slightly unusual compared to other frameworks in that it has many CMS like qualities, such as predefined user/permission models and the admin (not surprising, given its origins). Rails, Pylons, CodeIgniter, etc. provide you with an ORM, but it's up to you to define your schemas.

Not that this is necessarily a bad thing - it can save a lot of time in certain projects. Sometimes though it's overkill for projects that don't need these features.

So I'm glad that Django isn't the one and only framework in Python, as it's good to have a choice. Multiple frameworks also mean competition and more ideas and approaches, which benefit everyone.

I rock it old school and still prefer web.py over anything.

Well, let me just comment that I don't know a single line of Python, and yet I know what Django is (loosely) and have heard of it a lot.

Never heard of Pylons.