It would be convenient to have "Suggest a change" link in the bottom of
documentation pages to allow people propose fixes and additions. 

This can be a simple feedback form with an optional "patch" field that
submits mail to a list or simply adds documentation comment.

More complex approach would be construct patch right in the browser.
Then add it for review into "documentation queue". "documentation queue"
is a part of documentation system to automatically submit patches after
review (and optionally incrementally-recompile existing output formats).
In case of page comments review status can be displayed in place
together with comments and patches. The whole edit/review process may
require single sign on system described in issue #2837

Yes, this is under consideration.

Any links to alpha code?

None yet. :(

An easier approach would be to open the "create a new issue" page of the
tracker when the user clicks on "suggest a change".
The title can be set automatically to something like "Doc suggestion on
Page XYZ". The "Documentation" component and the version could be set
automatically too, and the "Comment" field could include a link to the page.
If the user is registered to the tracker, he would just have to write
the comment and submit, possibly changing/adjusting the fields.

I don't know if it's actually possible to do something like this, but it
shouldn't be too hard to hack the tracker in order to pass this
information when the users click on the link.

Since this will be handled by Sphinx, it is not an issue for this tracker.

This functionality is specific request for python documentation, so adjustments should be made to Sphinx customization layer for docs.python.org if Sphinx itself doesn't provide such feature.

Is there a bug tracker for python.org infrastructure?

I am glad there is some response to this issue in r80388

I'm closing this again; I can see that nobody will implement this for Python if it's not available in Sphinx, and also I don't think it would be a good idea to do so.

Georg, if you are not going to implement it, it doesn't mean nobody will do this. In Sphinx or through disqus. I do not see the reason to close this feature request.

Also I do not understand arguments like:
"Yes, this is under consideration." in msg69834 and
"I don't think it would be a good idea to do so." in msg104084

Care to explain why you have changed your mind?

I haven't changed my mind.  The "it would not be a good idea to do so" refers to implementing this for Python exclusively.

Of course, if you volunteer to do so, I won't stop you :)  Otherwise, please trust my judgement that this issue will do nothing except eventually be closed again due to lack of movement, or because something has been implemented in Sphinx.  It may even be during this year's Summer of Code, if we're lucky.

Still actual, esp. with this - http://google-opensource.blogspot.com/2011/01/make-quick-fixes-quicker-on-google.html and this http://codemirror.net/

And still this is a thing that Sphinx needs to do (already does, but needs to be finalized.)