Some portions of the documentation are using the term "member" to mean "data attribute".  This appears to be an aberration at this time, but occurrences should be identified and corrected, and "Documenting Python" updated to note correct usage.

Example use:

http://docs.python.org/dev/py3k/library/xmlrpc.client.html#fault-objects

http://docs.python.org/dev/py3k/library/urllib.request.html#basehandler-objects
(paragraph starting "The following members and methods should")

"Members and methods" should just be "attributes".

> "Members and methods" should just be "attributes".

+1 

That substitution should be made almost everywhere.

Individual slot variable are still called members though
and their type is a member_descriptor.  And the C API
still has a tp_members entry.  Otherwise, it looks like
the term member is out-of-date.

It is 'attributes' instead of term 'members'. The term 'method' when it denotes methods can be left as such.

I grepped the documentation in the cpython repository and replaced all mentions of "member(s)" with "attribute(s)" where I felt appropriate.  I left mentions of "members" related to structs or any C documentation alone as I'm less confident of their terminology (I'm new around here).

I used "methods and attribute" in io.rst because BufferedIOBase included one attribute and several methods.  My terminology may be off, but I felt this was the correct replacement of "members."

Please review the patch and critique it as necessary.

> I used "methods and attribute" in io.rst because BufferedIOBase
> included one attribute and several methods.  My terminology may be
> off, but I felt this was the correct replacement of "members."

Nearly :)  See in the first message:
> "Members and methods" should just be "attributes".

I was always under the impression attributes and methods were mutually exclusive.  I've corrected the patch as requested.

Alexander, could you comment on the review page about datetime.rst changes?

So I'm -1 on using "attributes" to denote methods. It will actively confuse non-expert users. If you want to ditch "members", please consider using the more explicit phrase "attributes and methods".

Yes, I agree with you. Good Suggestion. Thanks!

I’m not -1 on using “attributes” to denote methods, since they *are* and it’s not difficult to understand, and I’m not opposed either to using “attributes” to replace “members” (i.e. attributes that are not methods.)

For those wanting to see the raging debate: http://mail.python.org/pipermail/python-dev/2011-June/112020.html

:)

New changeset d442c313536b by Senthil Kumaran in branch '3.2':
issue10403 - Let's not use members anymore. Use 'attribute' where it denotes attribute and 'methods' where it denotes methods. Context should clarify usage.
http://hg.python.org/cpython/rev/d442c313536b

New changeset d442c313536b by Senthil Kumaran in branch '3.2':
issue10403 - Let's not use members anymore. Use 'attribute' where it denotes attribute and 'methods' where it denotes methods. Context should clarify usage.
http://hg.python.org/cpython/rev/d442c313536b

New changeset b8f5da066782 by Senthil Kumaran in branch '2.7':
Fix closes issue10403 - Let's not use members anymore.
http://hg.python.org/cpython/rev/b8f5da066782

Senthil, I’m not sure you read Alexander’s reply on Rietveld before committing.

Hello Eric, I missed noticing Alexander's comments in the reitveld, I looked only at tracker then. I see that some of them can be addressed. Like using members (components) of the field, instead of attributes when it is not an attribute. Shall correct it. Thanks.

New changeset 05f0ffe4e0b3 by Senthil Kumaran in branch '3.2':
Fix Issue10403 - datetime documentation clarification based on review in the reitveld by Alexendar belopolsky.
http://hg.python.org/cpython/rev/05f0ffe4e0b3

New changeset 3935a1fb1db2 by Senthil Kumaran in branch '2.7':
merge from 3.2 - Issue10403  - datetime module documentation changes based on review.
http://hg.python.org/cpython/rev/3935a1fb1db2