test_ok2/py/apigen/todo.txt

* when listing the methods of a class: 
  special "__*__" methods should come last except for __init__ 
  which comes first 

  DONE

* the page header should read:
  
    py.path.local API documentation [rev XYZ]

  instead of 

    api documentation for path.local 

  DONE, title changed and if possible (read: if source file in SVN) rev is
        retrieved and added

* have the py/doc/ and apigen page layout have
  an api and source link in the menu bar 
  (e.g.: home doc api source contact getting-started issue)

  DONE

* function view: 

     def __init__(self, rawcode):
        docstring-in-grey-and-not-in-a-box

  and the "show/hide funcinfo" link could be underyling
  the full "def __init__(self, rawcode)" or be a link right after 
  (or maybe before) it. 

  goal: consume less vertical space and have the functions
  be "sticking" out (the show/hide info link IMO disrupts this
  and it's not visually clear it belongs to the function above it)

  DONE, but please review if you like it like this...

* can it be avoided that py.execnet.Channel shows up as a
  primary object but still have it documented/linked from 
  remote_exec()'s "return value"? 

  DONE: if you set an attribute __hide_from_nav__ to True on an
        object somehow, it is hidden from the navigation

* class attributes are not "properties". can they get their
  section? 

  DONE: renamed title to 'class attributes and properties'
        (as discussed)

* stacktraces: a lot are "duplicates" like:

    /home/hpk/py-trunk/py/test/rsession/hostmanage.py - line 37
    /home/hpk/py-trunk/py/test/rsession/hostmanage.py - line 37
    /home/hpk/py-trunk/py/test/rsession/hostmanage.py - line 37

  even though they may reference different stackframes, 
  i think we should by default strip out these duplicates,
  this would also reduce the generated html files, right? 

  DONE, although I'm not happy with it... I'd rather only display call sites
        from calls in the test somehow or something...

* allow for flexibility regarding linking from 
  py/doc/*.txt documents to apigen with respect
  to where apigen/ docs are located. 

  LATER, as discussed