v 0.7
-we're on sf.net :- kebasdata.sf.net
-added timeoutsocket.py.  default to 15.  change according to need.  didnt
 implement timeout per connection, since the timeout will not be of any effect.  from
timeoutsocket.py readme """
	
As an example, one might use the timeout feature to create httplib
connections that will timeout after 30 seconds:

    import timeoutsocket
    import httplib
    H = httplib.HTTP("www.python.org")
    H.sock.set_timeout(30)

Note:  When used in this manner, the connect() routine may still
block because it happens before the timeout is set.  To avoid
this, use the 'timeoutsocket.setDefaultSocketTimeout()' function.
"""
zopezen told me of TimeoutSocket.py from zope's CVS, but i havent had time to investigate that yet.  so, if any 
of you have time ..:)
-rejigged view method, to check for hasattr(self, self.render_method) and return something, instead of Attribs error
if there's no render_method, or if the render_method is not set. (as informed by Hans-Dieter Stich <hdstich@hdstich.com>)
-redid self.url statement bit.  also, chucked self.f for tmpf, and f. (Oliver Bleutgen <myzope@gmx.net>
suggested to use _v_ for self.f, i'm not too sure about that, so i made it local and passed it around)
-now, clicking the Test tab will call and return the url
-set start/end of re to _v_* (suggested to use _v_ by Oliver Bleutgen <myzope@gmx.net>)

v 0.6
-refactored _get_url to chk_proxy(), _parse_data() 
-redid _chk_time() to NOT use self. to stop modification of the instance resulting in zodb bloat (a lot of stuff in undo tab in ZMI)
-added Max Gotgilf <max@ramax.com>'s patch in main.dtml to correctly show data.
-item 1 and 2, thanks to hazmat and TresEsques in #zope

v 0.5
-better start and end pattern search and match.  Gregory Dudek (Gregory Dudek <dudek@abatonix.dyn.dhs.org>
said: """
Hi. I made some changes to your great zope Product KebasData.
Most important, I changed it so that the end-of-extract
marker is not looked for until AFTER the extract begins (i.e. the
point where the start-of-extract marler is found).
"""
-better error reporting (for start/ end pattern not found)

v 0.4
-proxy support.  Thanks to "Mark W. Alexander" <slash@dotnetslash.net>.
The KebasData instance will first try and get the proxy from your environment.
v0.3.0
-reset method to reset Kebas Data instance before exporting from Zope-2.3.x,
and importing to Zope-2.4.x.  Otherwise, traceback "cant import re object" or
somesuch.

v0.2.0
-we can now get the parameter list for POST method from the query string.  This will enable dynamic call on get_matched
-change security assertion from "View management screens" to "Access Kebas Data get_managed".
-added a HowTo - http://www.zope.org/Members/kedai/useKebasData

v 0.1.0
-added Test tab.  This view wil hopefully tell what's been grabbed, whether
the URL returns anything or not.  Meant for debugging.
-some security declaration

v 0.0.1a5
-added request method. we can now grab from url, url that support GET method, and url that use POST method.
 with POST erquest method, we need to supply the parameter lists one per line; key and value.
-removed unnecessary messages

v 0.0.1a4
-with TheJester's help, i finally got the best way to view Kebas Data instance.
-added index_html so that instance can be accessed thru url e.g http://myserver.com/kebasinstance/
accidentally removed it in v 0.0.01a3
-better error handling.  Now, errors/tracebacks will be returned.
-included demo zexp (dll.zexp) that can jump start your site.  be sure to initialize all Kebas Data instance.  Also, mail me your examples/demo so that i can include it.

v 0.0.1a3

-added freshness check.  you can now specify the object freshness, in seconds.
every call to view() will then check whether a get_matched() should be execeuted.

-change dtml/* stuff 

v 0.0.1a2
-you can pass a url to get_matched() call
-new attribute added (render_method) as a default