[Setup] Re: Permissions

Raphael Ritz r.ritz at biologie.hu-berlin.de
Tue Jan 3 12:23:39 UTC 2006


Philip Kilner wrote:

[..]

> Yup - as a Zope and Plone user, I'm painfully aware that I do not know
> enough about the CMF that lies between them.

Like most of us I'm afraid ... :-(

>>PS: If people think Plone content ought to have a
>>review_state I simply disagree.
> 
> Not wanting to be awkward, but just to understand your reservations,
> would you see a problem in requiring the "review_state" attribute but
> supporting a value of "null" or better "irrelevant" or "immune" or some
> such?

In Python this is None

> (Can't think of the right word - as an RDBMS man, "null" makes me
> shudder but I guess it is what I'm really trying to say!) IOW, is it the
> attribute or it's values that trouble you?

Neither one nor the other. It's simply that Plone should be
robust enough not to break if a content object has no
review_state. (I think it does so by now)

There are two common cases where this can happen:
1. objects that aren't subject to any workflow (like
discussed here already).
2. objects where the state variable is not called 'review_state'
but something else. This case is currently not handled
correctly in Plone in that transitions from those workflows
aren't offered in the UI as soon as no workflow provides
the 'review_state' state variable. (but I admit that this
might be a rare situation and people messing with workflows
to the extend that this happens typically can live with it
or work around; if you want to see this check out my LockingWorkflow
and assign it once together with one of Plone's workflows and
once exclusively and you'll see the difference - or simply
rename the workflow state variable: this can even be
done in ZMI).

Raphael




More information about the Setup mailing list