----- Original Message -----
From: "Brion Vibber" <brion(a)pobox.com>
Individual revisions can be deleted manually, and if there's a need to do
it frequently, a function could be made for it. However there are broadly
three classes of things we'd want to remove:
* A newly created article that's cut-n-pasted from
a non-FDL source
without permission. If no good material has been added it can simply be
deleted; if the cut-n-paste was _replaced_ bodily with good material, it's
simple to delete the 'bad' revisions off the beginning and let the
'good'
article's history stand on its own as if it were created separately.
Yes, there are lots of pages that are/have been "blanked" because of
"copyright violations" and then they converted into a stub. Getting
rid of the original page with the violation info would be a useful thing
to do routinely. A page that has a "impure" page history could be
listed somewhere when someone finds it, and then the original page
could be blanked but the rest of the page history remains without the
violation (but possibly the mention that it happened). There must be
_thousands_ of pages like this already as one look at the VfD/copyvio
page shows that people are replacing violations with stubs all the time.
It may be a job finding all of them.
* An existing article that has had cut-n-paste added
to it, and
immediately reverted. Individual revision(s) can be blanked easily.
Can we do that right now as administrators (I was wondering what the
delete this page function means when it show up on prior versions, is
this what it is for or does it delete the whole page, not just the current
revision on screen?) or is that only something that developers can do?
I know that the folks on the Japanese Wikipedia are concerned
about that too, see Tomo's comment at the bottom of this page:
http://www.wikipedia.org/wiki/Wikipedia:Copyright_violations_on_history_pag…
* A longstanding problem, where cut-n-pasted text has
been intermingled
with original text over many revisions.
This third case is the tricky one. How best to handle it?
That is the one where the copyright problem may be removed. If Wikipedia
is about facts then if the text is edited so that it can't be
differentiated, then
it may not be an infringement. Probably best to keep editing it and leave
the "paper trail" to prove it has been edited beyond recognition. Probably
pages that have had only some but significant editing should be listed
somewhere to encourage more editing. It would be a shame to remove
text that has been edited over many generations. The diff function can
certain help one rewrite the article so that the content remains but the
form and structure has changed.
Is there any possibility of putting a notice on the history pages to show
they are there for _archival purposes only_? Do we really want to
encourage people to use older versions of articles anyway? If they link
to the article (as they should) then the link will bring one to the newest
version, not an infringing history version of it. This might be a much
simpler solution than creating a delete function (though it is encouraging
people to look into the histories for the infringing material) and the
delete
function will certain clean up these old infringements.
(Do we need a wikilegal-l list?)
I hope not! But maybe it is not a bad idea.
alex756