I just blocked 3 IPs of a vandal similar to if not the same as the previous
one. Both events involved more than one IP being used by a single individual.
Please see http://www.wikipedia.com/wiki/Special:Ipblocklist
One of the messages this person left was:
"ADMIN, you gunna block every IP im on? ROFL.. don't waste your time ;) <3
proxies ;)"
I know, we could whois the IPs and complain to the ISP but my partner is a
tech support supervisor for Earthlink and he says that Earthlink does not act
on these types of complaints. I don't know if this vandal's ISP will be so
ambivalent. Most ISPs seem to not care what their users do so long as EULAs
are not violated.
I'm not sure what to do -- this person will almost certainly come back again.
Any suggestions?
Jimbo you are an ISP owner, what do you think?
--mav
On Monday 29 July 2002 12:01 pm, you wrote:
> I actually do already have a "lock the database" button available
> to developers; maybe I should make that available to sysops as well
> (as long as "unlock" is as well, of course)?
>
> I also need to start thinking about some back-end stuff like the cron
> job for making more frequent backups.
Cool. This would of course have to be approved by our benevolent dictator,
er, benefactor Jimbo. Is this feature bug-free enough to trust mere sysops
with?
Hopefully such a panic button will /never/ have to be used. But having a
"protect all pages" button will give me, and I suspect others, some peace of
mind.
However, there will have to be severe limitations on any use of this function
though. If a sysop and the community can reasonably keep up with such an
attack this feature should /not/ be used. But 10 vandalized articles or more
a minute over multiple minutes would begin to be very taxing, esp. if a bunch
of different IPs are being used by the vandal.
--mav
On Monday 29 July 2002 01:58 am, Imran wrote:
> Most EULAs are so broad that it's hard not to violate them, bring to
> the ISP the exact sections of their AUP that has been violated,
> include all IPs involved with timestamps of access.
>
> Also follow up with a phone call if necessary.
>
> If that fails we should set up a mechanism such that anyone
> accessing the webpage from that ISP has a note attached to the
> top saying "Due to a failure by $ISPName$ to take action against
> an attack against wikipedia we are considering removing write
> access from users coming from this ISP. Please help us to avoid
> doing this by contacting your ISPs abuse department and making
> your views on the matter felt to them".
This sounds reasonable to me. I hope we will never have to make good on such
a threat though -- I know I get majorly pissed when I can't send an email to
somebody because my ISP has been blacklisted by my recipient's ISP because
somebody on my ISP is a low-life spammer.
--mav
On Monday 29 July 2002 01:58 am, you wrote:
> Thinks:
> It would be interesting to know what fraction of Wikipedia users come
> from ISPs with/without valid reverse DNS lookup set up for their IP
> addresses. If the fraction from clueless ISPs is small (and it should
> be), then we could
>
> * ban customers of crap no-reverse-delegation ISPs from editing by
> default (which will capture a lot of "grey" IP addresses, too), and
> * offer suitably privileged users the option to issue the appropriate
> warning (as above) on a per-ISP basis, based on the reverse lookup domain.
>
> Of course, things should never get this far in the first place, but
> having the tools in reserve would be nice.
>
> Neil
Whoa! I don't think we are at that point yet, but I will file your email away
for future reference. Our vandal problem is simply not bad enough yet to
warrent such drastic counter-measures. However I see no harm in looking ahead
to that eventuality, doing some research, and preparing the systems needed --
but I don't think there is any urgency yet.
What I do fear is some script kiddy with a couple dozen rotating proxies and
a ship-load of bots flooding the database with junk and overwriting 20
articles a minute. A panic button to lock-down the site would then be nice
(Sorry, I can't protect pages fast enough). Then that would give a sysop the
time needed to block all the IPs of the vandal. But again, I don't think we
are at that point yet.
We should think about different ways to optionally protect wikipedia from
these types of more sophisticated attacks. Although I would be /very/ wary of
categorically excluding ISPs when there have been no vandals from that ISP
(although that may be part of any panic button). ISPs that fail to act on our
complaints of documented vandalism are fair game as far as I am concerned
though.
--mav
On Monday 29 July 2002 01:58 am, you wrote:
> I would not hesitate here. If you have two notations for the same thing
> people who want to write have to know them both because they are going to
> encounter them and therefore have to know what they mean and if they are
> the same or not. Having one notation is simpler for everybody.
>
> -- Jan Hidders
I agree -- we should eventually move away from duplication and just have wiki
markup in cases where we now have both that do the /exact/ same thing
(although I don't change <b> to ''' whenever I see them like some people do).
I would also like to see an ultra easy and human readable and intuitive wiki
table markup enacted for Wikipedia (maybe the software can automatically
convert HTML to exact wiki equivalents when a page is saved... Then new
contributors with an HTML background can start off running.). .
The trouble with raw HTML is that human eye skips over HTML tags since they
add visual complexity and confussion to text. I have less sympathy for HTML
hackers who have to learn a simple new markup than for the non-HTML coding
majority of contributors who are intimidated by HTML and are unable to
effectively read heavily HTML tagged text (myself included -- even though I
am responsible for a number of HTML tables in wikipedia).
Wiki is dead easy to learn -- HTML coders of all people will quickly adapt.
But all this is really not nearly as important as clearing the bug list.
Here is a link to the wiki table feature request:
http://sourceforge.net/tracker/index.php?func=detail&aid=584459&group_id=34…
--mav
Concerning the page :[[Wikipedia:Article deletion log]] and several other
willy-nilly uses of the word "article".
Why in the world would we ever want to delete an article? I know we don't
delete articles (as in encyclopedia articles) but others may not be aware of
this. It is already difficult enough to explain the difference between
certain types of wikipedia-specific pages, mere definition pages, vandalized
pages and an actual encyclopedia articles without added confusion.
As Larry often said, every page in wikipedia is a page but not every page is
an article (or something like that). We needn't add un-needed confusion by
loosely using the word "article" for every page. This also effects the
proposed new wording of the front page which will read "anyone can edit any
article" -- which is a completely true statement if we stick with Larry's
definition and the criteria used by the statistics to estimate the number of
articles in the database.
Our statistics currently reads (with added emphasis by me):
"There 63416 total /pages/ in the database. This includes "talk" pages, pages
about Wikipedia, minimal "stub" pages, redirects, and others that probably
don't qualify as /articles/. Excluding those, there are 35128 pages that are
probably legitimate /articles/."
Why then does the save button always say "save article"?
--mav
I am against the idea. People seeing a tagged article on some topic they feel
strongly about would be more likely to try and insert their own views to
"balance" the topic. Rather than resolving any controversies, tagging
articles as controversial would probably serve to exacerbate the controversy.
Danny
Eclecticology wrote:
>What I really miss is not being able to use the back button to get back
>to where I was. It now forces a reload whenever I do so. I've noticed
>it most with Recent Changes, but I've also had it come up while editing
>a page. My habit has been to use search to go to verify a page name
>when I'm trying to create a link from the page I'm editing. Now when I
>back-button from the search results I get a brand new edit page with my
>partial edits deleted.
I've learned to *always* press "Preview" before leaving the edit page.
Then the only problem is if the connection fails for some reason
(probably a problem on my end, since it occurs with other sites),
in which case backtracking from the error page loses the most recent edit.
-- Toby Bartels
<toby+wikipedia-l(a)math.ucr.edu>
In response to my sourceforge feature request, LDC has already changed the
name of the deletion log and to [[wikipedia:deletion log]] and has replaced
"save article" with "save page". Thanks LDC!
So my previous post is probably mostly if not totally moot. Are there any
remaining inconsistencies that should be fixed in this regard?
--mav
Daniel wrote:
>Also, the existence of such a page, and some links to it, doesn't mean all
>possible non-NPOV pages at Wikipedia would have to ble tagged, in some giant
>operation...
>
>So:
>
>* A controversy flag doesn't really impede an article.
>* It doesn't add work for anyone.
>* It saves work and time for both sides in a dispute.
>* It "lightning rods" away undue emotional, immediate rewrites of articles.
>* It signals to the audience that not everyone on Wikipedia is satisified
> with the way the particular article is presented. (IMO, very valuable
> information.)
>* It would affect a very limited number of articles.
>
>Is it still such a bad idea?
>
>-- Daniel
I think so, sorry. :-)
*It's more complicated than what we're doing already, and for uncertain benefit.
*People will just have to get over their immediate emotional responses if they wish to write NPOV about controversial topics--that or avoid writing about those topics. Vigilance in this area is required, and yes, I've had my own transgressions on this front.
*So far, talk pages have been used to indicate dissatisfaction with an article's content; that has worked with considerable success.
*It would affect a very large number of articles, as my and your ideas of what is controversial are probably quite different. Suppose I cite "gun rights" and "capitalism", User:pRobertson cites "abortion" and "separation of church and state", someone else cites "taxes" and "Indonesia".... I fail to see how it's useful. It seems likely to become a convenient axe-grinding tool for partisans. And, really, does wikipedia need another list?
Just my opinion, of course, and--as always--feel free to disagree & elaborate. :-)
kq