Over the past few weeks, OTRS has seen quite a few messages concerning companies that are putting information about themselves onto Wikipedia for advertising purposes, insisting that it is their right to do this. An article in an online SEO (search engine optimization) magazine described how to mine wikipedia to get web traffic. We have had emails from such diverse groups as talent agencies (we will take the copyright off our own website, as long as it is included in Wikipedia), a Dominatrix, a vaporizer (I have no choice but to keep inserting my links on your site so as to fend off the competitors), and many others. In fact, this appears to be a growing trend in Wikipedia, as is evidenced by similar phone calls to the office (I did not write the article about my, my PR firm wrote it, and I paid them good money so you can't take it off). Shoppingtelly.uk has written that as long as we allow links to the BBC, they will insist on their "rights" to put links to their site on Wikipedia.
This is a worrying trend on the English Wikipedia which raises issues of POV, notability, and verifiability. Ironically, we do not allow paid advertising, but we are buckling when people use our site in order to get free advertising.
I do not know the solution to this problem--several have been raised, but in my mind none is completely satisfactory. I am simply posting this here in the hope that it will elicit discussion and, perhaps, a real policy decision to counter this worrying trend.
Danny
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
daniwo59@aol.com wrote:
Over the past few weeks, OTRS has seen quite a few messages concerning companies that are putting information about themselves onto Wikipedia for advertising purposes, insisting that it is their right to do this. An article in an online SEO (search engine optimization) magazine described how to mine wikipedia to get web traffic. We have had emails from such diverse groups as talent agencies (we will take the copyright off our own website, as long as it is included in Wikipedia), a Dominatrix, a vaporizer (I have no choice but to keep inserting my links on your site so as to fend off the competitors), and many others. In fact, this appears to be a growing trend in Wikipedia, as is evidenced by similar phone calls to the office (I did not write the article about my, my PR firm wrote it, and I paid them good money so you can't take it off). Shoppingtelly.uk has written that as long as we allow links to the BBC, they will insist on their "rights" to put links to their site on Wikipedia.
This is a worrying trend on the English Wikipedia which raises issues of POV, notability, and verifiability. Ironically, we do not allow paid advertising, but we are buckling when people use our site in order to get free advertising.
I do not know the solution to this problem--several have been raised, but in my mind none is completely satisfactory. I am simply posting this here in the hope that it will elicit discussion and, perhaps, a real policy decision to counter this worrying trend.
One good solution is liberal usage of the spam blacklist and an increased awareness that the spam blacklist exists. We can shut these guys down cold ... if anytime a spam URL is added it is acted upon and added to the blacklist.
The problem is that the blacklist is on Metawiki, not enwiki, so administrators such as myself can't do anything about it on our own and have to go running to a Meta admin. Maybe we could add some sort of spam blacklist queue on Enwiki that is regularly viewed by meta admins? Or more liberally giving out meta adminship might help too.
- -- Ben McIlwain ("Cyde Weys")
~ Sub veste quisque nudus est ~
On 4/30/06, Ben McIlwain cydeweys@gmail.com wrote:
One good solution is liberal usage of the spam blacklist and an increased awareness that the spam blacklist exists. We can shut these guys down cold ... if anytime a spam URL is added it is acted upon and added to the blacklist.
This is necessary but not sufficient.
In many places our editors are asleep at the switch when it comes to external links... once there are 20 externals on an article, no one.. not the readers, not the editors, are going to follow all of them. Some of these articles have no regular editors, and passing editors often don't want to anger a regular editor by removing a favorite link... and when there are regular editors, they sometimes ignore external additions.. additional links are a lot less cut and dry than "penis penis penis nazi batman" vandalism.
An example, I removed about 45 external links from [[Genealogy]] the other day after an admin pointed out that someone needed to clean it up. Another 45ish still remain in the external link section. Some were removed because they were listed 3 or more times. Some were removed because they were just sites of advertisements with little to no information. One obviously tried to install spyware. One was a mirror of another site (with lots of added adsense). Others simply required you to pay before you could determine if they had anything useful. ... just about all of these links (minus the spyware perhaps) could have been easily found using google.
I probably managed to remove a few good link as well, but it was clear that there had been almost *zero* edit oral oversight of these links. When we allow articles to have large lists of links without solid editorial oversight, we are no more useful than google... and worse: we are more subject to manipulation than the authors of the links.
It is my belief that almost every page with more than 10-20 externals is in the same position of low to no oversight on the externals, and many with more than 5 externals.
About 38,000 pages on enwiki have more than 10 externals, 14,000 have more than 20. About 118,000 have more than 5.
The problem is that the blacklist is on Metawiki, not enwiki, so administrators such as myself can't do anything about it on our own and have to go running to a Meta admin. Maybe we could add some sort of spam blacklist queue on Enwiki that is regularly viewed by meta admins? Or more liberally giving out meta adminship might help too.
There are many meta admins in the admin channel on IRC, I promise that you are never far from one. :)
The problem is that the SBL isn't useful for an infinite number of one-off spammers.
On 4/30/06, Gregory Maxwell gmaxwell@gmail.com wrote: [snip]
I probably managed to remove a few good link as well, but it was clear that there had been almost *zero* edit oral oversight of these links. When we allow articles to have large lists of links without solid editorial oversight, we are no more useful than google... and worse: we are more subject to manipulation than the authors of the links.
It is my belief that almost every page with more than 10-20 externals is in the same position of low to no oversight on the externals, and many with more than 5 externals.
A list of pages which I suspect of having low edit-oral-ness (whatever that is) of their external links is now available.
http://en.wikipedia.org/wiki/User:Gmaxwell/externalmania
The page doubles as a browser stress test. Enjoy.
On 01/05/06, Gregory Maxwell gmaxwell@gmail.com wrote:
A list of pages which I suspect of having low edit-oral-ness (whatever that is) of their external links is now available.
Looking at that page i notice that lots of them are lists (in fact the majority). Many of these lists are complete and will not need editing ever again (for historical things) or very rarely into the future. Perhaps a lost of these should be semi-locked or is that unwiki?
paz y amor, -rjs.
-- DO NOT SEND ME WORD ATTACHMENTS - I *WILL* BITE! http://www.gnu.org/philosophy/sylvester-response.html
Hit me: http://robin.shannon.id.au [broken] Jab me: robin.shannon@jabber.org.au Upgrade to kubuntu linux: http://releases.ubuntu.com/kubuntu/breezy/ Faith is under the left nipple. -- Martin Luther
On 4/30/06, Robin Shannon robin.shannon@gmail.com wrote:
On 01/05/06, Gregory Maxwell gmaxwell@gmail.com wrote:
A list of pages which I suspect of having low edit-oral-ness (whatever that is) of their external links is now available.
Looking at that page i notice that lots of them are lists (in fact the majority). Many of these lists are complete and will not need editing ever again (for historical things) or very rarely into the future. Perhaps a lost of these should be semi-locked or is that unwiki?
I didn't see any lists that were complete, but I haven't looked closely. I know that a good number of pages have already been fixed based on this list.. even the ones that have lots of link that look good, many of them still have a bad link or two lurking, you have to check them all.. and no one is, which was my point. :)
Gregory Maxwell wrote:
On 4/30/06, Ben McIlwain cydeweys@gmail.com wrote:
The problem is that the blacklist is on Metawiki, not enwiki, so administrators such as myself can't do anything about it on our own and have to go running to a Meta admin. Maybe we could add some sort of spam blacklist queue on Enwiki that is regularly viewed by meta admins? Or more liberally giving out meta adminship might help too.
There are many meta admins in the admin channel on IRC, I promise that you are never far from one. :)
Agree totally; I'm a meta admin, and I'm happy to put stuff on the list, as long as it is really spam. I wouldn't have a problem with having a "Requests for blacklisting" page, or with checking it regularly (though I'm sure it'd be overwhelmed in short order; perhaps one time [just one] that we could make a request page sysop-only [via protection]?).
Also, I support Brion's suggestion; I trust that after all the work he's done, he has both the best interests of the site and the best understanding of what's going on (technically). Go get 'em, Brion <ties scarf around Brion's sword as he rushes into battle with the spammers> :-D
Essjay
Hoi, Just saying that we should liberaly ban "spam" is too easy because the devil is in the details. A few days ago there was an article on slashdot suggesting that an American chain manipulates its Wikipedia article. Some people will agree some won't. When somebody calls and says that he has "bought" his wikipedia article, it is easy; you cannot buy a particular type of content on Wikipedia. It does however not necessarily mean that such an organisation does not have its relevancy. A kneejerk reaction is not necessarily always the right thing.
On the other hand you have people who have a beef with an organisation and insist on being negative about an organisation and do not allow for a NPOV article about such an organisation. In my opinion this is another side of the same coin.
There are also the people who do nothing but adding links to their info site and, as always there are people who think this a good idea and people who opine that it detracts from the Wikipedia article ....
I would in conclusion say, in case of obvious spam at least remove the external links and when the article is plain advertisement, make it NPOV or delete.
Thanks, GerardM
On 5/1/06, Ben McIlwain cydeweys@gmail.com wrote:
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
daniwo59@aol.com wrote:
Over the past few weeks, OTRS has seen quite a few messages concerning companies that are putting information about themselves onto Wikipedia for advertising purposes, insisting that it is their right to do this. An article in an online SEO (search engine optimization) magazine described how to mine wikipedia to get web traffic. We have had emails from such diverse groups as talent agencies (we will take the copyright off our own website, as long as it is included in Wikipedia), a Dominatrix, a vaporizer (I have no choice but to keep inserting my links on your site so as to fend off the competitors), and many others. In fact, this appears to be a growing trend in Wikipedia, as is evidenced by similar phone calls to the office (I did not write the article about my, my PR firm wrote it, and I paid them good money so you can't take it off). Shoppingtelly.uk has written that as long as we allow links to the BBC, they will insist on their "rights" to put links to their site on Wikipedia.
This is a worrying trend on the English Wikipedia which raises issues of POV, notability, and verifiability. Ironically, we do not allow paid advertising, but we are buckling when people use our site in order to get free advertising.
I do not know the solution to this problem--several have been raised, but in my mind none is completely satisfactory. I am simply posting this here in the hope that it will elicit discussion and, perhaps, a real policy decision to counter this worrying trend.
One good solution is liberal usage of the spam blacklist and an increased awareness that the spam blacklist exists. We can shut these guys down cold ... if anytime a spam URL is added it is acted upon and added to the blacklist.
The problem is that the blacklist is on Metawiki, not enwiki, so administrators such as myself can't do anything about it on our own and have to go running to a Meta admin. Maybe we could add some sort of spam blacklist queue on Enwiki that is regularly viewed by meta admins? Or more liberally giving out meta adminship might help too.
Ben McIlwain ("Cyde Weys")
~ Sub veste quisque nudus est ~ -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2.2 (MingW32)
iD8DBQFEVTpBvCEYTv+mBWcRAj0FAKCLBuJQkwRQDTeRq8yrJeSTkKvJrwCglSB0 NGzWKJubvHhu2MQh20aU0rE= =ncLH -----END PGP SIGNATURE----- _______________________________________________ foundation-l mailing list foundation-l@wikimedia.org http://mail.wikipedia.org/mailman/listinfo/foundation-l
daniwo59@aol.com wrote:
Over the past few weeks, OTRS has seen quite a few messages concerning companies that are putting information about themselves onto Wikipedia for advertising purposes, insisting that it is their right to do this. An article in an online SEO (search engine optimization) magazine described how to mine wikipedia to get web traffic. We have had emails from such diverse groups as talent agencies (we will take the copyright off our own website, as long as it is included in Wikipedia), a Dominatrix, a vaporizer (I have no choice but to keep inserting my links on your site so as to fend off the competitors), and many others. In fact, this appears to be a growing trend in Wikipedia, as is evidenced by similar phone calls to the office (I did not write the article about my, my PR firm wrote it, and I paid them good money so you can't take it off). Shoppingtelly.uk has written that as long as we allow links to the BBC, they will insist on their "rights" to put links to their site on Wikipedia.
This is a worrying trend on the English Wikipedia which raises issues of POV, notability, and verifiability. Ironically, we do not allow paid advertising, but we are buckling when people use our site in order to get free advertising.
I do not know the solution to this problem--several have been raised, but in my mind none is completely satisfactory. I am simply posting this here in the hope that it will elicit discussion and, perhaps, a real policy decision to counter this worrying trend.
Thank you for sharing this problem with us.
"Vaporizer" sounds too much like a Dalek; "You will be exterminated" :-)
Not even dedicated and recognized Wikipedians have a "right" to dictate content, so someone who is wotking through a PR firm shouldn't expect any better. If the firm told him such a thing there are issues that he will need to work out with them.
The falacy in shoppingtelly's arguments is based in the fact that the BBC does not itself add links to itself. If shoppingtelly has anything useful on its site someone someday will make a proper unbiased link there. It's just unlikely to be to their advertising material any more than it is to advertising material on the BBC.
The premise in your inquiry is that we are dealing with clear cases of spam. There is probably no one way to identify this stuff ahead of time, and most links will still take exploring to learn what's going on. I don't see any need to "buckle" when the spamming is obvious.
Ec
2006/5/1, daniwo59@aol.com daniwo59@aol.com:
I do not know the solution to this problem--several have been raised, but in my mind none is completely satisfactory. I am simply posting this here in the hope that it will elicit discussion and, perhaps, a real policy decision to counter this worrying trend.
I think that the solution should be to remove many more links than are done now. When I read that Gregory Maxwell removed 45 links from a page and kept another 45, in my opinion he has been rather reserved - rare is in my opinion the page that needs more than 3 links. Wikipedia is not a link collector.
I'll be going through pages on my watchlist and see whether there are any inappropriate links and remove them, then check whether they exist elsewhere on Wikipedia if they appear to be not on the most likely page (which would make it seem they have been spammed).
-- Andre Engels, andreengels@gmail.com ICQ: 6260644 -- Skype: a_engels
Andre Engels wrote:
I think that the solution should be to remove many more links than are done now. When I read that Gregory Maxwell removed 45 links from a page and kept another 45, in my opinion he has been rather reserved - rare is in my opinion the page that needs more than 3 links. Wikipedia is not a link collector.
I'll be going through pages on my watchlist and see whether there are any inappropriate links and remove them, then check whether they exist elsewhere on Wikipedia if they appear to be not on the most likely page (which would make it seem they have been spammed).
Perhaps this is a good time to push reenabling rel="nofollow" on en.wikipedia.org.
Domas and I visited the Google campus last week after the MySQL users conference, and chatted a bit with some folks from the search quality team. They assure us that, yes, rel="nofollow" *does* help prevent linkspam from having its intended effect and that, yes, at least some of the SEO folks end up giving up on sites because of it (though of course not all, and not immediately).
While this automated protection doesn't prevent eyeballs from seeing or following links, it does keep them from autospamming the search engines that we all rely on in our daily surfing.
The argument for turning it off centered on the idea that "good internet citizens" help "good sites" by linking to them, and that on en.wikipedia.org there's enough eyeballs to immediately remove linkspam.
Apparently that's just not so. Good internet citizens need to recognize this and keep our communal resources clean. Wikipedia is not a link farm!
-- brion vibber (brion @ pobox.com)
On 5/1/06, Brion Vibber brion@pobox.com wrote:
Perhaps this is a good time to push reenabling rel="nofollow" on en.wikipedia.org.
I don't much like the idea of effectively punishing thousands of perfectly fine sites because of a few people who pee in the proverbial public pool. How feasible would it be, in your opinion, to - add a timestamp to the externallinks table - only add nofollow to links which have been added recently (<7 days)?
That might be a reasonable middle ground. Yes, some spammers will get through, but that's to be expected. If we can keep the quality of links above the average on the web, then surely we should be treated like any other site on search engines.
As for the present situation, I think that more closely cooperating with existing link directories like DMOZ might help. If we can say, in the style of sister project links, "DMOZ has more hyperlinks about Foo", then we don't have to host these link lists on Wikipedia. I often find even large link lists very useful, but I would prefer more structured and carefully maintained ones than Wikipedia has, and that is exactly what projects like DMOZ are for.
This is more relevant to en.wikipedia.org than some others; de:, for example, has a guideline stating that five links should normally be enough, and that all links should be of the highest possible quality: http://de.wikipedia.org/wiki/Wikipedia:Weblinks
As a more gentle alternative, a substitution policy of link lists vs. links to link directories might make sense.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
I don't much like the idea of effectively punishing thousands of perfectly fine sites because of a few people who pee in the proverbial public pool.
How does turning on nofollow punish anyone? Nobody is entitled to free pagerank just because they've been listed on Wikipedia.
I see no good reason not to turn on nofollow for en. I also note that de has been nofollow for quite a while now.
Kelly
On Mon, May 1, 2006 15:11, Erik Moeller wrote:
On 5/1/06, Alison Wheeler wikimedia@alisonwheeler.com wrote:
But isn't that a "they did it so I will" argument that we all rv over the whole of WP already? If we are serious about sorting some of our problems
- and over-use of external links is but one of those - then we need to
start somewhere and put downa marker that we are dealing with the issue. We ahve to start somewhere ...
True. I remain unconvinced that indiscriminate use of the nofollow tag is the best way to do that, though. "If all you have in a hammer, everything looks like a nail." If all you have is a spam blacklist and a nofollow option .. I think Gregory's systematic queries point in the direction of the right solution. Perhaps a few new special pages could help us deal with external link issues more systematically. Hmm .. I think this is a job for Magnus or Rob! ;-)
Erik
(ignore previous ... mail uissues! response was actually ....)
On Mon, May 1, 2006 15:11, Erik Moeller wrote:
True. I remain unconvinced that indiscriminate use of the nofollow tag is the best way to do that, though. "If all you have in a hammer, everything looks like a nail." If all you have is a spam blacklist and a nofollow option .. I think Gregory's systematic queries point in the direction of the right solution. Perhaps a few new special pages could help us deal with external link issues more systematically. Hmm .. I think this is a job for Magnus or Rob! ;-)
ok. Let me come at this from the other direction then! What *benefit* does it do the Wikipedia and WMF projects to supply 'free' link value to other sites. It doesn't help us sfaict, indeed it probably harms us as we can be seen as detrimental to the quality of search engine results as we - in effect - promote spam and other sites.
No-one has a "right" to a link fom us, just as no-one has a "right" to an article about them (or for that article to read the way they would like it to read.) I see this growing use of external links as bad for WP and something to be seen to in whatever way possible.
Alison
On 5/1/06, Alison Wheeler wikimedia@alisonwheeler.com wrote:
ok. Let me come at this from the other direction then! What *benefit* does it do the Wikipedia and WMF projects to supply 'free' link value to other sites.
If you take a look back in history, Wikipedia benefited (and still benefits) extremely from links and their impact on our PageRank. So it might be a little bit unfair to add nofollow tags to *all* of our links, just because it seems to be a simple solution to one of our many problems. The world wide web is not a one-way street ;)
It doesn't help us sfaict, indeed it probably harms us as we can be seen as detrimental to the quality of search engine results as we - in effect - promote spam and other sites.
This might be true, but only if there are more bad links than good ones. Without having counted them, I'm sure that this is not the case.
Btw, the German Wikipedia still has to deal with link spammers, even though nofollow is activated.
-- Arne (akl)
On 5/1/06, Arne Klempert klempert@gmail.com wrote:
If you take a look back in history, Wikipedia benefited (and still benefits) extremely from links and their impact on our PageRank. So it might be a little bit unfair to add nofollow tags to *all* of our links, just because it seems to be a simple solution to one of our many problems. The world wide web is not a one-way street ;)
No one is proposing all our externals be nofollowed. Only ones submitted by users.
But when talking about gains in readership, ... we get more by simply failing to do a good job (i.e. incorrect bioagraphies) ... So to that extent, I can agree with you: perhaps no follow will bring us lots of attention some day, when the actions of some spammer end up causing the next media-blitz. :)
But really, what is the goal of Wikipedia? It is to make a free content encyclopedia. It is not to be most popular website, or even the most read encyclopedia. It's not to help SEOs or search engines...
It doesn't help us sfaict, indeed it probably harms us as we can be seen as detrimental to the quality of search engine results as we - in effect - promote spam and other sites.
This might be true, but only if there are more bad links than good ones. Without having counted them, I'm sure that this is not the case.
Btw, the German Wikipedia still has to deal with link spammers, even though nofollow is activated.
Yes, it's not a complete solution.. no one suggested that. De Wiki's external density is 1/2 that of enwiki (De has an average of 1.3409 externals per article, while en has 2.4671 externals per article).
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
On 5/1/06, Arne Klempert klempert@gmail.com wrote:
If you take a look back in history, Wikipedia benefited (and still benefits) extremely from links and their impact on our PageRank. So it might be a little bit unfair to add nofollow tags to *all* of our links, just because it seems to be a simple solution to one of our many problems. The world wide web is not a one-way street ;)
No one is proposing all our externals be nofollowed. Only ones submitted by users.
Hairsplitting? Okay, then let's say "almost all".
But really, what is the goal of Wikipedia? It is to make a free content encyclopedia. It is not to be most popular website, or even the most read encyclopedia. It's not to help SEOs or search engines...
Let's try it the other way around. Links (real ones, without nofollow) are essential for the WWW. Does it comply with our goals to sabotage the web, only because we're too lazy?
-- Arne (akl)
On 5/1/06, Arne Klempert klempert@gmail.com wrote:
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
On 5/1/06, Arne Klempert klempert@gmail.com wrote:
If you take a look back in history, Wikipedia benefited (and still benefits) extremely from links and their impact on our PageRank. So it might be a little bit unfair to add nofollow tags to *all* of our links, just because it seems to be a simple solution to one of our many problems. The world wide web is not a one-way street ;)
No one is proposing all our externals be nofollowed. Only ones submitted by users.
Hairsplitting? Okay, then let's say "almost all".
A little, but not really... You talk about Wikipedia benefiting from links, but most of the places places who link to us have linked to us because the site operator has deemed Wikipedia to be useful.
This isn't the same as our links.
But really, what is the goal of Wikipedia? It is to make a free content encyclopedia. It is not to be most popular website, or even the most read encyclopedia. It's not to help SEOs or search engines...
Let's try it the other way around. Links (real ones, without nofollow) are essential for the WWW. Does it comply with our goals to sabotage the web, only because we're too lazy?
Oh, are you part of the misinformed who participated in the prior vote on this thinking that setting no follow makes links useless?
In any case, sabotage is a very strong word. Can you provide evidence that doing so would sabotage the web? Or is this just random conjecture? I think that nofollowing on Wikipedia will actually improve search results, but I can't support that with concrete evidence. What I can say with confidence and back with evidence is that nofollow will reduce the incentive to spam, it will mostly eliminate some sources of spam, that it with not impact the search rankings of already popular sites, and that it will reduce distrust on enwiki to some extent.
On 5/1/06, Arne Klempert klempert@gmail.com wrote:
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
But really, what is the goal of Wikipedia? It is to make a free content encyclopedia. It is not to be most popular website, or even the most read encyclopedia. It's not to help SEOs or search engines...
Let's try it the other way around. Links (real ones, without nofollow) are essential for the WWW. Does it comply with our goals to sabotage the web, only because we're too lazy?
-- Arne (akl)
That's an utterly misleading line of thought. If everyone in the world added nofollow to their links tomorrow, nothing terribly drastic would happen to the web. Maybe some search engines would break for a day or two. In the longer term, everyone turning on nofollow would be equivalent to everyone turning it off. The web got along just fine before nofollow was invented.
Google invented nofollow to provide them with additonal information about certain types of links. Apparently they'd prefer Wikipedia to employ it (which would also imply they don't think it's going to "sabotage the web"). There doesn't seem to be any suggestion that turning it on would hurt the encyclopedia, and some are suggesting it would help the encyclopedia.
Anthony
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
I don't much like the idea of effectively punishing thousands of perfectly fine sites because of a few people who pee in the proverbial public pool.
How does turning on nofollow punish anyone? Nobody is entitled to free pagerank just because they've been listed on Wikipedia.
I see no good reason not to turn on nofollow for en. I also note that de has been nofollow for quite a while now.
I'm with Kelly on this one.
It seems to me that if a sites pagerank is dramatically impacted by their addition to wikipedia, then the addition of that link to Wikipedia is to some extent a form of original research. Wikipedia does not exist to improve the popularity of other websites.
We also must consider the social impact: The knowledge that being linked from Wikipedia so dramatically impacts google results causes users to distrust the motivations of people who have added a link more than they would otherwise. It's poisonous.
Erik's concept a time delay isn't a new idea... it's one that has already been disregarded, at least for this application: even if we ignore the technical fun of using external links table at realtime, we're still left with the fact that it's pointless. For links that are removed in a short span of time the SEO gains no advantage (mirrors haven't had a chance to mirror, google hasn't had a chance to spider), our concern stems from links which remain due to a lack of editorial oversight.
Furthermore, On several Wiki's I'm involved with no follow has been a godsend, dramatically cutting spam in a short span of time.
I don't think that no follow will solve all problems, but it's a start. If it reduces external link spamming by 10%, then it will have made an improvement larger than is possible with any other simple method.
There are many technical measures which can and will be applied, but I don't see them as solving the root problem... because the root problem is editorial, it's social, it's not easy, and it's not something that the computers can wave away for us.
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
Erik's concept a time delay isn't a new idea... it's one that has already been disregarded, at least for this application: even if we ignore the technical fun of using external links table at realtime, we're still left with the fact that it's pointless. For links that are removed in a short span of time the SEO gains no advantage (mirrors haven't had a chance to mirror, google hasn't had a chance to spider), our concern stems from links which remain due to a lack of editorial oversight.
That depends entirely, of course, on how long the time delay is. It could be as long as 30 days. A link which survives that long, continuously, may very well be a good link that deserves to be there.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
Erik's concept a time delay isn't a new idea... it's one that has already been disregarded, at least for this application: even if we ignore the technical fun of using external links table at realtime, we're still left with the fact that it's pointless. For links that are removed in a short span of time the SEO gains no advantage (mirrors haven't had a chance to mirror, google hasn't had a chance to spider), our concern stems from links which remain due to a lack of editorial oversight.
That depends entirely, of course, on how long the time delay is. It could be as long as 30 days. A link which survives that long, continuously, may very well be a good link that deserves to be there.
You've fallen for the myth of editorial review...
What you say might be true if significant numbers of people people were checking links other than at the moment of addition (i.e. when they were the top most diff), but they don't.
If a link survives it's initial addition due to inattentiveness of the editors it will stay a long time. Must I beat this one into your head with wikiwide statistics on link lifetimes as well? :)
In the past I've computed the density of lifetimes of links which were inserted then later removed. It's quite computationally expensive to do so, which is why I haven't updated it any time recently. ... but I can, if I must do so to convince people that links are either removed right away or mostly not at all.
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
What you say might be true if significant numbers of people people were checking links other than at the moment of addition (i.e. when they were the top most diff), but they don't.
I would be interested in your data, even if it is old, on what the the average lifetime of removed links is. That might be a good cutoff point for nofollow. I should also note that just because this is currently the case doesn't mean that it will have to stay that way. There's probably lots of ways to generate useful reports for editors from the externallinks table.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
I would be interested in your data, even if it is old, on what the the average lifetime of removed links is. That might be a good cutoff point for nofollow. I should also note that just because this is currently the case doesn't mean that it will have to stay that way. There's probably lots of ways to generate useful reports for editors from the externallinks table.
Okay.
Externallinks is nice but of somewhat limited usefulness... It's already been used to despam enwiki of spam that can be detected from lots of duplicate links. We often use it to scan for SBLed stuff.. But it hits its limits.
To find link removal times I have to scan all revisions. Weee!
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
How does turning on nofollow punish anyone? Nobody is entitled to free pagerank just because they've been listed on Wikipedia.
I see two possibilities:
1) Wikipedia specifically has a very high influence on a site's ranking. In this case, turning off nofollow will alter the shape of the web in search engines which respect it. If the average quality of links in Wikipedia is higher than the average quality of links outside Wikipedia, the quality of these search engine results as a whole will deteriorate. This is not about entitlement, it is about using the influence we have responsibly.
OR
2) Due to mirrors, enabling or disabling nofollow has hardly any visible effect on search engine rankings. In that case, it would be a placebo effect.
As I said, I think timing the addition of external links and enabling nofollow only for recent additions might be a reasonable compromise.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
OR
- Due to mirrors, enabling or disabling nofollow has hardly any
visible effect on search engine rankings. In that case, it would be a placebo effect.
This hyphothesis would appear to be invalidated by our expirence removing spam from english Wikipedia. It is fairly common that a spam site will vanish off google right after removing it from enwiki, far too quickly for most of the mirrors to catch up.
For the few mirrors that we have a good enough relationship with to send a live OAI feed to, we can probably suggest that they nofollow externals coming in from Wikipedia. It's doubtful that they'd have a problem with doing this.
When it comes down to it though, most of the mirrors that have significant placement on the search engines are engaged in spamming SEO behavior themselves. If someone wants on one of those sites, there are probably even easier methods than inserting their content in wikipedia and waiting for it to be mirrored.
As I said, I think timing the addition of external links and enabling nofollow only for recent additions might be a reasonable compromise.
Only if someone can demonstrate some evidence that we are really good about removing bad links once they go in, and that anyone gives a darn if they get impact during the 24 hous their ill fated link survives. ...
Erik Moeller wrote:
- Due to mirrors, enabling or disabling nofollow has hardly any
visible effect on search engine rankings. In that case, it would be a placebo effect.
I can't agree here.
1) Mirrors running MediaWiki on a database dump already get rel="nofollow" by default unless they explicitly switched it off.
2) Screen-scrapers would be getting rel="nofollow" in their copies too -- something they currently DO NOT GET, thus multiplying the effect of linkspam.
While it's hypothetically possible for them to strip it, why would they bother? It makes no difference to them, they just want text content to bring search engine users to *their* ads and links.
-- brion vibber (brion @ pobox.com)
On 5/1/06, Brion Vibber brion@pobox.com wrote:
- Mirrors running MediaWiki on a database dump already get rel="nofollow" by
default unless they explicitly switched it off.
- Screen-scrapers would be getting rel="nofollow" in their copies too --
something they currently DO NOT GET, thus multiplying the effect of linkspam.
I did a quick sample of some mirrors. 2 out of 10 used nofollow. 5 out of 10, however, used some method of link obfuscation (such as using a JavaScript that gets the URL in reverse spelling as a parameter) or even link stripping. So it appears that at least some mirrors are actively trying to _prevent_ links to external sites from being indexed, probably as some method of SEO.
Erik
On Mon, May 01, 2006 at 04:40:23PM +0200, Erik Moeller wrote:
I see two possibilities:
- Wikipedia specifically has a very high influence on a site's
ranking. In this case, turning off nofollow will alter the shape of the web in search engines which respect it. If the average quality of links in Wikipedia is higher than the average quality of links outside Wikipedia, the quality of these search engine results as a whole will deteriorate. This is not about entitlement, it is about using the influence we have responsibly.
The quality of Wikipedia should be given more weight than to task of responsibly influencing search engines. If turning nofollow on leads to decrease of spammers effort, it would be good for Wikipedia quality.
- Due to mirrors, enabling or disabling nofollow has hardly any
visible effect on search engine rankings. In that case, it would be a placebo effect.
Most highly ranking mirrors are used for SEO spam and turn external links off completely (except their own links).
As I said, I think timing the addition of external links and enabling nofollow only for recent additions might be a reasonable compromise.
I don't think so. Trying to think like the enemy... What I would do if I wanted to insert spam link to e.g. Prague (I remove spam from that page frequently).
1. spam on the Prague page wouldnt not survive long enough, too many people watch it... thats futile 2. its necessary to insert the link to less visible pages, preferably watched by nobody. what about Prague districts, Prague public transport 3. still, there is the risk of RC patrol. solution? spam indirectly e.g., create "useful" and "noncommercial" pages which won't be removed by RC patrollers on sight (what about list of useful telephone numbers... yet another weather service... seen in practice ) 4. link from such lower-profile pages would be less effective because lower keyword density & lower inter-Wikipedia page rank 5. so... let's do some inter-wiki SEO, create wikilinks so the spammed wiki pages get higher interwiki rank, optimize the text for keywords...
results: -> spam will be on pages with little editorial review -> masked as noncommercial / informative pages -> editing with malicious agenda of influencing interlink structure
Jan Kulveit - User:Wikimol
On 5/1/06, Jan Kulveit jk-wikifound@ks.cz wrote:
On Mon, May 01, 2006 at 04:40:23PM +0200, Erik Moeller wrote:
I see two possibilities:
- Wikipedia specifically has a very high influence on a site's
ranking. In this case, turning off nofollow will alter the shape of the web in search engines which respect it. If the average quality of links in Wikipedia is higher than the average quality of links outside Wikipedia, the quality of these search engine results as a whole will deteriorate. This is not about entitlement, it is about using the influence we have responsibly.
The quality of Wikipedia should be given more weight than to task of responsibly influencing search engines. If turning nofollow on leads to decrease of spammers effort, it would be good for Wikipedia quality.
Ultimately Wikipedia is big enough that the search engines could easily manually override its nofollow suggestion anyway.
Which leads to a question - what would Google and the other search engines *want* Wikipedia to do? Anyone think they'd have a good shot at getting an official answer?
Anthony
Anthony DiPierro wrote:
Which leads to a question - what would Google and the other search engines *want* Wikipedia to do? Anyone think they'd have a good shot at getting an official answer?
They want us to turn on nofollow; at least their anti-SEO team does. :)
-- brion vibber (brion @ pobox.com)
On 5/1/06, Jan Kulveit jk-wikifound@ks.cz wrote: [snip]
- still, there is the risk of RC patrol. solution? spam indirectly e.g.,
create "useful" and "noncommercial" pages which won't be removed by RC patrollers on sight (what about list of useful telephone numbers... yet another weather service... seen in practice )
A point here.. this is a common form of spam. For example, at one point in time someone made mirror of all the linux documentation project content, slathered it with ads, and replaced links in Wikipedia. When I found it months later, their LDP mirroring was out of date... still no one else had fixed them.
We see people doing the same sort of thing with Wikipedia content.
- link from such lower-profile pages would be less effective because
lower keyword density & lower inter-Wikipedia page rank 5. so... let's do some inter-wiki SEO, create wikilinks so the spammed wiki pages get higher interwiki rank, optimize the text for keywords...
results: -> spam will be on pages with little editorial review -> masked as noncommercial / informative pages -> editing with malicious agenda of influencing interlink structure
And even when no one is doing this, editors are left paranoid about people doing it... because the incentives for corrupting out site in this manner are so great.
It really is miserable.
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
And even when no one is doing this, editors are left paranoid about people doing it... because the incentives for corrupting out site in this manner are so great.
It is highly unlikely that disabling nofollow would significantly affect such paranoia, since it is also unlikely that it would drastically cut spam.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
It is highly unlikely that disabling nofollow
should be, of course, "enabling nofollow". Damn double negatives.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
How does turning on nofollow punish anyone? Nobody is entitled to free pagerank just because they've been listed on Wikipedia.
I see two possibilities:
- Wikipedia specifically has a very high influence on a site's
ranking. In this case, turning off nofollow will alter the shape of the web in search engines which respect it. If the average quality of links in Wikipedia is higher than the average quality of links outside Wikipedia, the quality of these search engine results as a whole will deteriorate. This is not about entitlement, it is about using the influence we have responsibly.
OR
- Due to mirrors, enabling or disabling nofollow has hardly any
visible effect on search engine rankings. In that case, it would be a placebo effect.
As I said, I think timing the addition of external links and enabling nofollow only for recent additions might be a reasonable compromise.
None of Wikipedia's mirrors has anywhere near the pagerank weight that Wikipedia itself has. With the exception of a small number of already high-prominence sites, Wikipedia linkage is almost certainly bound to substantially alter the rank of a given page in search engine results.
We are writing an encyclopedia, not seeking to influence search engine result quality. It should not be a serious concern of ours whether our actions have a positive or negative impact on the value of Google's product; at the very least, we should not eschew making the a decision which is correct for us because that choice might have a detrimental effect on the quality of Google's product. Since leaving nofollow off encourages SEO spamming, and turning it on discourages SEO spamming, I see no reason why we should not disable SEO spamming.
The "delay" mechanism not only requires additional technical complexity but just encourages gaming by SEO people. Just turn it off.
Kelly
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
None of Wikipedia's mirrors has anywhere near the pagerank weight that Wikipedia itself has. With the exception of a small number of already high-prominence sites, Wikipedia linkage is almost certainly bound to substantially alter the rank of a given page in search engine results.
I agree with this now that it is also clear that mirrors are deliberately reducing external links. It is clear that Wikipedia's impact on the ranking of the sites it references must be significant. I continue to maintain that tagging all links indiscriminately as "nofollow" is essentially an admission of defeat; it is like saying: None of our links are worth spidering. We're just a wiki. Please, search engines, go to some other trusted sites, not to us.
I don't think our links are bad enough to justify such an explicit statement with all the consequences it has. And the continual pounding of "Wikipedia is an encyclopedia" does nothing to change the fact that it is a resource used by millions of people, that it is a resource which is embedded into a larger network known as the WWW, a network which itself is accessed by most people through portals and search engines before they arrive at a content source. It is in our interest, for our own project and for moral reasons, to help people to find high quality educational resources, as opposed to those which are promoted from within typically corporate, "trusted" frameworks such as AOL or MSN.
The fact that Wikipedia maintains community-created lists of topical links has always been, to me, one of its strongest points, one of the best ways to further explore and find relevant information on a topic. When I find an interesting and useful link, I often add it to Wikipedia so I can look it up again later, no matter where I am. The fact that any link to any resource has a chance and is treated, hopefully, consistently and fairly according to community-drafted policies distinguishes it radically from those "trusted" resources which only allow links which benefit the owner of the resource, or at least do not threaten to damage them or their advertisers.
Do we, essentially, want an invisible wall of separation between user-generated content and content from "trusted" sites? Or do we want a better solution? I don't like nofollow. It is a hack. It lacks imagination.
Erik
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
None of Wikipedia's mirrors has anywhere near the pagerank weight that Wikipedia itself has. With the exception of a small number of already high-prominence sites, Wikipedia linkage is almost certainly bound to substantially alter the rank of a given page in search engine results.
I agree with this now that it is also clear that mirrors are deliberately reducing external links. It is clear that Wikipedia's impact on the ranking of the sites it references must be significant. I continue to maintain that tagging all links indiscriminately as "nofollow" is essentially an admission of defeat; it is like saying: None of our links are worth spidering. We're just a wiki. Please, search engines, go to some other trusted sites, not to us.
I'd say it's more like saying "Don't assume this link is worth spidering just because we link to it." That's quite different from what you're saying.
I don't think our links are bad enough to justify such an explicit statement with all the consequences it has. And the continual pounding of "Wikipedia is an encyclopedia" does nothing to change the fact that it is a resource used by millions of people, that it is a resource which is embedded into a larger network known as the WWW, a network which itself is accessed by most people through portals and search engines before they arrive at a content source. It is in our interest, for our own project and for moral reasons, to help people to find high quality educational resources, as opposed to those which are promoted from within typically corporate, "trusted" frameworks such as AOL or MSN.
Ultimately I think the search engines are in the best position to determine how best to accomplish the goal of helping people find the resources they're looking for. And I think it's a very open question whether treating links from a wiki differently from links on a site with more tightly controlled access, helps a search engine guide people to high quality educational resources.
Of course if Wikipedia had the time and resources to dedicate to such a problem, a better solution would be to have a relatively small group of individuals go through and mark links as followable or not followable (high quality educational resource, or spam) manually. But I don't think that's reasonable.
Maybe putting a time-delay on the attribute is a good enough compromise. I don't know. IANA search engine guru.
The fact that Wikipedia maintains community-created lists of topical links has always been, to me, one of its strongest points, one of the best ways to further explore and find relevant information on a topic. [snip]
Sure, Wikipedia is great for that. But I really don't see how that's relevant to the nofollow tag, which is a tag for bots, and more specifically search engine spiders, not humans.
Do we, essentially, want an invisible wall of separation between user-generated content and content from "trusted" sites? Or do we want a better solution? I don't like nofollow. It is a hack. It lacks imagination.
On this point I agree. Maybe Google or one of the search engines looking to compete with them would be willing to consider a better tagging system. <a href=blah type=wiki firstadded=20060501195702> or somesuch. "Nofollow" is definitely a poor terminology. The tag should describe the link, not try to dictate to search engines what to do with it.
Anthony
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
Do we, essentially, want an invisible wall of separation between user-generated content and content from "trusted" sites? Or do we want a better solution? I don't like nofollow. It is a hack. It lacks imagination.
My personal opinion is that /currently/ I think nofollow has merit; I think we're extremely bad at handling linkspamming and website promotion in Wikipedia, and thus reducing the benefit might help.
However, if we became better at keeping track of what Wikipedia links to and getting rid of spam, I'd support turning nofollow off.
What would be useful steps to take to keep better track of these? Is there anything that can be done in software? Or procedure/process? External applications?
-Matt
On 5/1/06, Matt Brown morven@gmail.com wrote:
My personal opinion is that /currently/ I think nofollow has merit; I think we're extremely bad at handling linkspamming and website promotion in Wikipedia, and thus reducing the benefit might help.
However, if we became better at keeping track of what Wikipedia links to and getting rid of spam, I'd support turning nofollow off.
What would be useful steps to take to keep better track of these? Is there anything that can be done in software? Or procedure/process? External applications?
I've ductaped curl to spamassassin and can now rate the spammyness of websites. :) Now I just need to replace all the email header centric SA rules with web centric ones.
Because there are ~2,800,000 external links from the main namespace alone on enwiki (there is now a patch to mediawiki to set nofollow per namespace, I think no matter what we should nofollow outside the main ns), I see automated scanning as the only reasonable step right now to improve our link quality.
So I'm working on it..
On 5/1/06, Gregory Maxwell gmaxwell@gmail.com wrote:
Because there are ~2,800,000 external links from the main namespace alone on enwiki (there is now a patch to mediawiki to set nofollow per namespace, I think no matter what we should nofollow outside the main ns), I see automated scanning as the only reasonable step right now to improve our link quality.
So I'm working on it..
Excellent news! I'd hoped you were doing something like this.
And yes, I see no reason to not set nofollow in all namespaces except the article one.
-Matt
On 5/1/06, Kelly Martin kelly.lynn.martin@gmail.com wrote:
On 5/1/06, Erik Moeller eloquence@gmail.com wrote:
I don't much like the idea of effectively punishing thousands of perfectly fine sites because of a few people who pee in the proverbial public pool.
How does turning on nofollow punish anyone? Nobody is entitled to free pagerank just because they've been listed on Wikipedia.
I see no good reason not to turn on nofollow for en. I also note that de has been nofollow for quite a while now.
Kelly
I tend to agree, though I don't think nofollow will have much of an impact either way. Maybe it'll slightly decrease the incentive for vandals to add links, but probably not much.
One definite advantage I did see during the times that nofollow was on was that there was less fighting over what links to include, what not to include, and what to include through some kludge where the link wasn't clickable. Some people are on a mission to make sure Wikipedia doesn't link to certain websites which they find socially unacceptable (diploma mills are one example). Anything that reduces the temptation for people to engage in these wars is a good thing.
In the end, nofollow has little impact on the goal of Wikipedia - to create an encyclopedia.
Anthony
On 5/1/06, Brion Vibber brion@pobox.com wrote:
Domas and I visited the Google campus last week after the MySQL users conference, and chatted a bit with some folks from the search quality team. They assure us that, yes, rel="nofollow" *does* help prevent linkspam from having its intended effect and that, yes, at least some of the SEO folks end up giving up on sites because of it (though of course not all, and not immediately).
Also note that Wikipedia is different from most blogs and even wikis in that it's replicated across the web. Most of these mirrors aren't likely to turn on nofollow anytime soon.
Erik
*Most* sites don't permit drive-by addition of external links; our projects are different because we are open and, as such, we need to be more considerate of the general browsing public out there (imho) and nofollow would be a great assist to that. If we 'patrolled' the external links which we permitted to be active rather than passive there might be a different solution, but we can't (the workload would be enormous plus carrier liablity issues) and shouldn't.
On Mon, May 1, 2006 13:23, Erik Moeller wrote:
Also note that Wikipedia is different from most blogs and even wikis in that it's replicated across the web. Most of these mirrors aren't likely to turn on nofollow anytime soon.
But isn't that a "they did it so I will" argument that we all rv over the whole of WP already? If we are serious about sorting some of our problems - and over-use of external links is but one of those - then we need to start somewhere and put downa marker that we are dealing with the issue. We ahve to start somewhere ...
Alison
On 5/1/06, Alison Wheeler wikimedia@alisonwheeler.com wrote:
But isn't that a "they did it so I will" argument that we all rv over the whole of WP already? If we are serious about sorting some of our problems
- and over-use of external links is but one of those - then we need to
start somewhere and put downa marker that we are dealing with the issue. We ahve to start somewhere ...
True. I remain unconvinced that indiscriminate use of the nofollow tag is the best way to do that, though. "If all you have in a hammer, everything looks like a nail." If all you have is a spam blacklist and a nofollow option .. I think Gregory's systematic queries point in the direction of the right solution. Perhaps a few new special pages could help us deal with external link issues more systematically. Hmm .. I think this is a job for Magnus or Rob! ;-)
Erik
Andre Engels wrote:
I think that the solution should be to remove many more links than are done now. When I read that Gregory Maxwell removed 45 links from a page and kept another 45, in my opinion he has been rather reserved - rare is in my opinion the page that needs more than 3 links. Wikipedia is not a link collector.
Maybe not, but one of our Prime Directives is "cite your source". This often requires multiple external links because there are multiple sources.
One hopes that you would class this type of link as "necessary" but there seem to be people to whom this kind of thing is like a red rag to a bull and would set upon a radical link-ectomy without bothering to consider whether there is any conflicting policy.
Indeed a small minority of them might even consider rewriting policy to imply that they were correct all along.
2006/5/2, Phil Boswell phil.boswell@gmail.com:
Maybe not, but one of our Prime Directives is "cite your source". This often requires multiple external links because there are multiple sources.
One hopes that you would class this type of link as "necessary" but there seem to be people to whom this kind of thing is like a red rag to a bull and would set upon a radical link-ectomy without bothering to consider whether there is any conflicting policy.
Sure, if you cite your source, and this source is an internet page, it should be linked. That seems clear to me. However, such a link I would then expect to be recognizable as such - at least specifying that it is a source rather than just a link, but preferably also which part of the page it is a source for. I would not expect them to be between the other external links; they also go with different rules (for example, if a site with letters to and from presidents were linked on [[Harry S. Truman]] as an external link, I would expect the link to go to an overview of the available material from Truman, whereas if it were linked as a source, I would expect to be taken to the specific letter that was used as a source).
-- Andre Engels, andreengels@gmail.com ICQ: 6260644 -- Skype: a_engels
On 5/2/06, Phil Boswell phil.boswell@gmail.com wrote:
Andre Engels wrote:
I think that the solution should be to remove many more links than are done now. When I read that Gregory Maxwell removed 45 links from a page and kept another 45, in my opinion he has been rather reserved - rare is in my opinion the page that needs more than 3 links. Wikipedia is not a link collector.
Maybe not, but one of our Prime Directives is "cite your source". This often requires multiple external links because there are multiple sources.
One hopes that you would class this type of link as "necessary" but there seem to be people to whom this kind of thing is like a red rag to a bull and would set upon a radical link-ectomy without bothering to consider whether there is any conflicting policy.
Indeed a small minority of them might even consider rewriting policy to imply that they were correct all along.
Um. In this case I was only talking about links in the external link section of the page.
The main body has another 24, most of which should also be checked and possibly removed, and the ref section has 3.. all of which should probably stay.
I agree that 45 externals in the external section was too many, but removing things (even externals) on wikipedia is always a dangerous game... I'd rather remove too few, and have my changes stand... then too many and run afoul of a regular editor.
Hi. I was thinking a long time about this ... because on one hand Wikipedia is neutral, so why should companies like Siemens and Crysler have more right for an article than small start-ups? Who knows what this start-up is going to be in some years?
Why not restrict information to: Company xyz is in abc and produces bla bla bla ....
Leading then to a project like http://wikicompany.org?
I wish to point out that I am not associated with Wikicompany and that I write this on a personal title - I only see his project as a very valid one for the way it is being built. At least on such a project there would be no problem of a company describing its services and inserting links. Another advantage is that there would not be that "Logos are not allowed"-problem with Commons - it is an outside project, a good and free one. And in that way: no company gets a direct link on wikipedia, but just one link to the outside project where their contents then of course must comply with the rules there.
Well, this could solve the problem ...
My 2 cts.
Best,
Sabine Chiacchiera con i tuoi amici in tempo reale! http://it.yahoo.com/mail_it/foot/*http://it.messenger.yahoo.com
On 4/30/06, daniwo59@aol.com daniwo59@aol.com wrote:
Over the past few weeks, OTRS has seen quite a few messages concerning companies that are putting information about themselves onto Wikipedia for advertising purposes, insisting that it is their right to do this. An article in an online SEO (search engine optimization) magazine described how to mine wikipedia to get web traffic. We have had emails from such diverse groups as talent agencies (we will take the copyright off our own website, as long as it is included in Wikipedia), a Dominatrix, a vaporizer (I have no choice but to keep inserting my links on your site so as to fend off the competitors), and many others. In fact, this appears to be a growing trend in Wikipedia, as is evidenced by similar phone calls to the office (I did not write the article about my, my PR firm wrote it, and I paid them good money so you can't take it off). Shoppingtelly.uk has written that as long as we allow links to the BBC, they will insist on their "rights" to put links to their site on Wikipedia.
This is a worrying trend on the English Wikipedia which raises issues of POV, notability, and verifiability. Ironically, we do not allow paid advertising, but we are buckling when people use our site in order to get free advertising.
I do not know the solution to this problem--several have been raised, but in my mind none is completely satisfactory. I am simply posting this here in the hope that it will elicit discussion and, perhaps, a real policy decision to counter this worrying trend.
Danny
Are the current solutions failing? Do you have some examples of violations of Wikipedia policies which have remained in the encyclopedia for any significant amount of time before being removed?
It seems to be my take on things that while the phone calls might be annoying to you the rest is pretty well handled by the normal wiki processes, and will be best handled by improving those normal wiki processes. Closing down VFD so people have more time to focus on ensuring that the information in the wiki is NPOV and verifiable would probably increase the efficiency a lot too.
Anthony
wikimedia-l@lists.wikimedia.org