Wikipedia:Nofollow/Vote
Vote
[edit]- A remove vote supports no longer adding "rel=nofollow" to external links in the English Wikipedia.
- A keep vote supports the continued use of this attribute.
Keep: 55 (39%)
Remove attribute: allow external links from WP to affect Google PageRanks of their targets
[edit]- --SPUI (talk) 00:58, 12 Feb 2005 (UTC)
- Slowking Man 01:04, Feb 12, 2005 (UTC)
- Ævar Arnfjörð Bjarmason 01:07, 2005 Feb 12 (UTC) [Comment & responses moved to Comments section below by Jerzy(t) 22:33, 2005 Feb 12 (UTC)]
- brlcad 01:17, 2005 Feb 12 (UTC)
- Khym Chanur 01:41, Feb 12, 2005 (UTC)
- — Itai (f&t) 02:27, 12 Feb 2005 (UTC)
- Jordi·✆ 02:28, 12 Feb 2005 (UTC)
- bdesham 02:36, 12 Feb 2005 (UTC)
- If the system discriminated in any reasonable way between "good" and "bad" links, I would support. As it is now, I cannot. This is a not yet complete implementation of a not yet proven technology. - RedWordSmith 02:59, Feb 12, 2005 (UTC)
- Which unproven "technology" are you referring to? It's just standard HTML. -- Curps 03:15, 12 Feb 2005 (UTC)
- No it isn't. There is no "nofollow" in the W3C XHTML standard. Other people may use this value with a different meaning. Implementing Google's "nofollow" in Wikipedia means that me may break other bots that use this value in a different way. NSK 05:43, 12 Feb 2005 (UTC)
- Neither is there a "stylesheet" attribute defined for element rel—the value specified by the XHTML 1.0 Transitional DTD is unqualified CDATA. rel="nofollow" is perfectly valid XHTML1 (run it through the validator), and as the three most widely used search engines (not to mention nearly all content management systems) have committed to supporting it, it's unlikely that someone will implement "nofollow" with different semantics before it's codified by the W3C. ADH (t&m) 11:19, Feb 13, 2005 (UTC)
- No it isn't. There is no "nofollow" in the W3C XHTML standard. Other people may use this value with a different meaning. Implementing Google's "nofollow" in Wikipedia means that me may break other bots that use this value in a different way. NSK 05:43, 12 Feb 2005 (UTC)
- If it isn't standard yet, it very soon will be. Yahoo and MSN Search will support it. [1] -- Curps 14:41, 12 Feb 2005 (UTC)
- Which unproven "technology" are you referring to? It's just standard HTML. -- Curps 03:15, 12 Feb 2005 (UTC)
- Jerzy(t) 03:41, 2005 Feb 12 (UTC)
- FrankH 03:57, 12 Feb 2005 (UTC) Wikipedia links have to be one of the highest quality sources of link that can be used to improve search results. Let's help improve search results and find some other way to stop the edit bots.
- What Ævar said. This "solution" is ridiculous. BLANKFAZE | (что??) 04:37, 12 Feb 2005 (UTC)
- Geni 05:12, 12 Feb 2005 (UTC)
- Why exactly are we doing this? It's stated right there that this action won't deter spammers, so why ruin the PageRank of good resources? Mo0[talk] 05:18, 12 Feb 2005 (UTC)
- As annoying as the Russian spambot has been this week, nofollow is not the right solution. Rhobite 05:24, Feb 12, 2005 (UTC)
- Remove this attribute. It doesn't solve spam, it doesn't help Wikipedia and it affects good external links. NSK 05:40, 12 Feb 2005 (UTC)
- Spammers won't know about the nofollow, and even if they did they'd want to get their link mirrored on all the sites that use our content (and on Wikipedia itself) for the direct traffic. Tuf-Kat 05:52, Feb 12, 2005 (UTC)
- We should boost the reputation of good sites we link to. Spam is quickly removed. —Morven 06:19, Feb 12, 2005 (UTC)
- It doesn't prevent link-spamming and as vandalism normally gets reverted pretty quickly anyway the chance of a search-engine bot indexing the page during the time the spamed link is there is small. Thryduulf 06:21, 12 Feb 2005 (UTC)
- Eloquence* 06:24, Feb 12, 2005 (UTC) - better solutions are needed (recently added links are less valuable than those which have persisted for months, for example)
- Johan Magnus 06:41, 12 Feb 2005 (UTC)
- Academic Challenger 08:37, 12 Feb 2005 (UTC)
- Remove from very active Wikipedias only, leave on those with less than say 20,000 articles. -- The Anome 10:30, Feb 12, 2005 (UTC)
- --Silverback 11:32, 12 Feb 2005 (UTC) Wikipedia's external links are as good as the articles and the rest of the internet including users of google should benefit from them.
- Fred Bauder 13:10, Feb 12, 2005 (UTC)
- JYolkowski 17:08, 12 Feb 2005 (UTC)
- It's way too early to be using nofollow. If the problem becomes critical for Wikipedia, we can always put it in later. At the moment, using nofollow has too many disadvantages for us. -- nknight 18:49, 12 Feb 2005 (UTC)
- We should relish the chance we have to be a positive contributor to the Web as a whole, instead of being isolationist. foobaz·✐ 21:18, 12 Feb 2005 (UTC)
- I don't see a benefit in doing this, and a see a lot of negatives. It seems petty. Easy choice, don't do it. (Sam Spade | talk | contributions) 01:20, 13 Feb 2005 (UTC)
- Jiang 02:25, 13 Feb 2005 (UTC)
- I refuse to support it because it is not part of any written standard. If there was something like a nospam, I would support it. -- AllyUnion (talk) 09:20, 13 Feb 2005 (UTC)
- Matthew Woodcraft 14:08, 13 Feb 2005 (UTC)
- The external links that persist in en Wikipedia are generally at least as likely to be relevant as say links on a random webpage. I would much rather revert the occasional spam edit than keep google from using the information. Jrincayc 16:41, 13 Feb 2005 (UTC)
- BesigedB (talk) 17:20, 13 Feb 2005 (UTC) Linking from here is a good reflection on the content of a page, reverting bad links as they're come across is easy enough.
- Wikipedia external links are generally good quality and it would be a shame if google pagerank didn't use them just because we're worried about spammer abuse. - McCart42 17:43, 2005 Feb 13 (UTC)
- However, it would be nice if it could be enabled on a per-page basis. Nofollow serves no purpose for, say, History of the Soviet Union, but might be a good idea for Viagra. Fredrik | talk 20:53, 13 Feb 2005 (UTC)
- Jmabel | Talk 21:08, Feb 13, 2005 (UTC)
- --Henrygb 21:58, 13 Feb 2005 (UTC) Remember to cull dubious links
- —Ben Brockert (42) UE News 00:25, Feb 14, 2005 (UTC)
- gcbirzantalk 02:25, 14 Feb 2005 (UTC)
- If link spam is an urgent problem, a better solution would be captchas for adding links --Ellmist 02:44, 14 Feb 2005 (UTC)
- Onco p53 10:43, 14 Feb 2005 (UTC)
- Obviously. Dunc|☺ 12:16, 14 Feb 2005 (UTC)
- nofollow will certainly help reduce the spam in google's search results, but I doubt it will actually deter spammers from posting anytime soon. ~leif ☺ HELO 15:04, Feb 14, 2005 (UTC)
- Grue 18:14, 14 Feb 2005 (UTC)
- Thoken 22:25, 14 Feb 2005 (UTC)
- ✏ Sverdrup 22:27, 14 Feb 2005 (UTC)
- -- Scott eiπ 23:04, Feb 14, 2005 (UTC)
- Spammers don't care if technology is in place to reduce their effectiveness; let's not break the web with yet another misguided attempt to solve the wrong problem. --Plek 23:18, 14 Feb 2005 (UTC)
- ABCD 03:04, 15 Feb 2005 (UTC)
- Jtatum 05:11, 15 Feb 2005 (UTC)
- Andre (talk) 13:36, Feb 15, 2005 (UTC)
- Gamaliel 18:10, 15 Feb 2005 (UTC)
- Beta m (talk) 18:25, 2005 Feb 15 (UTC)
- Trilobite (Talk) 21:52, 15 Feb 2005 (UTC)
- Olve 21:56, 15 Feb 2005 (UTC)
- Dcoetzee: See my comments below.
- SimonP 05:26, Feb 16, 2005 (UTC)
- Thue | talk 12:30, 16 Feb 2005 (UTC)
- cesarb 15:27, 16 Feb 2005 (UTC)
- Decumanus 16:24, 2005 Feb 16 (UTC). Should Wikipedia affect PageRank? That I can't answer. But I do know that I do enjoy rewarding good, authoritative sites (especially other wiki sites) in the "External links" section. For me, it makes the hunt for solid sources more enjoyable, knowing we can share the PageRank boost with them a little.
- Eugene van der Pijll 17:23, 16 Feb 2005 (UTC)
- cfp 21:28, 16 Feb 2005 (UTC)
- Too early to start using "nofollow", but perhaps we could allow admins to mark links from certain contributors as nofollow or force newbies and anons to mark their links that way, allowing other users to remove the attribute later. Bart133 (t) 02:22, 17 Feb 2005 (UTC)
- It's an interesting idea, but I suspect it's technically infeasable. Finding a way to implement this would be very hairy. --Improv 05:17, 17 Feb 2005 (UTC)
- Bcat (talk) 02:28, 17 Feb 2005 (UTC)
- The bleeding edge, bleeds for a reason. While the idea of controlling what we do to site's pagerank is attractive, there needs to be a balance. Like maybe a differnt wiki syntax like maybe {{nofollow}} on the page (and all the unpleasnt database implications) ... for now error on the side of standard and pattroling pages. Maybe a automatic Nofollow ONLY for stubs? There IS a middle ground, until then, Remove it. --Dbroadwell 05:33, 17 Feb 2005 (UTC)
- Squash 09:43, 17 Feb 2005 (UTC)
- (Comments below.) Uncle G 16:42, 2005 Feb 17 (UTC)
- Etimbo | Talk 17:01, 17 Feb 2005 (UTC)
- IMO, the implicit "punishment" for good sites we want to link far outweights the theoretical benefits. Ideally, a nofollow could be added to links that are new for a few days or a week, because after a few days at most, spam is usually removed. But as long as that is not possible, I don't want nofollow at all. -- AlexR 18:47, 17 Feb 2005 (UTC)
- --Eurotrader 18 Feb 2005 (UTC) Nofollow should be removed because 99.99% of Wikipedia's external links are pointing to high quality informational articles. These sites deserve to receive credit, and Wikipedia gained its Pagerank in the same exact manner. I am a webmaster and I link to Wikipedia pages because they are useful to my visitors. I have a high Pagerank and I am happy to share it with well deserving sites. Informational articles need links to show up high in search engines. If Wikipedia is removing link credit, these quality pages won't show up nearly as high and some irrelevent spam site will show up higher. Nofollow is proposed to reduce spam, but it really is "throwing out the baby with the bath water." Please remove nofollow.
- Fewer than 100 edits. Possible sockpuppet. --Slowking Man 00:26, Feb 26, 2005 (UTC)
- 99.99%? From a quick survey of my watchlist, I'd say 50% at best, and at least 30% garbage links. --Carnildo 23:18, 18 Feb 2005 (UTC)
- Frankie Roberto Remove it. We don't need this to combat linkspam, the community editing system itself can manage.
- Bad idea, Nofollow in Wikipedia. --Sweets 22:35, 19 Feb 2005 (UTC)
- Nofollow must go: it is a "poor quality" sign --BozMo|talk 23:13, 19 Feb 2005 (UTC)
- r3m0t 00:52, Feb 20, 2005 (UTC)
- someoneinmyheadbutit'snotme
- Fewer than 100 edits. Possible sockpuppet. --Slowking Man 00:26, Feb 26, 2005 (UTC)
- Remove and be bold editing the links Audiovideo 11:59, 20 Feb 2005 (UTC)
- Remove, --Monoet 14:02, 20 Feb 2005 (UTC)
- Remove, not very useful and way too many disadvantages for Wikipedia. --Mononoke 17:22, 20 Feb 2005 (UTC)
- Remove, Arguments see above --Leopard 22:43, 20 Feb 2005 (UTC)
- Remove. Maybe I missed a link to it, but I don't see anyone referencing Sunir's comments. I agree with him. -- Hex 23:40, 20 Feb 2005 (UTC)
- Remove, leave on for unattended wikis. Agree with BozMo: Nofollow is a "poor quality" sign. — mark ✎ 10:06, 23 Feb 2005 (UTC)
- Remove. It won't stop linkspam, and it reduces the value of other resources (such as Google). -- Dominus 13:57, 23 Feb 2005 (UTC)
- Yes, finally remove it again, at least for the popular Wikis as was discussed in the previous vote. This is a ridiculous attribute and doesn't help Wikipedia and the Web at all. Let Wikipedia help the search engines and thereby Web users worldwide with this very well-edited link lists. Patrice Neff 21:11, 23 Feb 2005 (UTC)
- Eriathwen 22:38, Feb 23, 2005 (UTC)
- Fewer than 100 edits. Possible sockpuppet. --Slowking Man 00:26, Feb 26, 2005 (UTC)
- — Stevie is the man! Talk | Contrib 03:53, 24 Feb 2005 (UTC)
- Remove. It's not doing a good job as it is. No need to keep complicating pages or trying to mess with other search engine's algorithms. Let Google sort links on its own. I would support adding nofollow in specific places, like sandboxes, user pages, or article stubs... but Wikipedia is more than just some huge guestbook full of comments and unfiltered links. It is only natural that search engines should be able to pick up on useful third party sites that get linked to for informational purposes from articles. Schmiddy 06:47, Feb 24, 2005 (UTC)
- Remove This is an issue for google, et al to sort out, not wikipedia. Let links be as they are. (Spammers are likely to come here, regardless). EggplantWizard 05:07, 25 Feb 2005 (UTC)
- Remove Dragons flight 20:36, Feb 25, 2005 (UTC)
- .:Ajvol:. 08:22, 7 Mar 2005 (UTC) Wikipedia links have to be one of the highest quality sources of link that can be used to improve search results.
- Remove --Chrisjustinparr 22:24, 22 January 2007 (UTC) Why should we be bullied by spammers when if all we need is for people to correct the spam links. Just lock the articles more often
Manta 17:16, 20 Aug 2008 (UTC)
Just one other comment from the peanut gallery: Some of us from the old school would just like to get people to visit our sites because we are trying to build something good and just in the face of thousands, if not hundreds of thousands, of commercial scams that threaten to destroy the usability of the internet just as they have slammed their adverising grafitti into the everything but our dreams! Sure I advertise, we all do, we have to support the cost of our efforts some way, but everything has its place and the place is not the onslaught of sites that contain nothing more than google ad links and clever code to make you go there in the first place in hopes you will use one of there links rather than hit the back button, or infomercials that do nothing more than sell pyramid schemes to sell more informercials!
OK Enough of my soap box, soory, back to what I wanted to say...
We need to allow links to sites so the cream to rise to the top. We need to find ways to educate the public in ignoring the bad sites, the scams that arrive in their emails and the infomercials that are almost as plentiful as quility programming these days or we will end up like the cops in the movie Demolition Man, Thinking that commercials are kewl!?
I have invested 10 years of my life in trying to provide content that is useful (this is the part where I am supposeed to plug my websites CC2k (look it up if you are interested) but I am not going to. I just couldnt resist speaking my piece to raise awareness in battling the commercial graffitti going on in our lives.
It is not ok when some 14 year old wanna be gangster spray paints your wall, it should be just as NOT OK for corporations to spray paint our TV, Stereos, Computers, and Minds!
- 'Nough Said!
Keep attribute: don't allow external links from WP to affect Google PageRanks of their targets
[edit]- Carnildo 01:05, 12 Feb 2005 (UTC)
- — Catherine\talk 02:12, 12 Feb 2005 (UTC) - better to have the rep that "you can't spam Wikipedia"
- Curps 02:14, 12 Feb 2005 (UTC)
- silsor 02:20, Feb 12, 2005 (UTC)
- DavidWBrooks 02:43, 12 Feb 2005 (UTC) - I encounter enough unnecessary external links as it is.
- —Korath (Talk) 04:52, Feb 12, 2005 (UTC)
- Less gaming = more content. - Nunh-huh 05:03, 12 Feb 2005 (UTC).
- We're an NPOV encyclopedia, we shouldn't be increasing other sites page rank regardless of if they are a good reference or not. — マイケル ₪ 05:34, Feb 12, 2005 (UTC)
- Sam Hocevar 11:34, 12 Feb 2005 (UTC) I don't think Wikipedia should reward "good" sites. If the information in the site is good, then just put that information in Wikipedia.
- I wholeheartedly support this initiative, and don't think we should be engaging in the politics of search engine ranking outside Wikimedia projects. ADH (t&m) 12:08, Feb 12, 2005 (UTC)
- I see no merit in giving editors control over this feature. It should be a matter of complete indifference to Wikipedia whether Google ranks are affected, and those to whom it does matter, for whatever reason, should be denied the opportunity to abuse the ability to affect Google rankings, since those rankings can be of no legitimate value to Wikipedia itself. --Tony Sidaway|Talk 12:45, 12 Feb 2005 (UTC)
- Stirling Newberry 14:08, 12 Feb 2005 (UTC) Given the three revert rule and the incentives to article spam, nofollow makes even more sense with big wikipedias.
- I don't agree with the argument that we should help boost the page ranking of "good" sites. What's good as an ext link in our article will often differ from what's good for a Google searcher. For example, we might well have no reason to link to a good general (encyclopedic) summary of a subject (we'd just cannibalize the information), but we'd link to a fringe site that gives detail on some minor aspect or expounds an unorthodox theory. Also, I'm less worried about spambots than about human users' normal disputes concerning links. Those disputes could be even harder to resolve, because, in addition to wanting to shape the article, editors want to game the Google rankings. JamesMLane 14:47, 12 Feb 2005 (UTC)
- The flaw in this argument is that you should be citing your sources by linking to informative pages you "cannabalize". Deco 03:09, 16 Feb 2005 (UTC)
- How is that either a "flaw" in the argument, or pertinent to the question at hand? Having "nofollow" interferes in no way whatsoever with citing sources. - Nunh-huh 04:37, 16 Feb 2005 (UTC)
- The argument asserts that we don't link good sources, because we'd rather copy material from them. I assert that such sources are linked, precisely because we do copy from them, and cite our sources. I don't see the trouble. Deco 05:36, 16 Feb 2005 (UTC)
- This is, in part, wishful thinking. While we should cite our sources, that does not mean that everyone does. It is indeed true that, for lots of data, sources go uncited. As a typical example, take Earth. The entire post of:
- The interior of Earth, like that of the other terrestrial planets, is chemically divided into an outer siliceous solid crust, a highly viscous mantle, a liquid outer core that is much less viscous than the mantle, and a solid inner core. The liquid outer core gives rise to a weak magnetic field due to the convection of its electrically conductive material.
- ...does not have any sources cited for its data. The data is considered publically available and simple consensus, and is therefore not cited. When people are Googling, they are more often looking for general data. So, when we fail to cite these things, we bias Google somewhat against what it's supposed to be doing. Just my two cents. --Chris Drostie 04:20, 19 Feb 2005 (UTC)
- I agree with Drostie. I'd add, in response to Deco's point, that I have no problem with citing a general summary if the editors use it for the Wikipedia article, but they won't always use it. If I see something in a general site, I might well confirm it with a more specialized resource, then cite the latter. Examples of proper reasons for ext linking would include "This site supports one particular assertion in the article" or "This site has more detailed information on some aspect that's not worth elaborating on at length in the article". There's no strong correlation between such reasons and "This site deserves a high ranking on a Google search for the overall topic." JamesMLane 09:45, 24 Feb 2005 (UTC)
- This is, in part, wishful thinking. While we should cite our sources, that does not mean that everyone does. It is indeed true that, for lots of data, sources go uncited. As a typical example, take Earth. The entire post of:
- The argument asserts that we don't link good sources, because we'd rather copy material from them. I assert that such sources are linked, precisely because we do copy from them, and cite our sources. I don't see the trouble. Deco 05:36, 16 Feb 2005 (UTC)
- How is that either a "flaw" in the argument, or pertinent to the question at hand? Having "nofollow" interferes in no way whatsoever with citing sources. - Nunh-huh 04:37, 16 Feb 2005 (UTC)
- The flaw in this argument is that you should be citing your sources by linking to informative pages you "cannabalize". Deco 03:09, 16 Feb 2005 (UTC)
- I don't think that WP should be in the business of helping or hurting a site's Google PageRank. If a good site is relevant to an article, adding an external link helps Wikipedia users, which is the whole point of WP. Carrp | Talk 15:26, 12 Feb 2005 (UTC)
- Keep it. For same reasons listed by Carrp. — FREAK OF NURxTURE (TALK) 10:02, Feb 15, 2005 (UTC)
- Use nofollow. See main proposal page for justification. --Improv 16:17, 12 Feb 2005 (UTC)
- Keep nofollow. See comments for justification.Superm401 00:32, 13 Feb 2005 (UTC)
- The upside is reduced incentive to spam. I see no downside. Wolfman 05:37, 13 Feb 2005 (UTC)
- Keep the Nofollow JFW | T@lk 07:12, 13 Feb 2005 (UTC)
- Datrio 12:04, 13 Feb 2005 (UTC)
- WP shouldn't be in a position to affect PageRanks, whether intentional/indirect or not.--jag123 14:20, 13 Feb 2005 (UTC)
- alteripse 17:25, 13 Feb 2005 (UTC)
- Keep nofollow. Kaibabsquirrel 18:19, 13 Feb 2005 (UTC)
- Weft 19:18, 13 Feb 2005 (UTC)
- munnin Keep
- Fewer than 100 edits. Possible sockpuppet. --Slowking Man 00:26, Feb 26, 2005 (UTC)
- --AStanhope 21:23, 13 Feb 2005 (UTC) Agree with Carrp. Wikipedia articles should exist in a vacuum as far as Google is concerned. Every functioning disincentive to spam should be employed for as long as it is able to serve its purpose.
- Keep. dont see why we should be pushing article editors POV about which sites are good or bad. I dont think WP should be pushing one site over another, whatever the reason. Plus a large number of articles are watched over by just 2-3 people at most, and nothing prevents them from pushing one site over another. kaal 21:41, 13 Feb 2005 (UTC)
- Agree with DavidWBrooks. Taco Deposit | Talk-o to Taco 22:44, Feb 13, 2005 (UTC)
- Keep, I think removing this feature goes against our NPOV policy. Mentioning a useful outside site is a good thing as it helps our users, but we should not be seen to be promoting them beyond this. --Rje 03:29, Feb 14, 2005 (UTC)
- I think we get enough linkspam as it is. If it were generally known that Wikipedia now removed the nofollow attribute, we can expect to be flooded with even more linkspam. --Deathphoenix 05:36, 14 Feb 2005 (UTC)
- Keep nofollow, if the choice is keeping it vs. removing it everywhere on wiki.riteme.site. I would, however, support removing the nofollow tag for article pages only, because I believe that pointing to good resources (which is what our external links should be) and therefore having them ranked more highly in Google and other searches is a good thing, and I'm persuaded that the community policies those sufficiently well to combat linkspam in the article namespace. However, when it comes to talkpages, we can't revert bad links on sight because that stifles discussion and runs contrary to our policy, and Wikipedia is viewed highly enough by Google and other search engines that linkspam on talk pages could become a *major* problem quickly. -- Seth Ilys 13:21, 14 Feb 2005 (UTC)
- Keep. ElBenevolente 17:37, 14 Feb 2005 (UTC)
- Keep. --Ryan! | Talk 18:12, Feb 14, 2005 (UTC)
- Keep. Agree with JamesMLane. --David Iberri | Talk 14:18, Feb 15, 2005 (UTC)
- Keep. Andrew pmk 04:40, 16 Feb 2005 (UTC)
- Keep. Warofdreams 15:05, 16 Feb 2005 (UTC)
- Keep. RSpeer 23:35, Feb 16, 2005 (UTC)
- Keep.SYSS Mouse 02:45, 17 Feb 2005 (UTC)
- Keep. Shoecream 06:20, Feb 17, 2005 (UTC)
- Keep. Squallwc 08:51, 17 Feb 2005 (UTC)
- Keep. Sorry for the good links, but link spamming has gone out of control. andy 09:40, 17 Feb 2005 (UTC)
- Keep. Link spamming is a problem, even on en, and I'm sure it will be even more of a problem in the future. Shanes 00:51, 18 Feb 2005 (UTC)
- Keep. --Hedgeman 06:56, Feb 18, 2005 (UTC)
- Keep. Circeus 13:52, Feb 18, 2005 (UTC)
- Keep. Matthewmayer 04:39, 19 Feb 2005 (UTC)
- Theo (Talk) 20:11, 19 Feb 2005 (UTC)
- Keep Dan100 13:31, Feb 21, 2005 (UTC)
- Keep. Wikipedia isn't here to improve other sites' Google rankings. – Quadell (talk) (sleuth) 00:34, Feb 23, 2005 (UTC)
- Keep. Alereon 00:58, Feb 23, 2005 (UTC)
- Fewer than 100 edits. Possible sockpuppet. --Slowking Man 00:26, Feb 26, 2005 (UTC)
- Keep Link spam is not being edited out fast enough hike395 01:56, 23 Feb 2005 (UTC)
- Keep for the reasons stated by other Wikipedians. JoaoRicardo 01:58, 23 Feb 2005 (UTC)
- Keep. A reference work that describes the objects to which it links should try hard not to disturb the context those objects reside in. This is along the lines of the observer observing atypical behavior and attributing it to the object as normal because this is what is observed. By altering the PageRank of the sites we link to, we affect the context of those sites by altering the traffic through them. Let us observe but not influence. Courtland 03:30, 2005 Feb 23 (UTC)
- Keep. Mikkalai 21:20, 23 Feb 2005 (UTC)
- Keep. Wikipedia's presence on the Internet is getting ever more heavy, and I am uncomfortable with allowing Wikipedia, an NPOV encyclopedia and reference source, to boost specific sites. Sites should stand on their own. —Lowellian (talk) 08:06, Feb 24, 2005 (UTC)
- Keep. Removes a reason for spammers to target us. Lan3y - Talk 15:58, Feb 24, 2005 (UTC)
- Keep. Pjacobi 16:47, 2005 Feb 25 (UTC)
- Keep. More appropriate for an encyclopaedia to try to comment on the world and only to change it indirectly. Also agree with regard to effect on linkspammers. Dbiv 23:51, 25 Feb 2005 (UTC)
- Strongly Keep--Trichnosis 14:24, 22 January 2007 (UTC)
Comments
[edit]- (response to DavidWBrooks' vote)
- Which breaking the links with rel=nofollow won't stop. Jordi·✆ 02:48, 12 Feb 2005 (UTC)
- This does not "break" the links in any way, they work just fine. It merely removes one very significant incentive for link spamming, especially automated industrial-scale link spamming by spambots as we have recently experienced; it does not and cannot remove all incentives to do so. -- Curps 03:11, 12 Feb 2005 (UTC)
- It "breaks" how link relationships work. 'The linked document contains "nofollow" information for the current document'. So what is "nofollow" information? Rel=nofollow is a typical hack which should never be implemented: it runs counter to the standard and is implemented by a single party (google), not a standards body. It's no better than <blink> or <marquee>. Jordi·✆ 03:16, 12 Feb 2005 (UTC)
- What Google does, in this area, is more of a std than a stds body can impose. --Jerzy(t) 03:57, 2005 Feb 12 (UTC)
- It's a hack which for once does not break any standards, but it's still a hack. And your argument is the same one Netscape used for <blink>, Microsoft for <marquee>, etc.. Look where that got us now!
- Technically, rev=nofollow would maybe make sense to some ('the current document contains 'nofollow' information for the linked document'), but rel=nofollow is absolute nonsense. In either case it is not appropriate imho. Linkspammers should be dealt with in other ways than crippling search engine functionality. Jordi·✆ 04:09, 12 Feb 2005 (UTC)
- How does using a feature which Google themselves implemented cripple the functionality of Google's own search engine? -- Curps 05:51, 12 Feb 2005 (UTC)
- I'm presuming that the comment relates to search engines other than google and other web tools. Thryduulf 07:47, 12 Feb 2005 (UTC)
- Nevertheless, it is unreasonable to suggest that these other search engines' functionality will be "crippled" by this proprietary Google tag. HTML parsers just ignore tags they don't understand. Yahoo's search engine technology won't crash and burn when it encounters a "nofollow". -- Curps 07:54, 12 Feb 2005 (UTC)
- Yahoo and MSN Search will also support "nofollow" [4]. So if this isn't standard yet it very soon will be. It will not "break" links or "cripple" search engines. -- Curps 14:43, 12 Feb 2005 (UTC)
- How does using a feature which Google themselves implemented cripple the functionality of Google's own search engine? -- Curps 05:51, 12 Feb 2005 (UTC)
- What Google does, in this area, is more of a std than a stds body can impose. --Jerzy(t) 03:57, 2005 Feb 12 (UTC)
- It "breaks" how link relationships work. 'The linked document contains "nofollow" information for the current document'. So what is "nofollow" information? Rel=nofollow is a typical hack which should never be implemented: it runs counter to the standard and is implemented by a single party (google), not a standards body. It's no better than <blink> or <marquee>. Jordi·✆ 03:16, 12 Feb 2005 (UTC)
- This does not "break" the links in any way, they work just fine. It merely removes one very significant incentive for link spamming, especially automated industrial-scale link spamming by spambots as we have recently experienced; it does not and cannot remove all incentives to do so. -- Curps 03:11, 12 Feb 2005 (UTC)
- Which breaking the links with rel=nofollow won't stop. Jordi·✆ 02:48, 12 Feb 2005 (UTC)
- I agree with DavidWBrooks that there seem to be a quite a few unnecessary and low quality external links on the Wikipedia—especially in some of the less-traveled area where a link on the Wikipedia might have a much greater effect upon a website's PageRank when searching under that topic. I've also seen quite a few questionable links in articles on different cities, especially those that are tourist destinations. However I am not yet convinced that the nofollow tag is the appropriate or best solution to the problem. Even with the nofollow tag, the Wikipedia will still get link spam because they also want people to visit their websites and see the advertising on the websites. They may also be trying to get people to visit websites that try to do stealth spyware installations. The solution may end up being something more mundane and tedious, such as a special Wikipedia Link Patrol to try to weed out link spam and low quality links, plus questionable and inappropriate external links. BlankVerse ∅ 10:02, 12 Feb 2005 (UTC)
- Having unpaid manual labor against paid automatic labor is a losing proposition. Stirling Newberry 15:32, 12 Feb 2005 (UTC)
- I agree with DavidWBrooks that there seem to be a quite a few unnecessary and low quality external links on the Wikipedia—especially in some of the less-traveled area where a link on the Wikipedia might have a much greater effect upon a website's PageRank when searching under that topic. I've also seen quite a few questionable links in articles on different cities, especially those that are tourist destinations. However I am not yet convinced that the nofollow tag is the appropriate or best solution to the problem. Even with the nofollow tag, the Wikipedia will still get link spam because they also want people to visit their websites and see the advertising on the websites. They may also be trying to get people to visit websites that try to do stealth spyware installations. The solution may end up being something more mundane and tedious, such as a special Wikipedia Link Patrol to try to weed out link spam and low quality links, plus questionable and inappropriate external links. BlankVerse ∅ 10:02, 12 Feb 2005 (UTC)
- [Vote copied, & comment & responses moved from Remove votes by Jerzy(t) 22:30, 2005 Feb 12 (UTC)] Ævar Arnfjörð Bjarmason 01:07, 2005 Feb 12 (UTC) The pros vastly outweigh the cons.
- If the pros vastly outweigh the cons, did you mean to vote in the other section? silsor 02:29, Feb 12, 2005 (UTC)
- The pros of removing it, sorry for being vague. —Ævar Arnfjörð Bjarmason 04:30, 2005 Feb 12 (UTC)
- The Russian spambot that hammered us only a few days ago isn't enough of a con? Curps 04:39, 12 Feb 2005 (UTC)
- The spambot you point out spammed Wikipedia while the nofollow tag was in use (and had been for some time). What exactly do you think it accomplishes with regards to spambots when the very example you gave shows that it had no effect at all. —Ævar Arnfjörð Bjarmason 05:22, 2005 Feb 12 (UTC)
- I think "nofollow" was only added in the latest very recent Mediawiki release. The spambot operator was likely unaware that we recently started using "nofollow". -- Curps 05:28, 12 Feb 2005 (UTC)
- It was added on 2005-01-19 and the bot started spamming on 2005-01-31 which as I understand it was the first page to be spammed, correct? —Ævar Arnfjörð Bjarmason 05:40, 2005 Feb 12 (UTC)
- The PHP page was first attacked January 31 2005, and it was probably the first page (other pages attacked were mostly linked from it; the bot followed links). -- Curps 05:47, 12 Feb 2005 (UTC)
- The Russian spambot that hammered us only a few days ago isn't enough of a con? Curps 04:39, 12 Feb 2005 (UTC)
- The pros of removing it, sorry for being vague. —Ævar Arnfjörð Bjarmason 04:30, 2005 Feb 12 (UTC)
- If the pros vastly outweigh the cons, did you mean to vote in the other section? silsor 02:29, Feb 12, 2005 (UTC)
- (Responding to Curps) I was referring to the fact that Google released this to the public recently, and we do not really know how spammers will react to it yet. - RedWordSmith 04:07, Feb 12, 2005 (UTC)
- I'm in favor of keeping nofollow . First of all, I'd like to refute the argument, made by [[User:An%E1rion|Anárion]] and probably agreed with by others, that this violates standards. This is sanctioned by the W3. For proof, see W3 information on rel. This says that additional types,(i.e. nofollow and others) are allowed, but that a profile should be defined. Luckily, it has been! See the draft standard linked to by Google's blog post.
- Note: I strongly recommend that MediaWiki uses this profile system. This means it should automatically implement a profile. It should do so in the manner detailed in my prior link to W3's profile information. This only need by done for the pages that have external links. But, it might be easier just to do it for all pages, which would be fine.
- However, if Mediawiki does not implement this, it will not affect my vote against removing the feature. (This is because I don't want to have unfair conditions not specified in the proposal).
Above I have described why it can be used. We should use it because it will eventually completely eliminate rankings-based spam linking by most expert spammers. I believe there will probably originally be a small group of spammers that seek revenge by massively vandalizing Wikipedia pages. However, given that this is "unpaid labor", eventually the tide will subside. The benefits to our links will be worth the wait. There is one last caveat. Even after that vandalism trend ends, we should remember that there will still be some spamming by those who simply want visibility, regardless of rankings advancement. I believe this is probably a smaller portion of the spam. The obvious cost to the "nofollow" procedure will be the sacrifice of gains in Google rankings for "good sites" because of links from Wikipedia. Compared to the improvement in Wikipedia external links' quality, this will be less important. We must remember that these external sites should be inherently valuable, and will therefore be linked to by enough non-wiki authors. If they are not valuable, they should not be linked to! Superm401 00:59, 13 Feb 2005 (UTC)
- I'll bite. Especially what is a nofollow relationship to the parent document? If I have a link <a href="foo.bar" rel="nofollow">, what I am telling the User Agent is that the foo.bar document has a nofollow relationship to the current document, so if I go there I get extra information (nofollow information) about where I am now. This is by analogy of existing link relationships like rel=next (the linked document is next in a series to which the current document also belongs), rel=copyright (the linked document has copyright information for the current document), or rel='shortcut icon' (the linked document is a shortcut icon for the current document). 'nofollow' is not a rel relationship, as the point of the attribute is that the link is not followed. In actuality rel=nofollow is telling the user agent that there should be no link from the linked document to the current document!
- If done at all, it should be done in other ways, such as a rev (reverse) relationship: rev=nofollow tells the user agent 'the linked document is a 'nofollow' for the current document'. My claim it violates the standard is because of this misinterpretation of what a 'rel' relationship is. Google has a nice idea and they should be commended for trying to make it parse well instead of inventing something dreadfull like <nofollow />, but it is not standards compliant as the attribute is semantically meaningless. Jordi·✆ 13:24, 13 Feb 2005 (UTC)
- First, I disagree that W3-supported tags and attributes have to be human-readable, as you imply. For an example, consider the BDO tag? What's a BDO? Well, you don't know until you read the definition. And that's okay, because human readers should look at the definition before implementing it in a program. There are also rel attributes that you might not understand without knowing the definition as well. What does it mean to have a "contents" relationship? It means that the page is a Table of Contents. Someone might need to look that up, but the W3 doesn't care. That's why they have references. A rel nofollow relationship makes sense as well, with a definition. A "nofollow" relationship signifies that the page linked to should not be advanced in search engine rankings through this link. The pages are related in that users will go between, but search engines will not go. This has a clear analogy with the "next" relationship. That signifies a page that human users will probably follow the link to. It also tells user-agents that it might be beneficial to pre-load the page. That establishes a precedent for rel relationships that only computers need consider. I disagree that "rel=nofollow" tells the user-agent that there should be no link. Attributes mean only what their definition says they mean, and the definition clearly doesn't say that.
- It should NOT be a "rev" relationship. According to the W3, "rev = link-types [CI] This attribute is used to describe a reverse link from the anchor specified by the href attribute to the current document. The value of this attribute is a space-separated list of link types."
- When we create a link with a rev attribute, we are not referring to a "link from the anchor specified by the href to the current document." That would mean the page linked to links back. We have no way of guaranteeing that and don't care if it's true. W3 defines rel as follows:
- "rel = link-types [CI] This attribute describes the relationship from the current document to the anchor specified by the href attribute. The value of this attribute is a space-separated list of link types."
- We are referring to "a relationship from the current document to the anchor." Specifically, we are saying that the link is related in that we want users to follow the link but don't want search engines to advance the page by following the link and adding ranking because of it. Superm401 16:00, 13 Feb 2005 (UTC)
- Are you seriously arguing against employing a 100% effective method of combating Googlebombing because you don't like the syntax?
- Like nearly all uses of rel, this hack makes little semantic sense—just like the meta element, defined only after the various browser implementations became so ubiquitous that there was no way around it, or br, which defines presentation rather than context. Compared to the utter crap that's resulted from over a decade of standards being descriptive, rather than prescriptive, this is pretty benign.
- The fact is, all uses of rel are special exceptions, to be handled on a case-by-case basis by user agents, where applicable. The attribute has always been vaguely defined as a catch-all for miscellaneous metadata, which the W3C acknowledges today in intentionally leaving it open to individual implementations.
- For my part, I'm tickled pink that Google has taken responsibility for the problem they inadvertently created and provided the community with a solution. What's more, it doesn't break the standard, and I don't think we could ask for much more.
- And that's not even the point.
- The English Wikipedia deciding that it's going to drop support for this solution isn't going to change the fact that it's the one everyone else is adopting. No, it won't stop linkspam, but it will stop one particularly prevalent form of linkspam, and that's better than nothing. Contrary to what others have said, this will not adversely affect the ranking of any site linked to in an article, but rather, it won't affect it at all. I happen to think that a wise policy, irrespective of the linkspam problem.
- Spam and POV linking are two things we can do without. The hidden argument here is whether editors should use Wikipedia's link equity to push sites they like. The answer should be "no", wikipedia's link equity is the result of the common effort and should not be exploitable by individual editors. Remove nofollow and every new link becomes suspect, and we have enough linking of questionable sites as it is. If you think a site is good, link to it from your own page. Stirling Newberry 19:54, 13 Feb 2005 (UTC)
- A large number of articles on WP have only 1-3 people watching them, and there is no reason to believe that they are not going to push one site over another based on their POV. I dont think WP should be in the business of promoting good site, as that would be the editors POV as mentioned by number of people above. The problem of link spammers will only keep getting worse as WP expands at the rate it does and becomes more and more visible. kaal 21:48, 13 Feb 2005 (UTC)
- In response to "dont see why we should be pushing article editors POV about which sites are good or bad". Why should we be seen as pushing anything? We are just doing wikipedia. It is google that is assigning weights. We should operate true to our purposes, google and other users on the internet are responsible for their own actions and usage of wikipedia. They use us at their own risk. The should read the disclaimers.--Silverback 12:41, 14 Feb 2005 (UTC)
- I really don't understand and follow all the HTML/W3 discussion above. But what seems clear to me is that having NOFOLLOW will NOT discourage spam-link-bots since the vandalism they add will very quickly be reversed by the human editors of wikipedia BEFORE any Google crawler will follow the links they created. Don't the Google crawlers only do it on a weekly or monthly basis and isn't vandalism cleaned up on an hourly or at most daily basis by all the unpaid labor on Wikipedia? So spam-link-robots WILL NOT get any Google Page Rank boost from their vandalism - so not having NOFOLLOW will not encourage them (too many negatives, but I think I got it right). The only gain they get is that somebody might click on the link before we revert it. So, somebody, please tell me what is wrong with this argument (as if I have to ask for feedback on this page). FrankH 18:25, 13 Feb 2005 (UTC)
- Although we're fond of touting our editing process as a revolutionary and miraculously effective idea, the fact is, it's not infallible. Even blatant vandalism sometimes slips through the cracks, and once it makes its way past Special:Recentchanges it can stay in an article for weeks. Linkspam is even harder to catch, as the vast majority of it is not the work of malicious spambots, but real people adding their sites to articles where it's often overlooked as relevant to the topic (as anyone who has IRC on his watchlist can attest). All nofollow does is make Wikipedia a neutral party, irrelevant in the eyes of search engine ranking systems. Removing the attribute is to say that we should take an active role in boosting the rankings of certain sites, and even if that weren't open to abuse (by, say, allowing anyone to introduce external links into article text, or something silly like that), it's not a policy I think we should adopt. ADH (t&m) 23:25, Feb 14, 2005 (UTC)
- I don't know how often Google crawls Wikipedia, but in general Google crawls high-ranked sites every day. Stephen Turner 19:11, 25 Feb 2005 (UTC)
- Another idea: Could the software that generates the HTML from the a pages source code insert the NOFOLLOW attributes for links UNTIL the page has not been edited for some period of time (like a day?) Then any spam-links inserted by robots would not have any effect until the page is stable for a day. Thus the human editors would have a day to revert edits. In fact links inserted by humans would also have NOFOLLOW for a day to allow other humans a day to remove inappropriate or POV links. Is this suggestion technologically possible? Will it help? -- FrankH 21:04, 14 Feb 2005 (UTC)
- It's possible and was brought up on one of the mailing lists, but, of course, is up to the developers. --Slowking Man 07:44, Feb 15, 2005 (UTC)
- Yes [DelayAction] could be a good compromise. This approach assumes that spam links do get cleaned up fairly soon after being added. On wikipedia this a reasonable assumption. You suggest switching to nofollow for all links on the page, after any edit to the page, for a period of a day. Even better would be if the software could somehow time it for each link, i.e. when link is added a timestamp is kept somewhere, and when that individual link has been left on the page for a period, it drops the rel=nofollow attribute. Not sure if this is technically feasible though. -- Nojer2 21:41, 5 Mar 2005 (UTC)
- Even if it is feasible, it doesn't mean it's a good idea. It soulds like a recipe for an unwanted level of complexity. I'm all for adding nofollow, but if we don't do it for all external links, I'd prefer not doing it at all rather than something like this that would make the system as a whole more complex. --Improv 23:21, 6 Mar 2005 (UTC)
- Yes [DelayAction] could be a good compromise. This approach assumes that spam links do get cleaned up fairly soon after being added. On wikipedia this a reasonable assumption. You suggest switching to nofollow for all links on the page, after any edit to the page, for a period of a day. Even better would be if the software could somehow time it for each link, i.e. when link is added a timestamp is kept somewhere, and when that individual link has been left on the page for a period, it drops the rel=nofollow attribute. Not sure if this is technically feasible though. -- Nojer2 21:41, 5 Mar 2005 (UTC)
- It's possible and was brought up on one of the mailing lists, but, of course, is up to the developers. --Slowking Man 07:44, Feb 15, 2005 (UTC)
- There are many good arguments here, on both sides, but my viewpoint is simply that Wikipedia is a valuable source of good links, and although we're not obligated to share that value through search engines, it is a great opportunity for us to contribute further to the community we serve. Despite our high pagerank we don't make or break any one site — but just as we're willing to fight vandalism rather than moderate edits, I think it's worth it to fight linkspam rather than withdraw our contributions.
- I also don't believe the continuing linkspam since nofollow was turned on is an ephemeral phenomenon. Spammers aren't technical wizzes, don't have a strong community, and don't pay attention very well. No matter what "gets out", we will continue getting linkspam — it's not revenge, it's just ignorance.
- Finally, I think there are a number of excellent compromises between the two extreme all-or-none viewpoints above. The most promising of these is to tag only young links with nofollow. This negatively impacts caching, but it would probably suffice to update the tags only when a page is edited — then, at least, we can force a refresh if necessary. This is a simple technical solution that could easily be implemented and deployed in only a short time, and I don't think we should withdraw from the community in the interim.
- Thanks for listening. Deco 03:40, 16 Feb 2005 (UTC)
- I have taken the liberty of sending an email to some friends I have in Google asking them to either contribute to the discussion, or to consider doing something on their end to deal with how effective linkspam is at affecting pagerank on Wikis. Depending on what they decide, nofollow might not actually affect pagerank regardless of nofollow. --Improv 19:33, 16 Feb 2005 (UTC)
- The effect of external links from Wikipedia on PageRank is how I came to be here in the first place. (What I thought was an obscure article that I had published elsewhere on the web consistently came out at the top of Google searches. I eventually traced the effect to the fact that my article had been listed as a reference by some Wikipedia articles.) I'm sure that I'm not the first and I'd like to think that I was not the last person to start contributing to Wikipedia because of this. (And, mindful of the effect, I've neutered my share of external links in advertising articles with <nowiki>...</nowiki> since coming here, too.)
Furthermore, Google is shooting itself in the foot with this one. Wikipedia and its siblings are exactly the sorts of sources of good-quality hyperlinking to the rest of the world wide web that it doesn't want to lose. If Google wants to solve this problem, then it should provide Wikipedia editors with search tools to locate spam hyperlinks easily, so that they can edit them out, rather than a tool that effectively removes all of the good that Wikipedia does for Google (which outweighs the bad by a wide margin, in my estimation). Uncle G 16:42, 2005 Feb 17 (UTC)
- A question, not relevant to my vote, just idle curiosity: Does the Google boost from a Wikipedia ext link depend on which Wikipedia article includes the link? Wikipedia as a whole has a high page rank but some articles are almost never visited. Does Google's algorithm take account of the number of visits to that particular article? JamesMLane 01:12, 18 Feb 2005 (UTC)
- In vote #70 of the "Remove NoFollow" votes, user:AlexR assumes it is not possible for "a nofollow [to] be added to links that are new for a few days or a week". Maybe i missed something by not following this discussion closely enough:
- The current software does indeed impose an all-or-none decision on each Wiki, e.g. on en:, and that decision is what the vote is on.
- However, the portion of the software that renders the wiki markup into an HTML page knows when the page was last edited. although it doesn't currently know when the ext lks were last changed, that may be a SMOP that could be implemented much as Alex would like, in a future release.
- --Jerzy(t) 03:31, 2005 Feb 18 (UTC)
- This has come up many times — I brought it up some time ago on meta. It seems like a natural solution. If I were to design it, it would go something like this:
- Create a table LinkLastModified mapping each externally linked URL in an article to its last-modified date/time. Whenever a page is updated, remove any URLs from LinkLastModified no longer present in the article, and add any new URLs not present in LinkLastModified and assign them the current time. When initially rendering, compare these times against the current time to determine whether to include nofollow. When determining whether or not to rerender a cached version, simply find the oldest link in LinkLastModified which was below the threshold at the time the cached version was rendered, and determine if it is above the threshold now. If not, the cached version is still up-to-date. If this results in too-frequent re-renderings, we can simply have it update on edit, which at least allows users to force an update if necessary.
- The only question being who would implement all this. :-) Deco 07:07, 18 Feb 2005 (UTC)
- In addition to arguments I made in other places, as well as the other arguments here, I would like us to give Google a little slap in the face. They've become too accustomed to saying "jump!" and everyone else saying "how high?" (before long, they'll get the same big head that Microsoft has). Instead, we just need to do what's right for the Wikipedia and the web community. We haven't found it _that_ difficult to manage our links here. — Stevie is the man! Talk | Contrib 11:18, 24 Feb 2005 (UTC)
- Stevie, mind telling us what you're talking about? I don't understand what you mean by them bossing around the internet community. --Improv 15:38, 1 Mar 2005 (UTC)
- Here's another argument for removing nofollow: we depend on Google traffic for a large proportion of our readers, and from them comes new editors. Of course nofollow doesn't change this, but the question is, is it right for us to take the benefits of our Pagerank without giving in return? How many people here would willingly give up our Pagerank altogether? We can say we want to remain separate from Google, but in truth we have a certain debt to pay. Deco 08:01, 1 Mar 2005 (UTC)
- Pagerank is not currency, and we shouldn't feel obliged to provide benefits to people that are propspective editors. There is no debt. In any case, with luck, google has already cut wikipedia out of contributing to pagerank, regardless of this discussion. --Improv 15:38, 1 Mar 2005 (UTC)
- Hmm, when I think about it this particular argument doesn't really make sense anyway — it's not like we asked Google to direct traffic to us. Still, it just seems like a nice thing to do to give value to searchers in return. Deco 23:30, 1 Mar 2005 (UTC)