Indicators
Defer discussion:
Defer to RS/N
Defer to WPSPAM
Defer to XLinkBot
Defer to Local blacklist
Defer to Abuse filter

Drexel Dragons men's lacrosse[edit]

Drexel Dragons men's lacrosse (edit | talk | history | protect | delete | links | watch | logs | views)

The external link for "Drexel Lacrosse A History of the Heart - William Thayer" seems to be setup so that the pdf file it's linked to downloads to your computer, instead of opening up in your web browser. Not sure if this kind of thing is acceptable per WP:ELNO. It could just be the way my computer is set up, but even so not sure if something needs to be added to the link's description so that the reader is aware that the file may start to download. Is there a way to tweak the url address so that clicking on the link does not automatically start to download the file? -- Marchjuly (talk) 00:17, 16 February 2016 (UTC)

Update: The link in question was removed by another editor for another reason, so this problem appears to have been resolved. -- Marchjuly (talk) 22:24, 26 February 2016 (UTC)

Use of external links in lists of datasets[edit]

Hello everyone,

I was referred here by another Wikipedia user concerning the usage of external links on pages like List of datasets for machine learning research and Comparison of datasets in machine learning. First, some background:

Machine learning is a field where automated methods are used to find patterns in data. It has been argued ^[1] that the availability of high-quality training datasets is actually far more important than the availability of new machine learning algorithms. It appears that in many/most cases, great technological leaps in machine learning performance come when a new high-quality dataset is made available, not a new algorithm (some examples of this given in ^[1]).

For these reasons, Wikipedia should have some record of the groundbreaking datasets that have helped to significantly advance machine learning research. Since the vast majority of individual datasets do not have their own articles (and despite their importance in the field, probably shouldn't), articles like List of datasets for machine learning research and Comparison of datasets in machine learning can serve to aggregate the most important among them.

To maximize the utility of these pages, they contain external links to the webpages where the datasets mentioned can actually be downloaded and used. In both of the example articles given, these links are placed in the rows of the tables. I wanted to open up a discussion here about the usage of these links and their compatibility with WP:EL. And, if they are not compatible, begin the process of discussing an exception. I see in WP:EL that "lists themselves should not be composed of external links." However, it also says "This section does not apply if the external link is serving as a citation for a stand-alone list entry that otherwise meets that list's inclusion criteria."

In closing, I believe pages like these will become great resources to the machine learning community as it continues to develop. I look forward to discussing with everyone. DATAKEEPER ✉ 18:49, 26 February 2016 (UTC)

WP:EL is very clear on this. There are no external links for the list entries in a standalone list, much less being inside the body of the article. These links are to datasets and are not citations for inclusion in the lists. From WP:EL: "These lists are primarily intended as providing direct information and internal navigation, not a directory of sites on the web." List of datasets for machine learning research with external links is solely a way to provide links outside of Wikipedia and not for internal navigation. I see no need for an exemption from how Wikipedia treats external links for this article. The best way to have an external link to one of these datasets is to add it as an Official Website link at the bottom of the article about that dataset. But, none of these datasets have an article to add an external link, which further casts doubt on if this standalone list should be included in Wikipedia. Please remove these links immediately or revert back to the edit that removed them previously. Thanks, Stesmo (talk) 03:15, 27 February 2016 (UTC)

@Stesmo: - Hello again, and thank you for clearly stating your stance on this. For everyone's information, I came to this page and posted this question at the request of Stesmo, which can be found on their talk page. Since Stesmo was the one that referred me here, I was hoping to get an impartial 3rd party opinion on the matter before laying it to rest. Thanks again everybody. DATAKEEPER ✉ 18:56, 29 February 2016 (UTC)

@Datakeeper:I don't agree with Stesmo's opinion (on his talk page) that these links cannot be added to the references. Most of the linked webpages provide all the info listed in the table, so they are a reference for that info. Removing them would leave an article about an IT topic with only 3 of 180 references available online. If they can't be in the table, make them references or add them to the references. You may want to get additional opinions before making such changes (I trust you can write a script to make such changes, would be rather tedious doing it manually). Prevalence 04:11, 2 March 2016 (UTC)

@Datakeeper, Prevalence, and Stesmo: - First, the way they are currently linked is inappropriate, at best they should be turned into references, or in a far-to-the-right column with 'link to dataset'. And for the latter I would still disagree that these links serve Wikipedia's purpose. They are in a way nothing different from direct links to the websites of restaurants in New York on the article 'restaurants in New York'. And both of those fail WP:NOTYELLOW.

Now, regarding the first entry in the first table on List of datasets for machine learning research there is an external link to FERET. That is a notable entry in itself, as witnessed by our internal article FERET. The external link there should hence be converted to the internal link to the internal page - and on the article FERET a link to the dataset (or better, the 'homepage' of the dataset) IS appropriate (and that is there). It does not belong in the list. I haven't checked any other cases, but I presume that there are many which do have internal links possible.

So shortly, all external links in the first column should go, and should all turn into internal Wikilinks. For items that are not notable in Wikipedia terms (hence, are redlinks) sufficient references should show some notability (as in, reliable independent sources must at least have 'noticed' the database), and could contain a reference to the 'homepage' of the dataset. For items that fail that notability test, they should simply be removed from the list (what we see on other similar list-pages is that people just add their self-crafted example to such lists - anyone can make a database of something, post it online and link it here, that is not enough for being mentioned in Wikipedia). --Dirk Beetstra ^{T C} 05:30, 2 March 2016 (UTC)

@Beetstra, Prevalence, and Stesmo: Thanks everyone for the comments and thoughts. I know everyone wants the best for Wikipedia, and I appreciate any time spent on this. Beetstra, I wanted to address a few of your comments and concerns. (1) A small number of the datasets listed have their own articles — these I will link in. Others are noteworthy and included due to their repeated mention in peer-reviewed machine learning research. FERET is sort of a special case because it is extremely famous. I've made sure that every dataset included on the page has at least one independent, peer-reviewed reference proving its notability. (2) I'm fine removing external links from the first column, but this will leave a large number of the datasets without internal links.

How about I write a script to move all of the external links to the references that are already there for each dataset? Simply appending a "link to this dataset" or "dataset homepage" to each item? Would this be a solution everyone can accept? Thanks again. DATAKEEPER ✉ 20:32, 2 March 2016 (UTC)

As explained above, the links violate WP:EL and WP:NOTYELLOW. Changing the links to references only makes the problem worse by pretending that the links are references when in fact they are there to serve as a directory. The solution is to write the article first, for each notable entry, then link to the relevant article in the list. --Ronz (talk) 23:53, 2 March 2016 (UTC)

@Datakeeper: - I think it would be good to see what happens when you try to link the proper articles. I do hope that the references for the then redlinked ones do show reasonable notability over being just mentioned as 'yet another dataset'. As I said above, everyone can create databases of whatever, that does not mean that it should be in the list. I think that the bar should be that the dataset is being used in some reasonably visible or important way. --Dirk Beetstra ^{T C} 06:31, 3 March 2016 (UTC)

@Ronz: Thanks for the input. As was discussed above I'm fine moving the links out of the table altogether. I suggested what others suggested, moving the links to the end of each reference as a link to "dataset homepage." This could easily be done with a script. I do not plan to change the links in the first column of the article to internal Wikipedia references, because most datasets (even those used in a great deal of research) do not really warrant their own articles and so they don't exist. @Beetstra: this applies to your comments too. I do want to address your concern that "everyone can create databases of whatever." This is definitely true! I wanted to assure you that, as evidenced by the references, every single item in this article is not only of value to machine learning research, but has been referenced in one or more academic journal publications on the topic. This list is not just an indiscriminate group of datasets from various places.

So, to resolve this and avoid violation of WP:NOTYELLOW & WP:EL, there are two options I see:

1. Remove all external links.
2. Move external links from table to tail end of each reference as a "link to the dataset homepage" as was previously preposed.

Because it could be argued from WP:EL that these links are indeed serving as citations, I think the second option is the better of the two. I've heard opinions on both sides and I think the second one will result in a better and more useful article (while abiding by WP:EL) as it grows as a resource for the machine learning community. How about I try it out? I'm fine personally making these edits. Thanks everyone. DATAKEEPER ✉ 19:02, 3 March 2016 (UTC)

Any link that isn't clearly a reference would be blatant linkspam/refspam. Assuming that all links would be appropriate as references ignores our polices and guidelines. This is a common problem Wikipedia has with spammers. Let's be sure to steer clear of what could be seen as an attempt to circumvent Wikipedia's policies and guidelines in in order to promote non-notable and unencyclopedic information.

You're overlooking the obvious solution: delete the article as inherently unencyclopdic per WP:NOT. --Ronz (talk) 19:25, 3 March 2016 (UTC)

@Ronz: Each link clearly is a reference. I do not assume this. Additionally, I assure you that there is no hidden agenda to "circumvent Wikipedia's policies." It is for compliance with these policies that I opened up the discussion here. I appreciate your concern for the quality of Wikipedia and adherence to its policies - I share this concern with you and so have made sure to open up a discussion here when the question of external links was raised on the article under discussion. Also, I strongly disagree with the notion that the article should be deleted. This discussion is centered around external links, not the deletion of the article as a whole. Thank you for stating your opinion on that matter. If the consensus seems to be that the external links should go, I'll be glad to remove them. DATAKEEPER ✉ 19:45, 3 March 2016 (UTC)

I looked at the first link in List of datasets for machine learning research, and it doesn't look like a reference to me. Is it an exception or the rule?

Lists of non-notable entries are problematic on a number of levels, comparisons worse. They need to meet WP:N and WP:NOT to start. --Ronz (talk) 20:40, 3 March 2016 (UTC)

As discussed above FERET is actually one of the very few datasets with its own article. As for the link being a reference: the given link verifies the contents of several of the columns in that row. DATAKEEPER ✉ 22:12, 3 March 2016 (UTC)

As I said, I'm fine removing the links if that is the consensus. It seems to be the majority. I'd be happy to resolve this. DATAKEEPER ✉ 22:12, 3 March 2016 (UTC)

@Datakeeper and Ronz:. I have now looked at the second item in the list (that the first one is notable is already established, we have an article on it). We are talking there about Pose, Illumination, and Expression (PIE), which is a database created by Ralph Gross, with the help of PhD student Terence Sim and under supervision of Simon Baker. You reference that to "Sim, Terence, Simon Baker, and Maan Bsat. "The CMU pose, illumination, and expression (PIE) database." Automatic Face and Gesture Recognition, 2002. Proceedings. Fifth IEEE International Conference on. IEEE, 2002.", which is obviously not an independent reference. The next 4 references clearly state the same names as people who created the database in question. I state above "I do hope that the references for the then redlinked ones do show reasonable notability over being just mentioned as 'yet another dataset'. As I said above, everyone can create databases of whatever, that does not mean that it should be in the list. I think that the bar should be that the dataset is being used in some reasonably visible or important way." For the first handful of items below FERET you have not been able to show that (you merely show existence, which is not enough), nor do I think that these databases are having their own Wikipedia article. I will tag the article with {{primarysources}} and {{notability}} - and I do hope that you will resolve the issues. --Dirk Beetstra ^{T C} 03:41, 6 March 2016 (UTC)

@Beetstra: Thanks for looking the article over. I appreciate the keen eye and feedback. The topic of that article - datasets used in machine learning research - is a notable one - this is established in the introduction. Data plays a key role in shaping the discoveries in this field, and so famous datasets are ones that contribute to significant discoveries in the field. As machine learning and artificial intelligence continue to advance, the availability of high-quality datasets for research will be one of the key drivers of innovation. Due to the unarguable importance of groudbreaking datasets to this field as evidenced by the literature, the notability of this article is established. High quality datasets have significant coverage in reliable (peer-reviewed and other), independent sources.

The inclusion criteria for this list is: included datasets have been used in academic peer-reviewed literature or books. This is to prevent inclusion of various, not-noteworthy "yet another dataset[s]." For them to make it into literature and be cited, they have to be significant. The purpose of the reference column in the article is to show that it is worthy of inclusion in the list. As for your example of "The CMU pose, illumination, and expression (PIE) database," a Google Scholar search (here) will shows that the reference included has been cited 2429 times in academic literature, which makes it extremely significant. The reason I included the primary source is so people can find later find the dataset, but perhaps this would be better to be left Google searching. I'll make sure to go through the sources and add secondary sources to the places where primary sources are used. Since this discussion no longer centers around external links, let's take any further discussion to the talk page of the article. Thanks again everyone for the great discussion. DATAKEEPER ✉ 05:40, 6 March 2016 (UTC)

@Datakeeper: and that is exactly where the problem lies - 'included datasets have been used in academic peer-reviewed literature or books' alone is not enough. Notability means that independent parties have used the database for some independent research or real-life use. The references that are there are not showing that, and even having 2429 articles referring to that paper is not enough. If all of those are 'look, there are more databases with faces' then that is not an independent use of the information, let alone that that 'use' is notable enough to be included in Wikipedia. Moreover, they may be referring to the techniques used in the articles, not to the database itself. I will grant you, with 2429 references referring to the article there are prone to be a couple which show independent importance, but the current list of references does not show that at all - and hence you are not showing notability of the individual items.

I am not questioning the notability of the subject, but a list article is not a proper place to show notability of the subject. Machine learning is the place for that. --Dirk Beetstra ^{T C} 05:57, 6 March 2016 (UTC)

@Beetstra: Thanks for the clarification, I will cite secondary, independent sources that use the datasets for their real-life research. Continuing with our example of the second dataset, you will see that many of those sources are real-life applications of the dataset to train machine learning algorithms to perform tasks. Additionally, this article is distinct from Machine Learning in that it is a list of the most important datasets available in the field. As the coverage of machine learning and its sub-topics increases on Wikipedia, I expect many of these datasets could have their own articles. I'll be sure to put in some edits soon to address your concerns. DATAKEEPER ✉ 06:42, 6 March 2016 (UTC)

@Datakeeper: What I meant was, that the base subject is notable, and that the databases are of importance is quite clear. However, if that can be covered by 3 or 4 databases that show significant real life notability then there is no need for the list-article (that can be part of the base subject). There can be 2429 references to the paper of #2 .. but if none of those scientific papers (or mainstream reports outside of this count) show that that specific database is currently in use in some real-life form, then the notability may not be sufficient for Wikipedia to include them in the list. Being used in the 'grander scheme of things' does not necessarily make the parts notable either. We'll see what happens to the 200 items in the list now. --Dirk Beetstra ^{T C} 07:59, 6 March 2016 (UTC)

List of company registers[edit]

Well, this is my first time behind the curtain here at Wikipedia (long-time user and financial contributor) so not sure how to go about this. Someone, and I won't mention whom because apparently I have to notify them with some string of characters I don't understand; someone keeps removing the links on this very useful page that lists all government registry websites and when you click on the name it takes you there. Apparently there are some pedantic reasons why the links shouldn't be there but it kind of makes the article useful to people who use it. I read the WP:EL thingy and I really think there is enough wiggle room to allow the links to stay. Especially considering that every time someone takes out the links someone puts them back in. Now I wonder is it the people who use the article that are taking them out or is it someone who really has no use at all for an article about company registers? Its locked now so I'm here trying to get it unlocked and the links put back in. — Preceding unsigned comment added by The Westernerer (talk • contribs) 19:13, 9 March 2016 (UTC)

• List of company registers (edit | talk | history | protect | delete | links | watch | logs | views)

Wikipedia is not a directory. The links don't belong.

Given Company registers doesn't exist, why does this list? --Ronz (talk) 21:23, 9 March 2016 (UTC)

• Well, how about instead of deleting nearly 40,000 carefully gathered and useful links they get moved to External Links? — Preceding unsigned comment added by The Westernerer (talk • contribs) 21:49, 9 March 2016 (UTC) The Westernerer (talk) 21:51, 9 March 2016 (UTC)

Exactly the same reasons as datasets discussion above. --Ronz (talk) 01:45, 10 March 2016 (UTC)

I don't see the reasons you are referring to. It seems that 'the dataset is being used in some reasonably visible or important way' to me. Especially registries where there is only a single source per nation/state. The Westernerer (talk) 16:05, 13 March 2016 (UTC)

There are a couple threads in that discussion. One being an interesting "how to improve the article/list entry Notability requirements" (the discussion about sources, etc.). Another being "Can the Lists of datasets article have external links in the list", which resulted in a No. Datakeeper then removed all of the external links from the article and has been working to improve the article. Stesmo (talk) 18:30, 13 March 2016 (UTC)

• I'm here to disagree again with the narrow reading of WP:EL. Specific articles as indicated in the datasets entry, maybe, but I don't see how this applies everywhere. I think List of company registers more closely resembles this list: https://en.wikipedia.org/wiki/List_of_newspapers_in_Sri_Lanka and should have the external links included. Or, again, if you, with the keys to but not the need of this list, could allow the external links to be moved to External links instead of erasing them out right a list that serves the public good can be saved. Considering how often the edits were reversed we can see the clear usefulness of this list for any one other than a wikipedia editor.The Westernerer (talk) 22:36, 19 March 2016 (UTC)

Thank you for pointing out the newspaper article with external links in the body of the article / in a stand-alone list. I've fixed that article as well. You seem to be ignoring that the List of company registers article *already has external links in the EL section* pointing to lists of company registers. There is no reason to add hundreds more to the EL section or the body of the article. Stesmo (talk) 17:31, 20 March 2016 (UTC)

Ah, instead of 'Thank you,' it seems you actually just meant that you're thankful in general (unless you misunderstood my intentions there?). Anyway, thanks for only mostly dismissing my thoughts, I understand I've no standing here and you think you already have a large enough body of discussion to justify editing and locking the lists (or articles as you keep calling them) as per the pertinent panel's conclusions. Oh, one last thing, is this related to the bitterness about the search engine brouhaha that's got everyone pissed here or did this list just get unlucky at random? The Westernerer (talk) 07:19, 21 March 2016 (UTC)

Furthermore, those links at the bottom are far from comprehensive and are hardly a substitute for the comprehensive volunteer maintained list here.The Westernerer (talk) 17:23, 21 March 2016 (UTC)

Pinging recent-ish contributors: @1.187.250.208:@106.220.72.36:@121.242.29.87:@139.149.1.231:@161.10.40.41:@168.168.33.250:@181.58.19.21:@186.72.109.243:@212.91.12.4:@62.190.147.220:@66.44.40.169:@78.155.36.74:@90.61.182.254:@91.196.215.202:A455bcd9Abhisheksingh8747Ale-sandroBD2412CharybdiszDavid8302DewritechDumbBOTFrenchmalawiGOLDLOANS1Ktr101Mean as custardMelanieNOnel5969Robin of locksleyRprprSERutherfordStephenM.S.LaiStesmoTamilMuthuWillyinnorway Bazj (talk) 14:14, 8 April 2016 (UTC)

Keep the link for each register where as yet no article exists for it. It seems to me the whole purpose of WP:EL is to stop the addition of a whole load of promotional crud. Company Registers or Corporate Registries or whatever they're called in each jurisdiction are government agencies performing a legal function, not businesses plying for trade. Short of creating a stub article for each corporate registry, how else is this material to be presented?

The article is useful for verifying articles about companies. It's too valuable a resource to be eviscerated just for a zealous adherence to WP:EL. Bazj (talk) 14:29, 8 April 2016 (UTC)

Could you address WP:NOTDIRECTORY? --Ronz (talk) 18:13, 8 April 2016 (UTC)

Ronz, That would be an argument against keeping the article at all rather than addressing the external links within the article. But since you ask, certainly I can...

Government agencies: Lists of diplomatic missions which links to 200ish lists of diplomatic missions. I think 200² diplomatic missions dwarfs the scope of 200 company registers.

List of central banks.

Others: Lists of hospitals in Africa, Lists of hospitals in Asia, Lists of hospitals in Europe, Lists of hospitals in North America, Lists of hospitals in Oceania, Lists of hospitals in South America, Lists of military installations, Lists of wind farms by country, Lists of universities and colleges, Lists of banks, Lists of companies, Lists of corporate headquarters by city, Lists of cathedrals, Lists of mosques, Lists of magazines, Lists of newspapers, Lists of radio stations, Lists of television stations in North America, Lists of restaurants, Lists of curling clubs... ad nauseam. Bazj (talk) 19:14, 8 April 2016 (UTC)

WP:OSE, but thanks for the list of articles. No, it's not about keeping the article specifically, though that is indeed an option here.

None are relevant. None have links like List of company registers. All you've done is show that the article under discussion is the outlier. So why does this directory of links deserve to be an exception? --Ronz (talk) 21:52, 8 April 2016 (UTC)

Two had some minor problems with their External links sections, which I removed [1] [2], though they are not directly relevant to this discussion. --Ronz (talk) 21:45, 8 April 2016 (UTC)

• Keep, per the rationale presented by Bazj above. Kevin Rutherford (talk) 16:51, 8 April 2016 (UTC)

Per WP:ELBURDEN, I've restored the article to a version without the links. I don't see consensus swinging toward inclusion, and editors have had a month to make a case. --Ronz (talk) 22:01, 8 April 2016 (UTC)

MetroLyrics songwriters miscredits[edit]

The featured article about "Something", a popular Beatles song written by George Harrison, contains an external link to MetroLyrics, an online song lyrics provider. Along with the lyrics, MetroLyrics includes "SONGWRITERS MAX GREEN, CRAIG EDWARD MABBITT, BRIAN MONEY, ROBERT ORTIZ, JOHN FELDMANN".[3] This is not an isolated error. MetroLyrics miscredits songwriters for several well-known songs, including:

• "Stairway to Heaven" (Led Zeppelin) – HOWARD GREENFIELD, NEIL SEDAKA[4]
• "Good Vibrations" (Beach Boys) – HARTMAN, DAN/WAHLBERG, DONALD EDMOND/WAHLBERG, MARK[5]
• "Johnny B. Goode" (Chuck Berry) – CHARLIE DANIELS, BRENDAN COKER[6]
• "L.A. Woman" (The Doors) – JAMES[7]
• "Money" (Pink Floyd) – GOAPELE K. MOHLABANE, JAMES RYAN WUIHUN HO[8]
• "Green Onions" (Booker T. instrumental) – with lyrics by KRIS KRISTOFFERSON[9]

MetroLyrics stresses that it licenses lyrics: "Leading The Way – MetroLyrics was the first lyrics site to provide users with licensed song lyrics and to compensate copyright holders for the content through its partnership with Gracenote."[10] However, their links seem to go against WP policies and guidelines. WP:ELNEVER states "material that violates the copyrights of others per contributors' rights and obligations should not be linked." How are the copyright holders rights being respected (and compensated) if they are misidentified? WP:ELNO provides "one should generally avoid providing external links to: ... 2. Any site that misleads the reader by use of factually inaccurate material ..." Links to inaccurate information, especially in WP:Featured articles, reflects poorly on WP.

Most of the external links to MetroLyrics were added to song articles by User:LyricsBot (the "Something" link was added 7 September 2013)[11] and were not subject to fact checking. Do song lyrics links meet WP policies and guidelines if the lyrics are mostly correct, but the songwriters are miscredited?
—Ojorojo (talk) 21:11, 12 March 2016 (UTC)

• They do seem extraordinary errors for the site to be making, and one naturally queries its quality and integrity as a result. The statements included at MetroLyrics would testify to its worth, though. Is there a way to contact the site perhaps? (he asks without bothering to look), because these are all major errors but ones that would be easy for an admin there to fix. JG66 (talk) 03:47, 13 March 2016 (UTC)

• MetroLyrics appears to lack a system of fact checking and oversight. I started noticing problems with links to the site about two years ago while preparing song articles for Jimi Hendrix GANs. I stopped counting after finding the wrong songwriters credited with about 20 of his songs (identified with the LyricFind logo, which is supposed to indicate proper licensing). Many more songs without the LF logo, usually supplied by site users, were also incorrect or missing credits. I thought that this might be particular to Hendrix, but found many more miscredits for other songwriters as well. Two years later these errors seem to remain. WP volunteers might be able to point out some specific mistakes, but the underlying problem needs to be addressed. MetroLyrics is a commercial (and apparently profitable) enterprise and should have the resources to ensure the quality of their product. —Ojorojo (talk) 17:34, 13 March 2016 (UTC)

• Those credits for "Something" appear to be the credits for the Escape the Fate song titled "Something" from their album This War Is Ours, and the "Good Vibrations" example are clearly the credits from "Good Vibrations" from Marky Mark and the Funky Bunch. It could just be a glitch in their system where the most recent entry of a song with the same name messes up older entries or something. Might be worth looking into before removing these links from 20,000 pages. Have you tried their Contact Us page? Fezmar9 (talk) 15:54, 17 March 2016 (UTC)

• Not sure why the writers get mixed up (is a Neil Sedaka release of his "Stairway to Heaven" more recent than one from LZ?) The song lyrics links on used WP are supposed to be properly licensed (indicated by the LyricFind logo). Anyway, no one is suggesting that all MetroLyrics links be removed, but I don't think that song links which have been identified as having the wrong songwriters should be used in WP articles. I brought this problem up with the bot operator and his final comment was "Unfortunately most of the sources with correct lyrics are illegal, and most of the sources with legal licensed lyrics are incorrect! It is a tragic state of affairs and I don't see any resolution."[12] I assumed he was in communication with ML and the problem wasn't going to get fixed. Maybe someone in an official WP position could raise the issue with MetroLyrics, otherwise it would probably get ignored. —Ojorojo (talk) 17:29, 17 March 2016 (UTC)

Aren't you saying that MetroLyrics is no longer a reliable source? Glitch or not, the same mistakes haven't been corrected in over 2 years (AGF on above). --Richhoncho (talk) 11:29, 19 March 2016 (UTC)

• With all the errors I've come across, MetroLyrics should not be considered a reliable source for songwriter credits. I haven't compared many of the actual lyrics for accuracy, but sometimes their lyrics are for a different song (for Hendrix's "Stepping Stone", the lyrics and writers are for another song, but they get the album right[13]). A recommended list of lyrics providers (similar to WP:GOODCHARTS and WP:ALBUM/SOURCES) that have been checked for accuracy as well as licensing is a possible solution. Then article editors could choose to include a link or reference it if it appears correct. —Ojorojo (talk) 19:13, 19 March 2016 (UTC)

Is Goodreads an appropriate EL?[edit]

Hi. I have been discussing with @Srich32977: about his additions of a Goodreads EL template to BLP articles. See our discussion here. I think this is a judgement call, what do others thinks?

I'm inclined towards it failing Links normally to be avoided on points 1, 10, 14 and 17.

Example of some of the links added:

• George_Stigler#External_links
• Adam_Smith#External_links
• Bruce_Yandle#External_links

Thanks. --Escape Orbit ^(Talk) 12:14, 30 March 2016 (UTC)

• I think it's a judgement call. If the article is about an author, it may be a good way to include a bibliography without going to the effort of creating a full bibliography in the article. Obviously a full bibliography and the Goodreads EL should not exist in the same article. In the case of George Stigler, it serves no purpose since there is already a more complete Bibliography in the article. In the case of Adam Smith, it seems to be redundant with other ELs and the Further reading section. In the case of Bruce Yandle, I would exclude the Goodreads link and simply add the missing few publications to the article.- MrX 12:34, 30 March 2016 (UTC)

Could someone point out a case where it should be kept, other than as a placeholder for better information in a low quality article? If anyone thinks it is reliable, could they explain why? --Ronz (talk) 16:32, 30 March 2016 (UTC)

• Comments: 1. There are two variations of the {{tl:Goodreads}} template. One is '{{Goodreads author}}' and one for particular '{{Goodreads book}}'. The 'author' Goodreads pages do not have links to vendors. The 'books' Goodreads pages have links to vendors and WorldCat. 2. In the case of George Stigler, above, Goodreads lists 27 distinct books by Stigler, whereas the WP article lists 13 books and other articles. With this in mind the Goodreads page is useful for any editor who wishes to expand the article and/or verify ISBNs or other publishing data. (@MrX: would expanding the WP listing on Yandle include all of the same data that Goodreads conveniently provides?) 3. The data contained in the Goodreads listings is certainly reliable – the project lists 125 staff members who manage it. (Also, any WP editor can cross-check the data by clicking the ISBN listing.) 4. Editing of Goodreads pages can be done by Goodreads members (non-staff) if they achieve "Librarian" status. "Super Librarians" have more editing privileges. 5. With these comments in mind, I posit that the use of both templates (books and authors) is WP:NOTEWORTHY and desirable. – S. Rich (talk) 20:47, 30 March 2016 (UTC) Modified to supply template links. 12:42, 31 March 2016 (UTC)

Could you please provide some best-case examples?

It's user-editable with qualifications? Any discussions at RSN about this or similar cases?

We don't add external links to help an article's expansion, though I certainly don't mind for the case I already noted, low-quality articles. --Ronz (talk) 22:17, 30 March 2016 (UTC)

@S. Rich: I would not necessarily list everything on Goodreads in a Wikipedia article. Some authors are very prolific, for example, Edwy Searles Brooks who authored more than 800 published works. In fact, a case like this would be a perfect application of using an external link to a Goodreads bibliography, except that I see they only list one of his works.- MrX 22:55, 30 March 2016 (UTC)

I'd say Brooks supports my point about the usefulness of the links. He's got one book listed under Brooks, but the page links to his various pen names. Thus the site qualifies as one "that contain neutral and accurate material that is relevant to an encyclopedic understanding of the subject and cannot be integrated into the Wikipedia article due to ... amount of detail (such as professional athlete statistics, movie or television credits, interview transcripts, or online textbooks), or other reasons. (WP:ELYES No. 3. Emphasis added.) Thus it can serve to help with an article's expansion. – S. Rich (talk) 00:01, 31 March 2016 (UTC)

So maybe on a case-by-case basis, for low quality articles that have no better sources for the same information. --Ronz (talk) 15:50, 31 March 2016 (UTC)

Goodreads is a wiki and what's worse, it presents quantitative rankings that aggregate the "votes" of site visitors. There's no expert editorial control and no authority who's responsible for the content. Moreover, while there are numerous sources for highly notable authors such as Adam Smith, there are few such RS lists of the publications of less notable or borderline writers. Thus, the use of Goodreads or similar sites distorts the encyclopedia by adding WP:UNDUE content with the implication that the Goodreads page lists noteworthy works and valid commentary on the authors. Goodreads should never be used as an external link. SPECIFICO talk 13:05, 31 March 2016 (UTC)

I agree. The material is neither neutral nor accurate enough to meet ELYES#3 criteria. If it's being added to help article expansion, add it to the talk page. --Ronz (talk) 15:50, 31 March 2016 (UTC)

These comments are not correct. The basic (bibliographic) data in Goodreads is controlled by the 125 staff members. It is not a wiki because the only content that members can edit is their own commentary. Those user comments and ratings are but one aspect of Goodreads. There is no implication that only noteworthy words are listed. The commentary provided by members is no more valid that the commentary we see here on this talk page. (The members are simply people who have read particular works and given their reactions.) So what if "less notable or borderline" authors are listed? WorldCat lists a lot of writers too, and it provides a system for rating and reviewing works. The big difference is that a number of books and writers listed on Goodreads have such ratings whereas WorldCat users do not generally contribute. – S. Rich (talk) 21:18, 31 March 2016 (UTC)

• I'm undecided on this. Goodreads is fairly easy to edit and users only need to shelve 50 books before they can petition for librarian status. It's really not difficult to get this. Heck, I have librarian status. With this status I can edit just about anything on there. There are very few pages that can't be edited in some fashion and the only ones you can't edit easily would be author pages where the author has signed up for an account. Only the author can edit those and they can put whatever they want in those infoboxes. If they wanted to say that they're the second coming, they can. Goodreads is relatively easygoing and a user has to really, really give off an impression of ill will to not get approved as a librarian if they've otherwise done everything they should.

When it comes to how correct Goodreads is, there have been cases of people adding incorrect information to pages, sometimes on purpose and sometimes not. Most of the time it's not and it's just a case of people changing something because they heard a rumor somewhere about a book plot. Other times (like Scott Sigler's Nocturnal) it's well-intentioned librarians merging two completely separate books because they share a title and an author. (In the case of Nocturnal Sigler had a rough draft podcast that was 80-90% different from the printed novel, yet people kept merging them with the assumption that they were the same work when they weren't.)

Review and opinion-wise, this is where it gets slightly iffy. Many of the reviewers on Goodreads give honest opinions, but you also get a lot of people who deliberately try to sway reviews one way or another. Sometimes it's authors trying to raise their star rating (which happens relatively frequently) and other times it's people trying to lower ratings because the author did something insanely stupid like attack a reviewer. (This is actually the center of the controversy section on the Goodreads article.) This is something to take into consideration to some extent, although what I'm more worried about is whether or not the site is correct enough to be listed.

In the end I'm kind of halfsies on this. I typically don't use it in articles because it's so often used in order to promote something and people will occasionally try using it in the article as a RS. It's sort of problematic in the way that IMDb is problematic, however the big difference between IMDb and Goodreads is that all changes need to be approved by someone on the IMDb staff before it's implemented. You don't have this filter with Goodreads and I could make a change right now and have it instantly get posted. IMDb is frequently problematic because their quality checks are terrible (the whole Chaneyverse debacle is a good example of IMDb not checking properly, as is Gretel Ashzinger), but they do have just enough of a checking process to where it isn't completely problematic.

I'm not against this being used as an external link, but I do want to make sure that everything is taken into consideration. Tokyogirl79 (｡◕‿◕｡) 08:18, 1 April 2016 (UTC)

• Honest reviews is one issue but the aggregation of uninformed public reaction is not consistent with WP editorial standards which require us to use the opinions of qualified authorities on the subjects for which we cite them. Just look at the winners of Goodreads' annual polls for best-in-show of various categories. Plenty of cats and dogs. Not a source for an encyclopedia. We should not direct WP readers to quantitative rankings by the hoi polloi. SPECIFICO talk 21:22, 3 April 2016 (UTC)
  • Quite judgmental to be calling Goodreads members hoi polloi and cats and dogs. And the comment misses the point. These are simple and useful External links related to the authors and/or books. No one is suggesting that the links be used as a reliable source for material in the project. – S. Rich (talk) 02:36, 4 April 2016 (UTC)

• I would in general say that these links fail our inclusion standards - WP:EL: "Some acceptable links include those that contain further research that is accurate and on-topic, information that could not be added to the article for reasons such as copyright or amount of detail, or other meaningful, relevant content that is not suitable for inclusion in an article for reasons unrelated to its accuracy." (my bolding) - this information is very easy to include into our articles, and would enrich our articles. --Dirk Beetstra ^{T C} 05:09, 4 April 2016 (UTC)
  • Hence, as others have said, it is a judgment call. If the Goodreads link (for either author or book) does not contain information already included in the article then it will not be helpful. But if it has additional useful info, then adding as an EL in the EL section will be/can be helpful to readers and editors. – S. Rich (talk) 05:16, 4 April 2016 (UTC)

This is why the guideline says that the usefulness of the EL should be determined not against what information the article has currently, but against what the article could/should contain if it were of FA quality. If an EL may be a useful source for editors, then that's what the talk page is for.

I understand completely your thinking and the rational of adding Goodreads to less than perfect articles; for the benefit of readers. But we need to be careful to keep in mind that Wikipedia is an encyclopaedia, not a links directory. And, when all is said and done, Goodreads is a commercial operation. It does use advertising and gets referrals income from the book retailers it links readers to. --Escape Orbit ^(Talk) 16:15, 4 April 2016 (UTC)

• • • Hence, if there is additional info in the article, then that can mostly be incorporated in the article. Looking at the pages, I don't think there is a lot of additional information on these pages that cannot be incorporated in our pages, and I would argue that most of the encyclopedic information (e.g. the lists of books) should be incorporated, leaving nothing that is not needed to be incorporated. --Dirk Beetstra ^{T C} 05:31, 4 April 2016 (UTC)

• But if an editor is creating or upgrading a stub, without wanting to spend a lot of time adding all the individual books, is the reader better served by a link to Goodreads' listing or not? Sure, the book details "should" all be added to the article, but an editor may not have the time or enthusiasm to do so. PamD 09:44, 4 April 2016 (UTC)

If we were to go by this criteria, there are dozens of possible external websites on every article that a reader may find useful. But Wikipedia is not a directory of useful internet links. And its usefulness shouldn't be measured against a stub article. The guideline specifically, for good reason, makes a point of evaluating the EL against a Featured Article. --Escape Orbit ^(Talk) 16:15, 4 April 2016 (UTC)

Well, that would also be against our inclusion standards per WP:EL ("If the website or page to which you want to link includes information that is not yet a part of the article, consider using it as a source for the article, and citing it." - though this as a reference is probably unsuitable, and you would have to find better references to show that the list is correct, which again in itself reflects on it's suitability as an external link - WP:ELNO #12, open wikis) .. and as to the point of the three articles in this thread, none of them are stubs (and many of which I checked in this aren't - I have yet to actually find a stub in that list of additions). Moreover, in some cases, it amounts to plain linkfarming (e.g. Noam_Chomsky#External_links - disregarding the article contents, what does this add over the other external links that are already there?). If you find good information that is not part of the article, and don't have time to include it, then the time spent on adding them to the external links section is the same as putting a note on the talkpage suggesting the link as a suitable source of information, and generally just as effective (external links tend to stay and hardly being used).

Moreover, if you have hours to spend updating external links sections on 100s of pages (many of which may not even need this external link), you also have time to actually upgrade a handful of articles that really need a proper biography. --Dirk Beetstra ^{T C} 11:46, 4 April 2016 (UTC)

• No. Goodreads is not a reliable source and really adds nothing of encyclopedic value. Drmies (talk) 04:02, 5 April 2016 (UTC)
• No, Goodreads is not an appropriate external link. There may be one or two exceptions, but some pages I checked were clear fails of WP:EL because Wikipedia is not a directory of links that someone thinks are handy, and the linked pages had nothing of value. If someone wanted to expand an article they would try Google, and that would find Goodreads and any other potential clues, so there is no need to include such a dubious link. Johnuniq (talk) 04:43, 5 April 2016 (UTC)
• No. Purge. Srich, your efforts were well-intentioned, but I suggest you remove any such links you have added to WP articles. SPECIFICO talk 13:49, 5 April 2016 (UTC)

• Again, folks are missing the point. NO ONE suggests that Goodreads is a reliable source. It is simply a resource. Here is an example: the article about early writer Mary Hunter Austin has a listing of books with WorldCat links posted. The links give some publication data (based on the feeds that WorldCat gets from libraries). Mary Austin's author listing at Goodreads shows 10 books, and none of them have member ratings or reviews. Goodreads lists her works as culturally significant, and the books listed have more up-to-date publication data. (The listing for Isidro at Goodreads gives us a newer edition (2015) that is not listed by WorldCat.) Now compare this listing to the WorldCat listing: OCLC 861553018. Hmmmmm! WorldCat allows users to review and rate the book. (Shameful!) And WorldCat has a link where users can buy the book. (Evil!!) Do you need more documentation? Look at OCLC 921035182 where we see a user review and links to Amazon and B&N. Come on, per the comments above WP should eliminate WorldCat links from its pages. Is that what you're suggesting? I think not. Having Goodreads links is simply what Lorcan Dempsey calls an "addressable knowledge base". E.g., WP "makes it easy to include in any online communications a pointer to more knowledge on any topic...." Goodreads, like the OCLC, is another pointer. – S. Rich (talk) 15:02, 5 April 2016 (UTC)15:09, 5 April 2016 (UTC)
  • But the moment you say "Goodreads lists her works as culturally significant" you ARE suggesting that it is a reliable source, not just a "valuable addition". I mean, if it's not reliable (and it isn't), then how can it be valuable? Drmies (talk) 19:37, 5 April 2016 (UTC)
    • Again, no, you miss my point. I do not say we can use Goodreads as a WP:reliable source as in "Austin has been found by scholars to be culturally significant".^[1] But it is more or less valuable to editors who wish to use it to track down publication data such as dates, publishers, ISBNs etc. In the context of the External links sections, it is helpful. Nothing more. – S. Rich (talk) 22:14, 5 April 2016 (UTC)
    • FYI: I have posted a notice about this discussion to two of the editors who created or modified the {{tl:Goodreads}} templates. – S. Rich (talk) 22:56, 5 April 2016 (UTC)

Srich, To avoid any appearance of canvassing, I suggest you also post it to RSN and other neutral venues. SPECIFICO talk 23:20, 5 April 2016 (UTC)

@Srich32977: I know you read the suggestion of people complaining that this is not a reliable source, but the basic point here is that this is NOT a proper external link on many of the pages where you added it to, it simply does not give any additional information beyond what is already there, or what can not already be included. You even argue that "Goodreads, like the OCLC, is another pointer", you are hence plainly linkfarming there. It may only be a good temporary external link on some minor stubs where it can be used to expand the article further, but even that is a stretch (we have talkpages for those suggestions).

I think the overall conclusion is here, that you are being asked to remove the links again, and then get first proper consensus before inclusion of this link (per WP:EL). Can you please remove the links and get a proper consensus for the use of these templates? --Dirk Beetstra ^{T C} 03:48, 6 April 2016 (UTC) (ping edit: @Srich32977: --Dirk Beetstra ^{T C} 04:51, 6 April 2016 (UTC))

• In a sense everything in WP is temporary because anyone can edit. If editors don't like the addition of a {{Goodreads author}} or {{Goodreads book}} or {{OCLC}} or {{imdb}} or {{ISSN}} or any one of the hundreds of Category:External link templates, they are free to revert such additions. Doing so is part of the WP:CONACHIEVE process, and we do not go to the talk page to first ask if we can add links. We simply WP:Be bold and use good judgment and add them in accordance with policy and guidelines when we seek to improve articles. We go to the talk page only after another editor objects and we (in good faith) disagree. Editors who promote a blanket prohibition against Goodreads links – because Goodreads allows member comments and ratings and because Goodreads has links to Amazon, Barnes & Noble etc. – are missing the point. More importantly, if their "do not use" rationale was applied to WorldCat-type or Goodreads-type link, then every external link template would be subject to a WP:TFD. – S. Rich (talk) 06:59, 6 April 2016 (UTC)
  • Sure, but editors who persist in adding material that consensus has shown is not suitable are eventually sanctioned. Johnuniq (talk) 07:26, 6 April 2016 (UTC)
    • What?! I have not "persisted in adding material". In fact, no editor has reverted my edits. This is simply a general discussion as to whether adding the author or book Goodreads templates to various articles is appropriate. – S. Rich (talk) 07:38, 6 April 2016 (UTC)
      • @Srich32977: you have used WP:BOLD to add these links, and there is nothing wrong with that (you could have started a discussion and get consensus, or be bold and see what people think - you chose the latter). The point is that there is now a consensus against these inclusions (see it as a WP:BRD cycle where we first discuss and then revert), and we ask you to revert them. And as this thread is now developing (with quite a number of editors supporting the removal), there is no reason why not anyone c/should do it. --Dirk Beetstra ^{T C} 10:55, 6 April 2016 (UTC)

Let's not get things heated here. S. Rich's additions were entirely proper, good faith and bold. He also graciously stopped doing them when I questioned them, so that they may be discussed. No-one has reverted his work because, I would hope, they didn't want to be jumping the gun on this discussion. However, it looks like consensus is broadly against the blanket additions of these ELs. There may be some cases where an EL is justified, and these can be evaluated on their individual merits. --Escape Orbit ^(Talk) 12:49, 6 April 2016 (UTC)

• • • • Also, this is not a blanket objection to Goodreads - there may be cases where the link is appropriate, but in the majority of cases where you added the link the link is superfluous over others (and there may be other cases where the link is there where it is also superfluous, or where it has become superfluous) or over the information that is/can be included in the article. That there are other links (WorldCat etc.) there that are also superfluous in the same reading of WP:NOT/WP:EL is not an argument to add even more of them. That is the whole point of WP:NOT#REPOSITORY and (the intro of) WP:EL. --Dirk Beetstra ^{T C} 10:55, 6 April 2016 (UTC)