83

A search for domains like mysite.com provides thousands of results on Stack Overflow and hundreds on Drupal Answers. I sampled some. Each and every one I checked was a post by someone ignorant about http://example.com, http://example.org and .example TLD, as defined in RFC 2606. Other "natural examples" domains are abused pretty much too.

My request is - disallow these:

  • mysite.TLD
  • site.TLD
  • abc.TLD
  • xyz.TLD
  • [one letter repeated].TLD

When they are detected, propose http://example.com or http://example.org. With old questions, use robot or find other way to edit them.


Even if SEO is not a problem, professional sites about Internet and associated technologies should not ignore RFC on such a large scale. Not to mention linking porn like about thousand links to xxx.com - I was already detected as one who visits "sites related to content forbidden here" by a firewall or two thanks to that.

I can assure you: you don't want to explain to your boss why are you searching for "adult content" in working hours, or agrotourism farm owner why that happens under his roof. Even WSODs are not nice. Now I think I have SE whitelisted, but anyway... Sadly, firewall was right, I did have xxx.com mentioned in my search results here and on Drupal Answers.

9
  • 1
    Would programmatic links to images posted on SO, for example, be OK?
    – trashgod
    Jun 28, 2013 at 14:17
  • 3
    @trashgod - links to images and links to related sites should be OK. Problem is with domains that belong to commercial entities and have nothing to do with the question. Especially ones that seems to take short, "natural" name on purpose, as only these are misused in significant amount to care. Also, now all images seems to be served from imgur.com anyway and in no way I'm proposing to disallow imgurl. It's related as it hosts material used in question, so it deserves each and every SEO benefit it may have from that.
    – Mołot
    Jun 28, 2013 at 14:28
  • 13
    I would also suggest to disallow xxx.com. I've had an unpleasant experience.
    – HamZa
    Jun 28, 2013 at 14:37
  • 1
    @HamZa updated, in a more general way. Good for you?
    – Mołot
    Jun 28, 2013 at 14:39
  • 3
    While I think it's a good idea to get people to use the domains specifically set up to be used as examples and not inadvertently drive up the Google Juice of actual sites, I don't see how this will be possible in an automated way. There's just too much chance that something will be changed that shouldn't. Sentient intervention is required. This is probably better off as a clean-up project by a team of volunteers.
    – ale
    Jun 28, 2013 at 15:52
  • 3
    I'm not clear on what the problem is here. You're saying that someone might accidentally arrive on Stack Overflow because they googled for mysite.com? How is that hurting the world?
    – user102937
    Jun 28, 2013 at 17:02
  • 4
    @RobertHarvey I'm saying someone may go to xxx.com from a Stack page. Pretty embarrassing at work. See HamZa's comment. And I'm saying Stack gives traffic and reputation (not mechanical one, just human) to sites people don't even know they are mentioning - on Drupal Answers mysite.com happens to be mentioned as a dummy link also in questions that touches hosting problems - giving them unintentional credit. Last but not least, professional page breaking RFC thousands of times does not look professional.
    – Mołot
    Jun 29, 2013 at 7:14
  • 1
    @HamZa seems you told me why I experience WSODs. Thank you... I guess ;) (question updated)
    – Mołot
    Jul 2, 2013 at 14:23
  • Slightly off-topic question: are there any well-known example domains that have MX records?
    – skeggse
    Apr 30, 2014 at 16:20

2 Answers 2

49

The "adult content" rationale is pretty compelling here. I've blacklisted the following URLs:

...along with variations that include the "www." subdomain, and the .org and .net TLDs.

This doesn't remove existing examples from the site, but will require that they be changed if the post is edited - and will block all new instances from being added. Folks trying to edit any post containing these will now see:

Body cannot contain ... Please use example.com (or .org or .net) for fake URLs.

Which links to this page: Why are certain example URLs like http://site.com/ and http://mysite.com/ blocked from post content?

January 9, 2015: this blacklist is now live on all sites EXCEPT for meta.stackexchange.com.

27
  • 3
    There is a domain which is chosen as a standard example name... example.com (en.wikipedia.org/wiki/Example.com). Maybe we could mention it in a help page?
    – Sklivvz
    Nov 13, 2013 at 2:00
  • 3
    I figured we'd mention it the first time someone asked about the error here on meta, @Sklivvz. If we can get a hook for a help-link into the blacklist though, it'd be worth writing up something for that.
    – Shog9
    Nov 13, 2013 at 2:09
  • 1
    Thank you, gentleman. @Sklivvz I think pretty good explanation of example.com is directly on example.com ;) Simple, direct, and surely comes from authorized source. But link to Wikipedia might work, too. Depends of the detail level you want to give your users.
    – Mołot
    Nov 18, 2013 at 11:27
  • Is it implemented on Drupal Answers? And what about using it in code blocks?
    – Mołot
    Dec 10, 2013 at 9:02
  • Unfortunately without some specific text that says "use example.com" this solution is just confusing. It just tells you that you cant add http in your answer. It look several attempts and a post on meta before I found this info.
    – steve cook
    Mar 20, 2014 at 2:40
  • 2
    We currently have someone going on a suggested edit spree trying to change all the "site.com"s to "example.com"s. Should we approve them?
    – Undo
    Mar 20, 2014 at 5:05
  • @Undo I decided to go ahead and focus the edits on hunting down instances of xxx.com I see
    – jdphenix
    Mar 26, 2014 at 4:21
  • @jdp Yeah, that's much better. I would strongly suggest, though, editing something else in the post to be better - many will reject otherwise.
    – Undo
    Mar 26, 2014 at 4:22
  • @Shog9, please have a look at meta.stackexchange.com/questions/226793/… and let us know what your views are Mar 26, 2014 at 12:58
  • 1
    Done. @Michael. If you need the pattern extended, post a request on meta.sf
    – Shog9
    Jul 23, 2014 at 19:32
  • 1
    Doesn't look like it, @tepples. The network site - abc.go.com - isn't blacklisted.
    – Shog9
    Jan 9, 2015 at 17:47
  • 1
    if they drop the http:// part, then it won't be automatically converted to a link, @Michael - which was the problem we set out to solve here. I should take https links into account though - done.
    – Shog9
    Jan 10, 2015 at 18:49
  • 1
    Try it now, @braiam
    – Shog9
    Jan 11, 2015 at 14:37
  • 1
    I think domain.com should also be blacklisted. Feb 28, 2017 at 15:57
  • 1
    That's a good idea, @Donald. Done.
    – Shog9
    Mar 1, 2017 at 0:54
9

Any links that are actually created accidently like http://mysite.com include a nofollow, so I don't think there are any SEO benefits for those domains anyhow. The only possible problem is someone clicking on said link thinking it goes somewhere useful, but that sounds like an unlikely and insignificant problem to me.

The nofollow is removed if the post is considered reputable enough, so if there truly are a number of "reputable" posts linking to mysite and whatnot, it may warrant a modification of the nofollow removal logic to skip those links.

2
  • 3
    Mołot, although that report indicates that the nofollow attribute might not prevent search engines from indexing a mentioned address, indexing is not Stack Exchange's concern. Stack Exchange has no interest in concealing the existence of the linked page. Rather, the attribute is applied to avoid influencing the ranking of the linked page by having Stack Exchange's link be interpreted as some sort of endorsement. Jun 29, 2013 at 13:10
  • 1
    @RobKennedy OK, even if I'll accept everything you say my non-SEO arguments still apply - porn links like xxx.com and human reputation given to sites authors did not meant to mention are bad. Breaking RFC on page that claims is for pros is not good. So let's say for now I accept your explanation and my request is due to all other reasons provided, OK?
    – Mołot
    Jun 29, 2013 at 20:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .