Duplicate content and PageRank loss

[nodecode]

It is not good for search engines to use when scrolling URLs like https://community.nodebb.org/topic/72/we-need-your-help/18. As soon as a user this URL copied, and they linked from a website, Google finds duplicate content. In addition, the PageRank goes to different URLs, and not to one single.
See also: https://support.google.com/webmasters/answer/66359?hl=en

There are two ways to resolve this problem:

• The easiest and in my opinion better solution would be to use URLs like https://community.nodebb.org/topic/72/we-need-your-help#18.
• The second solution would be to use rel="canonical". (https://support.google.com/webmasters/answer/139066?hl=en)

[psychobunny]

For your second solution, in that link there is a section under the canonical urls section that say that sitemaps would be another solution. NodeBB does ship with a built in sitemap (found at root/sitemap.xml) which looks something like this:

<url><loc>https://community.nodebb.org/topic/23/youtube-links</loc><lastmod>2013-08-09T08:33:35.285Z</lastmod><changefreq>daily</changefreq><priority>0.6</priority></url>

Adding rel="canonical" isn't a bad idea though, thanks for reporting

[svishi]

@psychobunny somehow this still keeps happening. The latest source code doesn't have the rel=canonical tag on topics. Can this be fixed by adding another JSON block in src\controllers\topics.js at line 290 res.locals.linkTags like so:

	res.locals.linkTags = [
		{
			rel: 'alternate',
			type: 'application/rss+xml',
			href: topicData.rssFeedUrl,
		},
		{
			rel: 'canonical',
			href: nconf.get('url') + '/topic/' + topicData.slug,
		},
	];

[psychobunny]

reopened, RFC

[julianlam]

If we don't already have the link to the canonical topic URL, then we should definitely add one

[svishi]

It's not there. Could we add it in the next release?

[BenLubar]

@barisusakli what about multi-page topics?

[barisusakli]

I think this code was added to handle that.

[BenLubar]

I'm pretty sure Google honors rel="canonical" on pages even if they're linked to from a <link> tag. I would suggest setting the canonical URL to the current post rounded down to a multiple of the page size. That's why I currently have WTDWTF set to use pagination for logged-out users.

[barisusakli]

So canonical url should be as follows when posts per page is 20?

http://forum.com/topic/1/title on page1
http://forum.com/topic/2/title/20 on page2
http://forum.com/topic/2/title/40 on page3 
...

[BenLubar]

Yeah, that would work, assuming the page ending in /20 includes posts 20-39 on load (without any scrolling to load content required)