CyberGene crawl through all the posts in each discussion because the posts are part of the discussion, so it can directly crawl through it.

No, each posts reply hyper link is set to noFollow by flarum that needs to be dofollow. Other wise spider will just ignore going through discussion since it is mentioned as noFollow

When we scroll down the URL will change like 1, 2, 20th post..etc which is technically a completely different page. Though flarum considers as post link according to SEO it is not.

      Hari there seems to be a contradiction between what you say and what matteocontrini said. Seems like Google is able to crawl through all the posts in a discussion. If that’s the case, then there’s no problem.

          What @matteocontrini says is there’s some Flarum bug and Google mixes posts but is still able to go through all of them. Which means it can do so without dofollow.

          • Hari replied to this.

            Hari I think there are two separate issues here. One is regarding the search bots crawling discussion pages (should be fixed). The other is whether internal links should be dofollow. I’m still not convinced about the latter but I don’t care, it won’t hurt, so whatever Flarum devs decide.

            • Hari replied to this.
                7 days later

                General

                I will say this not every forum in the world is going to rank top and perform as each one of us is expecting to become. The time flarum developers have consumed developing the core there's no one here can actually be considerable near to pay for their work ✌️ They have setup almost everything that is needed or you can dream off

                Don't sleep on your work and expect magic to happen over night. Even a worse/bad created script can receive much help from you, this could be how you portray either an article or a service that is in demand, and geo targeting.

                If its an article, check whether does the title correspond with your message, are there similar words that can be found? Are there keywords inside sentences. Actually these studies are not for free that's why the upcoming find it hard in the game. They either trying to form their own new keywords or bad grammar. Internet is vast, don't accept that to happen to you.

                When you earn money do you usually eat it all? 😅😅 Or invest it? Where do you invest it? It matters a lot if you don't consider to invest it in your website. Again investing in your forum is a huge topic! You can immediately know a good performing forum just by how you found it. Its not easy. There many out there which are seeking the same position day and night.

                When you start a marathon posting discussions just know also you'll have to go deeper in your pocket. Even crawlers do get tired ☹️ And with geo it doesn't mean you would rank worldwide, you have to bounce with those graphs there no favors

                Let them come and ask how you did that? 💚

                Hari the PR to core was merged, making the changes available for 1.4+. I'll probably make my test extension available for Flarum v.1.4 and then move it to another team with more options.

                I've now also published the example extension: https://packagist.org/packages/luceos/flarum-ext-dofollow. This extension only works on Flarum v1.4 and up. It only acts as an example how the new Link extender in core works, there's no configuration, no settings. But you can mimic what I've done in a local extender for complete flexibility and other extensions might provide an actual ux for this feature.

                The extension works like this, in case the link is from the same domain as the community it:

                • sets target to _self
                • sets rel to ugc

                Otherwise it:

                • sets target to _blank
                • sets rel to noopener ugc nofollow

                    thanks a lot for stepping into this and solving it 🙇🏼‍♂️

                    luceos same domain

                    you mean domain, sub domain, sub directory as same domain right?

                    🚀 ♥🧡💗💓❤

                        Hari you mean domain, sub domain, sub directory as same domain right?

                        Just a very quick review of the code (didn't dive to deep) it would be whatever the URL of the forum is under, so for discuss it will match discuss.flarum.org but not flarum.org (assuming I remember how the functions work)

                          CyberGene I just don’t understand why the crawler needs internal links. The crawler already has all the discussions from the sitemap and will crawl through all the posts in each discussion because the posts are part of the discussion, so it can directly crawl through them. It doesn’t need internal cross-post links to function properly.

                          This is not about indexing, but positioning in search results. Internal links increase "strength" of linked pages, which increase its position in search results. Google may be able to index all forum with "nofollow" on all internal links, but position of some links may be lower than without "nofollow" on internal links.

                              rob006 has this been confirmed by Google? To me it seems very easy for a website to create fake ranking by creating zillions of internal cross-links to trick Google into thinking the website links are high-ranked but I highly doubt Google would be so stupid to work this way. Only external links should matter for ranking.

                                  CyberGene has this been confirmed by Google?

                                  I'm not sure if Google officially confirmed it (they rarely share how their internal algorithm works), but I feel like this is a common knowledge, you can read about it on any SEO blog: https://yoast.com/internal-linking-for-seo-why-and-how/, and I've seen how improvements in internal linking improve position in search results.

                                  To me it seems very easy for a website to create fake ranking by creating zillions of internal cross-links to trick Google into thinking the website links are high-ranked but I highly doubt Google would be so stupid to work this way.

                                  That would be true only if you implement it in dumb way. But it does not work as simple "one internal link = 1 point, 1 million links = 1 million points". Every page shares its own link juice, which is shared between all links. The more links you have on one page, the less value each link will have. Also "amount" of link juice depends on strength of particular page. If you have garbage page with millions of inks, you get nothing from internal links. But if you have regular page with natural linking, this allows you to bump some pages and suggest which pages are more important. For example if you have support-like forum, where you have one canonical discussion with question and answer, and 10 duplicated discussions with the same question, you close duplicates with link to canonical discussion. In this way you suggest, that if someone search the same question in google, search results should promote canonical discussion over one of 10 duplicates (because canonical discussion have more internal links).

                                  Only external links should matter for ranking.

                                  External link are not that much more reliable - you still could have farms of links on external domains, so Google needs to mitigate linking manipulations anyway.

                                    6 days later

                                    CyberGene I just don’t understand why the crawler needs internal links

                                    Because that's how link juice flows. It's a part of on-page SEO. It can't be ignored.

                                      8 days later

                                      So, I installed the extension after the upgrade to 1.4. I have a 1.3 copy of my forum. I opened one and same discussion on both and compared page sources. They are the same. For instance, if someone replies to another member, the link to that is not having “nofollow” both in 1.3 and 1.4. I really can’t see what the original problem was.

                                      Well, I checked that if someone explicitly posts a link to another discussion, then indeed in 1.3 there’s a “noopener” whereas in 1.4 there isn’t.

                                      So, was that the big deal? Posting explicit links to other discussions? If that’s the case, this seems like a very minor issue for my forum since I have probably 10 such links on my forum out of 15k posts. Or am I missing something here?

                                      @Hari and the rest who really wanted this one. Let’s clarify what exactly was needed because I’m confused.

                                      I have registered a sitemap with Google. All discussions are listed there. When the crawler starts its job, it will go through each discussion URL from the sitemap. Then in order to crawl all the posts (since there’s paging) it will have to follow the links to the other pages. My understanding was those links to the other pages are “noopener” and as a result Google is unable to crawl through all the posts. Hence your request here. However it turns out that is entirely not true. Am I right? Or am I missing something?

                                      On a further inspection it seems like the only applicable scenario is when a user copied a link to another discussion/post and pasted it explicitly as link in a discussion/post. Is that correct? If so, then I really misread the stuff and I don’t need the new extension at all. On my forum we discuss pianos and sometimes we may say “look at what we already discussed in that other discussion” by linking to it but it’s so rare it could never have happened as well for the entire life of the forum.

                                      @luceos what is your opinion?

                                        CyberGene I strongly recommend to read this thread, it was already explained what it is about. And no, it is not about "noopener".

                                            CyberGene There is no link, because you are already in discussion I was talking about. I already explaied why it matters. It also does not cost anything and it could improve SEO. So amount of posts advocating against this is just mind boggling. It really should be straightforward bugfix included in core - since core is responsible for adding nofollow rel tag, it should do it properly, without compromising SEO.