Future of Canonical urls / JM Canonical plugin / Forums


  • I have some reservations about the usefulness of this plugin. Canonical link should appear when another page is served depending on the server request. This is not the case.
     

  • @net1313 You could be fully right. I could get it wrong, but now I have no problems with double content.

    by the way - where option split topic went?

    My sources:
    https://moz.com/learn/seo/canonicalization
    https://yoast.com/rel-canonical/

    With e107 you have the same content displayed on different URLs. And with sef-urls, you can have almost any text in URL and your content is still displayed.

    301 redirection is just a bonus, not the purpose of this plugin. The same with sitemaps.

    This was the reason why I started with it and I tried to understand how now e107 URLs work. You can't relay on e_PAGE anymore. Almost every part works a different way...

    I couldn't get my homepage indexed by google:
    Submitted URL not selected as canonical: The URL is one of a set of duplicate URLs without an explicitly marked canonical page. You explicitly asked this URL to be indexed, but because it is a duplicate, and Google thinks that another URL is a better candidate for canonical, Google did not index this URL. Instead, we indexed the canonical that we selected. The difference between this status and "Google chose different canonical than user" is that, in this case, you explicitly requested indexing.





     

  • This is a vast subject of discussion "Canonical link". As I mentioned before, not only user can request a page, but more important - bots.
    And the request can be something like this: (e.g. news) - "default", "extended", your_site.com/news/default0.0/news_item9999
    but e107 will render your_site.com/news_item.

    Both request will be valid and no canonical link appears.
    In your case, with Canonical plugin, only: "your_site.com/news_item" could show "rel canonical" but not when requested another "bad" news item.

    Don't forget about pdf or email url of "news". Your plugin will not generate canonical links.

    My approach to solve this issue is different. But it is only mine, cause I'm paranoid about solicitation of DB. One more database request with others plugins and with a core is too much for me.

    I solved this in v1.
    Each canonical page/news has its own approved url as a line in text file (generated automatically). The server request was compared and if no matches was found then canonical link appears in header.
    Even more. But I don't want to discribe all details.

    Maybe we can find a solution in V2. For example by comparing core prefs url settings with a request.
    I mean to compare with a pattern. Something like this. I dont' know now.
     

  • Example for this news item, please (I am just trying to understand where is the problem - except next db query):

    official URL: https://www.e107.sk/news/view/42/gaia-theme/openmind-theme-demo-available
    (works)
    canonical URL: (with the backslash)
    https://www.e107.sk/news/view/42/gaia-theme/openmind-theme-demo-available/
    (works)
    https://www.e107.sk/news/view/42/anytexthere
    (works)
    legacy URL
    https://www.e107.sk/news.php?extend.42
    (works)

    All URLs have the same canonical URLs because they display the same content.

    URL https://www.e107.sk/news.php?default.42 - doesn't exist and it redirects to news homepage with core e107, nothing with canonical URLs.


    Don't forget about pdf or email URL of "news". Your plugin will not generate canonical links.
    I forgot. It will be added some new versions :) Maybe, so little time.

    For example by comparing core prefs URL settings with a request.
    Yes, but there must be some system (logic) in URL settings, each part works differently. Nightmare.
    With version 1 you could get always correct content with combination e_PAGE +e_QUERY, now not. And as soon you solve news + pages, then you will go on with the forum and you can start from the beginning. Not mention gallery plugin.

    For me, this was just workaround. They will not solve this in the core and official plugin solves only news. And it didn't worked before.

    I will add a new relational table between canonical URLs and gsitemap table. Why? Gsitemap table doesn't know ID of related record so you don't know if title or URL weren't changed. Or automatically add new records to the gsitemap page. It is planned for 2020 :) And of course, I asked for this to add in core (just 2 fields - source + ID), no it will not be added.

    I need those tables in admin area mainly, but it doesn't mean that that txt file can't be generated from them (in fact it should be very easy) but the problem (at least for me) still left how to from requested URL get the correct canonical link.

     

  • I will add a new relational table between canonical URLs and gsitemap table.
    Yes! exactly what i thought. Only this way.
     

Moderator(s):
Search just forum
Chatbox
You must be logged in to post comments on this site - please either log in from the Login box or from here.


  • Jimako
    Jimako  2 months ago

    @wrxxy I am ill, sorry cant log in gitter from mobile. Try ask at main room, nothing except update was changed.

  • Jimako
    Jimako  3 months ago

    @wrxxy I got notification from gitter but I am off until Monday. I am not able to log in gitter. Answer is add it to existing gitjub issue.

  • Jimako
    Jimako  5 months ago

    I will be on holiday until end of November. So no answers from me until then.

  • Jimako
    Jimako  5 months ago

    After 2 days I was able to create demo of BS Magazine 01 theme:
    https://v2demo.e107.sk/magazine01/

  • Jimako
    Jimako  5 months ago

    My antivirus software blocked and disabled gitter.