You are viewing a single comment's thread from:
RE: Communities vs. Blogs
Another thought just occurred to me. I remember at one point, Steemit had configured condenser to report the canonical URL so that the site where an article was posted would be recognized by search engines as the original source of the article... I guess here and here?
Maybe we could make use of the same technique so that condenser would insert the post's 2nd tag into the canonical URL, instead of the hive-######/community tag?
Either that, or we could use the same sort-of lookup mechanism that Steemit used in the first case, and no metadata would be needed. I guess community owners/admins would just need to get their canonical category into github somehow.
cc: @the-gorilla
My understanding of "Canonical" is that it's a boolean value. If canonical = true, then that page is the original. If more than one page says it's the original, I believe there's a "published date" precedence - although I don't know how this works.
So, my understanding of canonical is that it wouldn't index an alternative URL.
Your idea of URLs ignoring the hive-xxxxxx tag could potentially be the solution but I don't know how that would work with the existing page routing. It would have to somehow store the "actual" page URL (the indicator to the content in Hivemind) and then redirect the user to a friendly URL - with the friendly URL knowing that this is what's happened.
I've probably explained this badly but this solution feels feasible. Continuing that trail of thought...
So: /hive-123456/@the-gorilla/my-day-watching-football, would automatically redirect to /football/@the-gorilla/my-day-watching-football (assuming football's the 2nd tag) with the page knowing to load the content from the original location... which it CAN do as this URL (https://steemit.com/steemit-quiz/@the-gorilla/steemit-quiz-2025-week-1) proves.
So when loading the content, steemit completely ignores the 2nd part of the URL... so I'd need to implement a 301 redirect when routing these links.
@steemchiller - does this sound feasible to you? What am I missing?
0.00 SBD,
0.00 STEEM,
5.95 SP
My brain is currently entirely into steemd and SDS development, as I want to finally finish this version over the next few weeks. I guess I would still tend more to a complete/clean solution (adding a community field in the json_metadata) instead of a workaround, but I need to give it more thoughts when I find some more time for this...
0.18 SBD,
0.53 STEEM,
1.21 SP
No problem, there's no rush - I will test if the idea's feasible.
I don't think this idea is a work around. It's essentially returning the URL back to how it was before communities existed - where the first tag defines the internal (and therefore external) link.
Instead of the internal link being:
/(hive-xxxxxx)/(username)/(permlink)
It's
/(1st-tag)/(username)/(permlink)
If the community makes more sense, it could probably be:
/(communityName)/(username)/(permlink)
without too much effort.
It feels like a fairly elegant and straightforward solution.
A question for me is what happens to the SEO value of these posts if a community decides to change name. Which makes my current preference using the 1st tag (non hive-xxxx).
0.00 SBD,
0.00 STEEM,
1.16 SP
@remlaps
This works:
This is the feeds code.
Instead of always using the category, if it starts with 'hive-xxxxxx', it replaces this with the 1st tag.
Returning Steemit to the old SEO URL value in 4 new and 1 edited lines of code (+ 1 comment + 3 empty lines for readability).
0.00 SBD,
0.00 STEEM,
0.34 SP
Not sure, but I think they used the "rel" method that's described here for indicating the canonical URL. But that page says that redirects take precedence, so maybe a redirect would be preferable(?)...
I'm out of my depth here, but I'm pretty sure it's possible with no changes to hivemind (which is what I had been thinking before)... I just don't know the level of effort.
0.05 SBD,
0.14 STEEM,
0.31 SP