Redirect to wiki page? about github-wiki-see-rs

[gingerbeardman]

I use PageExtender.app for Safari on macOS which allows user-defined JS and CSS per domain. There are other similar solutions for other browsers and platforms.

So I create a file: github-wiki-see.page.js

Contents:

if (document.getElementById('header_button')) 
    document.querySelector(".visit_url_button").click();

That's it!

The code will run only on pages from the domain github-wiki-see.page

Now when I land on a see.page the visit button is automatically clicked for me. Not as good as a proper HTTP redirect but good enough for me.

[nelsonjchen]

Hmm, I think I might make a bit of a contract to provide some elements via ID to click.

[nelsonjchen]

Ah sorry this is a WONTFIX. Or rather if it was implemented, it would risk getting the content unindexed or penalized.

I don't do this because if a search engine crawler visits such a link and gets automatically redirected, it won't index the content. Search engines crawlers will see the content at the link as a redirect, and then just ignore the content at the link and then not index it. In fact, the search engines crawlers will actually land on the original GitHub Wiki page and sees that it's unindexable, making the whole crawl pointless.

I also do not selectively try to redirect only search engines crawlers and not users. That would be cloaking and it might violate the guidelines. I'm sure that large search engines "secret shopper" the links to check for this automatically. I don't want to risk any penalizing of the content so I err on not doing this approach either. It's probably debatable if the current approach is cloaking "because it provides our users with different results than they expected" but as we all know, some of the search engines are uncaring inhuman organizations and I prefer to err on the side of pleasing the dumb, human-made computers by having a dumb interstitial for all, robots and humans.

The only wikis/pages with this behavior are the decommissioned wikis that GitHub allows direct indexing of. I actually want this service to fall off search engines for those Wikis and I have the service configured to automatically redirect. Because search engines crawlers will be automatically redirected to GitHub's content and GitHub does allow that content to be indexed on GitHub.com, the content is not at risk of being unindexed.

[nelsonjchen]

In place of this, I do try the following:

• Make the buttons and URL to get to the original GitHub page as obvious and easy to tap or click as possible.
• Also make said buttons and URL use window.replace() instead of a plain anchor link to make the service's content page disappear in history if the user goes to GitHub. When they press back or look in their history from GitHub, the service's content page won't be there contaminating it and will go back to their search engine result page or however they got to the service's content page.

[nelsonjchen]

Actually, I think I might even back out of the window.replace thing. Looking in Google's search console, that might be tripping Google's redirect nose though not as strongly or consistently.

[nelsonjchen]

You can see a fight I'm having like in #137.

[gingerbeardman]

no worries, i will implement the redirect myself using a browser extension in my setup.

feel free to close

[nelsonjchen]

If you do a write up on it, I would love to link to it. Keep me posted!