Hyperlinks database

From WhyNotWiki

Jump to: navigation, search

Hyperlinks database  edit   (Category  edit) .


Contents

[edit] Main

I want to be able to put lists of hyperlinks on various/multiple article pages. There may be overlap: I may want the same link to appear on two different pages.

Solution: Keep all my hyperlinks in a database. Allow a hyperlink to belong to multiple tags/categories. Then on article A, I would just query for al hyperlinks belonging to tag/category "A" and do the same for article B.

Advantages:

  • allow intelligent reuse/duplication (intentionally showing hyperlinks in more than one list / on more than one page)
    • change a hyperlink in one place and it changes it in all places
  • Broken link checking is easy: just run through all the links, do an HTTP request, and check the response
  • Caching is easier
  • Add more metadata to each link:
    • media type of target (text, HTML, audio, video, etc.)
    • long description (not just title)

Drawbacks: Dependence on the database. I would feel like I have to do a database query every time I simply wanted to display a link on the page (to make sure I had its current title, URL, etc.), but I think the benefits outweight any drawbacks!

Alternative idea: Just use hyperlinks inline in the wikitext (as a primitive) and do the categorization/metadata on the page(s) that contain the links. Would still want to silently (automatically, but without user intervention) keep a database of links and cache those pages.

[edit] The question it comes down to: do hyperlinks need to be "first-class objects"?

They're not in the context of MediaWiki -- they're just a bit of markup in the wikitext.

They are first-class objects in the case of Bookmarks or perhaps Digg, etc. But Digg, etc. still have associated text, so you could in that case just make the associated text be the first-class object.

I'm leaning towards just having the pages/sections containing links be first-class objects. This is because I want to move away from "bookmark"-style thinking, where you bookmark a URL without any context. I think we would benefit from always putting links in some kind of context/comments.

[edit] Blogs

Keep track of blogs that I find interesting. View by category or by interesting-rating (so I can use my limited time only to keep up-to-date on the best ones.)

What about RSS feeds? If I just subscribe to the feeds for all my favorite blogs, do I also need an entry in this database? Maybe.

  • the blog's website has more features (like browsing by category, blogrolls, etc.) than the rss feed

But the list of feeds in my rss aggregator is duplicating the contents of this database. And that's bad.


[edit] List them on a big list and under specific categories

For instance, I want all my "Life-hacking"/"Getting things done" links listed on my "getting things done" page. But I also want them listed (along with many other links) on my "favorite blogs" page.


[edit] Low

[edit] Are hyperlinks outdated?

Maybe hyperlinks are on their way out, as RSS/pull starts to replace manual web browsing/HTTP requests/push...

No, I don't think hyperlinks will be dead any time soon!

[edit] Related projects

Wrinkle

Applications written in Ruby#urlmarkr

Personal tools