Automated disclaimer: This post was written more than 16 years ago and I may not have looked at it since.
Older posts may not align with who I am today and how I would think or write, and may have been written in reaction to a cultural context that no longer applies. Some of my high school or college posts are just embarrassing. However, I have left them public because I believe in keeping old web pages alive—and it's interesting to see how I've changed.
What if web surfers could edit any page on a website? What if webmasters could get webcorrections from users? What if readers could fix typos in blog posts, without leaving nitpicky comments? I've got a plan...
Tim Berners-Lee envisioned the web as more of a rewriteable medium, a system with read/write capability built in. And yet only with the advent of blogs and wikis have we really seen a resurgence of this extremely important concept. This goes beyond correcting typos and outdated links, useful as that may be -- this is about the participatory reader. I hope to spark some ideas and interest with this very rough sketch of a universal read/write framework.
The user as editor. You, J. Random User, come across a web page that has a typo in a paragraph. As you move your mouse over the paragraph, it highlights, indicating that it is editable. You double-click, switching the formatted text to a textarea with "Save" and "Cancel" buttons, a la edit-in-place. After correcting the typo, you hit save. The editable area is replaced with a message thanking you for your correction, which will be reviewed in due time. The message fades away, reverting to the original paragraph.
The server as streamliner. You, a plug-in for a CMS, receive a number of blocks of text at a variable rate. You collate them by page URL and block ID, then retrieve the cached page and compare the new text against the old, using the well-known diff command. Using a sloppy algorithm, you correlate heavily edited regions and similar suggestions, building a weighted list of regions that likely need author attention. Identical or highly similar edits receive a large weight, as well as edits that all have roughly the same boundary. Edits are grouped into "buckets" of similarity, and presented upon demand to the author as an ordered list with color or images as importance indicators. When the author marks a bucket as "ignore" or "done", it is hidden from the list.
The author as lead editor. You, the author of the page, periodically stop by the "Suggested Edits" page of your CMS. Seeing an egregious mistake at the top of the list, you click the "Fix" button to the side of the entry. Since the plug-in is sufficiently integrated with your CMS, you are taken to the appropriate editing area, or given instructions on how to find the offending text. After fixing it, you return to the Suggested Edits page and click "Done". You happen to notice that a number of well-meaning readers have attempted to correct "Slashdot" to "SlashDot". You click "Ignore" next to that entry, knowing that future corrections of that nature will be hidden from view.
- The user's involvement in the process needs to be highly streamlined, so as to get the maximum feedback possible. We need a large input base to get that nice "long tail" effect that is so prized.
- Users whose suggestions are more often accepted would be given more weight in the future (especially those users whose accepted suggestions are less common).
- Prizes or other incentives to best users?
Who tracks the data?
A site manages its own database, using a site plugin, and class-marked regions of the page.
- Private data is kept internally
- URLs/page IDs are canonicalized
- Content on parts of a page
- Users would have to learn each site's system
A central tracking system manages a database: Users and site owners both log in centrally.
- Higher recorded volume per user allows better identification of trusted users
- Unified system suggests a browser plugin
- Less control by individual sites
- A site manages its own database, using a site plugin, and class-marked regions of the page.
Should there be a notification system?
Notify author whenever 2 similar corrections come in for the same word or sentence.
- Speedy correction
- A little filtering of spam and other low-grade meat
- Poor spellers get bombarded with notifications (is this really a disadvantage?)
- Might miss more subtle corrections
No notification at all.
- No notifications to deal with.
- Important corrections may be missed
RSS feed of all "significant" corrections, instead of email notification.
- No need to delete corrections you want to ignore -- just let them drop off the list.
- Like the del.icio.us/popular list, the most popular items will float to the top. (Until dealt with.)
- With high volume, important corrections might be lost.
- Notify author whenever 2 similar corrections come in for the same word or sentence.
Download? What download? This is entirely vaporware, and I currently have no plans to implement this idea. However, as with all the content on this site, this software/system proposal is under a free license. I'd like nothing more than to see this proposal turn into a project. Go for it! (And when you've got something working, let me know. I'll post a link.) A system like this could be nothing less than the next step in the evolution of the web.