Re: Wiki (re)volution

76 views
Skip to first unread message

Tobias Beer

unread,
Jul 14, 2012, 10:24:09 AM7/14/12
to tiddly...@googlegroups.com
I would say the main problem of current web semantics is the following:
Hyperlinks are not qualified (enough)!

If hyperlinks were qualified and the relation could be voted on, further qualified or simply flagged as inappropriate, wrong, falsified (by giving a reason)... al of which could as well be voted on in order to keep  down attempts of libel and slander... well that would be good.

To me that's really something that happens in the browser. Whenever I click on a link, it would really help if I could give feedback as to the quality of the link and the actual semantic relation between source and target.

The thing is, that this semantic relation in fact is completely independent from source and target. What I mean is this...

source: news article
target: product page for a book at amazon

Now, the semantic relation might actually look somewhat like this:

source-object: "Albert Einstein [scientist]"
relation: "visited [1906]"
target-object: "Cologne [City, Germany]

...whereas I would not only want to be able relate entire documents but in fact only highlighted sections or even words.

Whether or not such semantic relations with respect to one entity hyperlinking to another are actually meaningful with respect to the semantic objects and their relation (as in the example above) would be up for evaluation by any other user who discovered this semantic relation and found it to be meaningful (or not) for whatever he was actually looking for.


This state of affairs of "unqualified linking" also applies to TiddlyWiki. From my perspective it would be majorly awesome to have qualified WikiLinks or qualified tags.

Today, if I had a tiddler...

[Proposition A]
This is supported by [Evidence X].

...I would have to read the whole damn thing to figure out that A of type proposition is supported by X of type evidence. If only the relation were qualified as "supported by"... now we could do a lot more with that, couldn't we?

The same goes for tags.

For example, I could tag something [[done|12.06.2012 by Roger]]... now there's a whole lot more that we could do than just figure out that something has a tag of "done". Of course this could be decoupled and we could introduce a myriad of extended fields... but I would think that the ability to define such semantics would make things so much more readable, let alone condensed to the very degree we actually want it to be.

Cheers, Tobias.

TonyM

unread,
Jul 19, 2012, 1:18:26 AM7/19/12
to tiddly...@googlegroups.com
Have you reviewed tiddlySnip firefox plugin for a similar feature - Snip text into an offline tiddly wiki ?

On Thursday, 19 July 2012 05:41:32 UTC+10, SpiderX wrote:
You can already link to paragraphs or chapters. It's called an anchor. Anything after the hash mark in the url will scroll the page to the element with the ID in it.
So, the problem is that website designers don't always give IDs to all objects. Wikipedia is the exception though...
From a random page:

Just putting #Cast after the URL makes it jump to that part.

On Friday, July 13, 2012 11:06:55 AM UTC-7, Yoann Babel wrote:
Recently I've been thinking about how (great but) frustrating wikis are.
It seem's I'm not alone : http://www.youtube.com/watch?feature=player_embedded&v=mX0M19AqKQw

My idea of the next wikis is quite different.
When I'm searching through wikipedia some kind of historic or scientific point of view on a partucliar subjet, I'm not reading all the articles, I'm collecting information, assembling it, and creating a new one.
A piece of this article, a piece of that one, etc...
And I create a new wiki, which is barely a "view" on the core wiki (pedia).

What would be very interresting would be a tool (and I think it could be an evolution or a plugin of tiddlywiki) that allows to collorate or anotate a wiki.
A wiki page could be an anotation of other pages parts. (A kind of Aspect oriented programming like for wikis if you want).
I'm not sure about the best way to do it technycally speaking. It would look like hierarchical hyperlinks maybe ?

Not just word would be linked, but paragraphs, entire chapters, and a paragraph could be link to multiple subjects.
We could represent it by différent colors.
So, we could click on tags like "economics" on an article about houses construction for example.

And with something like partTiddler we could also compose by drag and droping new tiddlers from micro-tiddlers ... and tag paragraphs/chapter inside them.

I'm not sure if it's quite clear. I don't know if it's easy to implement. But I know it would be very useful for many people, not just like me.
We now have to much information to handle with current tools, we have to think of new tools to handle the growning complexity.

This is a kind a semantic web, but with just a new kind of link (call it tag/context/contexutal tag ... )
Reply all
Reply to author
Forward
0 new messages