reference labels

Labels inside the references is pretty kewl such as [macrovu is referenced by: item 627] and indeed we have the argumentation style in Touring AI Argument off the shelf . I couldn't find where the labels are documented .
On the other hand this spawns the idea of label rooms and so forth.
[label label1:,label2:,label3:] and for the Touring Argumentation [label is supported by:, is disputed by:] or something like that.

I think label rooms are needed to support Touring style argumentation & possibly the group UnhackTheBrain project.

Please discuss...

Comments

Mark de LA says
And then someday somebody will want a plugin to output (print) the maps themselves (;-)).

Mark de LA says
A label room might require a special purpose sort so that the thread is output from top to bottom of the argumentation route so that no matter how often you edit the items they appear in some kind of logical order.

Seth says
The thing to remember about the label is that the colon is not part of the label. The colon just serves to separate the labels from the obejcts in the reference. So the colon defines that whatever name is to the left of it is a label.

Mark de LA says
OK, so you would be able to parse [label is supported by, is disputed by] easy enough ?

Seth says
I am not sure what a label room would be. Is it a room all of who's objects contain the designated label ? For example, this item contains a label "referenced by" so this item would be in the room [labeled_with referenced by]. Is that your idea of a label room ? Or is it a room full of items whose titels are labels. If an item can be typed like so [type: label], then we could have a room of items which all contain the reference "[type: label]". Presumably you could reference that room with a nested reference like [all [type: label] ].

I don't know ...

Mark de LA says
"label is contained in"

Mark de LA says
Here is also a point I thought about the other day. It would be nice to have labels in the about box ... thus making the about box more general. & in that case we could move the label room to a special case of the about room. Or keep both ways of doing things.

Mark de LA says
Part of the group UnhackTheBrain project is getting a chain of "is supported by" & "is contradicted by" reporter notes & interviews.

Seth says

OK, so you would be able to parse [label is supported by, is disputed by] easy enough ?

Mark from Marketing on 2005-09-22 05:58:57

That parses to "label" as the key word of the reference and a comma separated list ("is supported by", "is distributed by")

The question is what does "label" mean in that construct ? Presumably it could mean that the objects were labels. But what relationship does it have to the item on which the reference hangs ?

Seth says
The way references are working now, the first non part of the reference types the object. [LABEL : TYPE OBJECT].

Presumably all of the objects of the reference have the same type. So the parser might grok [my blog roll: group seth, mark, elaine] as implying that the subject item had the relationship "my blog roll" to the following groups named "seth", "mark", and "elaine".

I doubt that the current parser is that advanced but it should be. It parses based on the knowing the ":" and "group" and ",". Since in that grammar there is only one slot for TYPE, all objects in the list are assumed to be of the same type. Iow, they are groups, or permalinks, or titles, or indexes, etc.

Seth says

... [my blog roll: group seth, mark, elaine] ...

seth from seth on 2005-09-22 06:26:09

Apparently i can't just define the syntax here and it be instantly implemented. The system is trying to make (seth, mark, elaine) all one hyperlink or arrow. It needs to grok that that is a labeled list and hyperlink those separately.

I don't know how far i want to go into enhansing this grammar now.

Mark de LA says
...thanks for looking at it. It's not a now thing. Just wanted to get in the future things. At some point we might have the links be the Cybermind like programs of the FastBlogIt, i.e. 123, ve label2: memvance sym etc.

Seth says
I think a doton would start with a "!" so that might be [! at (item 123) ve (label2) memvance sym ing] ... somewhere i scoped that we would use function syntax ... f().

Seth says
Note that the comma delimited list of tags works now ... prolly because that is what i designed it for here .. s rte,package contains two rte packages plus one that got lost.

Mark de LA says
prolly [f(x,y,z)] - should be distinguishable from people who want to copy code into their items, though.

Seth says
... yep that is definitely something to put into the hopper to think about when we go to extend this grammar.

Mark de LA says
This is deep is deep CyberMind stuff not a wish on my list anymore (yet?)

louie vuitton says
good url!

reference labels

Tags

Comments

See Also