Hi Dani, hello everybody.
I'd like to point another aspect of the issue.
> Than, one day you figure out: this piece of content is related to more than
> only one category. And my data model (directories tree) is not expressive
> enough to reflect this fact.
I also think it is an important point. I use symlinks on both PC and
handheld and I'm going to introduce hardlinks in both ones (the second
was changed to a one working on Maemo) though I haven't found a tool
for Windows which is good enough (extend context menu, assigne hot
keys and allow to create multiple hardlinks; not mentioning copying
hardlinks from one disk with NTFS to another, say, external hard
disk).
> Than tags step in - you can tag a piece of content by many tags. If it only
> belongs to one - than great, tag it with only one tag. And here we have a
> more expressive data model. Great.
Because of this I rarely use tags, mostly for "pinning" big notes to
~topics and for technical uses (for instance, I mark tiddlers with
relative html links with a special tag, so that if I need to rename or
move something, it's not an unsolvable problem).
> But (!) on the way we lost something important . Hierarchical structure
> hides from us the vast majority of our content, exposing on each step only a
> small part of relevant content. This follows one of our biggest natural
> cognitive limitations: *attention*.
This is why I think the "hardlinks" approache would be the most
powerfull one. It also has "backward compability" in file systems and
any tree systems, but we don't have any tree structure in TW and don't
have any "converter" which would load a part of file system with txt
files, for instance. So, the first questoin is how to implement the
"hardlinks" approache in TW.
It is quite simple. I have one TW which is built almost totally in
this concept frames. It is called "info-node" and is used for
aggregating info from web -- links to info resources, communication
flows, notes and links for searches in progress and others. It relies
heavily on PasteUpPlugin [1] which supports the workflow (but not the
appearence of a "finished" document, so is not important for reading).
I have a "Root" tiddler with specific CSS applied. It is the only
tiddler in the DefaultTiddlers. And next thing is Root contains
<<tiddler>> macros and <<slider>> macros used along with sections,
like in the example below. Finally, I can use multiple <<tiddler
tiddlerA>> macros (in different tiddlers). I have to note, though,
that this is more "symbolic links" rather "hardlinks": if I want to
rename my "tiddlerA", I have to see the references and change the
links there.
== in the Searches tiddler, imaginary example ==
...
<<slider "" [[Searches##mobile phone]] "New mobile phone" "">>
.../%
..
!mobile phone
{{indent{
<<tiddler [[New mobile phone]] edit>>}}}
!..
...
!end
%/
== end of the example ==
the "edit" option can be skipped since the latest versions of
PasteUpPlugin has a global chkPasteUpEditAll option thanks to Eric.
Though, to my shame, I haven't introduces use of this in my documents
yet.
A side note: this approache shows that a corresponding extension of
the slider macro would be of considerable value. Also, I'd note that
transclusion is not the only possible approach -- links can be used as
well; all this is "manual" and "custom" and I have no idea of
"automatic" analogues. Keep in mind that I use "new tiddler" button
*very* seldom in this document: first I create <<tiddler [[..]]
edit>>, then I follow the link to create and sometimes add a slider to
hide the new content. More than that, I do so in most of my tw-
documents -- create <<tiddler>> macro or a pretty link first. The "new
tiddler" button is heavily used in documents which aggregate highly
unstructured information, like news.
One though can see the lack of autogenerated "contents" to be a
shortcoming of the approach. I can see some ways to implement it but
no really beatiful one (aside creating some big script or plugin).
------------------------------------
Now let's move to the work with highly "unstructured" data (I'll refer
to the news aggregating which I do from time to time). This is where I
use tags heavily and where things about "categorizing tags" really
apply. Actually something goes wrong here, I must admit. The intention
of using tags here is to define the value of the news to provide
further rereading flow. Acting in this frame I face the fact that many
things are interesting very indirectly -- as materials for
~statistics, as precedents, or only if I become interested in a narrow
branch of science, or they are announcements which will be interested
only if they don't become reality (otherwise they are replaced with
the actual news). In fact, many news pieces remain in the RSS
aggregator; and I'm not sure whether I can say something substantial
in this aspect.
[1] http://www.tiddlytools.com/#PasteUpPlugin
--
You received this message because you are subscribed to the Google Groups
"TiddlyWiki" group.
To post to this group, send email to tiddlywiki@googlegroups.com.
To unsubscribe from this group, send email to
tiddlywiki+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/tiddlywiki?hl=en.
