Eric Shulman has bookmarking code similar to what your describe. You make a new bookmark from the side panel. Bookmarks look like this:
/% |URL:|Enter a URL| |Description:|enter a description| |Author:|enter author/moderator info| %/ enter your notes here and are rendered by his TaggedTemplateTweak to display the entire page. But your own template could display something else. I feel this kind of bookmarking is much more natural in the context of TW. XML requires a lot of extra overhead characters and a whole bunch of extra parsing. I've adapted Eric's bookmarking system, and the file drop plugin, to begin creating my own contentment management system, with bookmarks that look like: /% Name:%name% URL:%url% File:%file% Author:%author% Thumb:%thumb% %description% %/ With my bookmarks, I can provide a thumbnail image to represent the bookmark. Also, I have a special page for viewing the tag "parent" of a group of bookmarks, so that I can see all the bookmarks tagged "SolarCooking" for instance. Bookmarks can be created from local content when a local file is dropped onto the TW. I believe the mozilla engine behind firefox may now have the ability to capture dropped text, so it should be possible for someone to create a system that would capture that. A bit of work I imagine. But I'm not sure information such as source url is simultaneously captured. That might require a separate FF plugin, such as TiddlySnip already does (or at least would, if it would stay working ;-( ) Mark On Mar 25, 4:53 pm, xen <[email protected]> wrote: > Hi there, > > I'm wondering whether there has been development or ideas about > something that you would call a "data application". I'm not sure TW is > the right space for that, but I can envision some kind of integration > between data input, formatting and display. > > There is some tension in TW between giving everything its own tiddler > and aggregating various entries into one bigger tiddler. Maybe this > would be the reason that TW is not suited for incorporating an > "application" for data display. A tiddler is meant to contain text and > formatting, simple note-taking that receives its formatting on-the- > fly. Where to store unstyled (yet structured) data? > > Let me give you an example of what I mean. > > Suppose you would want to use TW to maintain a list of bookmarks. You > envision several categories of bookmarks. There are categories for > "politics", "economics", "spirituality". You don't want to just add > the bookmark, but you want to add some summary description about what > the bookmark is about, perhaps it links to an article and you want to > record your views about what was being said. A summary. You want to > add a "topic" description that doesn't have to be unique and you want > to add a "today" in some universal format (yyyy-mm-dd) that describes > the day the bookmark was added. Perhaps you also want to include a > "date" that pertains to the publication date of the article. > > In XML the data structure would look like this: > > <entry today="2008-06-23" topic="Elite education in USA" > category="education"> > <source>http://www.theamericanscholar.org/su08/elite- > deresiewicz.html</source> > <title>The Disadvantages of an elite education</title> > <summary>People are being taught to conform but not to think > independently</summary> > </entry> > > You also imagine expanding this system to include scans you make of > magazine articles that you store in whatever format (jpg, pdf..). So > the <source> tag may refer to a local store "/documents/uwv- > perspectief-column-nuttig.jpg". > > It becomes a bit of a problem to envision unique tiddler names for > every entry you add. You decide to add a WikiWord name matching the > topic "EliteEducation" and to add new bookmarks on the same topic to > the same tiddler. You tag the tiddler both with "bookmark" and the > category "education" and possibly "scannedDocument". Tagging it with > "bookmark" makes the tiddler eligible for inclusion in your soon-to-be > automated system. All the other tags become selectors in your > application-tiddler that allow you to filter the list of bookmarks. > Another filter would be "today" - the bookmark date (you would filter > on year and/or month). Today would be the default sort criterium for > the displaying application. > > Now the only problem is to imagine and develop a way for tiddlers to > contain readable text /and/ to exhibit the data structure mentioned in > the XML fragment. > > We would need a new action in the options panel called "new bookmark" > that would make it easier to add the bookmark and the format needed on > every new entry. Perhaps even a little dialog box that would faciliate > importing a live bookmark on the internet, filling into a new (or > existing) tiddler the title, today and source, as well as the tag > "bookmark", and providing also a choice of existing categories (tags) > in the system. > > The format could be as simple as a definition list > > ; today > : 25 march 2010 > ; topic > : cultuurrelativisme > ; title > : Cultuurrelativisme; 2003; Wim van Rooy > > Or it could employ a new css class for easy parsing and recognition > (and styling as well) > > {{bookmark{ > |''today''|2006-12-14| > |''topic''|cultuurrelativisme| > |''category''|filosofie| > |''title''|Cultuurrelativisme; 2003; Wim van Rooy| > |''source''|http://www.eurozine.com/articles/2003-08-12-vanRooy- > nl.html| > |''summary''|blablabla| > > } > > It could even be a little javascript-based form application that > wouldn't require editing the tiddler directly but rather would involve > directly entering and selecting from the various form controls (edit > fields, drop down lists), although the form would still need to save > the data to the tiddler itself. I think a basic easy-to-parse format > such as the table above would be very appropriate. > > All that would be required then is a way to specify a template or > format for the data structure that can be used to generate the > application (the macro?) that is going to do the actual displaying of > the aggregate bookmark list. Perhaps all that is needed is to create a > template that reflects not the data format but the actual html > structure for displaying the bookmark list. > > <div application="bookmark" macro='view data list' selectors='today > category scannedDocument'> > .<table> > ..<th><td>today</td><td>topic</td><td>category</td> etc </th> > ..<tr> > ...<td class="date" field="today" macro="view data field"/> > ...<td class="selection" field="category" macro="view data field"/> > ...<td class="text" field="topic" macro="view data field"/> > ..</tr> > .</table> > </div> > > The basic macro (?) would simply list all of the tiddlers it finds > that have the specified tag ('bookmark') and parse all the > "{{bookmark{}}}" sections that it comes across, filling in the values > it finds into the table. It can then hide the subsections of the table > based on selections of tag and year/month. The table would also > provide clickable links to both http sources and local files. The > macro might even wikify the text it finds, such as the summary text. > > It might become a bit troublesome to include ALL the tags as possible > selectors for the list. It is not necessary for the tiddlers to have > tags beyond 'bookmark' and 'scannedDocument'; the category field in > this case could use a multiple comma/space separated format > ("spirituality economics"). On the other hand, all tags that are > present in any of the matched tiddlers could be shown in a single drop- > down list for single-selection, much like the 'category' field in this > case would provide a selector populated with the values found in all > the tiddlers. Perhaps the additional 'tag' selector would exclude the > tags already in use in other selectors. The purpose of TiddlyWiki tags > would be to cross the border between "data application" and the total > wiki, enabling all of the tagging features normally found in TW. > > It would be very easy to customize this database system for other > needs, as long as the records are simple. All you need to do is edit > the template for the list-application, and also provide the basic > template for the {{bookmark{}}} sections such that there can be a "new > bookmark" button. > > So, this is then what I would have in mind when talking about database > application of TiddlyWiki. Of course, it would be just as easy to fire > up a PHP/MySQL/web server, write a single php script, write a single- > table SQL definition, and be done. But it's probably really just as > much work as doing this in TW for those that have the knowledge and > you would be doing it in a much richer environment. > > What do you think? Does something like this already exist? > > Greetings, Xen. -- You received this message because you are subscribed to the Google Groups "TiddlyWiki" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/tiddlywiki?hl=en.
