Dido---a tiddlywikified database application

[David]

I'm posting to announce Dido, a tiddlywiki-inspired tool for managing
small databases in a web page. Dido is intended for managing small
databases of information---an address book, a stamp collection, a
shopping list, a video collection. It contains an editable structured
database (stored in the page, tiddlywiki-style) and layers rich
interactive visualizations (sortable lists, grids, maps, timelines)
and filtering (faceted browsing and text search) over the data. You
can find Dido at
http://projects.csail.mit.edu/exhibit/Dido
Both the data and the visualization can be edited using the WYSIWYG
editor that's part of the document.
As with tiddlywiki, you can download it and start playing, saving your
file to persist any changes (we use the tiddlywiki file save code;
thanks for that!).

Dido is based on our exhibit visualization framework; the page at
http://simile-widgets.org/exhibit shows some examples of what can be
created using that framework.

Dido works well on firefox and chrome; it triggers certain IE bugs so
has gaps on that platform.

Dido is still very rough, But I continue to develop it actively and
am happy to get feedback. The project is BSD open sourced. The
subversion repository will be open for anonymous checkout in a day or
two.

[Jeremy Ruston]

Hi David

Congratulations, Dido is a considerable piece of work, and fabulous to
see you both reusing so many useful components, and yet coming up with
something that's really quite novel. Like many in the community, I've
long been a fan of your Simile widgets.

Best wishes

Jeremy

> --
> You received this message because you are subscribed to the Google Groups "TiddlyWikiDev" group.
> To post to this group, send email to tiddly...@googlegroups.com.
> To unsubscribe from this group, send email to tiddlywikide...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/tiddlywikidev?hl=en.
>
>

--
Jeremy Ruston
mailto:jer...@osmosoft.com
http://www.tiddlywiki.com

[David]

Thanks for the kind words.  I remain pretty enthusiastic about the SIMILE vision.

I developed Dido to prove out the idea, and doing so independent of tw was the easiest. But having done it, I think it would be interesting to discuss whether some of the same ideas --- holding structured data and offering visualizations --- might be worth considering for future use in tw.  We'd of course be happy to contribute and help with the use of the simile widgets visualizations.  Note that we have also integrated these visualizations into a standard wiki framework --- mediawiki: http://projects.csail.mit.edu/wibit

[Jeremy Ruston]

Hi David

There has been quite a lot of experimentation within the TiddlyWiki on
both structured data and visualisations. Also, some of our recent work
in Osmosoft on TiddlySpace is exploring how structured data sets can
be published and reused.

In terms of structured data, TiddlyWiki is still based on tiddlers
being simple name/value pairs with a title string and a body string.
On top of that it allows for structured data in two ways:

* tiddler fields are custom name/value string pairs that can be
attached to tiddlers. Some of these fields are used internally (eg as
metadata for the synchronisation scheme), but can be extended by
plugin writers. Custom fields would be used for geotagging a tiddler

* tiddler slices and sections are ways of addressing substrings within
the body of a tiddler. Slices allow data to be extracted either from
name/value pairs in tables or lists, while sections can be used to
extract chunks of content based by the name of their containing
heading or subheading

There is a frequent and interesting debate within the community as to
how to choose between the two approaches in different circumstances.
There is a tension between ones desire to simplify relationships by
encapsulating related information into a single container, and the
realisation of one of TiddlyWikis guiding principles: that information
is more reusable, and therefore more useful, if you chop it up into
the smallest semantic units. Both approaches are well supported by
plugins - Eric Shulman's TiddlyTools has a extensive support for
forms, tabular editting, and "tag-grids"

In terms of visualisations, there's been quite a few plugins to
integrate various hyperbolic tree views and so on. Dawn Ahukanna's
adaptation of your SIMILE Timeline plugin into TiddlyWiki attracted a
lot of interest in the community back in 2007.

Osmosoft is currently involved in a couple of projects that bear upon
these topics:

* Open Britain, a project to build an open data store of information
for disabled visitors to Britain, where we are integrating geocoding
within TiddlySpace

* Preso, a project to allow people to construct 2D spatial maps of
content that exploit instinctive conventions for communicating
relationships and relevance

* TiddlySpace, which includes the addition of native SVG support to
TiddlyWiki, partly prompted by an interest in using SVG for data
visualisations

* Jon Robson has given us http://charts.tiddlyspace.com/, which shows
an approach for rendering dynamic charts from tabular data, and
http://mindmaps.tiddlyspace.com/

I believe that TiddlyWiki and TiddlySpace have a useful niche for
these types of applications. The ability to construct standalone data
capture forms with TiddlyWiki is cool, and the ease with which one can
archive a fully functional, interactive self-contained visualisation
that doesn't require a server is potentially disruptive.

Best wishes

Jeremy

--
http://jermolene.com
http://tiddlywiki.com
http://osmosoft.com

--

[David]

Thanks Jeremy; I'll look through these.

[David]

As promised, the svn repository for Dido is now open for anonymous checkout:
https://svn.csail.mit.edu/editme

[PMario]

Hi David,
Dido looks great. But it would be nice to have some more download
info. May be some kind of TiddlyWiki's download button [1]. The
default way: Browser menu: File: Save as ... fails. (FF 3.6.13,
ubuntu10.10) It saves a ~370 kByte file.

If you save it the TW way. Rightclick the link. Save Target as ....
downloads the needed 1.4 MByte file, which works offline.

===
I personally like the data structure editor (lens), and simple data
input mechanism.

-mario
[1] http://www.tiddlywiki.com/

On Dec 16, 12:31 am, David <kar...@csail.mit.edu> wrote:
> I'm posting to announce Dido, a tiddlywiki-inspired tool for managing
> small databases in a web page.  Dido is intended for managing small
> databases of information---an address book, a stamp collection, a
> shopping list, a video collection.  It contains an editable structured
> database (stored in the page, tiddlywiki-style) and layers rich
> interactive visualizations (sortable lists, grids, maps, timelines)
> and filtering (faceted browsing and text search) over the data.  You
> can find Dido at
>  http://projects.csail.mit.edu/exhibit/Dido
> Both the data and the visualization can be edited using the WYSIWYG
> editor that's part of the document.
> As with tiddlywiki, you can download it and start playing, saving your
> file to persist any changes (we use the tiddlywiki file save code;
> thanks for that!).
>

> Dido is based on our exhibit visualization framework; the page athttp://simile-widgets.org/exhibitshows some examples of what can be

[David]

Hi Mario.  Executing file save-as once you visit the dido page fails because some browsers save the modified, dom-manipulated page instead of the original downloaded page.  On the other hand, right clicking on the link and selecting "save as" will work just fine.  I'll try to add the download button as you suggested.

[David]

OK, the page now has a download link that does the right thing.  Thanks for the suggestion!