Re: [Treebase-devel] ABI proposal for phyloinformatics

[Andrew H. <and...@gm...>]

Hi all,

Need to throw in my two cents here. I imagine these comments might
represent a slightly different pov than has come up.

Near-future grant proposals probably should not spend heavily on the
development of tree visualization tools with goals of long-term
sustainability. Instead, the next wave of proposals (I'm only talking
about viz here) should focus on a few pre-tool advancements. My idea
of what these advancements should be is based on

1) the future/now of viz is on the open-web (html5 etc).
2) the best data viz people are probably not in our community, they
are in their own or are unsuspecting tinkerers who will stumble upon
it.
3) the coolest innovations in tree viz aren't going to be wrapped into viz tools

So, with that, here is what I propose needs to happen,

1) PhyloJSON notations. XML on the web is slow and unnecessary. JSON
is quickly moving in on XML dominance. PhyloBox and other projects all
moved down their own paths for creating json phylogeny notation. This
is the normal way to do it, but our community could be developing some
common notations that we could all start using/expecting.

2) Advanced REST tree queries (TreeBASE) for strictly web-client based
consumption. What I mean here, is the use of cross-domain rest queries
to find, trim, scope, and combine trees. On top of that, wrapping the
results into the product of (1) above. This will allow web mashups and
tools to emerge that combine the data in unexpected ways.

3) Proof of concept tools and documentation. All I mean here, are
examples of how the products of (1) and (2) can be used to merge
phylogeny with external sources of data (wikipedia, eol, flickr, etc),
using the modern web, html5, css3, and javascript. This will provide
the foundations for a much larger pool of producer/consumers to do
interesting things with the data. Documentation is king. It will
enable those not at the table to understand the decisions made.

Each of those have clear integration points with treebase (especially
the meeting of 1 and 2) and other projects. More importantly, none of
them have been correctly explored with sufficient resources and
planning. On top of that, developing the three in coordination will
allow much better development of use-cases for phylogeny viz that feed
back into each of the three.

best,

a



On Wed, Jun 15, 2011 at 11:14 AM, Robert Guralnick
<Rob...@co...> wrote:
>  I see TreeViz and UI are partly overlapping.  There are a lot more
> interesting ways to explore and annotate trees that push back to
> overall UI design.  Also, it might be wise for us to both look
> internally to experts within the community, and also externally to
> those who do design in different biological (or other) arenas.  We
> want to develop for a broad set of users, and not necessarily simply
> for card carrying systematists.  I am very sympatico to the idea that
> we really do need to bring in HCI expertise, but am even more keen to
> find great and visually talented artists and designers can carry us a
> ways forward too, without getting too "academic".   In developing some
> UI and TreeViz software, our lab has found great value in having a
> talented visual designer (Sander Pick;
> http://plebeian.tv/#information) to help us think through some of the
> UI/Viz challenges.  Sander's skills transfer well across domains (he
> is now doing design work for clean energy).
>
> Best, Rob
>
>
> On Wed, Jun 15, 2011 at 10:49 AM, Rutger Vos <R....@re...> wrote:
>>> In talking with Rob G, we
>>> are both really interested in UI, and we both have programmers in our groups
>>> who are working on various UI features for Tree Viz, so that would be an
>>> area where I think we would be interested to help the project as the interop
>>> details are hammered out.
>>
>> Tree visualization and interaction design of web applications are two
>> different things. There is a lot of community expertise of the former
>> (and there are other people we might involve in this, e.g. Tamara
>> Munzner), but not enough of the latter. People complain about TreeBASE
>> in part because the workflows and visual metaphors aren't informed by
>> good HCI design, and this is not going to be addressed by better tree
>> viz. People will complain less about TreeBASE if we know why facebook
>> is nicer than myspace and apply those principles.
>>
>> --
>> Dr. Rutger A. Vos
>> School of Biological Sciences
>> Philip Lyle Building, Level 4
>> University of Reading
>> Reading, RG6 6BX, United Kingdom
>> Tel: +44 (0) 118 378 7535
>> http://rutgervos.blogspot.com
>>
>



-- 
Ecology and Evolutionary Biology
University of Colorado

http://biodiversity.colorado.edu
http://biodivertido.blogspot.com