cyberinfrastructure

What I've learned from building iPlant

In nature you've got a continuous very-small-feedback loop going on, which is why things get to be harmonious. If it weren't for the time dimension, it wouldn't happen.

--Chris Alexander

be a persistent volunteer

young projects aren't good at accepting contributions yet, but you can shape them

data analysis IS still limiting

let's work on fixing this together

put your analysis software someplace everyone can use it and can build on it

Physical sciences have had a national cyberinfrastructure for decades.

Once upon a time, modernism reigned supreme. Straight lines and uncluttered vistas were seen as the ultimate aesthetic virtue, and science reflected that. The age of modernism was the age of reductionism.

But in a post-postmodern world, where the interplay of artful jumbles is the dominant aesthetic mode, and interpretation has taken over from imposition, networks are the new paradigm -- and science now reflects that.

add value, not only novelty!

please make your methods available in a form others can use

encourage this in annual reports

Lessons learned:

Don’t just build it

and expect people

to use it.

Instead, get people to help build.

Cycles are cheap, easy access is harder (the ‘last 100 feet’ problem)

It’s a success if it’s used for things you never predicted.

building iPlant:

How did I get started in CI?

FATAL: Insufficient memory to execute data step program. Aborted

during the EXECUTION phase.

NOTE: The SAS System stopped processing this step because of

insufficient memory.

So I went over to the statistics and

computer science department...

Better Data Analysis for Biologists:

Help Build the Factory, Not Just the Tools

Ann E. Stapleton

Department of Biology and Marine Biology

University of North Carolina Wilmington

interesting CI questions to consider

An important aspect of design is the degree to which the object involves you in its own completion. Some work invites you into itself by not offering a finished, glossy, one-reading-only surface.

I think humans have a taste for things that not only show they have been through a process of evolution, but which also show that they are still a part of one. They are not dead yet.

--Brian Eno

the facilities manager reports to the president

any responsible person in the building can send a work request and get a quick response

what we learned from Gridnexus:

we can and should do this

1.5 years for just North Carolina is not the right scale

software evolution, deliberate simplicity

hierarchical modularity dynamics

federation

design patterns for manycore

interesting neighbors, strong floors, freedom

The romance of maintenance is that is has none. Its joys are quiet ones. There is a certain high calling in the steady tending.

quotes from Stewart Brand's

How Buildings Learn

While all buildings change with time, only some improve.

High road: durable, independent, respected

Low road: quick, responsive, street-smart

What software or protocol is like this?

In organizations and in buildings, evolution is always and necessarily surprising.

You cannot predict or

control adaptivity.

All you can do is make

room for it.

gopher, xml, json...what has adapted well?

wabi sabi

in a beautiful thing there is always some part which is lovingly and carefully done, and

some parts which are roughly done, because

the compensation between the two is necessary in a real thing

we tend to focus inordinately on the happiness

we think we will derive from big changes in our lives

and discount the pleasure to be had by eliminating

minor and frequent annoyances

both big and small projects over time

piecemeal growth, not the idea it will be perfect when built

Mark McCahill: keep it simple, clean, make sure it does what people expect...

we've got several different entry points in iPlant (edu, expert, in-between)

overbuild structure

provide excess services capacity

go for oversize

your research question can serve multiple ends

what do you suggest as examples of multi-objective research?

building designers should take on problem transparency--

use materials that smell when wet, build in inspection windows

How could what you like and what you do be leveraged?

Spend the right amount of time thinking about what you do as well as doing it.

leverage what's already available

consider who else would learn from what you have accomplished