WATERS 2012, program and panel discussion

[Giuseppe Lipari]

Dear all,

The WATERS program is now on-line. Check it out at http://retis.sssup.it/waters2012/program.php

Every presentation lasts for 25-30 minutes, including questions. 

At 16.30 we will also have a panel open to all participants.

The panel will start with a short presentation by myself, followed by discussion. The topic is: "Software and Academic Research: are we going in the right direction?". 

Here is a short summary of what I would like to discuss with all of the participants. Please comment, and/or send me additional topics you would like to discuss. 

--------------------------------------------------------------------

Problems in our academic domain

  - Sometimes too much emphasis on theory

  - Too much emphasis on "new" research

  - (Too often this is done just to publish yet another paper)  

  - No emphasis on good engineering and design

  - Throw-away programming

  

Consequences of this approach

  Local consequences

  - every time you need to implement your algorithm *and* the other ones

    often trying to understand the algorithm from a loose description 

    in the paper

  - no real data is used/collected, benchmarks are few and kept secret

  - research is finely pulverized in thousand of papers on hundreds of 

    technical journals

  - Hence, difficult to interact with real engineers (we cannot pretend they

    spend so much time reading papers)

  - Too little connection to industry

  - Low impact

  

  - If we (professors) do got give value to software design, neither our PhD 

    students will do

  - For "applied" computer scientists and computer engineers this is terrible, 

    because it means that we are educating generations of future academics and 

    engineers that do not give value to software!!

  - Also, there is less possibility for them to find a job in industry

  

Main reason for this sad state of things

  - Researchers are professionals that try to maximize their ROI (Return On Investment)

  - Revenues depends on publications, h-index, etc.

  - Thus, researchers seek to maximize the number and the quality of 

    their publications

  - Any other activity is a waste of time (from our point of view)

  - Nobody wants to waste time on "programming", something that brings 

    no immediate return on investments

    

    

What do we want? 

  - We would like to change the academic culture, at least a little

  

  - We would like academics to give value to good SOFTWARE

    * PhD students write software for simulation, analysis, comparison

    * They should be acknowleged for their good work (when it's good, of course)

  - We would like academics to give value to not-so-new research that is 

    well done, well designed and well implemented

  

  

How?

  DISCUSSION

--------------------------------------------------------------------

I hope to see you all in Pisa!

Giuseppe Lipari