Comments on: Scientific Software and UI Design http://leefalin.com/blog/2008/06/27/scientific-software-and-ui-design/ Thu, 20 Nov 2008 08:30:08 +0000 http://wordpress.org/?v=2.5.1 By: Nick http://leefalin.com/blog/2008/06/27/scientific-software-and-ui-design/#comment-508 Nick Sat, 12 Jul 2008 19:03:26 +0000 http://leefalin.com/blog/?p=127#comment-508 As an Industrial Design graduate turned software developer, I totally agree with your points. Apple is successful because they focus on industrial design which is a discipline that marries form *and* function. Both are equally important and failing at one is failing altogether. A problem I see in scientists is that they see designers (and the general public) as intellectual "lightweights" whose opinion is unimportant. Whenever I read comments on Slashdot about why another Apple product is inferior to Gimp/Inkscape/(insert any OSS Linux program), I know that some people will never get it. Yeah, they should definitely make this a new discipline, preferably as part of CS programs. - Nick - As an Industrial Design graduate turned software developer, I totally agree with your points.

Apple is successful because they focus on industrial design which is a discipline that marries form *and* function. Both are equally important and failing at one is failing altogether.

A problem I see in scientists is that they see designers (and the general public) as intellectual “lightweights” whose opinion is unimportant. Whenever I read comments on Slashdot about why another Apple product is inferior to Gimp/Inkscape/(insert any OSS Linux program), I know that some people will never get it.

Yeah, they should definitely make this a new discipline, preferably as part of CS programs.

- Nick -

]]> By: leefalin http://leefalin.com/blog/2008/06/27/scientific-software-and-ui-design/#comment-505 leefalin Wed, 09 Jul 2008 11:37:17 +0000 http://leefalin.com/blog/?p=127#comment-505 Ntino, I agree with you on your depiction of the "does-it-all" graduate student, but there are <a href="http://www.red-sweater.com/marsedit/" rel="nofollow">plenty</a> of <a href="http://flyingmeat.com/acorn/" rel="nofollow">examples</a> of <a href="http://flyingmeat.com/voodoopad/" rel="nofollow">software</a> <a href=http://www.newsgator.com/Individuals/NetNewsWire/" rel="nofollow">created</a> by just one person that works well AND looks good. I think the real problems is as you said, for most software created in an academic setting the polished product is the paper and not the software. Ntino,

I agree with you on your depiction of the “does-it-all” graduate student, but there are plenty of examples of software created by just one person that works well AND looks good.

I think the real problems is as you said, for most software created in an academic setting the polished product is the paper and not the software.

]]> By: Ntino Krampis http://leefalin.com/blog/2008/06/27/scientific-software-and-ui-design/#comment-503 Ntino Krampis Sun, 29 Jun 2008 02:10:49 +0000 http://leefalin.com/blog/?p=127#comment-503 I agree with all the points above. Concerning points 1,2,4 simply the model has to change, with funding agencies giving more weight to UIs; as they now in a multi-million $$ grant they require some standards for say the microarray data you generate or the procedures you follow generating them, they also have to require specific standards for the software (and not only standards for the UI, but also following software development practices such as documentation etc). Great part of the problem is that people deciding for bioinformatics grants in funding agencies or being PIs in funded projects have no idea on software development practices. On point 3, just think about the does-it-all graduate student that goes from the lab to grind tissue to the cube to write code (because most of the scientific software out there is one man's story - or thesis). And for your comparison with the Apple software or any commercial software, I think it cannot stand cause there you have teams of people developing a single application, and each of the developers perfects his own little part of the software (well we have to give credit to the student which usually produces as much as 3x developers, but the final application goes over one man's powers to achieve perfection). What I see for point 4, is that the problem is with how the whole academic system is structured, where the polished product is the paper and not the software. I agree with all the points above. Concerning points 1,2,4 simply the model has to change, with funding agencies giving more weight to UIs; as they now in a multi-million $$ grant they require some standards for say the microarray data you generate or the procedures you follow generating them, they also have to require specific standards for the software (and not only standards for the UI, but also following software development practices such as documentation etc). Great part of the problem is that people deciding for bioinformatics grants in funding agencies or being PIs in funded projects have no idea on software development practices. On point 3, just think about the does-it-all graduate student that goes from the lab to grind tissue to the cube to write code (because most of the scientific software out there is one man’s story - or thesis). And for your comparison with the Apple software or any commercial software, I think it cannot stand cause there you have teams of people developing a single application, and each of the developers perfects his own little part of the software (well we have to give credit to the student which usually produces as much as 3x developers, but the final application goes over one man’s powers to achieve perfection). What I see for point 4, is that the problem is with how the whole academic system is structured, where the polished product is the paper and not the software.

]]>