|
|
Bioinformatics.org
|
![[?]](https://www.bioinformatics.org/images/icons/info.png)
|
|
Research
|
Online databases
Online analysis tools
Online education tools
|
|
Development
|
|
|
Forums
|
News & Commentary
Jobs Forum
(Career Center)
|
|
|
News & Commentary - Message forums
|
         
|
 |
|
URL: C# in developing bioinformatics programming object model
Submitted by Deying Kong; posted on Thursday, October 03, 2002 (17 comments)
Currently most of software tools and programming models were developed with Perl or Java under the Unix system. Millions of Windows programmers are blocked from accessing these programming objects. Do they need to learn Unix and Perl? Many biologist are familiar with the easy-to-use Windows system. Can they spend less time learning Windows solutions? I am exploring the power of the Windows platform and the modern, elegant language C# (a cross platform .Net solution development language) and using them to develop a proper hierarchical object model and analysis tools.
You can visit this web site to check on the advancing project:
http://globetrack.dnsalias.com/bioinformatics/
You are very welcome to post your ideas and comments on this web site.
|
|
Expanded view | Monitor forum | Save place
Post a followup to this message:
You have to be to post a reply.
Thread view
|
|
You are missing the point about Perl, *nix and Windows. What should be important to the bioinformatics community is that the source-code of the programs we use be open. As a scientist, I will not use reagents whose components I cannot have access to; I want to know the composition of the protease inhibitor cocktail I use making nuclear extracts and what regents are in the coctail I use to label cRNA (kits are a fact of life, but I won't use a kit that isn't based on a protocol that is published somewhere that I can see it, preferrably in a peer reviewed journal).
The same holds true for the software I use to analyze data; having access to the algorithms used to classify the results of an array experiment or to design a set of unique oligos to print on an array is a requirment before I will use the programs.
Microsoft has not been a supporter of open source software, and has, according to memos released from inside Microsoft, actively tried to stop this movement. Becuase of this attitude, Microsoft likes to use propietary processes that make porting to a different os difficult.
That said, there are open source C compilers and java runtime environments available for windows, and programs written in java run on windows, if you install the latest Sun Java and not the old version supplied by Microsoft (another example of Microsofts efforts to squelch open source, see the news articles about the Sun suit against Microsoft that is taking place right now).
And there are perl, python, and tcl interpretors/compilers for Windows (see activestate website). More importantly, there are a many software developers who are working on the bioperl, biopython, and biojava projects whose goal is to facilitate the development of portable code, rather than having to reinvent code for everyday tasks over and over again.
You say that most software is written for Unix OS in perl and Java. I think that if you look, you will find that much of the commercial, proprietary bioinformatics software is written for windows. This is a good business plan because companies know that Microsoft has the lions share of the desktop market. It only makes sense to write software for the most common desktop OS so you can sell a license for every desktop that needs to use the program. That's a good business plan, but not necessarily good for developing bioinformatics projects.
While I wouldn't think of discouraging your investigations into C# and .Net, I think you ought to take a close look at the opensource projects (like bioperl, biopython, biojava, bioConductor, R, genex and many many more), and see what people are actually doing in bioinformatics rather than to assume that because it isn't based around a microsoft format it is either too hard or done badly.
btw, this little form will only allow me to choose the value "Anonymous" for "Post as". I'm sure that this is just a problem with browser compatibility, but its another example of why we need to have some common ways to do things, whether we use linux, MS, solaris, or a macintosh.
srlasky
|
|
|
|
|
|
|
|
|
![[University of Birmingham]](https://www.bioinformatics.org/images/ack_banners/Birmingham-728x90.gif)