[Bioperl-l] Auto-method caller proposal

David Messina dmessina at wustl.edu
Tue Jan 9 12:51:25 EST 2007


On Jan 9, 2007, at 10:50 AM, Sendu Bala wrote:
> My preference at the moment is to use my proposed method, if only
> because I'm lazy. I want someone to tell me there is significant value
> in using AutoClass before I spend time on it.

I am reluctant to get in the way of someone threatening to code, and  
I'll admit I don't know enough about the particulars of this problem,  
but in general the problem I see with a custom solution to an already- 
solved problem is that:

- the custom code will be brand-new and examined by few pairs of  
eyes, whereas the existing code (AutoClass) has been already vetted  
and debugged by the Perl community. Based on Nat's design notes,  
there are some subtle gotchas to implementing auto-classing which  
have to be considered carefully.

- the custom code has not been used before in BioPerl, whereas  
AutoClass has been already -- why have two sets of code in BioPerl  
doing the same thing? That makes the BioPerl codebase harder to  
maintain and unnecessarily complicated.


Again, this is just how it looks to me sitting in the cheap seats, so  
my apologies if I'm way off base here...


Dave


More information about the Bioperl-l mailing list