>>>>> "Malay" == "Malay" <curiouser at ccmb.ap.nic.in> writes:
>> what I'd really like to see is some sort of certification by a
>> trusted third party.
Malay> While I agree that source code should be available for all
Malay> softwares used in "mission critical situation", such as
Malay> "publishing a paper", I guess bringing a third party policing
Malay> will stifle the amateurs.
I don't remember the last paper I saw published by an "amateur".
If you are referring to software produced by non-profit types of
organizations, I think the obvious thing to do would be to make the
certification process free for them. Payware, on the other hand, would
incur a fee for certification.
Malay> And any policing/reviewing committee members will be consisting
Malay> of human, thereby making it a fertile ground of poor code
Malay> certified by *such and such committee*.
Perhaps I'm being unrealistically optimistic, but I think an
appropriately objective review carried out in an open manner would be
relatively free of bias, especially since the reviewers would be in
some sense staking their reputations on their opinions.
Malay> The best way is to stick to the old adage of *tried and
Malay> tested*. Sheer number of users should be the only criterion to
Malay> judge any *free software*.
Quality is not the issue. _Correctness_ and _lack of bugs_ is the
issue.
If I pick up a molecular biology paper, I can read the materials and
methods section and repeat the experiments described in the paper
(with some amount of difficulty). Using my own reagents, and my own
preferred protocols for common steps.
For a bioinformatics paper (when no source code is available), I can't
easily repeat the experiment, because I lack details of the way the
algorithms in the paper were implemented. All I can do is look at the
results the authors provide, and hope or trust that the programmer(s)
were competent.
This is like reading a yeast genetics paper without having the
genotypes of the strains given, or a biochemistry paper without
details of the experimental conditions.
john.
--
----------------------------------------------------------------------------
[ John S Jacobs Anderson ]------><URL:mailto:jacobs at genehack.org>
[ Genehack: Not your daddy's weblog ]------><URL:http://genehack.org>
