Biomedical Engineering Reference
In-Depth Information
like libraries for reading and writing standard fi le formats who do not wish to
open source their entire software offering.
14.6
CONSTRAINTS ON OPEN - SOURCE SCIENCE
Open source and open data are making serious inroads into scientifi c research
practice, but the cultural transformation is not complete:
•
Fear, Uncertainty, and Doubt
Researchers still have lingering suspicions
that making source code publicly available before a paper on a project
is published might somehow constitute “previous publication” and derail
the paper in review. Many scientifi c academies and journals have taken
steps to clarify their policies around this, but there is still confusion
among authors, particularly when they have not yet chosen a journal for
submission, and this can have a chilling effect on openness.
•
Data Privacy
There are also potential privacy issues that may make
publishing some data more trouble than it is worth. A prime example of
this is the health data security and privacy provisions of the Health
Insurance Portability and Accountability Act of 1996 (HIPAA), which
have been seen by many as having a chilling effect on research [13].
•
Data Value
And, of course, raw data are the ore from which researchers
extract their nuggets of publishable work, and many are reluctant to
share with potential claim jumpers.
•
Bureaucratic Barriers
Even researchers who see that more ore means
more nuggets and agree to share and share alike may discover that grant-
ing would-be collaborators access to one's computing resources from
outside the institutional fi rewall is an administrative nightmare. Happily
(to paraphrase Internet pioneer John Gilmore), the Internet treats this
as an error and routes around it in the form of cloud computing.
14.7 USING CLOUD COMPUTING TO ELIMINATE
BARRIERS TO COLLABORATION
Say you wish to allow a colleague at another institution to analyze some of
your data and share the results with you in an iterative, collaborative fashion.
You have the data and you have chosen or developed the analysis software.
You even have secured the blessing of your supervisors to work on this col-
laboration. All that remains is getting your institution's IT powers to allow
your collaborators access to your network. This is probably going to be more
bother than you would like, and may be even impossible, what with getting
them visiting scholar status and other administrative details, negotiating your
institution's IT policies just to open your fi rewall, and so forth. An attractive
Search WWH ::
Custom Search