Two thoughts: 1) What I think is interesting about this discussion is that there seems to be an underlying issue of the importance of a data format in human or machine context. In some research communities such as chemical imaging a research paper (human readable) with a graphics plot of an experiment (human interpretable) are absolutely necessary for studies. Its a slow but trusted process. However the researcher is also highly dependent on web based capabilities search based on certain citation and keyword criteria. From a data usage vocabulary perspective I can't help but think that we need "FOR HUMAN CONSUMPTION ONLY" labels (okay silly label, but just trying to make a point) to describe data usage techniques available for certain datasets. 2) As we think about formats such as PDF fifty years into the future, we do need to think about the best practices to interpret them. There are efforts such as http://en.wikipedia.org/wiki/Data_Format_Description_Language that have been addressing what you do with specialized forms of data. Last year I just got done recovering scientific algorithms written in FORTRAN77 from a magnetic tape reel written 25 years earlier. Even text isn't so easy recovering, thank goodness I knew I was looking for FORTRAN77. Eric S On Fri, Mar 27, 2015 at 9:50 AM, Ghislain Atemezing < auguste.atemezing@eurecom.fr> wrote: > > Le 27 mars 2015 à 17:07, Steven Adler <adler1@us.ibm.com> a écrit : > > So, does our BP document only apply to data published in the future in the > file types we bless? > > > It could also help to make *that* future more closed to the present…as > simple as in the case of NYC, say the next 1001 document produced will not > need anymore the Ruby code made by Chris, but would be access directly > without any transformation by anyone and for free ;) > > Ghislain >
Received on Friday, 27 March 2015 17:51:54 UTC