Apologies for being out of touch... > 1. Is anyone out there using GSL besides me?
Since defending, I've not been using GSL on a consistent basis. Both I and my advisor's next student made heavy use of the B-spline routines. He went on for a postdoc and is still employing them constantly. > 2. What functionality would you like to see added to GSL? I can't sign up for the work, but I think making GSL follow something like the gnulib module system could be very powerful. For background, gnulib is comprised of a collection of modules with programmatic dependencies. You say you want some capability imported into a project via gnulib, and gnulib plugs it into your build system along with any of its prereqs. No external compilation/linking is necessary. > 3. Are you willing to develop and contribute the features you want? No, sadly. Always thought that modularization would be fun but the day job eats up too much time. > 4. Would you like to see a quick release of GSL v2.0, or are you content to > work off the git repository? 2.0 is worth getting into the wild. It's been a long time since a formal release. - Rhys On Fri, Oct 23, 2015 at 10:09 PM, Mohammad Akhlaghi <akhla...@astr.tohoku.ac.jp> wrote: > 1. Is anyone out there using GSL besides me? :-) > > The GNU Astronomy Utilities (Gnuastro) extensively use GSL, > for example the the FFT routines (for convolution), the > random number generators (for making simulated noise or > random sampling). GSL is a great asset and we hope to use > it much more as Gnuastro grows and evolves, it is still > in its infancy. Thanks for the great work. > > > > 2. What functionality would you like to see added to GSL? > > Nothing special in the short term. > > > > 3. Are you willing to develop and contribute the features you want? > > Ofcourse. > > Gnuastro also has internal libraries for managing common > functions to more than one utility. We are planning to > convert them (and other useful functions) to shared/installable > libraries to be used by the astronomers (currently Gnuastro only > installs executables) in their separate programming projects. We > will use GSL's model for the creation of those libraries. > > If the libraries might be useful for the larger (non-astronomer) > community of scientists, then we would be happy to move those > libraries to GSL after they have been used/tested in Gnuastro and > passed the tests of the GSL maintainers. Since Gnuastro links to GSL > anyway, it won't make any difference for us where those libraries are > positioned. > > > > > 4. Would you like to see a quick release of GSL v2.0, or are you content > to work off the git repository? > > In the short term (while Gnuastro is still under heavy development), > the official release of GSL is the best for us. Gnuastro's programs > link to GSL which has to be installed by Gnuastro's users separately. > So as others have already mentioned, we will have to rely on the released > versions, not the git repo. So more frequent releases would be better > for us. > > Thanks a lot for the great effort in managing this wonderful and very > fundamental library. > > Mohammad >