Excellent idea! Myself, I use the R statistics software, ImageJ and MIPAV, but I'd really like the ease of use of searching for an IPKG package and downloading it...
Nowadays I find I mainly look for java apps for the ease of portability. R is available, although old, from CSW. Bryan On Fri, Nov 23, 2012 at 6:16 PM, Aurélien Larcher < aurelien.larc...@gmail.com> wrote: > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi, > I would like to enquire whether someone would be interested in > developing some kind of scientific software packaging initiative ? > > I am a post-doc researcher in applied and computational math and I am > using OpenIndiana on my workstation and personal computing nodes. > To the question "Why ?", I would reply that it's mainly about having the > possibility to setup a RAIDZ on the workstation and then benefiting from > better data "safety" (few horror/corruption stories with ext3...), using > snapshots for my data sets and results, as well as the possibility to > rollback if a new version of numerical libraries breaks something (or > even just an upgrade). > Another good point is the good support of Nvidia graphic cards and the > fact that something on the system is not breaking after every single > upgrade... > > To this regard OpenIndiana is a very good working environment for a > researcher. > The main drawback is that it requires lots of "manual" installations > which any researcher/engineer cannot afford. > Before considering using optimized libraries, one is mainly concerned > about just doing his research work :) > > As the compilation and upgrade of numerical libraries and application > tends to be tedious and as I need to upgrade PETSc and friends, I am > taking the time now to package any software I am using at work. > > I am currently packaging TeXLive, LaTeX Gedit plugin, Scilab, Paraview > and linear algebra and PDE solver software which have been using in the > past years. > > I am using pkgbuild and SFE and here is a link to a sandbox where I > started pushing spec files I wrote this week: > > https://bitbucket.org/alarcher/oi-scientific/wiki/Home > > Related to this, I have then few questions: > > - - What can be a good way to collaborate on packaging scientific software > ? > - - Should we provide the packages through the OI SFE repository or a > separate one ? Then how ? > - - Should we focus on using GCC 4.6.3 provided by SFE ? (I personally > don't have the time to handle SUNCC gotchas if any :S) > - - Where should the "OI Scientific" project by hosted ? > - - How should we handle dependencies to OI packages ? (for instance Ant > from OI is too old for some software I am using). > > Moreover, if there are enough people interested I would advocate having > a system similar to Debian Sciences with few maintainer per package and > for each a separate repository with spec files, copyright, patches, > etc... (something like Redmine ?). > > So if you are interested in a joint effort in that direction, I would be > glad to hear you suggestions. > > Best regards, > > Aurélien > > > > > > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > Comment: Using GnuPG with undefined - http://www.enigmail.net/ > > iQEcBAEBAgAGBQJQr69qAAoJEA/MOW1tOCCga/8H/i6anp+u01ZtrcGRXymouxmv > 5DodCaOUPMXsflEnRMnkmQnup7dPPoEBvMGOS1GtMeHzWIrQfATst5dxLyxetwNL > zPX83g9ychjAq/8xaHMt/ADt0isjhGXoofKR0UAvzPyy5eXhbvIfB2wxVHe9e/GC > zJbwAd984K5iAtEyp/AE7TpqGlx1Z1Tdt5aTyYs3BMxPSIEp7CUbjVecp/0kShoh > UmnqErmpTUEe+cVRw9EV5R4LTpBDcdahNzcm0F2cogMbKxxMn2EgGeqHh8Jkd2I6 > zvHp4ROUfkqE6x1Sfr4yzXHyMnw30Ul58ye5JUYOxChY2JRc5MFDhqbP3YZjkms= > =nKLb > -----END PGP SIGNATURE----- > > > _______________________________________________ > OpenIndiana-discuss mailing list > OpenIndiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss > _______________________________________________ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
