On 2010-12-22 14:34, Cedric wrote: > I love SAGE but then there is one flaw in it that I find one of the > most severe in software-design of all. From my layman point of view > (which I'm sure is wrong and unjustified) a software that bundles > every single of its dozens dependencies has either a fundamental > design flaw or is extremely badly managed.
The "bigsagerant" actually mentions several points: 1) doctests. The Sage source code and documentation is full of examples. The examples are tested to make sure that a patch to Sage doesn't break anything. In some cases, using a different version of software will give slightly different results. These results are mathematically still correct, but are considered different when doctesting. 2) patches. Sage ships a lot of its dependencies is non-vanilla form, with patches. In many cases, it is hard or even impossible to make Sage work with non-patched upstream packages. 3) using development upstream packages. For PARI, we don't use the stable release (which is now quite old), we need a development version. Jeroen. -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
