Le jeudi 01 mars, Harald Schilly a écrit:
> On Thursday, March 1, 2012 12:01:46 AM UTC+1, Snark wrote:
> >
> > Could you tell me which magical property of sage makes impossible
> > what
> >
> > is possible for other complex systems, and huge sets of packages?
> >
>
> Put simple: Sage is turing complete, a "video editor" (office package
> [*], or whatever) is not.
> A bug in such a productivity software results in an immediate error
> or something that's plain visible. A wrong calculation is just a
> number as any other.
I'm sorry, but I still fail to see how situation :
sage contains all its deps (even if they're already there)
will be any better for subtle bugs than situation :
sage has the same deps, but doesn't manage them itself more than by
declaring them (and having the system packaging satisfy them).
But I very plainly see how the second situation has the following
advantages :
(-) no risk of a bad interaction between some system packages and what
is basically a copy within sage (termcap issues) ;
(-) the burden of maintenance between sage deps is now external to sage
(why would you need to care about the deps of say moinmoin? Just
declare it a dep and let the system packaging handle that constraint!) ;
(-) security concerns get treated by the security teams of the various
distributions ;
(-) no duplications means it takes less room (sage is as big, but
sys+sage is smaller)
(-) faster to build, since it becomes possible to use more binary
packages as a base.
Snark on #sagemath
--
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
