Comment Re:Continuity (Score 1) 891
I use pylab and scipy as a replacement for Matlab. But it's really frustrating because sometimes you do an update and everything can bust because this or that lib won't compile with your current compiler or this or that dependency is not available or it wont work with X or aqua term or whatever.
On OS X, matplotlib (pylab), numpy, and scipy are all single click installers now for python 2.5, which is itself a single click installer. I put all these through their paces daily and the newest installs are rock solid.
Just as you wouldn't buy OS X 10.6 and install it before the second point upgrade (e.g. 10.6.2), you shouldn't try to upgrade to python 2.6 before all of the other scientific packages are ready.
Maintaining a python 2.5 scientific environment has never been easier on OS X and ubuntu. I think it's even easy on windows with enthought.