As one of the people that reviews physics PR's, I'll admit despite our best efforts this can be a problem. There are certainly a couple obvious examples (Dagger [1], TensorProduct, HilbertSpace's) and likely a few more things that are less obvious. A few things I'll note related to this problem:
- I think one possible issue here is in some cases, there can be behavior used in doing the physics that doesn't really generalize, a good example being fermi surface considerations from the second quantization code which is still floating around in the KroneckerDelta class. - Another problem comes with things that, tho they could be used more generally, don't really fit anywhere. Here the Hilbert space stuff is a good example. The problem comes when someone starts implementing things with vector spaces and they aren't aware of the physics code. I'm not sure of a good way to remedy this. We can open issues for things that could be generalized, but it doesn't seem very efficient to have a list of things other people would have to check to ensure things don't get duplicated. More review, both review of physics PR's and people that review physics PR's (myself included) reviewing outside the physics module, seems like the best sustainable answer, the difficulty being exactly what Joachim noted on the struggle of reviewing material outside your knowledge base. Sean [1] https://github.com/sympy/sympy/pull/1158 On Thu, Jun 14, 2012 at 4:01 PM, Joachim Durchholz <[email protected]> wrote: > Am 14.06.2012 22:46, schrieb Aaron Meurer: > > Any thoughts on how to prevent this issue? I know it's not an easy >> problem, and it's probably one that every large project faces, but I >> feel that we can do better in many cases. >> > > It's definitely something to be detected during review. > > I think it's the general problem of review. It's really hard to review > stuff that's giving people the MEGO effect. I barely understand maybe 20% > of what's being worked on; how am I supposed to review anything? At best, I > can help with bikeshedding (and I'm certain I have done my fair share of > that). > Now I'm more a programming than a math type, so this kind of problem is > more pronounced for me than for all you math types; still, I suspect much > of the review is hampered by similar effects. > > Just my observations, from a very specific viewing angle - I'm pretty sure > I'm getting just a small part of the picture. > > > -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to sympy+unsubscribe@** > googlegroups.com <sympy%[email protected]>. > For more options, visit this group at http://groups.google.com/** > group/sympy?hl=en <http://groups.google.com/group/sympy?hl=en>. > > -- You received this message because you are subscribed to the Google Groups "sympy" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/sympy?hl=en.
