Perhaps other things should be mentioned in the GSOC application
instructions. I do think that the PR requirement is reasonable as a
baseline for GSOC though.
To be clear opening pull requests is a good thing. Reviewing pull
requests is a lot of work for maintainers though. If lots of people
start opening PRs but not doing anything else to help by e.g.
reviewing those PRs then we will just have lots of unreviewed PRs
which is not so useful. Especially many of these PRs have very basic
problems like "a test should be added" or "the test should be for what
is actually fixed". These could easily be reviewed by many of the new
contributors.
This PR is a good example of the work problem that is created by
opening PRs rather than reviewing them:
https://github.com/sympy/sympy/pull/26117
What is needed is for someone to verify that the definition of
assoc_legendre as used in SymPy matches with the definition that is
used in mpmath. This is partly about reading the docs and partly just
a case of doing some numerical checks but it needs to be done
systematically paying close attention to any special values and to
what SymPy's symbolic evaluation handles. Writing the code and opening
the PR without doing that checking leaves all of the work to the
reviewer. Anyone else could review it though by doing that checking.
--
Oscar
> To view this discussion on the web visit
https://groups.google.com/d/msgid/sympy/CAP7f1AhzY_A2Rt9aMQy1M_6MzZ09L1uP0p_0RLcni1TpX8rwpw%40mail.gmail.com.