Message316858
| Author |
ned.deily |
| Recipients |
Anthony Sottile, bethard, bskinn, eric.araujo, gregory.p.smith, memeplex, ned.deily, paul.j3, wolma |
| Date |
2018年05月16日.20:06:06 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<1526501166.36.0.682650639539.issue33109@psf.upfronthosting.co.za> |
| In-reply-to |
| Content |
Several of the core developers here at the PyCon US sprints in Cleveland have discussed this issue. It seems like there legitimate arguments for either behavior. But, while none of us are argparse experts, we all were persuaded by Wolfgang's and Brian's arguments that preserving compatibility with 3.6 (and recent earlier 3.x releases) should be given more weight than attempting to restore 2.7 behavior at this point. We specifically did not get into the finer points of argparse behavior and any other proposed argparse change.
Unless someone comes up with a more persuasive argument, I intend to merge this change for 3.7.0rc1 in a few days. |
|