# Make code review mandatory? **URL:** \ **Category:** Committers\ **Created:** [May 12, 2020, 6:26pm UTC](https://discuss.python.org/t/make-code-review-mandatory/4174 "2020-05-12T18:26:47Z")\ **Posts on this page:** 1\ **Showing post:** 33
**Author:** ![vstinner](https://sea2.discourse-cdn.com/flex002/user_avatar/discuss.python.org/vstinner/32/15130_2.png) [@vstinner](https://discuss.python.org/u/vstinner)\ **Post date:** [May 14, 2020, 2:38pm UTC](https://discuss.python.org/t/make-code-review-mandatory/4174/33 "2020-05-14T14:38:39Z")
Sometimes I see the “expert list” and “code owners” as gate keepers: a PR should not pass unless they approve it. In theory, it’s a great idea to review PRs. In my experience, it’s quite hard to get a review in less than one week. Sometimes, if I don’t get a review in less than one week, it usually means that I will never get any review on this PR. I would love to get a review on all of my PRs, but I abandoned this idea long time ago. I defined some guidelines for myself to decide if I must wait for at least one review for a PR or not. For example, I really dislike touching ceval.c or gcmodule.c without a review since I’m not comfortable in this part of the code, and it’s easy to mess up when touching them. In general, I also prefer to get a review when I add a new public function (at the C level or Python level). For refactoring-only changes which don’t change the behavior in any way, I just go ahead without review. Development velocity also matters for me: since GitHub doesn’t support “patch series” (multiple chained PR), I like to merge many PRs in a row to get small “atomic” commits. Having to wait for the CI around 30 min to push a single commit is already very long. Pushing 10 small commits takes me 5 hours. Having to wait for a review which takes 1 week for a serie of 10 commits would take 10 weeks instead of 5 hours… To get reviews, sometimes I ping some developers on IRC, Google Hangout, WhatsApp, IRC, etc. Whatever works 🙂 Sometimes I wrote an email. In my experience, most core developers (if not all) ignore all GitHub notifications. I didn’t write “review by core developer” on purpose here. In the last 12 months, I got more and more often reviews from contributors who are not core developers. It makes me feel more comfortable to get a review from a complete foreigner to me, than having no review at all! I dislike pinging core developers too often, so I either try to not propose too many pull requests in parallel, or I skip the review step. What worked well for me is to work in pair. I did that sometimes with Pablo and Dong-hee for some specific changes. Text chat or video call, whatever works 🙂 It’s also a way to get more eyes on your code without going through to “formal” review process. Usually, we spent most of the time on investigating a bug or discussing options to solve an issue, rather than writing the code. Reviews can be done later “off-line” 🙂 --- _[View the full topic](https://discuss.python.org/t/make-code-review-mandatory/4174)._