-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
However, /sometimes /breking the rules helps maintain readability, and breaking the rules when it is neccessary is one of Wikipedia's main rules. :-) An example for long lines: URL in comment where the problem or solution is explained.
I fine with this and do not see why this should be a problem. As Antoine and Merljin explained breaking the rules ca be done on a global level for any rule wishes as well as on a specific case based level with # noqa. And I see it exactly as you do and would like to underline it with PEP20:
[...] Special cases aren't special enough to break the rules. Although practicality beats purity. [...]
To make it clear; I vote for automated style checks in order to act for us (the developers - humans) as an additional help and source of info (e.g. about the code "quality" or "conformity"). I AM STRICTLY AGAINST failing style checks to prevent us from merging the code! So the bot should run and inform us - as it does right now - and Antoine can help/support e.g. xqt with PEP8 changes as done in the past. This includes all changes that improve code readability, maintainability and do some good - but we should (of course) NOT insist in PEP8 for all kinds of code pieces and stuff like comments, docu and else. And NOT change how the bot behaves right now (e.g. block merges).
Furthermore this PEP8 "transition" or compilance check should go gradually, e.g.: 1.) we exclude ALL PEP8 rules violated by our current code in the '.pep8' file 2.) we take the first rule excluded and try to solve all issues 3.) we remove this rule exclusion from '.pep8' file 4.) we iterate 2.) to 3.) until all issues possible are solved 5.) we check what we have, if and how to proceed
Greetings DrTrigon