[navalsaini]

This is a really interesting story. Glad you shared it. I suppose, being a CEO, one is entitled to such privileges. :)

PR = pull request. Created so that your peers can review the code and catch any issues (architectural, logical or just styling). I think the value attributed to them should be 100x to 10x to 1x (or lesser than 1x).

But its not a commonly held belief. Many software engineers take pride on producing beautiful code. Its not about the product but also about wanting to be the best (say taking it from 99% to 99.9%). But it has its bad side - product can take a back seat to the new cool aid drink language on the block.

[sloaken]

So back in the dark ages ... 20 years ago, company I worked for had what they called 'SoftWare Intensive Inspections'. Code we were writting delt with Telecomm and if a system failed then the telephone company lost money and when it failed real bad, 911 would not work and people could get hurt.

Part of the proces was that when you reviewed stuff, you could comment on: 1) does not work 2) not understandable 3) not maintainable 4) failed to meet standard - but these were usually related to #2 or #3 5) excessively slow or large - like could be written to be 10 times faster, just double the speed was a STFU

During the inspection the author was the note taker. Another would read the code and explain it as best they could. If they could not explain the code, then it failed #2 & #3. One person, usually team lead would arbitrate if the problem was really an issue.

This process seemed to work real well, and found a lot of bugs before the code was fielded. Of course someone would always argue the tabs verses spaces, and where do the curly braces go. This was usually met with STFU. The most common standard fail was the use of magic numbers. I had one jerk who actually did a #define TEN 10. Needless to say that failed.

In all of this STYLE was never a valid complaint.