The notion that code must be reviewed is just odd to me. Even offensive. Most of code review comments I've seen are just about style, hardly ever about substance.
-
-
Take a look at “Facts and Fallacies of Software Development” Robert L Glass, where it is clearly shown based on multiple researches and examples how p2p code reviews are one of the most efficient ways to eliminate bugs and bad design earlier.
2 replies 0 retweets 0 likes -
The problem with this kind of "research", however, is that nobody has explained why there is so much terrible design and bug-ridden software shipped by companies that do exactly these practices.
1 reply 0 retweets 4 likes -
Replying to @cmuratori @meglio and
If the argument is that, say, Google is shipping "slightly less buggy and slightly better designed, but still terrible" software because they mandate lots of code reviews, I would say, well, then we need to figure out something else, because "slightly" isn't good enough.
1 reply 0 retweets 3 likes -
Replying to @cmuratori @wisam910 and
In the original context, i.e. accepting merge request to an Open Source product, what would be a better way to block low quality code from entering the codebase? Now that code reviews are questioned per se, what would be an efficient way to give feedback and ask for improvements?
1 reply 0 retweets 0 likes -
I guess I'm not sure what you are talking about now. You were citing "Fact and Fallacies of Software Development", which is not about open source software, as evidence for "P2P code reviews". I am simply saying, that is not a useful citation.
1 reply 0 retweets 1 like -
Replying to @cmuratori @meglio and
If you're asking me how I would organize an open source project where I had no idea who was contributing to it or why, I guess I would say, don't do that? The only software I know that is good is written by small teams of people who are all very good.
1 reply 0 retweets 6 likes -
Replying to @cmuratori @meglio and
I have never seen a low-bug, high-perf project that relied on code reviewing code from people who are not very good as a strategy. If it has ever worked, I am unaware of the project where it did.
1 reply 0 retweets 2 likes -
Replying to @cmuratori @meglio and
If someone wants to make the claim that "code reviews are essential to shipping large-team mediocre software", I would not necessarily challenge that assertion.
1 reply 2 retweets 6 likes -
Replying to @cmuratori @meglio and
If someone wants to make the claim that "code reviews are essential to shipping high-quality software" I would challenge that assertion, because the only high-quality software projects I can think of do not use code reviews.
2 replies 1 retweet 5 likes
And the flip side - I know hundreds of projects that require code review that are extremely low quality.
-
-
Replying to @cmuratori @meglio and
So it seems kind of self evident to me that, to the extent that one cares about high-quality software, code reviews as a required thing are probably irrelevant. They may be useful in certain circumstances, but as a rule, I don't think the actual observables support that at all.
3 replies 0 retweets 5 likes -
Replying to @cmuratori @meglio and
I agree. Code reviews are no panacea. I've only seen them be useful when they are done in person, where you have to explain your code to someone else, but it's mostly just 'rubber ducking'. They are rarely useful when there is no interaction.
1 reply 0 retweets 1 like - Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.