Tonight I rest well...

May I ask what this is?

@Mitch377 very true.

@tisaconundrum It's a code coverage report, it shows which parts of the code are tested with automated tests and which are not.

@kvsm you should try coding in type safe languages (read Haskell 😎) and see for yourself that you will be able to write such correct programs you won't need 100% test coverage

108 lines... Get back to us when you have this on 108k and tests actually assert something.

Coverage means shit without assertion metrics.

@addvilz hope you're enjoying working on your crusty old monolith I guess

@rayanon I've been doing a bit of PureScript recently, loving it. ☺👍

Still needs tests to ensure that your logic is right though 😉

First of all, large code base does not automatically imply legacy or monolith code. Apart from unit tests, there also exists specifications, integration tests, e2e tests etc. On a more philosophical note, micro services and strong segregation is also not a golden hammer for all the use cases. As with everything, common sense applies, especially on scaffolding level.

As for the random JS dude you just linked - fuck for? That talk is garbage, makes plenty of assumptions (incorrect ones, might I add) and author obviously lacks experience in design and maintenance of large scale systems. Shifting or splitting complexity does not prevent it, it introduces even more complexity. Separation of concerns can and should be maintained instead of blindly hacking source code apart and introducing extra complexity and points of failure in a form of common management of parts, thus reducing resilience of whole business logic.

@kvsm Cool! Re: tests for logic - very true.

I'm not going to debate your views, you are entitled to them, but you're out of order to dismiss the author of that talk so quickly. Greg Young definitely knows enough about composing maintainable systems for his thoughts on the matter to be worth listening to, by anyone.

@DLMousey Nah, no flame wars, arguing over opinions on the Internet is never a fruitful endeavour. ☺

I'm dismissing nothing quickly. Young's entire "expertise" comes from coining a term for a slightly dangerous practice that has been in use for decades, repackaging it and marketing it nicely, and getting some small endorsement from Fowler. He has little authority in the matter. Apart from just-out-of uni "superstar" developers who consider V8 the next best thing right after sliced bread and parallel garbage collectors, and have too much free time on their hands - nobody cares.

Well this hostility was uncalled for, honestly. Let the lad enjoy his good work, obviously at some point he will write codes in thousands of lines, but everyone starts somewhere!
#KeepDevrantPositive
#RantAboutMoronsNotEachOther

@Justananon Well said! I have worked on larger projects professionally, this is just a side project I started the other day. ☺ I wanted to investigate the performance differences between brute forcing a solution to a sudoku puzzle, and solving it by logic (there are brute force solvers on npm, but I wanted one that could also step through a solution as a human would solve it). Anyway, it'll be on npm when it's done. 👍

Don't feed the troll.

@kvsm thank you for this. I'm glad code rewrites are not a bad thing.

@Justananon bs. I said that code coverage alone without asserts is worth shit. Am I wrong? Then op assumed I am working on something I am not.

Literally everyone can write test with 100% coverage on couple of hundred lines. Not everyone goes on to vote-whore about it tho.

Then again, other devs can also be morons. This is how we end up with test suites worth nothing. If thats not worth the title, idk what is.

@addvilz Yaaawwwn.