Hacker News new | past | comments | ask | show | jobs | submit login

It is a pretty convincing argument for adding transformation (do this thing to every item in the list), and filtering (give me a sub list of everything that meets this condition)

It is a much weaker argument for adding summation, since you'd have to get into the down and dirty details of whatever operation your summation represents, e.g. is it associative? You can see how it get ugly fast in the discussion about the differences between foldl and foldr (presumably fold-left and fold-right respectively)

----

Now, Java had filtering in one of the libraries, I remember using filtering with folder operations many moons ago, so it's not like you can't roll your own. Likewise with a "do this thing to everything in the list" operation.

For loops have a use, specifically when you want to do things in a particular order. Reading lines from a file for instance you might want the lines you're putting into your data structure to be in the same order as they are in the file (as a specific counter-example, I once wrote a 'unique line' app where it didn't care about the ordering, of course if I'd been on unix there would have been some command line tool to do it for me, but it was fun and fast to write and hand tuning the grep would have been a PITA)

BUT, there are a lot of tiems I've used for loops where the sequential nature of the loop is irrelevant, and here I could see someone making an argument that the for loop is clumsy and random (from the article they object because it uses too many tokens! The horror! :D )

I could also see someone making the counter-argument that what that actually means is that the for loop is powerful (never mind that we have no objective standard for what that means. Perhaps 'powerful' in the sense of 'easy to blow your own foot off with' would fit here :D )




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: