*1) Slow startup speed.* Everyone dislikes the slow startup speed. Though it's been argued that it should be known as the Clojure slow startup speed. Since even though the JVM is slower to start then say python, most of the slowness comes from the Clojure overhead.
*I know this problem is being addressed by the Clojure supporters, but its a hard one to solve, all I can say is that its improved version after version.* *2) Slow dependency resolution* It doesn't sound like an issue with Clojure. Can you elaborate? Are you complaining about Leiningen being slow, or just that having to download dependencies is slow? Either way, this is probably more of a JVM issue, or even a tooling issue, you could look into alternate tooling. *Not a Clojure related problem.* *3) `empty?` doesn't behave as I intuitively interpret the name of it* One thing you should embrace in Clojure, is that part of why its so awesome is that you are not as restricted by the limitations of the language as with other languages. So if you don't like empty? throwing exceptions for Long and Keyword, you can re-def your own empty? which catches those and returns false or true (whatever behavior you want) instead. As some people have said before, a lot of the functions have good reason to behave as they do, but sometimes its not as intuitive from the name or form people's expectations. The doc mentions: "Returns true if coll has no items - same as (not (seq coll)). Please use the idiom (seq x) rather than (not (empty? x))" I'd be curious if anyone knows what is the idea behind empty? and why the doc mentions to use the idiom (seq x) when there is also a not-empty fn in core. As an example, I struggled with `contains?` for a while. Couldn't figure out why they decided to make this: `(contains? [:a :b :c] :a)` return false. Why they chose to make contains? check for the existence of the index on a vector. Then I learned that the use-case behind `contains?` is to verify that a call to `get` will succeed. Once I learned that I was like... make total sense. So I wrote myself a `contains-val?` for the use-case I was having and done. I'm sure there's similar reasoning for `empty?` *So I'd say this is not a Clojure related problem, though maybe there's a little issue with the names of functions having similar names then other popular languages and not behaving the same way.* *4) Maps are unordered by default* Maps are unordered from the theory itself. Don't ever expect Maps to be ordered, unless its very explicitly stated. Ordered Maps have different time/space characteristics too, which is why Maps are unordered by default. This isn't anything to do with Clojure. You should try and get more familiar with the formal literature on common data structures, as its what most languages build upon. *Not a Clojure related problem.* *5) Reloading has some quirks* Clojure gives you so much reloading-ability with its REPL centric development process, that it plays against it when it fails, we get really upset. Its almost like the uncanny valley of reloading. You can search for reloaded workflows though, some people have apparently figured out ways to organize your code so that everything can be reloaded without ever failing or having to restart the REPL. *This problem relates to Clojure, but also to Lisp in general, and I don't think its solvable without constraining the language too much.* *6) Cryptic official doc* I thought the official doc was very cryptic at first too, but over time, I got used to the way its documented, and I can now figure out how to use almost all functions just from the official doc. So I'd say this is just a matter of time and experience with the nomenclature and the way its explained. But I agree, its not meant for people just starting off with Clojure. I recommend clojuredocs for a more beginner friendly doc. *Clojure could try to add a bit of a friendlier doc to its official doc.* On Wednesday, 20 July 2016 05:54:27 UTC-7, Peter Romfeld wrote: > > I really love clojure over all, it makes maintenance/collaboration of code > such a breeze. its easy to get new employees start to work on it even > without any previous clojure knowledge! > > I do hate the JVM startup shit, i hate how you it takes forever to fetch > deps on a aws medium instance (you probably can fix it with uberjars) > Im frustrated with `empty?` throwing exceptions for Long and Keyword > Not complete clojure, but i hate datomic.api/log since you cant put into > your tests with a in-mem db > I had some issues understanding that maps of certain size are not in order > anymore because of the underlying java functions and optimizations > Im a bit upset about development reload with defrecords > most of the things that are not documented by the community are documented > very cryptic and hard to understand > the api docs are almost useless! - to get the arrity and stuff.. wtf i > still have no clue how to use this new function i never used before... > security features in most frameworks are just smoke and mirrors, functions > that dont actually do what they should do... > > anyways, if i remember more i can add... > just wanted to give you guys some input, and looking forward to make > development experience even better! > > cheers, > peter > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
