No reason to avoid single-segment namespaces.

[llasram]

Clojure namespaces with gen-class map to classes with the same full name as the namespace. This produced some concern about single-segment namespaces being equivalent to Java package-less classes and a cargo-culting avoidance of single-segment namespaces. In practice non-gen-class single-segment namespaces do not appear to cause any known problems, and have begun to appear in libraries by well-respected Clojure community members, such as Brandom Bloom’s backtick [1].

[1] https://github.com/brandonbloom/backtick

[bbatsov]

Clojure namespaces with gen-class map to classes with the same full name as the namespace. This produced some concern about single-segment namespaces being equivalent to Java package-less classes and a cargo-culting avoidance of single-segment namespaces.

Yeah, this is why we introduced this rule in the first place. What you're saying sounds reasonable, but I'd like to do some digging first (and perhaps hear more comments).

Btw, probably it still makes some sense to suggest that projects should use multi-segment namespaces when they have more than a single namespace. I also have a feeling people are abusing the xxx.core namespace, but we have lein to blame for that. :-)

[arrdem]

I have no particular objection to single segment namespaces. gen-class is a feature I almost never reach for, and I've found it expedient in the past, and for small namespaces.

[puredanger]

Namespaces exist to disambiguate names. Using a single segment namespace puts you in direct conflict with everyone else using single segment namespaces, thus making it more likely you will conflict with another code base.

So, for libs, I think it's a very good idea to use a first segment for context vs every other lib you might include. Within your own private app of course, you can do whatever you like. :)

[alexander-yakushev]

For what it is or isn't worth, Android doesn't support single-level packages, probably for good reasons. Besides, dot in a package name is a useful telltale sign for different tools (like clojure-mode) to know something is a package/namespace or a class.

[danielcompton]

Single segment namespaces have problems in ClojureScript as well.

[David Nolen] Compiler support for single segment namespaces has always been spotty and there's no intention to improve it. There are fundamental semantic issues - for example if you have a DOM element with a CSS id and you haven't supplied a namespace but you goog.require one with that name it will return the DOM element. Good luck debugging that.

Single segment namespaces are nearly alway representative of a problem waiting to manifest. If you want to use them then you can suppress the warning in your build options.

You'll get warnings in ClojureScript by default when compiling single segment namespaces.

IMO everything in the design of Clojure, the JVM, and ClojureScript seems to be pushing away from single segment namespaces. The only benefit from single segment namespaces (I think?) is a slightly more elegant ns declaration. While that's nothing to be sniffed at, IMO it's not a big enough gain. People can do whatever they want in their own code, but in the de facto style guide for Clojure, I think single segment namespaces should be avoided.

[bbatsov]

I guess the best we can do is to extend the rationale behind the existing rule then. In general a lot of rules in the guide could benefit from extended rationale, but I've been pretty short on time lately.

[dijonkitchen]

Should this PR be closed now given the above discussion? Maybe just have the default be something like name.core instead of just name.

[bbatsov]

I didn't close it simply because for me it served as a reminder to move some of this conversation to the style guide. Alas, I never got to doing this. :-)

[bbatsov]

And I finally converted this reminder to improvements in the guide. Thanks to everyone who participated in the conversation!