One of the trickier problems of language design is the syntax. Right now I am wrestling with the syntax of anonymous functions in Heron.

I know that syntax is treated as not worthy of study by many computer scientists, but out in the real world good choices in language syntax I truly believe make or break a programming language. 

I honestly believe that the s-expression syntax of Lisp and Scheme is a huge obstacle to its adoption, because it makes the language appear more exotic than they really are.

Right now, I am working on the syntax I want to use in my language Heron for passing anonymous functions to higher-order functions. 

The C# syntax for this isn't too bad: 

However, I find the obligatory "});" to be aesthetically displeasing. Also, it seems far too verbose compared to what I want to express. 

What I really want to write is: 

This is convenient, but could be ambiguous for the reader. There are a lot of things that happen behind the scenes 

In many curly-braced languages the "()" operator acts a function application operator. When we don't require the "()" operator in a language, we have implicit function application and usually with it implicit partial application. This is a feature of ML based languages (like OCaML and F#) which took me a long time to get used to. 

So another option I am considering right now is having a secondary operator for applying functions.

The above example would then look like:  

This would be read as: apply the function myHigherOrderFunction to the anonymous function. 

In the case of functions taking no arguments, it might make sense to allow omitting the "() <=".

So we would have:

But now I am unhappy because we are either doing some black magic or we have a group of weird symobls "-> () <=". The only left-over advantage of the "->" operator is when performing partial application.

Oh well, like I said: designing the syntax is hard.