[erlang-questions] Why isn't erlang strongly typed?

Toby Thain toby@REDACTED
Thu Oct 23 22:01:14 CEST 2008


On 22-Oct-08, at 3:51 AM, Joe Armstrong wrote:

> That's really nice, sometime, almost as a joke I've written code  
> like this
>
> in_degree_of(G, in, V) ->
>      ...
>


Ohhh! That's no joke! I think it's wonderful.

Pattern matched glue as internal documentation! I'm more in love with  
Erlang than ever.

--Toby

> But I think your syntax is nicer - it's also reallly easy to implement
> - worth experimenting with I think
>
> /Joe
>
>
> On Wed, Oct 22, 2008 at 4:47 AM, Richard O'Keefe  
> <ok@REDACTED> wrote:
>>
>>>
>>> This is a consequence of Erlangs incredibly unergonomic function/
>>> parameter
>>> naming conventions and has little to do with strong typing (or its
>>> absence).
>>
>> I have proposed in the past that Erlang should adopt
>> Paul Lyons' "split procedure names" idea.
>>
>> Instead of
>>
>>        call = [module:]name([expr{,expr}])
>>
>> we take
>>
>>        call = [module:]{name([expr{,expr}])}+
>>
>> For example, instead of
>>
>>        in_degree(G, V)
>>
>> you would have
>>
>>        in_degree_of(V) in(G)
>>
>> My experience of Smalltalk is the same as Holger Hoffstaette's;
>> the names tell you what the argument rôles are so that it is
>> very unusual to get arguments confused.
>>
>> _______________________________________________
>> erlang-questions mailing list
>> erlang-questions@REDACTED
>> http://www.erlang.org/mailman/listinfo/erlang-questions
>>
> _______________________________________________
> erlang-questions mailing list
> erlang-questions@REDACTED
> http://www.erlang.org/mailman/listinfo/erlang-questions




More information about the erlang-questions mailing list