duck!

Steven Schveighoffer schveiguy at yahoo.com
Mon Oct 18 05:03:42 PDT 2010


On Sat, 16 Oct 2010 18:04:23 -0400, Walter Bright  
<newshound2 at digitalmars.com> wrote:

> Steven Schveighoffer wrote:
>> On Sat, 16 Oct 2010 16:31:34 -0400, Walter Bright  
>> <newshound2 at digitalmars.com> wrote:
>>
>>> If I google for "adapt for D" I'll get a hopeless mess of irrelevant  
>>> links. "duck typing for D" should be much better. Remember that google  
>>> ranks pages by relevance, and searching for "duck" will give higher  
>>> ranking for pages with "duck" in the url, title, headings, etc. That's  
>>> just what we want.
>>  And it should come up with the page on digitalmars.com titled 'duck  
>> typing in D' which describes how to use templates or the adaptTo type  
>> to achieve duck typing.
>
>
> When writing fiction, it's a good idea to constantly shift which words  
> used to describe something. But when writing tech manuals, and when  
> making things search engine friendly, it pays to use a single term and  
> use it consistently.
>
> For example, once upon a time I read some article on arrays, and it  
> variously referred to the array "elements", "entries", and "values".  
> Really, that sucked, as the reader was left being not quite sure if they  
> meant the same thing or not.
>
> If you expect people to search for "duck typing" (and I do) then why  
> stick another level of indirection? Call it a "duck". When you find  
> yourself constantly saying "duck typing: see adaptTo", then you named it  
> wrong.

Is this a case of foresight is 20/20?  Look, you can't predict the future,  
and knowing what you will constantly be saying isn't possible.  You could  
constantly be saying 'yeah, I know duck isn't the only way to do duck  
typing in D, but we felt it was a good descriptive name.'  And not having  
to constantly say anything isn't proof that you have preemptively avoided  
it.

But it doesn't matter anyways.  My point is, marketing has nothing to do  
with naming functions.  People will not decide to use or not use a  
language based on a function name.  If the name describes the function  
properly, then it will not be noticed as a defect, and few will have  
problems with it.  All three suggestions (adaptTo, as, and duck) satisfy  
this IMO, so really this argument has nothing to do with the name.  It's  
all about the incorrect belief that the name itself will somehow be seen  
as a marketing benefit.

You're making a lake out of a duck pond ;)

-Steve


More information about the Digitalmars-d mailing list