Re: camelCase or under_score ?

[Philippe Lhoste <PhiLho@...>]

On 31/03/2011 08:41, Brian Maher wrote:
>   LexHTMLToXML
>   LexHtmlToXml
>   lex_html_to_xml

And why not lex_HTML_to_XML?
This example is a bit contrived, focusing on mixing uppercase abbreviations with normal words.
And, maybe I am too used to camel case, but I still find the first one more readable than 
the others... ^_^'
Readability is indeed subjective.
Some people find languages which are case insensitive easier to use (I disagree... :-)). 
And a language, Nimrod, even went as far as making the three (four with mine) examples 
above... addressing the same variable!

Otherwise, while I adopted camel case for my needs, I go with the other people, I just 
stick with conventions used in the code I hack...

About library design: it is indeed a compromise. With a very large API like Qt or, 
perhaps, OpenGL, it is both simpler and sensible to use the naming and idioms of the 
existing code base (with adaptations if made in a consistent, logical way like dropping 
type differences, etc.). It is easier to adapt existing code or too look up information.

With smaller APIs, like LuaSocket, it is nice to adapt the existing, probably too 
idiomatic API, to idioms of the language implementing it (multiple return values, 
iterators, tables, etc.).

--
Philippe Lhoste
--  (near) Paris -- France
--  http://Phi.Lho.free.fr
--  --  --  --  --  --  --  --  --  --  --  --  --  --