Re: RFC: error code strings

On Thu, Jan 17, 2013 at 3:01 PM, Andreas Tolf Tolfsen
<andreastt@opera.com>wrote:

> On 01/17/2013 03:56 PM, Simon Stewart wrote:
> >
> > Based on feedback from early drafts, we're moving from status codes being
> > reported as integers to using strings. I'd appreciate early feedback on
> the
> > proposed change to the spec. Notably, I'd like to know whether people
> > prefer the "lower case with spaces" approach or would prefer "EALLSHOUTY"
> > or "CamelCasedWithNoSpaces"?
> >
> >
> https://dvcs.w3.org/hg/webdriver/raw-file/tip/webdriver-spec.html#error-codes
>
> Part of the argument for moving to use strings for error codes was to
> make the wire protocol more obvious for humans.  Since we are going with
> strings, I don't see any particular reason why we'd need to use
> “variable name” syntax on them.
>

So "strings with spaces" then?


> The strings should be straight forward, and be able to be understood by
> a human inspecting the communication.
>

Most people who would need to read these strings can cope with CamelCase or
even thisisalloneword (thanks to URLs), but I quite like the legibility of
"strings with spaces"

Simon