Re: Declaration design preference: type or name first

On 1 Nov, 03:38, CBFalconer <cbfalco...@[EMAIL PROTECTED]
> wrote:
> Mike Sieweke wrote:
>
> ... snip ...
>
> > "Type first" tends to line up the types, leaving the names at
> > random columns, as in C:
>
> > int x
> > float y
> > typedef int t
> > const int c = 4
> > char a, b
> > int (*m)( int j, int k )
> > void inc( int *x )
> > {
> >    *x += 1
> > }
>
> > Now quickly find the definition of "c".  Sure it's easy here, but
> > it's not so easy if there are 20 variable definitions.
>
> While I agree with your preference, here is a counter-example:
>
>   int         x;
>   float       y;
>   const int   c = 4;
>   char        a, b;
>
>   typedef int t;
>   int         (*m)(int j, int k);
>
>   void inc (int *x) {
>     *x += 1;
>   }

Of course, this only works where the type names are short. It also
needs discipline when typing and maybe regressions to put in extra
space not originally allowed for. I don't think it is practical for a
language but recognise that wasn't the point you were making.

Just got me wondering. At the moment my syntax highlights the type as
in definitions:

<integer 32> i, j, k

and in conversions:

k = <integer 32> p

Which, I think, provides an appropriate analog between the two forms.
There has been a lot of talk about making it "very easy" to find
variable names in declarations lists. It is im****tant but is it really
that im****tant? Would "quite easy" to find not be sufficient?

I have a rule that declarations can only appear prior to executable
statements in any block of the code. To find an identifier, therefore,
one needs to check the first lines of the current block, then the
first lines of the enclosing blocks until the name is found. Functions
are not nestable but blocks are. I think this makes finding a name
reasonably easy without further syntax changes.

Anything can change, though....