Re: explicit null

[bobduff@world.std.com (Robert A Duff)]

In article <9605091727.AA01200@most>,
W. Wesley Groleau (Wes) <wwgrol@PSESERV3.FW.HAC.COM> wrote:
>The chances of the above search occurring under deadline pressure are
>'null' :-)  With no pressure, about 5%. ...

You're correct, of course.  In most projects, in most situations.
But it seems to me that this is an argument against having *any* coding
conventions beyond what the language RM actually requires.

>...  If what I've observed in twenty
>years of working with other people is typical, more than half will not
>notice the explicit null.  Of those that do notice it, most will think,
>"Whoever wrote this doesn't know Ada very well."  They will remove the
>"unnecessary construction" and maybe even search the file for other
>"unnecessary constructions."  Having the mindset to find such, they will
>fmisjudge things that ARE necessary and remove them too.

True, if people don't understand the convention, they won't obey, and
they'll change code to disobey it.  But this seems like a point against
*any* sort of coding convention.

See, for example, the way ACT works, for a situation in which people
actually pay attention to project-wide conventions.

>Me, neither.  One exception: I have run across a compiler that failed to
>initialize access fields in certain nested record situations.  But then
>I would take half a line AT EACH SUCH LOCATION to explain why there's
>an explicit null.

Indeed.  Every time I write code in a certain way just because some
compiler has a bug, you'll see a comment in the code to that effect.

- Bob