Re: [PATCH 23/25] syntax-check: enable sc_program_name.

[Gary V. Vaughan]

Hi Stefano,

On 16 Nov 2011, at 15:26, Stefano Lattarini wrote:
> On Wednesday 16 November 2011, Gary V wrote:
>> Thanks for the review.
>> 
>>>> 
>>>> # GPL_version: checks for GPLv3, which we don't use
>>>> -# program_name: libtool has no programs!
>>>> 
>>> But then, since libtool doesn't offer any "real" program, what is
>>> the point of enabling the `sc_program_name' check?
>> 
>> Quite arguably, there is no point at the moment.
>> 
> So at least I haven't misunderstood the scope/meaning of your change ...
> Good for me, I was suspecting to be missing something obvious :-)
> 
>> But at some point
>> we may grow a "real" program (ltmain.c has been on the horizon for
>> many years), and letting syntax-check run everything it has if at all
>> possible means we won't have to remember to go back and enable those
>> additional NOP tests if that day ever arrives.
>> 
>> If you object strongly, I can be persuaded to change my mind however.
>> 
> I don't object *strongly*; it just seems confusing to me to add checks
> for a feature/setup that is not yet implemented, nor is planned to be
> implemented in the near future.  I leave the decision of whether this
> is worth worrying about to you.

I'm not actually adding checks.  I added code to cfg.mk to disable all
the syntax-checks that provoked failures, without really looking into
what they were when I first introduced maint.mk into the tree... now
that I've gone through the failures, some of them having already gone
by virtue of changes to the tree since then, I'm removing the ones that
don't fail or can be fixed easily from the list of disabled checks.

Since these tests are not doing any harm, and might catch something
useful one day, there seems no reason to go to the effort of explicitly
turning them off.

Cheers,
-- 
Gary V. Vaughan (gary AT gnu DOT org)