[PATCH] ada/30614: Do not try too hard to give a good message when expander is off

[Robert Dewar]

Samuel Tardieu wrote:
>>>>>> "Robert" == Robert Dewar <dewar@adacore.com> writes:
> 
> Robert> Arnaud Charlet wrote:
>>> Here is Robert's answer:
>>>
>>>> This patch removes recursive calls from the diagnostic subprogram
>>>> when the expander is not active. It is better to issue the generic
>>>> "lhs must be a variable" message rather than an incorrect
>>>> diagnostic.
>>> That's true but I am afraid this is giving up too easily. I would
>>> prefer to give the same diagnostic for a case like this with or
>>> without expansion, and I doubt that this is so difficult to do.  So
>>> let's not make this change without further investigation.
>>>
>>> Arno
> 
> Robert> I will have a look into this ASAP, and see if it is possible
> Robert> to do a better job here, I don't like the idea of expansion
> Robert> on/off affecting error messages for sure, but I also don't
> Robert> like to lose interesting information.
> 
> Any news?
> 
>   Sam

Have not had a chance to look at this one yet ...
The issue is simply the quality of the message here, and the proposed
patch is not clearly an improvement, so I think we need to look for
something better here, but in any case this is not a critical problem.