Fabio,

There is no point to continue this discussion. You may continue to dislike the fact that fields take precedence over aliases in SELECT list.

Thanks,
Aleksey.

>>Fabio,
>>
>>There is a benefit - the fact of adding/removing GROUP BY or DISTINCT clause does not modify meaning of the ORDER BY clause.
>
>This is incorrect, with VFP ORDER BY search fieldname rules
>adding/removing GROUP BY or DISTINCT clause modify meaning of the ORDER BY clause.
>In some cases the result is invariant (is the cases to which you is thinking),
>but the meant one changes.
>
>

>* no ambiguous ORDER BY
>* mean: order S.fieldname, and after fill the fieldName
>SELECT S.fieldname AS fieldName FROM sourceTable S ORDER BY  S.fieldname
>* mean: fill the resultfieldname and after order resultfieldname
>SELECT S.fieldname AS resultfieldname FROM sourceTable S ORDER BY  resultfieldname
>
>* then, for having one coherente grammatical, the ambiguous have to be solved with:
>* mean: fill the fieldName and after order fieldname
>SELECT S.fieldname AS fieldName FROM sourceTable S ORDER BY fieldName
>
>* mean: fill the fieldName and after order fieldname
>SELECT S.fieldname AS fieldName FROM sourceTable S ORDER BY  1
>

>
>Add DISTINCT don't change the mean.
>
>Of course you can apply the invariance rules, and to change that what comes effectively ordered,
>but the meant one is that one written into the syntax.
>
>If you are sure, post a example with benefit, where my definition is lacked.
>
>>
>>If I remember correctly, you were so anxious to have similar benefit with the syntax for correlated UPDATE (thread #989905). I wonder what is the reason for your inconsistent position? Is it an opportunity to call something a bug?
>>
>>Thanks,
>>Aleksey.
>>
>
>My UPDATE syntax definition and ORDER BY are coherent, and aliasName or tableName correlation it is defined without ambiguity.
>
>Fabio