Hi Andy,

I thought about it and think that the changed query does not add much 
value. I removed the second query from the test and checked in the 
changed (see r330775).

However, I think the class misses to test the order of implicit 
parameters used in different query elements. E.g. a query might use an 
implicit parameter in the filter and another implicit parameter in the 
having clause. I discussed this with Michael W. and he will add some 
more test queries.

Regards Michael

> Hi Andy,
> 
> good catch! The second filter should be "lastname == :param1 && 
> firstname == :param2". Then it returns an empty result as expected. I'll 
> change this soon.
> 
> Regards Michael
> 
>> The TCK test OrderOfParameters has an error currently.
>>
>> The second query has "firstname == :param2 & lastname == :param1" (the 
>> opposite
>> way around to the params in the first query), and thye parameter 
>> values for both
>> queries are the same. As a result the result will be the same (the 
>> parameters are
>> assigned to the values in the order they are encountered in the test). 
>> You should
>> swap the values around if you want to get the expected result AFAIK.
>>
>> -- 
>> Andy
>>  
>>
> 
> 


-- 
Michael Bouschen		Tech@Spree Engineering GmbH
mailto:mbo.tech@spree.de	http://www.tech.spree.de/
Tel.:++49/30/235 520-33		Buelowstr. 66			
Fax.:++49/30/2175 2012		D-10783 Berlin