Lance,
I will try jConnect 6 tomorrow. Thank you for that insight into Sybase
internals. :-)
Have Fun
Markus
Lance J. Andersen schrieb:
> Hi Markus,
>
> Markus KARG wrote:
>> Lance J. Andersen schrieb:
>>> Markus,
>>>
>>> Which JDBC driver are you using? jConnect or the SQL Anywhere
>>> driver which uses the jdbc-odbc bridge?
>> Lance,
>>
>> the SQLAnywherePlatform that I implemented is targetting the usage of
>> the iAnywhere driver (Sybase's JDBC-ODBC bridge). While beeing not
>> platform independent, it provides superior performance compared to
>> jConnect, and is the official recommendation of Sybase.
> The bridge has limitations and yes I am aware of ianywhere pushing the
> bridge driver though they still support both. I spent many years at
> Sybase working on the jConnect driver implementation and working with
> the SQL Anywhere team to make sure that the TDS support on SQL
> Anywhere worked well. There are some advantages though to not using
> TDS with SQL Anywhere.
>
> Anyways, please run your tests so that we make sure jConnect 6 also
> works correctly.
>
> Regards
> Lance
>>
>> Regards
>> Markus
>>>
>>>
>>> Markus KARG wrote:
>>>> Tom,
>>>>
>>>> I had been using absolute(x) in the past with SQLAnywhere and it
>>>> worked well always. If you send me a few (at least abstract) code
>>>> lines then I will try them out.
>>>>
>>>> Thanks
>>>> Markus
>>>>
>>>> Tom Ware schrieb:
>>>>> Hi Markus,
>>>>>
>>>>> There has actually been some recent discussion on the mailing
>>>>> list about setFirstResult() and setMaxResult(). Here's how they
>>>>> work:
>>>>>
>>>>> The current implementation of setFirstResult() and setMaxResult()
>>>>> make
>>>>> use of JDBC constructs rather than RDBMS constructs.
>>>>>
>>>>> For setMaxResult() we call the java.sql.Statement.setMaxRows(int)
>>>>> method.
>>>>>
>>>>> For setFirstResult() we make use of the ResultSet that is returned in
>>>>> JDBC. If firstResult is set, we build query that returns a
>>>>> TYPE_SCROLL_INSENSITIVE ResultSet and call resultSet.absolute(int) on
>>>>> that result set.
>>>>>
>>>>> Is it possible the result for the SQLAnywhere driver is not what
>>>>> we expect?
>>>>>
>>>>> -Tom
>>>>>
>>>>>
>>>>> Markus KARG wrote:
>>>>>
>>>>>> When I was implementing the SQLAnywhere platform I remarked a
>>>>>> failing test: complexResultPropertiesTest. That test tries three
>>>>>> things: setFirstResult, setMaxResult and a combination of both.
>>>>>> The used table holds 15 row. The first test (setFirstResult(2))
>>>>>> return 13 rows (what is correct, since 15-2=13). The second test
>>>>>> (setMaxResults(15-1)) returns 14 rows (what is correct, since
>>>>>> 15-1=14). But the third test fails. The correct result would be
>>>>>> to return 12 rows (= 15-2-1). But the actual result is 13 rows!
>>>>>>
>>>>>> I have not seen any faulty SQL (in fact, it seems that TopLink
>>>>>> does this internally -- it does not apply any "SELECT TOP START
>>>>>> AT" range constraints to the SQL, but seems to process the range
>>>>>> constraints internally.
>>>>>>
>>>>>> So did I detect a bug in TopLink, or where is there a fault in my
>>>>>> platform implementation?
>>>>>>
>>>>>> Thanks
>>>>>> Markus
>>>>>>
>>>>>>
>>>>>
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail: dev-unsubscribe_at_glassfish.dev.java.net
>>>>> For additional commands, e-mail: dev-help_at_glassfish.dev.java.net
>>>>>
>>>>
>>>
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: dev-unsubscribe_at_glassfish.dev.java.net
>>> For additional commands, e-mail: dev-help_at_glassfish.dev.java.net
>>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: dev-help_at_glassfish.dev.java.net
>