Hello Antti,

you can fix the bug in the english resources and send the fixed files
to me and to the binary fastreport group. I didn't look at them for a
long time (only on the german resources).

When I work wiht FastReport I never use the automatic finding feature
of Datasets of FastReport.
I always create the Report Datasets as needed and work the alias
feature in FR.

Andreas

> Thanks Andreas - I think we figured it out. It seems that FastReport looks
> for the datasets based on just the dataset name, which I never thought of
> before because when you set the data sources in the FastReport editor it
> shows the form name as well.
>
> So what happened (pretty sure - still need to verify this) was that we had
> some forms auto-created (don't know why yet, usually we create the forms on
> the fly), and one of the forms had FastReport datasets *with the same name*
> as the problematic form! So, FastReport tried to get data from those, and
> naturally the queries were empty (they're set at runtime) -> hence the error
> message.
>
> Now, what I haven't yet figured out is why it behaves differently when the
> report is stored in the DFM. That was what threw me off. Anyway, I'm more
> than happy to say it was our own fault :-)
>
> P.S. I'm looking into your FastReport IBO components, and had some trouble
> compiling the English resources. Should I email you directly about that, or
> do you prefer some other place?
>
>
> Antti Kurenniemi
>
> ----- Original Message -----
> From: "Andreas Hesse" <ah@...>
> Cc: <ibobjects@yahoogroups.com>
> Sent: Friday, February 06, 2009 11:21 AM
> Subject: Re: [IBO] Cannot prepare a blank statement -error
>
>
>> So it is time for debugging.
>>
>> Halt on this exception and see, what object is causing it and what is
>> the stack trace. Compare the object that raise the exception with the
>> one that you expect.
>>
>> Post the information here, so we can help you much better with some more
>> informations.
>>
>> Andreas
>>
>>
>>> Hi all.
>>>
>>>
>>> I have a case which I can't figure out, and it's really hurting me right
>>> now. I have an app with about 50 reports, using FastReports. Just a few
>>> days
>>> ago I noticed that two of my reports are throwing "Cannot prepare a blank
>>> statement" error, which seems to come from IBO. As far as I know, I
>>> haven't
>>> changed anything in those reports in ages, and I checked an older
>>> compiled
>>> version of the app and it worked just fine.
>>>
>>> Here's where it gets weird: I set two queries (of TIBOQuery type) at
>>> runtime. There's a fairly simple SQL in one of them, and a bit more
>>> complex
>>> but nothing strange in the second one, just a few joins. If I do
>>> something
>>> like this after opening the queries:
>>>
>>> while not HeaderQuery.Eof do
>>> begin
>>> ShowMessage(HeaderQuery.Fields[0].AsString);
>>> HeaderQuery.Next;
>>> end;
>>>
>>> ....it works ok, I can go through all records without any trouble. But,
>>> as
>>> soon as the report opens, it throws the error mentioned above.
>>>
>>> Now, the report may be a culprit here, but I'm dumbfounded as to how to
>>> start searching for the cause here. The IBO source is way too complex for
>>> me
>>> to wrap my mind around, so I'm posting this here (as well as in the
>>> FastReport newsgroups) in hopes that someone could at least point me
>>> towards
>>> the solution. Like, in which exact case would that error get thrown
>>> usually?
>>>
>>> Any help greatly appreciated,
>>> Antti Kurenniemi
>>>