John Siracusa scribbled on 12/18/06 5:03 PM:
> On 12/18/06, Peter Karman <[EMAIL PROTECTED]> wrote:
>> I am using the svn trunk version rev1556.
>
> The fix was in Rose::DB::Object, in case that wasn't clear. So make
> sure you're using trunk rev 1556 of Rose::DB::Object. (Rose::DB was
> not c
On 12/18/06, Peter Karman <[EMAIL PROTECTED]> wrote:
> I am using the svn trunk version rev1556.
The fix was in Rose::DB::Object, in case that wasn't clear. So make
sure you're using trunk rev 1556 of Rose::DB::Object. (Rose::DB was
not changed for this fix, but the SVN version of Rose::DB::Obje
John Siracusa scribbled on 12/18/06 3:58 PM:
> On 12/18/06, Peter Karman <[EMAIL PROTECTED]> wrote:
>> I seem to get a memory
>> leak where the underlying Rose::DB object is never destroyed.
>>
>> while(my $o = $i->next)
>> {
>> ...
>> }
>> $i->finish; # must call this no matter wha
On 12/18/06, Peter Karman <[EMAIL PROTECTED]> wrote:
> I seem to get a memory
> leak where the underlying Rose::DB object is never destroyed.
>
> while(my $o = $i->next)
> {
> ...
> }
> $i->finish; # must call this no matter what
Fixed in SVN, I think. Try it and let me know.
-John
I'm using RDBO in a Catalyst app, the latest versions of both.
I have not tested the following under other Cat engines, but with the dev
server, I find that I must call finish() when using get_objects_iterator(),
even
if I successfully iterate over all objects. Otherwise, I seem to get a memory
