>
> random_ads = self._get_random_urls(ad_ids_to_skip)
> for new_ad in random_ads:
> ad_details = {
> 'index':start_index,
> 'id':new_ad['id'],
> 'url':new_ad['url'],
> 'valid':'',
> 'suburlcount':'',
> 'viewcanceled':'',
> 'timestart':'',
> 'timeend':'',
> 'rating':''
> }
> start_index += 1
> #print "New index", start_index
> unseen.append(ad_details)
> ads_viewed.append(ad_details)
> today_ads.update_record(adsviewed=cPickle.dumps(ads_viewed))
>
> I have it working in both frameworks, however I needed to include gluon in
> python path on the old framework. So I am just curious as to where is this
> code is causing it to use a storage object.
>
I don't know. Again, it might help to inspect the stored values in the
adsviewed field. What does self._get_random_urls() look like?
Anthony
