For example, https://www.ozbargain.com.au/deals?page=50 returns
404 Not Found
Page that you are trying to request cannot be found.
All pages from 50 onwards are not found.
For example, https://www.ozbargain.com.au/deals?page=50 returns
404 Not Found
Page that you are trying to request cannot be found.
All pages from 50 onwards are not found.
We have to deliberately set a hard limit on the maximum number of old pages, because in the past we have bots iterating through thousands of pages of past deals. What do you think is a sensible limit there?
I don't have anything in mind scotty. It was the first time I have seen such a limit imposed on a public forum. I needed to go past page 50 today to hunt down the reason a particular deal was unpublished on a date that I know (so I cannot search for it), found this limit and thought it was a bug.
In light of your response perhaps return a more informative message than a 404 error code? The limit you've set is not an error but an artificial restriction. Clicking on an URL to an unpublished deal, for example, does not return a 404 but a content removed message.
How about a code 418? :)
Or imposed a request rate limiter for pages beyond 50 to get rid of bots but allow human?
I remember having your issue a while ago but I thought it no longer applied with OzBargain's new infinite scrolling feature? Of course it's not very convenient if you're looking back a few months/years, but an iterative process would be labour intensive regardless.
For the record you can easily find the deals leading up to an exact epoch timestamp using the OzBargain API, but I don't know if scotty wants the URL publicised (since I imagine this compromises the former limitation).
Does the API return unpublished deals and the reasons? I can use search to find normal and expired deals but not unpublished ones.
@alvian: Yes.
@peterpeterpumpkin: That good to know. The only API mentioned in the help pages is for submitting deals: https://www.ozbargain.com.au/wiki/help:api
@alvian: Sorry for the confusion. It's not in the "published" API. Apparently there's no plans for a universal API.
Can you set it so that it has a robot identification challenge to see pages 51-100 and then another challenge for 101-150?
Pages 51++ are just eneloop deals :D
You reminded me: When infinite scroll fails, the button to load more deals says "Load More Result…", which isn't pretty English.
Anyone else notice OzBargain has been down a lot lately (i.e. twice in a fortnight)?
We have been under quite a few DDoS attack recently.
Last around an hour each time.
thats not a bug, thats a feature