Hi I’m Ryxku and while working on a game with one of my friends, we noticed some strange behaviour with one of our scripts.
For some reason, when searching the catalog using the AvatarEditorService and using the Recommended Category Filter, it can’t progress to the next page and it thinks that the first page is also the last.
If we change the Category Filter to any other one, it works perfectly fine and prints a page every second. Here is the simplified version of the script:
local AvatarEditorService = game:GetService("AvatarEditorService")
local Params = CatalogSearchParams.new()
Params.CategoryFilter = Enum.CatalogCategoryFilter.Recommended
local Catalog = AvatarEditorService:SearchCatalog(Params)
while task.wait(1) do
print(Catalog:GetCurrentPage()
print(Catalog.IsFinished))
Catalog:AdvanceToNextPageAsync()
end
I tried to look for solutions, some people think it’s a bug related to Roblox but I want to make sure it’s a bug so I can use report it!
Maybe it’s not a bug and I’m wrong, feel free to help, thank you!
You won’t solve issues by using ChatGPT, ChatGPT does a lot of mistakes, believe me. You have nothing to do on the roblox devforum if you bring no innovation to the community.
Anyway, at first glance it seems like your code is fine. My gut tells me this is intentional, but at the same time I feel like there should be multiple pages.
How about you try adding additional filters to the CatalogSearchParams and see if it works then?
First of all, I checked and there isn’t enough documentation under the service to know what the intended result should be (in terms of the number of pages). It could be that the recommended filter only returns one page at most. (though more of my thoughts on this can be found at the bottom of my reply).
Secondly, I don’t defend AI too often, but I wouldn’t be criticizing AI about mistakes too much. Using ChatGPT to answer questions (or really anything) shouldn’t be trusted blindly, but if a developer makes sure the code works as intended, and makes corrections if there are any errors, I think it can be a very helpful tool to everyone, including here on this forum.
Of course, if someone uses ChatGPT and provides completely wrong info, there should be penalties, just like there is normally.
@Darknessbussiness’s answer is what I would have responded with at first, because I thought you didn’t know how to stop advancing to the next page after you reached the end. Their answer also has no errors, so it was a good response.
I can’t say for sure if returning only one page is a bug. It is only returning one page. But since setting the limit to a higher number does increase the amount of info returned, it really may be a bug (I tested changing the limit in studio).
I don’t think there is any harm in just reporting it. If it’s a bug, it will be fixed. If not, then oh well. But it is good that you didn’t just recklessly report something that doesn’t work the way you thought it should.
Roblox at the very least should increase the documentation on the different filter types if their functionality differs at all from each other.
Thank you for trying to researches for the issue, but the ChatGPT’s answer is bad anyway, for 2 reasons :
1st : It doesn’t fix the problem
2nd : Invading the devforum with botted answers will be very bad if it’s used for anything, otherwise the forum will lose all its meaning.
Additionally, we sometimes test with chatbots to see if they can find an error. If I write my problem here, it is to get human help.
Agree to disagree, I guess. Just make sure not to degrade people that might think differently and try to use it in a way productive way that is helpful. Until it really is banned, at least.
I don’t degrade people that might think differently because they don’t think, they use Ai to answer instead. Also I’m on my right to be angry because it’s disrespectful, @darknesssbusiness doesn’t mention in his first message he uses Ai to reply.
I don’t want people to naively believe that we speak like AI, we are not all capable of distinguishing them, let’s not let ourselves be replaced by bot messages.
