Personal List Updates

[rectifyer]

Lists are getting a ton of upgrades on Trakt! Here's a rundown of what's change in the API.

VIP Enhanced

We're introducting the concept of 🔥 VIP Enhanced API methods. A 420 HTTP status code is returned when the user has exceeded their account limit. Signing up for Trakt VIP will increase these limits. If the user isn't a VIP, please open a browser to X-Upgrade-URL so the user can sign up for Trakt VIP. If they are already VIP and still exceeded the limit, please display a message indicating this.

Note: this only applies to personal lists, not the officially curated endpoints like popular, watched, recommended, etc.

---

List Notes

VIP members can send notes (255 maximum characters) for all list items. This works on the watchlist, recommendations, and personal lists.

Coming Soon

• Collaboration -- VIP members can invite other members to collaborate on their lists. The website has this, and the API will gain support for this soon.
• Improved List Pagination -- The website has switched to pagianted list pages and sorting. The API already supports pagination, but we are looking into porting more of this functionality to the API.

---

Get User Limits

The /users/settings method returns all limits for a user's account. These limits are enforced on the various list related methods and a 420 error is returned if they are exceeded.

"limits": {
    "list": {
        "count": 5,
        "item_count": 500
    },
    "watchlist": {
        "item_count": 1000
    },
    "recommendations": {
        "item_count": 50
    }
}

---

Enforced Limits

We'll be enforcing some limits around lists. There are several apps that create lists with way too many items (think 100k+) which negatively affects Trakt API and website performance for all members. We picked limits that shouldn't affect your normal usage, but will prevent these huge lists from being created.

These limits will be enforced starting Tuesday, June 14, 2022 May 31, 2022

VIP Members

• Unlimited lists
• 10,000 2,000 items on your watchlist
• 10,000 2,000 items on a personal list
• notes on each list item
• Collaboration

New Members

• 5 personal lists
• 1,000 250 items on your watchlist
• 500 250 items on a personal list

Existing Members

• 10 personal lists
• 2,000 500 items on your watchlist
• 1,000 500 items on a personal list

If you have more than 10 lists already, you won't lose any. It just means no new lists can be created. Same with list items, we're only planning to prune massive lists.

[antigravity83]

This is TERRIBLE.

2000 limit for paying customers is awfully low.

I’ll be cancelling my sub because of this and suspect many others will too.

[antigravity83]

My understanding from this post was the limit was on the api call, not the lists created.

Happy to be wrong though!

[rudf0rd]

Yeah, this is only about personal lists.

[antigravity83]

Ok thanks- is that confirmation from the Trakt team?

Might be best to clarify that in the announcement- it reads as if the limits apply to any API call- even from the Trending/Popular/Watched lists.

[rudf0rd]

Yeah. Justin, Tyson and I are the Trakt team.

And yeah, I do see the miscommunication as those 3 things are listed under lists in the api docs.

[antigravity83]

Awesome thanks.

[linaspurinis]

Limits are missing at least a single zero to be useful, make it at least:

VIP Members
20,000 items per list

New Members
2,500 items per list

Existing Members
5,000 items per list

[rectifyer]

I just realized you're the dev at mdblist.com 🤦 This is one of the sites allowing users to create massive lists and one of the main reasons we are implementing these limits. I appreciate your project for helping people be creative. Mblist itself has list and item limits, so I think you understand where we are coming from.

[YozoraXCII]

If there are only a few apps that are problematic, would it not be better to work with the developers of the apps in question than enforce rules on everyone?

[rectifyer]

Over the years, we've learned the limits need to be enforced or the problems will persist. We have a lot of "best practices" but ultimately we need to ensure people don't ignore those or make coding mistakes.

[TheUltimateC0der]

I have read the word "Performance" here quite a lot. Do you mind to give us some insight, what the problem really is?

[thezoggy]

what stops people from using some sort of list pagination, to just break out their 100k list into 50x 2k lists?
but also if a giant list breaks trakt, why isnt trakt just paginating the list to begin with? so said list size is a bit irrelevant?

[Logs22]

Does the 10 lists include liked lists, or just user created lists?

[rectifyer]

Liked lists aren't affected, no limit to how many you can like.

[YozoraXCII]

What happens to existing lists which are already over the 2000 limit, are you going to arbitrarily remove items from them or just forbid additional items from being added?

I for one use Trakt to track items on various streaming services, many of which are already over 2000, with others fast approaching the limit:
https://trakt.tv/users/plexmetamanager/lists/amazon-movies?sort=rank,asc
https://trakt.tv/users/plexmetamanager/lists/hbomax-movies?sort=rank,asc

As others have said, 2000 items is an extremely small amount, particularly for paying customers. I understand the need to shrink the amount of large lists - the main cause of this (in my opinion) is the complete lack of ability to meaningfully search/limit/sort/filter existing lists to find what you are looking for.

For Collaboration - will VIP members be able to invite non-VIP members to collaborate, or is it only VIP members who can collaborate with each other?

[rectifyer]

Justwatch.com tracks all the streaming providers and has an API. Might be better to migrate plexmetamanager to their API than have trakt lists.

JustWatch has an API, but I don't believe it's public unless you're a partner like Trakt. I believe TMDB has basic streaming availability info in their API.

[YozoraXCII]

We are running more performance tests with some of these lists you are attaching, thanks for those.

What limit do you have in mind if 2,000 is too limiting? Our thought was a single large list can become multiple smaller lists, such as by decade. We obviously want to find a balance between performance and usability.

I'll preface this by saying that I am not an app developer, or an expert on API.

I would be happy with 10,000 or even 5,000. I would much prefer a limited number of lists (for example 500) with a larger size (5,000+) than unlimited lists with 2,000 items. I also feel that splitting into multiple lists is going to be harder to manage for the end user. And from my very limited understanding of the API, there would be more API calls if there were 5 lists than 1 list, so what happens when API usage goes through the roof as a result of this change - do you start clamping down on API calls?

As a "last resort", even a second VIP level that offered increased limits would be better than having no way to go beyond 2,000 - I think the current approach will drive away some paying customers.

I also think that giving app developers 1 week's notice to such a drastic change is pretty poor and doesn't give developers much time to implement, test and deploy solutions.

Also just wanted to thank you for the open dialog around this, it's greatly appreciated.

Can you expand on the search/limit/sort/filter that are missing for you on lists? Or do you mean locating the list in the first place?

I mean locating the lists in the first place. If I go to The Mandalorian and select personal lists featuring it, I have the option to sort it by Popular, Most Likes, Most Comments, Most Items etc. But if I'm looking for a list called "Disney" then I have no such option to sort it, and I can't filter it for "contains The Mandalorian"

[rectifyer]

Thanks for the feedback.

I would much prefer a limited number of lists (for example 500) with a larger size (5,000+) than unlimited lists with 2,000 items.

The list size itself is the problem more than the number of lists. True it could be more API calls, but they are less intense. We have API rate limits already in place.

As a "last resort", even a second VIP level that offered increased limits would be better than having no way to go beyond 2,000 - I think the current approach will drive away some paying customers.

Interesting idea, we have VIP EP as a higher level VIP.

I also think that giving app developers 1 week's notice to such a drastic change is pretty poor and doesn't give developers much time to implement, test and deploy solutions.

Most Trakt users will continue using lists the same as they have without any difference. The app and list generation sites allowing these massive lists are what we're trying fix as soon as we can. I recently found out some of these sites enforce their own list and item limits depending on paid membership level, so the concept doesn't seem completely unknown.

I mean locating the lists in the first place. If I go to The Mandalorian and select personal lists featuring it, I have the option to sort it by Popular, Most Likes, Most Comments, Most Items etc. But if I'm looking for a list called "Disney" then I have no such option to sort it, and I can't filter it for "contains The Mandalorian"

Would a keyword search on the Mandolorian lists page help with that? It could do a text search on the list title for example.

[YozoraXCII]

Most Trakt users will continue using lists the same as they have without any difference. The app and list generation sites allowing these massive lists are what we're trying fix as soon as we can. I recently found out some of these sites enforce their own list and item limits depending on paid membership level, so the concept doesn't seem completely unknown.

I completely understand this, and agree that other services introduce tiered list limits - the problem is that Trakt seems to only offer one tier of "paid" limits, and as has been showcased there are valid use cases to go beyond that limit, so I would be in favour of having levels of VIP (assuming the pricing structure is fair) to go beyond the 2,000 limit. I think the vast majority of Trakt users won't really have a need to go beyond the 2000 limit, but "power users" would benefit from additional limits, so I support the idea of paying to go beyond it if it is only going to benefit a minority of the community.

Would a keyword search on the Mandolorian lists page help with that? It could do a text search on the list title for example.

The Mandalorian was just an example tbh - currently the easiest way to find lists is to find a movie/show that you think might be in it, and then sort/organize the lists in that way. Currently there is no easy way to directly search for a "LGBT Movies" list and meaningfully sort the results by most viewed/most items - instead I have to guess a movie that will be in such lists and then sort/organize the lists via that method. I hope this makes some sense.

[rectifyer]

In regards to the limits, we're compiling all feedback on that and should have an update tomorrow.

That's a great point on the general list search. I should be able to add some advanced filters to that search page such as item count and vote count. The search pages don't have a concept of sorting since it's all based around the query, so I'm not sure how/if I could do that yet.

[Terebi42]

Heres a super common use case I think :

Watchlist. I go through anticipated/trending and add things to my watch list. I probably don't need them there x months after it releases. But there is not a good way for me to prune that to jus things that are in the future, or which have been on my list for more than X months.

I'm not even super prolific with this compared to some, and am already hitting limits in the <300s. 500 doesn't give me a ton of breathing room. And I could easily see people who are more addicted than I having multiples of my list size.

[mcdallas]

A button to remove watched or collected items

[Terebi42]

Good point on the watchlist, we are considering having a higher limit for watchlists specifically.

But there is not a good way for me to prune that to jus things that are in the future, or which have been on my list for more than X months.

What tools would you need on a list to manage this?

@rectifyer My main use case is honestly syncing these over to sonarr/radarr. Id be ok with forking TV and movies into separate watchlists, or even watchlists per year or whatever. But your UI would have to help us do that. It really funnels us into "add to watchlist".

At the API level, if multiple API hits are ok, perhaps meta-lists that contain other lists? It would be a breaking change to support, but tools could read the meta list, then send requests for the sub-members. Lists could automatically shard on your side, based on time or size.

management :

For movies : bulk remove all watched. Bulk remove all released on DVD/streaming for longer than X

For TV : Bulk remove ended watched series
possibly bulk remove series with many watched episodes

Bulk remove stuff that has been on the watch list for longer than X days/weeks/months/years, especially already released. (If I haven't watched something after 2 years thats out, im probably not going to)

[rectifyer]

At the API level, if multiple API hits are ok, perhaps meta-lists that contain other lists? It would be a breaking change to support, but tools could read the meta list, then send requests for the sub-members. Lists could automatically shard on your side, based on time or size.

I looked into a "folder" concept for grouping lists, but I haven't figured out a good way to implement that yet. I still want to do it if I can work out the details so it's easy to use.

Bulk remove stuff

Will be looking into this. My initial idea is add a "delete all" option when in manage mode. That would delete all the filtered items (or all items if no filters applied).

[Terebi42]

The updated limits seem much more reasonable, and unlikely to be an issue for most users.

[JourneyOver]

I mean they are still pretty limiting :s but at this point I don't even care anymore, I've already moved most of my lists off the site onto much better sites that don't limit people to how many items they can have on a list or how many personal lists they can have.

[sereisoglu]

I support this limitation you brought to the lists because after a certain item these lists become meaningless. But i'm sure a lot of users won't like it.

[rudf0rd]

We've learned that api endpoints without limits (rate, count, pagination) are just ticking time bombs. :(

That said, we'll re-evaluate based on bother user and dev feedback, as well as sane numbers based on the overwhelming majority of what users actually use. We thought going into this that the limits were set high enough based on querying list item counts and grabbing the top 1% and then giving additional head room. We've gotten limited feedback so far, so we'll see how it goes.

Thanks for not going all doomsday.

[sereisoglu]

I've seen lists of tens of thousands of items being created using Trakt's API. I've been seeing on your site that this is causing performance issues. I thought this was wrong from the start, I think it would be better if you had set a limitation from the start. Trakt has a lot of pointless and duplicate listings. Letterboxd does this very well and they have some very good lists. If it's okay, I'd like you to share some statistics about the lists. For example, how many lists are in Trakt, how many are actively used, and what is the density of the number of items in these lists? Also, I think "watchlist" should not act like a list.

[rudf0rd]

Sure!

• Total lists: 20_264_613
• Average item counts:
  • All lists: 6.2 (unused watchlist and recommendations)
  • Excluding empty lists: 36.4
    • 90th percentile: 61
    • 95th percentile: 117

It hockey sticks pretty hard. The highest list item count is 570_812!!! Actively used is a bit harder to grab since it's not used and we only set an access flag on updates, not reads.

[sereisoglu]

Thanks for sharing these statistics, I hope the lists in Trakt will be better.

Not related to this topic, but Trakt's website says 62.1k movies, 12.6k shows, and 99.6k episodes on the homepage. Is this data up to date? If it's not up to date, can you tell me the current statistics?

[rudf0rd]

Yeah that hasn't been updated in a bit. It should be 748,131 movies, 116,250 shows, 5,294,664 episodes.

[Tiger3030]

I think the limits for personal lists are fine and won’t affect most users.
But because it seems to be too low for a few power users, I would consider raising the limit for VIP users to 5,000 or 10,000 items per list.

With the watchlist I agree with @Terebi42 and see the danger that even average users will quickly reach the limit of 250 items and then perhaps switch to another service. Therefore, I would raise the limit for new users here at least to 500.

The only real problem I see is the recommendation limit. In the example above it says 50, is that right?
I personally use the recommendation feature similar to the like feature on Letterboxd for movies/shows I really enjoyed regardless of the rating I gave them.
I just did a random check and came up with that a user usually recommends/likes one tenth to one fifth of the movies/series watched. This means the average user will already reach the limit of 50 recommendations after watching only 250 movies/shows.
Are you sure that such a low limit is needed for recommendations?

Overall I think you should not rush this decision and think carefully about the individual limits. As far as I know Letterboxd had a limit of about 3,000 items per list until 2018, but now allows unlimited items on unlimited lists for all users. Of course they don't have the API problem (yet) and don't allow shows/seasons/episodes or people on lists. But some people might move over, if they only care about movie lists.

My proposal would be to start with the following limits:

Personal lists
250 on 5 lists (new members)
500 on 10 lists (existing members)
5,000 / 10,000 on unlimited lists (VIP members)

Watchlist and recommendations
500 (new members)
1,000 (existing members)
2,000 (VIP members)

[TheUltimateC0der]

In my opinion (not knowing the exact software architecture behind Trakt) and from a technical point of view, the not further specified Performance problems seem to come from bad software architecture decisions.

There is (or should) be no difference between having 1, 10, 1000, or even 1.000.000 items in a list, of you implement it properly by using paginated responses with properly sized pages.

The first Problem is, that the end user is not able to open up a big list in the browser, because the lists are not properly paginated. Instead everything is loaded at once, causing the browser to choke by the sheer amount of HTML, images, and JS that integrates with it. So the argument I have seen on Twitter (https://mobile.twitter.com/trakt/status/1529215326738522114) "These lists never loaded anyway" is also caused by Trakt themselves, by not using pagination.

To further reduce API requests in general, it would be clever to have data exports like TMDb or IMDb. Since developers heavily relying on your service can build a cache server. Which will further reduce requests.

To me it looks more like they are trying to get rid of users fully utilizing their service, instead of working together with the devs of apps that cause these "performance problems". Since an API key is bound to a user account, they can simply identify these apps and send out emails to these people.

I haven't received a single email regarding this topic. Instead a nice user from reddit pointed this out to me.

[rudf0rd]

@TheUltimateC0der I agree, we did make some bad architectural decisions based on how we assumed end users would utilize lists. The above changes are the correction to our bad decisions. The web UI for lists has server side pagination now and we'll be enforcing pagination on the API endpoint at a later date (a large enough time for all devs to update any apps).

Regarding your question about performance in the other comment thread, pagination reduces the issue on the web servers for sure, but the database is now doing the filtering. So now the biggest issue is that in order to bring this all server side and allow for unlimited item counts means more indexing and then multiply that by the 20 million lists we have. This leads to increased database server costs to allow less than 0.000139% of our users to create, what I consider, unreasonable list sizes.

As for tmdb, imdb exports, we encourage you to use the same source as us rather than us being a middle man to your app's metadata. We also encourage all API developers to "watch" this repo so we don't have to email all API developers of changes. I believe this is on both our terms of service and our API docs page.

[rudf0rd]

@Tiger3030 We're still evaluating feedback on list sizes and appreciate your thoughts. We should have more to talk about tomorrow regarding that.

For recommendations, I believe we launched this poorly. We thought this up in the middle of the iOS app development and were going to formalize our thoughts around it later. Later is unfortunately now. What recommendations should be used for is your top of the top movies and tv. Think something like if you only had 50 movies and tv shows to bring with you on a desert island, what would they be? If we do this right, it opens up a really great opportunity to enhance a number of areas on the site.

[rectifyer]

Here's how recommendations are displayed in the iPad app. We plan to bring this to web, iPhone, and android too. But you can see we prominently feature your top 3, with the ability to see the rest too. The idea is to really make this 50 items stand out above things you've rated a 10.

[Tiger3030]

@rudf0rd @rectifyer

Think something like if you only had 50 movies and tv shows to bring with you on a desert island, what would they be?

I really like this idea (even a lot more than how I have been using recommendations so far) and think 50 is a good limit then.

You just have to communicate this better, so everyone knows what recommendations are intended for, otherwise people might get frustrated. The introduction of the limits is probably a good time to do this.

[kevincador]

I'm really excited and looking forward to get my hands on the notes and collaboration APIs ☺️

Some comments:

• The initial post mentions the notes but not the collaboration feature, are you going to open both in the public API (pleeeease 🙏)?
• I've always missed a "personal note" feature on trakt. I was more thinking about adding a privacy setting to comments but the note on a list item might also work. Do you see private/friends comments on media coming to trakt in the future?
• I didn't play with notes a lot but it seems like they do not support mention or markdown. I understand markdown is maybe a bit too much but combined with collaboration, mentioning a collaborator would have been super powerful.
• The limits on list items make sense for me because I don't have personal lists that big and I like to keep them manageable so they should remain under the normal limits even if I'm VIP. That said, I'm "following" some lists that may get close to 2k items because they will just grow over time. Maybe some lists could get a special treatment. I'm thinking about Academy Award for Best Picture - All Winners and nominees since 1927 (sort by year) for instance.
• There are also a lot of lists that add movies and episodes from the MCU or the DCU or the Star Wars universe, I think one problem here is that people add each and every episodes to the list which makes the item count grow rapidly. I'm not saying you should remove the ability to add episodes to list but maybe there's something to do to avoid this like add some display preference to expend or group TV shows.
• About the X-Upgrade-URL, I do not mind redirecting users to push them to get VIP but I think that might get some apps rejected by Apple as it links to an external purchase mechanism.

Cheers, and thank you for always improving trakt 👏

[rectifyer]

The initial post mentions the notes but not the collaboration feature, are you going to open both in the public API (pleeeease 🙏)?

Yes, collaboration will be added to the API. I think this means new endpoints for:

• get all lists the user can collaborate on
• invite other users
• get pending invitations and accept/deny
• update permissions check to allow list add/edit/delete for lists you collaborate on

I've always missed a "personal note" feature on trakt. I was more thinking about adding a privacy setting to comments but the note on a list item might also work. Do you see private/friends comments on media coming to trakt in the future?

Not sure, but I've thought about that too. It would probably make sense to have a private/friends setting and use the same commenting system. I'll have to think about this more, but using comments seems to make more sense then a whole new concept.

I didn't play with notes a lot but it seems like they do not support mention or markdown. I understand markdown is maybe a bit too much but combined with collaboration, mentioning a collaborator would have been super powerful.

Intersting suggestion, I'll need to think about this more too. You could always comment on the list and mention someone too, how would this be more useful?

The limits on list items make sense for me because I don't have personal lists that big and I like to keep them manageable so they should remain under the normal limits even if I'm VIP. That said, I'm "following" some lists that may get close to 2k items because they will just grow over time. Maybe some lists could get a special treatment. I'm thinking about Academy Award for Best Picture - All Winners and nominees since 1927 (sort by year) for instance.

Thanks for the example, my preference would be to have a limit that works for all scenarios like this.

There are also a lot of lists that add movies and episodes from the MCU or the DCU or the Star Wars universe, I think one problem here is that people add each and every episodes to the list which makes the item count grow rapidly. I'm not saying you should remove the ability to add episodes to list but maybe there's something to do to avoid this like add some display preference to expend or group TV shows.

Episodes do for sure bloat lists a lot quicker.

About the X-Upgrade-URL, I do not mind redirecting users to push them to get VIP but I think that might get some apps rejected by Apple as it links to an external purchase mechanism.

We've have some success with limited messaging about this in our own app. If Apple isn't cool with it, maybe a simple "Please contact Trakt support for more info" or something like that.

[dairefagan]

Very disappointed to learn of such unreasonably low limits for both free and VIP tiers and as they are I will have to look for an alternative. To improve performance limits are justified but they are too restrictive and this change is happening very suddenly. I urge you to take the feedback of this thread into consideration, and revise the limits upwards, which should still have the effect of negating abuse and also incentivising VIP subscriptions.

[rectifyer]

Please provide info on your lists and how many items they have.

[dairefagan]

I have multiple lists containing several thousand items which is automatically maintained though my most important list is manually maintained and contains approximately 1100 items.

[if1mad]

It's sad to hear the changes however I am glad you are taking feedback into consideration.

So far there hasn't been a ton of responses here however I'm sure once any new limits are rolled out that people will flock here wondering why.

I agree that lists of 100k is rather crazy, i've seen many many lists but never any that size! But please let's not go to extremes and introduce harsh limits in response.

I do not know how many users have <250 item lists however it does seem like such a tiny tiny limit and is forcing users to cough up for VIP (if they have tiny list requirements) or to go to another service, not a lot of choice!

I think 10,000 item lists would be fine for the vast majority of people - would Trakt be able to cope with these?

Would only serving the huge lists via API and not web be a temporary band-aid until Trakt can implement software to handle it as normal?

For me I just use the list and trending/popular functions of Trakt so paying $30/year to not even be able to do that with the current suggested changes would mean I will have to try and find an alternative.

When deciding on these new limits was it taken into account not just how many users have lists in the thousands of items, but also how many times that data is pulled? I imagine a lot of the big lists are used by hundreds if not thousands of people unlike some random person's personal 200 item list.

Appreciate any reponses

[rudf0rd]

I’m on my phone, so I’ll give a couple quick points and expand more later.

For some quick list usage stats, check out this comment: #350 (reply in thread)

As for trending, popular, etc, these lists aren’t included in the above. This is only for personal lists.

[if1mad]

I’m on my phone, so I’ll give a couple quick points and expand more later.

For some quick list usage stats, check out this comment: #350 (reply in thread)

As for trending, popular, etc, these lists aren’t included in the above. This is only for personal lists.

Hey thanks for the reply but that link doesn't say how many people are using other people's lists.

I'm not sure if Trakt tracks this information? I feel it's a vital part of deciding any limits.

[rudf0rd]

Sorry I must be missing something. Why would other peoples lists matter? For copying them?

Edit: Re-read your original comment. Unfortunately, due to what we can afford at our cdn, we won’t have those stats.

[if1mad]

I understand, it would be a lot of data.

I think a lot of trakt users (especially those that use large lists) will be copying/using other user's lists (why spend hours making something that is already made, e.g all movies from X studio or service) instead of making their own which will have not given you the whole picture of how lists are used.

[rectifyer]

FYI, we've increased the limits based on feedback and list examples. We also pushed back the enforcement date by 2 weeks to June 14, 2022.

[JourneyOver]

very very disappointed in these limits, I am not VIP as I can not pay for it and the "Existing user" limit is very restrictive.

[rectifyer]

Please provide details on your lists and how many items they have.

[JourneyOver]

@rectifyer I'm complaining about both the limit on how many personal lists you can have, on top of the limit to how many items you can actually have in the personal lists.

the 10 personal lists limit honestly seems extremely low, I use lists to keep track of various collections of movies/tv shows to be used with https://github.com/meisnate12/Plex-Meta-Manager (with 99% of my lists having less than 20-50 items in them) and I've already exceeded the limit that you have put on personal lists (currently sitting at 17 lists not counting watchlist and recommendations) so now your just going to force me to create more trakt accounts to be able to create more lists.

ontop of that the 1000 item limit for personal lists just seems really low, I currently have 4 lists of the 17 lists I use as lists for both sonarr and radarr to keep better track of what I add and remove on them, my movie list already exceeds (creeping up slowly to 2000 movies on the list) the number of items you would allow on a personal list, so it would force me to break the list into smaller lists but wait I can't do that because I exceed the list limit so it would force me to once again create more trakt accounts to be able to create more lists.

the limit on the watchlist is kind of whatever, it is pretty low but tbh I have yet to even exceed 1000 items on my watchlist yet (though it is getting kinda close) but things get removed as they are watched so 2000 limit is fine-ish on that.

Honestly I could live with the item limits (even if they are on the very low side) on the personal lists if the limit of personal lists was just completely scrubbed away to continue allowing unlimited lists because then I could just break my lists (when they start getting to big into smaller lists instead), otherwise I'm going to be forced to make multiple accounts to get around said limit.

[JourneyOver]

I guess since you never responded, you don't care about my feedback. It no longer matters though as I already moved most of my lists to a site that is much better and doesn't limit people to how many lists they can have nor how many items they can have on a list.

[rectifyer]

Yesterday was a holiday and we didn't check over the weekend, I'm just catching up on all messages now. Existing members won't lose any lists (unless they are massive with 100k+ items) or the items on them. Sean posted stats in #350 (reply in thread) and the vast majority of Trakt members will be unaffected by these limits. We'd of course love to keep you as a member, but I'm glad you found an alternative for your specific situation.

[dabutch3r]

I read through all the comments did I miss how this 10 lists will be determined if a user is OVER the limit of lists? Most used, newest, oldest? Just want to make sure I am telling my folks the correct information. THX

[rectifyer]

No lists will be removed if you're over the limit. It just means new lists can't be added. Same with items, we're not planning to remove items unless it's a really massive list.

[dabutch3r]

Thank you for clarifying that .. Fair enough

[eliehaykal]

I believe the 5 and 10 limit for personal lists is too restrictive. Many users like to create custom lists that they can share with friends and family so limiting the count will be a huge obstacle.
Limiting the number of items in a list might be a good idea, but 500 and 1000 items can be quite restrictive.
I have many users with 1000+ items so this will surely be a huge disappointment. The limit IMO should be above in the 2000 range for all as it becomes almost unmanageable forcing users to divide and further organize the list.

[JourneyOver]

So much this, the limits are just way to restrictive, it's only going to force people to either make multiple accounts to get around this issue (and since new users are only limited to 5 personal lists that means even more accounts even quicker) or they are going to start looking elsewhere to making lists..

Honestly I think the trakt devs are wanting to drive people away from using their site, but that's just me xD

[rectifyer]

I have many users with 1000+ items

@eliehaykal Can you provide more specifics how these lists are built? Are your app users creating those by hand or is the app automatically generating them?

The limits are just way to restrictive

@JourneyOver For the majority of Trakt users, the limits won't change how they use Trakt. We do have a path to increase the limits with a VIP membership. I understand you mentioned you're not able to upgrade and it sounds like you've found an alternative for your specific situation.

[eliehaykal]

I have many users with 1000+ items

@eliehaykal Can you provide more specifics how these lists are built? Are your app users creating those by hand or is the app automatically generating them?

The limits are just way to restrictive

@JourneyOver For the majority of Trakt users, the limits won't change how they use Trakt. We do have a path to increase the limits with a VIP membership. I understand you mentioned you're not able to upgrade and it sounds like you've found an alternative for your specific situation.

The lists are created either through the app or Trakt website. There are no automated ways to generate the lists however, I do provide users with custom created lists to allow users to select and add items to their watchlist. Let me know if I can assist you with anything else. Thanks

[dabutch3r]

After reading through the comments and seeing how many people use Trakt daily or weekly to keep their tv/movies lists organized. It surprises me the outrage people feel and entitlement they expect for FREE. I have used Trakt for 10+ I guess and until a month back never got VIP. I used it for lists that I used in many of my projects as does many people I know. We got a RELIABLE service and were able to piggyback off its servers, storage, bandwidth. We were able to tie these lists into many projects and pulls its lists and meta without cost to us. For any of us that have bought hosting we know of its costs and how bandwidth is not cheap. I appreciate the service I have enjoyed for years but now I have to pony up if I want that premium package. I am ok with that and have bought VIP. Trakt is a service I value and I see its value to me. So I will pay $30 a year to continue to use it and have my 25+ lists.

Trakt is not booting people off or deleting lists. They have gave people fair warning that moving forward the power users of trakt ( the ones making many or long lists ) need to help burden the load of server costs. We cant expect services we use all the time to be free forever. $30 a year = $2.50 a month = $.09 a day. This seems like a very small cost to a service I value.
Thank you to Trakt staff keep up the good work.

[eliehaykal]

I have been an advocate since Day 1 for having a monthly fee for any developer who wants to use the Trakt Service. Similar to the VIP program for users, it would be interesting to have something similar for Developers also. I shared my concern quite a while ago with Justin when the API was not stable as it was easy for anyone to abuse it.
Devs who are serious enough to take full advantage of the service will be more than willing to pay a subscription fee.
I would keep the service free with very limited API access (set some strict API call limits) for devs who want to simply "play" with the API. This approach will greatly filter out passive devs and those who intend to use the service will be assisting in one way or another in covering a portion of the costs.

[TheUltimateC0der]

I have been an advocate since Day 1 for having a monthly fee for any developer who wants to use the Trakt Service. Similar to the VIP program for users, it would be interesting to have something similar for Developers also.

A paid API? Funny thing. Others tried and failed so far with this. I mean sure the money has to come from somewhere, but monetizing those people that spend their freetime to create something that has more value for Trakt than the dev using the API makes absolutely no sense to me.

Every software that relies on the list feature of trakt potentially sends more and more users to Trakt every day. Some of these users may like other features of trakt that require VIP, and there you go, I brought you a paying customer just by using your API.

I think people would be shocked with our monthly AWS hosting bill 😮

If you show your AWS hosting bill, and the income that Trakt makes side by side, I guess no one would be shocked really.

I am not trying to be mean, sure money has to come from somewhere. But saying I think people would be shocked with our monthly AWS hosting bill 😮 and just how much it costs to run Trakt. is just stupid. Its simple business. You have to spend money to do business.

Maybe consider switching from overpriced AWS 😛

[rectifyer]

We haven't seriously considered charging for the API for those reasons. We appreciate all the developers who choose to integrate Trakt with their apps.

A developer might not be shocked, but I meant a Trakt member in general might be. We run pretty lean on AWS and optimize performance and costs as well as we can.

[xyzfre]

Well said, thank you. I think people would be shocked with our monthly AWS hosting bill 😮 and just how much it costs to run Trakt. VIP memberships directly pay for those server costs. It also allows us to expand the team. For example, we just hired an additional support person who is focusing on content moderation. They've been able to fix over 1,000 reports in just a few weeks. A few years ago, we were able to hire a full time iOS (and now Android) developer to create official Trakt mobile apps. We really appreciate all the VIP support to allow us to grow Trakt like this!

I don't understand because Trakt is a amazing feature to have so much potential to help us users keep track of what we watch what movie or episode comes next I can keep on and on but I definitely agree people should just become VIP members,, plus for the cost it is well worth it and some I mean I can spend more than that on just lunch lol much respect to the Trakt Team✊🏼✊🏼💯💯

[eliehaykal]

Fair enough. I am a big fan of the service for over 7 years now and try to support it however possible. Keep up the great work :)

[rectifyer]

List limits are enforced as of today, June 14. I'll have updates soon regarding new API methods for list collaboration.

[rectifyer]

Pushed some API updates today. More updates coming, but wanted to get these out there to start.

A new method was added to get all lists a user has permissions to collaborate on.
/users/:id/lists/collaboration

Any collaboration lists gives the user full access to add, remove, and re-order list items. It essentially works just like a list owned by the user, just make sure to use the correct list owner user when building the API URLs.

A new method to get all saved filters was added. VIP members can apply advanced filters and save them in the movies, TV, calendar, and search sections. You can now get these saved filters and use them to query the API. This eliminates the need to create a custom list in some cases.
/users/saved_filters

[michaldrabik]

@rectifyer Are "Hidden Items" affected by this in any way?

[michaldrabik]

Ok, thanks. Well they do exist though :) . Got a user that claims to have 6-8 thousand items hidden.

[rectifyer]

Wow, I'd be curious why they have that many hidden items. I can't guarantee things will work ok with that many items, I envisioned hiding a few hundred at most.

[michaldrabik]

Yeahh. Just wanted to make sure it's not connected to these recently added limits.

[michaldrabik]

@rectifyer By the way (jsut gonna quick ask here). Shouldn't GET /sync/watched/**** endpoint have an option to paginate? This is very easily returning thousands of movies/shows for long term users for sure.

[rectifyer]

Yes it should, we're planning on an updated method that is paginated. No ETA on that yet.

[Terebi42]

Use case for bigger lists, caused by a missing feature in trakt :

I use trakt primarily as a discovery mechanism, and as a syncing mechanism via lists for radarr/sonarr.

For discovery, I look through trending and anticipated quite often. There are Filters and Advanced Filters available to filter these lists. But those filters do not remain applied, nor is there a way to save them.

So as I am scrolling through anticipated, I either have to ignore all of the shows I definitely don't care about (genres I don't like, languages I don't speak, etc) or reset those filters every time.

As a workaround, I created an "ignore" list that I add shows to to make them go away, because the basic filter to "hide listed" does persist. But that requires a click per show/movie to make it go away, and i'm competing against the entire world for creating those shows/movies, so its a losing battle.

Then just recently I discovered mdblist, where I can build lists automatically, so I created one for the genres I don't want, and one for the languages I don't want, and filtered it to the current year through the next year. That does get rid of most of the shows/movies I don't care about.

But its also many thousands of shows/movies in a few lists now. And the only purpose of those lists is to hide things from the built in lists.

So, giving persisted advanced filters, or the ability to save and apply advanced filters would pretty much eliminate the need for me to populate these lists. and probably significantly reduce the thrashing hitting your api from mdblist.

[rectifyer]

I replied about 10 days ago with details about a new method to get saved filters.
#350 (comment)

Filters can be saved on the Trakt website (blog post), which places them into the sidebar for easy access. The new API method allows apps to retrieved those saved filters as well. Saved filters were created for the exact reasons you mentioned, you shouldn't have to maintain personal lists when saved filters will do the job.