API Version 2 changes in regards to SRS stages

WaniKani Announcements
1

Note: This topic originally presented an upcoming change to SRS timings. After reviewing your feedback we have decided to not go forward with the changes. You can find the original content of this post by reviewing the edits of this post

This information is mainly for the developers who use the API.

If you have worked with the API you may be familiar with the /srs_stages endpoint.

Due to some upcoming planned features we are going to change how this information is being delivered.

Right now the information is delivered as a report object type. This report delivers information on a singular spaced repetition system, which is used for all the subjects.

We are going to drop the /srs_stages endpoint and replace it with a /spaced_repetition_systems. This endpoint will mimic our other endpoints where one can retrieve a collection of spaced_repetition_system objects or a singular object resource.

The reason for this change is we are preparing support for multiple spaced repetition systems (systems with different timings, number of stages, and/or passing stage).

In addition to the new endpoint we will be adding a new field (most likely named spaced_repetition_system_id) to subject objects. This association is used to determine the timing calculations for the subject and if the subject has met conditions for passing or burning. In other words, each subject will belong to one spaced repetition system.

I know we have announced API version 2 moving out of beta. But given the drastic change and how newly it is out of beta we are going to do a breaking change without versioning. There is no good way to support multiple spaced repetition systems in the current “report” set-up. This will save us the headache of not having to address this down the road.

To make this transition easy we will have a period where both /srs_stages and /spaced_repetition_systems exist. There will be a hard date when /srs_stages will be dropped. Consider this an early warning.

Updated documentation on the new endpoint will be made publically at a future date.

We will keep you informed when the documentation has been updated and dates have been set. Consider joining the API mailing list to get the latest news and updates.

42 Likes
2

æ°—ă«ăȘă‚ŠăŸă™

26 Likes
4

Should I be embarrassed that I don’t know what “accelerated” intervals are?

20 Likes
5

I hope not! I don’t either haha

4 Likes
6

They apply only for levels 1 and 2

10 Likes
7

I’m kind of sad, as this will cost me 4 hours a level and thus make the 350 day record unbeatable. I was on pace for 344 days or so. I’m hoping this update isn’t launched soon.

Edit: I might just barely be able to squeeze it out if I lose 5 days for the remaining 30 normal levels and 36 hours from remaining 18 “fast” levels. But still, I feel like I have to define my final time in relation to what is optimal.

Also, I could get locked into a really sucky time like 1am, which would suck.

Edit 2: Given the response, some people may not realize, but the 42 levels that require at least one radical to have enough kanji will take an additional 4 hours because you have to go through stage 3/4 of the radicals and stage 3/4 of the kanji. So this will overall slow the maximum by 4x42 + 2x18 hours which is about 8 and a half days.

10 Likes
8

See here

They mention a different timing for level 1 and 2.

3 Likes
9

I think the 1hr subtraction is actually a feature and am pretty sad if it would be removed.

It allows you to gradually return to an original target review time without a significant impact on rentention (I don’t believe 1hr less makes a major difference).

This is beneficial if you are late for a review on a one-off occassion, but want to keep the times aligned.

90 Likes
10

Technically only 2 hours *for the fast levels, because the first two stages are unchanged.

Edited to acknowledge that the non-fast levels will be affected twice.

Also, Wani :clap: Kani :clap: is :clap: not :clap: a :clap: race.

32 Likes
11

Yeah, I really enjoyed having two hours (well, 1:59:59) of padding to keep my reviews aligned.

48 Likes
12

Yes, I have also found this aspect helpful. If you do your reviews a couple hours late you can get back to the proper timing after 2 review sessions. I understand the adjustment though.

24 Likes
13

I think this will have little to no affect on average users, but speedsters certainly benefit from the extra padding of that extra hour. Sad to see it go.

8 Likes
14

Also not a big fan of the change. The 1h padding is really helpful in preventing the late night squeeze in the long term. Especially if you’re going back and forth between being able to stick to a schedule and more hectic times.

54 Likes
15

This isn’t true, since it effects both the radicals stage 3 and 4 and the kanji stage 3 and 4 that rely on the radicals. Hence, there are 4 hours extra from every level except the 18 that don’t need the radical to complete. Those are an extra 2 hours.

2 Likes
16

I’m just gonna jump in here and address the seemingly large percentage of new users who seem hell-bent on turning WaniKani into a 100m dash.

You’re ultimately here to learn how to read kanji. Something that Japanese children take upwards of NINE YEARS to do at the same level WaniKani allows you to. Even if you took a full 4 years to finish, you’d still be over twice as fast as actual, native-speaking, Japanese people. I know it’s gameified, and that’s easy to get snared up in. It happened to me too at various points. However, even if you average 2 weeks a level (including the “fast levels”), you’d still be done in just over 2 years. 1 hour per review stage is really not a big deal, and in the long run will have literally zero effect on how well you remember kanji, and how good you actually get at Japanese.

Relax. Go at whatever pace you feel is right (whether that’s light-speed or snail’s pace) and just don’t get so caught up on your actual speed. Real life does happen, and as long as you can honestly say “I learned something!” at the end of WaniKani, it won’t matter how long that took you. /endrant

76 Likes
17

Well, I’ve been going full max speed (losing only 4 hours in the past 12 levels) and so far its been working for me just fine (not feeling like I’m burning out at all). At this point part of my motivation is to get the fastest time though, so this is bumming me out a bit. I obviously know it isn’t required (I recently tried to remind a new person of this that was stressing themselves out and weren’t having fun), but I legitimately enjoy the kanji learning process. Grammar on the other hand


3 Likes
18

I understand that. But maybe remembering why you started studying in the first place and looking back at how much progress you’ve already made and will continue to make will help with the fact that you might finish learning 2000+ kanji one week later than you originally thought.

Besides, when the new fastest time changes, if you do go at absolute full possible max speed the entire way, regardless of where you land on the “leaderboard” it would be impossible for anyone in the future to pass you on the board since you will have had the old SRS timings for the first X number of levels.

6 Likes
19

The missing hour in review intervals is one of WaniKani’s most subtly brilliant features.

It helps people keep a reasonably consistent review schedule, even when life events (or Daylight Savings) conspire to get in the way.

Removing it is a huge change, and it seems very strange to do so for no particular reason other than an apparent dislike of non-round numbers.

72 Likes
20

Could we rename Race to the Cake 🍰 to “Not a Race to the Cake”?

8 Likes