@dae
Thank you official Anki Dae, I appreciate your appropriate response.
@glutanimate
Thank you glutanimate, I appreciate your sincere response and consideration and I apologize for any inconvenience caused.
2 Likes
Thanks! I thought that displaying something like an official Anki certification badge on the Add-ons name of Anki for Desktop Contributors would encourage add-ons developers to participate in the development of Anki for Desktop.
1 Like
For the guidelines, one of my motivations was to see if there were already clear guidelines (i.e., if they had already been discussed). I understand approximately this, so I have no additional questions or concerns, thank you. Even if rating pop-ups are allowed, I basically want my Add-ons to be just a changelog pop-up (no rating) and just a rating button (Or something that does not affect the rating).
For special cases, such as L.M. Sherlock wanting to use pop-ups and wanting us to discuss guidelines that are not unfair, I can discuss that a bit, but other than that, there are no specific requests from me. (Cause I would rather writing source code than writing discussions)
To explain a little, I currently have about 300 add-ons, including 40 repaired add-ons. When I look up these add-ons, I also simply look up who makes them, why the add-on is broken, why the developer quit developing it, how it was funded, etc. In my intuition, your FSRSAnkiHelper is the add-on likely to be abandoned and broken.
I know that you have probably already received donations from fans and Anki famous people and you quickly added a rating button to your menu. However, in my opinion, that is not enough, I confirm that some developers, even advanced and diligent ones like you or Glutanimate, have given up or lost interest in the development.
Thanks again @dae 
No worries at all. Thanks a bunch also for your candor in dealing with all of this, and all your insightful and thoughtful responses throughout this thread. You raised a lot of interesting points, and I think there’s still a lot to explore in terms of finding more encouragement for add-on devs, and avoiding a situation where obtrusive vote boosting strategies are incentivized. Also in terms of how to communicate the consensus we arrived a in this thread to developers who might have not seen the discussion.
On other add-on rating strategies, I would say that once we move towards “less enticing” voting prompts and smaller effects on overall add-on ranking, things become more nuanced quickly, and both less feasible and less impactful to enforce.
I have some more thoughts on this, but have to be a bit more mindful of my time this week, so will have to come back to this later.
1 Like
For emoji, I think I’ll remove all of them, as I don’t think they are that important when I think about it. Text only is not so bad because it is more concise, and it seems that I am almost the only one who uses emoji.
1 Like
Thank you, and one last thing, I would like to explain the issue of the negative effects of low ratings.
There was a previous thread about adding emojis to this AnkiForums, and the conclusion there was that low rated emojis may not produce positive effects.
In my opinion, I suspect that this same competitive system of low ratings is having a negative effect of decreasing developer motivation for Add-ons and users becoming more arrogant, creating a vicious cycle that is decreasing the number of volunteer developers that are important to Anki for desktop.
I think developers already know, for example, that unreasonably low ratings like this often occur.
- If low ratings are given in error, they never go away again.
- The more years of activity, the more low ratings increase.
- The first rating on an advanced add-on gets an unreasonably low rating
- The problem has already been fixed, but the low rating never goes away
- When an author tries advanced features, the low rating increases due to errors
- Nothing changes when the author sincerely responds to the low rating comments
- Users only rate when there is error or dissatisfaction
- Users give orders or threats instead of asking authors to do so
- Users do not read descriptions by the author
- Author ignores all rating comments
- Author abandons development of Add-ons with many low ratings
- Author discontinues development and does not return
- Author deletes all Add-ons and quits development
Perhaps from a novice developer’s point of view, Anki users are a very unreasonable group of people, most of whom are extremely rude, only complaining without gratitude, and thinking that they deserve to be developed and repaired according to users’ requests.
I have also been threatened by a user in a rating comment to add new features if they want me to remove the low rating.
Even now, when a novice developer posts an add-on for the first time, the user unreasonably gives him a low rating, saying that there is no explanation, and directs him to another add-on, etc. It is quite a terrible situation.
Not only does this discourage developers from developing and volunteering, it is not a useful rating. It seems to encourage users to become arrogant.
5 Likes
I think it would be tricky in practice. Either we make the badge meaningless by allowing things like a simple wording change to count as “contributor”, or we need to get into the messy details of judging what makes a non-trivial change. And it would likely cause more PR noise, as people contribute changes that are not necessarily particularly useful/in line with Anki’s vision, because they want the badge.
2 Likes
I agreed, the badges are just an idea, I think it is either not needed because of the cost of development in practice, or another way that is easier would be better.
Hmmm, I am a little surprised that my guess about the increase in PR is different than mine. My guess was that there was always a shortage of Anki for Desktop volunteer developers and that more new volunteer developers and contributors were needed.
If the PRs are actively recruited, I expect that even if most of the PRs are not adopted, the developers will become more volunteer-minded and eventually it will be easier for advanced volunteer developers to participate in the problem solving and development of Anki for Desktop.
In other words, my guess was that recruiting and training volunteer developers was of high importance, even if there is a downside of increased PR noise.
Is there actually not a shortage of developers, or is there not a need for so many new developers?
Hmmm, I deleted all the Emojis and now I don’t know which Add-ons are which. I will put Emoji back in, sorry. But I will remove Emoji from the Add-ons with the high number of ratings.
Most of my Add-ons with Emojis have a rating of 1 or 2, so I don’t think there is much of an unfair advertising effect. If there is a problem, please tell me again.
2 Likes
The medical group Ankihub (AnKing, Ankipalace) is waiving monthly fees for Anki and AnkiDorid contributors. They have about 48,000 subscribers, so there may already be such PR noise effect.
But medical students who want to can use the scholarship (They can make it free without contributing in development), so I don’t know if this system is actually being used.
I downgraded Add-ons to the previous version, and completely removed the pop-up that caused the problem, adjusted the Add-ons page to a somewhat more considerate version for the original Add-ons author.
I think that probably completes all the processing of this my issue.
I am using my own judgment on how to create Add-ons, because there is not much information available, so please feel free to contact me again if you find any other problems. Thank you very much.
4 Likes
Currently mails are sent out regarding another popular add-on (different author), asking to post a review on AnkiWeb. In exchange „free credits“ for third party tools used by this add-on are offered.
I don’t like how reviews are influenced by being paid in one way or another. We have this elsewhere on the web and it is not in the best interest of users.
2 Likes
I just stumbled upon this thread, so I apologize for bringing this up again, but it’s a really interesting read. I noticed the insane boost your addons received during that timespan and actually thought you just bought the reviews because they were super low quality and within a short time frame, lol. The idea of making some add-ons stickies is interesting, but also a slippery slope. Who decides what addon “deserves” to be pinned? With the huge range of things people use Anki for, it’s pretty much impossible to find add-ons that are truly relevant to all.
I honestly don’t like the current ranking system whatsoever and brought it up half a year ago, but Dae said that bad actors already tried to cheat the old system (Cant sort add-ons by its ratings, on add-ons web site? - #4 by dae) and you just proved that the current system can be manipulated just as easily.
It’s also insane how discouraging downvotes can be. I have yet to see a useful downvote on any decent add-on. Most of the time it’s just users reporting a bug (“I got this error xxy pls fix”, “Doesn’t work on the latest version”) or users hating on the addon (AnkiHubs Addon Page is the perfect example). Neither of which has any positive effect on the community. It’s just anecdotal, but I received 3 consecutive downvotes when I first released my addon AnkiCollab, and I was on the verge of quitting all development for it. It’s not fun to work hard on something and then have people be mean about it.
I’m not sure what the point of the review system really is if it’s solely based on up/downvotes. Add-ons with large communities or a way to blackmail users get upvoted, others don’t. Outdated addons can stay at the top (https://ankiweb.net/shared/info/498789867) even though they haven’t been updated or useful in 7 years.
The ranking should be more complex and take more things into consideration. Some ideas:
- Time-based ratings (more weight to recent reviews rather than treating all reviews equally)
- Verified reviews (moderate the review pages or allow add-on owners to report reviews)
- Usage metrics (number of active users (if you uninstall the add-on right away, it should have an effect on the add-on’s ranking), update frequency (compatibility with the latest version?), downloads, etc.)
tl;dr
Sticky Addons is not great, because it’s impossible to say which add-ons are “important”
Downvotes are dumb because most have 0 meaning and super discouraging,
the current ranking system does not reflect the relevance of add-ons
5 Likes
If rewards for ratings are not recommended, I think there need to be official or unofficial guidelines.
For example, in my opinion, the best interest of users in Add-ons is to have a lot of Add-ons available. So Add-ons with rewards for ratings is better than no Add-ons being developed.
In short, even if the interest for users is important, each developer will come to different conclusions.
2 Likes
I would extend this request to not only include add-ons that trade goodies for upvotes, but any pop-up or visible request that requires user interaction (not a banner / button somewhere).
Some add-ons started to ask users for a rating with a popup with great success (https://ankiweb.net/shared/info/1908235722).
Imagine if all major add-ons would do that. The users would have to handle 50 pop-ups before they get to do what they wanted to do. Not great UX imo.
Perhaps some rules like what apple and google do for their playstores (at most 3 popups per year / only after a period of time and not right after add-on installation ,etc.) would make it “managable”?
Thoughts?
1 Like
I did not have time to discuss it in this thread, so I skipped the explanation.
In my opinion, there is nothing wrong with 50 popups at once.
-
Official Anki does not check or endorse Add-ons. Any problems are the responsibility of the developer and user and do not require official or community support. So just as with Add-ons issues and questions, there is no need for official or community support to address pop-up issues.
-
The pop-up is intended to download only one Add-on. Lots of pop-ups are due to interference with other Add-ons. If Add-ons interfere and cause problems, it is common to exclude compatibility and make them unavailable together. So as with the Add-ons interference problems, batch downloads are not a problem that should be addressed by the authors.
-
Organizations that offer batch downloads, like Anking, can negotiate with developers and request that pop-ups be disabled. If a developer does not want to remove the pop-ups then they should not remove them.
-
It is up to the user to decide if a Pop-up is offensive or not. If users find pop-ups offensive, they can rate them all low. Even if the rating system is in the wrong order, all users have not paid for Add-ons, so there is no harm done to them.
1 Like
In my opinion, the official Anki’s can be sorted in any order they like (benefit, contribution, friend, whim, etc.). Because Add-ons developers use Anki’s servers for free, and users don’t pay for add-ons. So if a developer is dissatisfied with AnkiWeb, they need to use their own server.
I think these are often the case for add-ons developers who are active even though they have a lot of low ratings.
[1] Ignore low ratings
The developers who have been active for a long time often seem to ignore low ratings.
Perhaps the authors have a strong mentality, or the authors are just using Add-ons for their own purposes.
[2] Putting Add-ons Behind a Paywall
If developers use a paywall, there will be zero trolls, and only dedicated fans, so developers will be more comfortable with development.
Reasonably, I think this is the best solution, because even if the developer improves add-on, the large number of low ratings once accumulated will not be improved.
However, I doubt that either will work in the long run, development will eventually get bored and the developer can make money from programming.
In this case perhaps it would be better to have a community like AnkiCollab’s discord. Some developers seem to have their own communities.
Perhaps the sudden increase in the number of ratings this time is just a coincidence, so I think it is difficult to reproduce.
The 4 add-ons that increased in ratings were the most popular add-ons out of about 40 that I repaired (Pokemanki, KillStreak, ProgressBar, Zoom23). In other words, those add-ons are quite popular from the beginning. “Always on top” has the same mechanism, but its rating has not increased like other add-ons.
So I do not think that using rewards for my original add-ons, or for regular add-ons, will increase the number of ratings rapidly. Maybe in some cases, the number of low ratings might increase. Those are not controllable.
Hey, so the pop-up mentioned in the addon above works just like you basically described:
(at most 3 popups per year / only after a period of time and not right after add-on installation, etc.)
It asks the users only once after they have progressed and experienced the addon by achieving a certain number of achievements.
(In the next update, the threshold for the number of achievements needed is raised.)
The user can choose not to rate the addon and won’t be asked again in future updates.
This makes it, based on your comment, more manageable and ensures it only asks after a certain amount of time using the addon, not directly after installation. Users who don’t want to rate the addon can simply skip the pop-up by clicking “I don’t want to rate the addon,” while those who have used the addon and might have useful feedback are encouraged to rate it. This could lead to high-quality comments, aswell as reducing the amount of low quality comments and enabling more interactions between users and developers.
However, as @Shigeyuki mentioned:
-
There is no guaranteed way to ensure that users will always write high-quality comments.
-
Creating extensive “rule books” for addons might discourage developers from creating or maintaining addons. Open source projects thrive because community members are free to interpret and create without strict rules.
For example, there could be a question raised about how and if it’s fine to monetize addons and in what way, especially if we are already questioning the rate system and its interactions.
Given that Anki is an open-source program, how the availability of addons and their code should be handled is also important. The fact that the code of addons is not always open-source and uploaded to platforms like GitHub could also be discussed when considering how addons are developed and shared.
=> But determining these factors through rules could again discourage the development of open source projects like anki - and its addons
-
There will always be differing opinions on how things are developed or created. As long as users don’t find the pop-up annoying or a poor UX, there’s no need to discuss hypotheticals. Issues are often corrected by developers based on user feedback, as @Shigeyuki demonstrated here. He noticed the issue and addressed it.
-
I don’t think batch downloading should be a thing, as it is not the intended way Anki is designed. For example, you wouldn’t download 20 different Anki decks and start using them all at once, nor do users typically download 20 or 50 different addons in one day (except due to influencers and other developers promoting bulk installations). This could confuse users if they start with many different addons at once and this kind of installation could promote the amount of conflicts between addons.
-
If there are issues between addons or too many pop-ups from multiple addons, users will clearly comment on these issues. (For example, I had a user comment just one hour after I uploaded the addon, mentioning that it doesn’t work with PyQt5 and I had forgotten to mention this.)
-
What deserves to be pinned should be decided by the community, not by rule books. Criteria like recency of updates, likes, comments, downloads, and interaction with comments should be balanced to determine which addon should be featured at the top.
Again this is just my opinion on this matter and does not correspond with the complete communities decision.
All-in-all I agree to @Shigeyuki’s comments and Im still open-minded to discussing this matter.
2 Likes