-
-
Notifications
You must be signed in to change notification settings - Fork 319
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Getting the same training puzzles repeatedly, no score updates #813
Comments
This was reported on reddit too, thought I maybe should leave it here |
And this was reported in lichess' forums as well, some examples: https://lichess.org/forum/lichess-feedback/bug-mobile-app-android-puzzle https://lichess.org/forum/lichess-feedback/tactics-bug-when-trying-to-retry-this-puzzle-ios https://lichess.org/forum/lichess-feedback/lichess-app-puzzle-bug-with-dodgy-internet-connection |
Ok. I know this issue has been reported several times, but I have not been able to reproduce it so far. It may be a client (app) issue but it may be also a server issue. The offline caching functionality adds complexity. Now, you must send at least the app version (some bugs have been fixed on latest version) and precise steps to reproduce the problem, or I can't do anything about it. |
Sure thing, I'm on app version 5.6.1 and android version 7 with a S7. Steps:
My example was executed while having a full internet connection. This is just one example though, in other cases the rating in the overflow menu does not update or in very rare cases I have duplicated puzzles after a while. For this I can't get you any steps to reproduce though as of now |
I see the same as @l3d00m (score stuck at 1492 now, with no puzzles I complete listed). No amount of data clearing or reinstallation fixes it, which seems to indicate a client-server problem rather than a caching problem. I'm currently on v5.6.1. Username is bilts if that helps in any way. Android 8.1.0 @veloce I'm also a developer with experience in both web and Android. I'm not in a situation where I can commit a ton of time right now, but I can get my dev environment set up and help with inserting / reporting diagnostics if you can point me at useful areas in the codebase to log or alter. |
I should also mention that I've noticed glitchiness for a while, but when I first noticed this particular set of problems being an issue, I thought the website may have changed to not allow people to solve the same problem twice. My hypothesis was that I was being shown repeats of problems I had solved before, so solving them wasn't affecting my score, so I wasn't able to get new problems. That said, I should reiterate what I said in my initial report: This all works fine for me on the website. |
Ok, I could reproduce duplicated puzzles too. Same app version as above.
I'm not sure if it can be always reproduced, I'll play around a bit further. This happenened while I had full internet connection |
I compiled the app and looked at network logs when logged in and have some more details here. Surprisingly, I see the same issue running the app on my computer as I see when running on my phone. First, my rating is 1443 as reported on the website and through the API, but 1492 in the app (don't want to hear it... it had been over 2000 and I got super lazy and stopped being able to see when it was tanking in the app) Interestingly, the graphs on the website and in the app roughly agree and show that the 1443 number is from several games ago: Here are relevant bits from loading a puzzle and playing it (pretty printed for sanity reasons): https://lichess.org/training/65447/load {
"game": {
"id": "rF8uN7Js",
"perf": {"icon": ")", "name": "Blitz"},
"rated": true,
"players": [
{"userId": "piet1pegr", "name": "piet1pegr (2044)", "color": "white"},
{"userId": "yonzon", "name": "yonzon (1799)", "color": "black"}
],
"treeParts": {
"ply": 30,
"fen": "r4rk1/pp3ppp/1n6/2qp4/6b1/2PQ1N2/P3BPPP/R3R1K1 w - - 8 16",
"id": "]A",
"uci": "c8g4",
"san": "Bg4"
},
"clock": "5+0"
},
"puzzle": {
"id": 65447,
"rating": 1366,
"attempts": 20074,
"fen": "r1b2rk1/pp3ppp/1n6/2qp4/8/2PQ1N2/P3BPPP/R3R1K1 b - - 7 15",
"color": "white",
"initialPly": 30,
"gameId": "rF8uN7Js",
"lines": {"f3g5": {"g7g6": {"e2g4": {"c5d6": "win"}}}},
"vote": 74,
"branch": {
"ply": 31,
"fen": "r4rk1/pp3ppp/1n6/2qp2N1/6b1/2PQ4/P3BPPP/R3R1K1 b - - 1 16",
"id": "8I",
"uci": "f3g5",
"san": "Ng5",
"children": [
{
"ply": 32,
"fen": "r4rk1/pp3p1p/1n4p1/2qp2N1/6b1/2PQ4/P3BPPP/R3R1K1 w - - 0 17",
"id": "YQ",
"uci": "g7g6",
"san": "g6",
"children": [
{
"ply": 33,
"fen": "r4rk1/pp3p1p/1n4p1/2qp2N1/6B1/2PQ4/P4PPP/R3R1K1 b - - 0 17",
"id": "/A",
"uci": "e2g4",
"san": "Bxg4",
"children": []
}
]
}
]
},
"enabled": true
},
"mode": "play",
"user": {
"rating": 1443,
"recent": [
[64880, -3, 1425],
[64886, 12, 1422],
[64889, 15, 1434],
[64892, -6, 1449],
[65011, 5, 1542],
[65020, 12, 1547],
[65024, -10, 1559],
[65028, -8, 1549],
[65031, 8, 1541],
[65032, -8, 1549],
[65033, 7, 1541],
[65035, 12, 1548],
[65038, -15, 1560],
[65258, -6, 1510],
[65259, -12, 1504]
]
}
} https://lichess.org/training/65447/round2 Request URL: https://lichess.org/training/65447/round2 Body: {
"user": {
"rating": 1443,
"recent": [
[64880, -3, 1425],
[64886, 12, 1422],
[64889, 15, 1434],
[64892, -6, 1449],
[65011, 5, 1542],
[65020, 12, 1547],
[65024, -10, 1559],
[65028, -8, 1549],
[65031, 8, 1541],
[65032, -8, 1549],
[65033, 7, 1541],
[65035, 12, 1548],
[65038, -15, 1560],
[65258, -6, 1510],
[65259, -12, 1504]
]
},
"round": {"ratingDiff": 0, "win": true},
"voted": null
} Notice that it is registering the win, but awarding no points and thus isn't showing up in the score history. It could very well be that I've played this particular puzzle before (I don't remember it) and the site is refusing to give points for replaying an app. That would be a clean explanation, but I can't say for sure. As I have said before, though, I've noticed a lot of duplicate puzzles appearing. If the puzzles selected are directed by the server, it would seem that this is a lichess.org issue. *edit ... I'm waffling on that last statement, because as I mentioned before, this all works flawlessly when I play on the lichess website. I'm only affected in the app. |
What do you want in the way of reproduction steps? I pretty much only play puzzles on Lichess so this problem is 100% reproducable every day, for months. Perhaps as a starting point add some logging to the app to dump my puzzle score/history there locally and compare against what's on the server. It's as if there are additional puzzle history locally which makes the score differ. Oddly, if I use two mobile devices they both stay in sync and have the same (wrong) score compared to the server. I just solved a few puzzles this morning; the server is correct but my score hasn't changed on my phone. I'm a developer and I can spend time on this issue. What do you want me to do? |
Poking around a bit in the Chrome console, I've noticed several concrete bugs that would explain some of the problems:
And one that's unrelated but should probably be fixed nonetheless:
|
Something new on my end: The duplicate puzzles happen on lichess' site as well. I finished this puzzle (https://lichess.org/training/68431) first in the app, then I got it again at lichess.org (after finishing rating + 0). I did a few other puzzles and exited lichess.org. After returning back I got the puzzle again, exactly what happened in the app too. So it might be a lichess web problem too. I'll open an issue there. |
As described in lichess-org/lila#4317, I could reproduce the other reported bug on lichess web as well. It seems like a general lichess bug and not one only in the app. |
After dealing with it for over a month. I'm no longer seeing this in the app. Don't know what changed, maybe something server side. |
Something changed server side indeed. I also made a couple of edge cases fixes on app side too. Closing this then. |
Still is happening. Even after reinstall of Android app. This issue should be reopened. |
This issue should be reopened as it still persists in 6.1.1. This bug can be completely avoided by using 5.3.1. Bug can ONLY be reproduced if
If clients are alternated - let's say first app (works!), then web (works!), then app (bug!) - puzzles get out of sync. All puzzles played on the web will need to be replayed in the app without the account getting any rating for it. Bug symptoms disappear after
We can provide documentation for this asynchronous puzzle behavior as described if needed, as we already tested several devices and app versions to narrow down the problem. We would love to provide any additional information to get this bug resolved, as it is an otherwise awesome, beautiful app! edit regarding version 6.1.2: While the underlying sync issue does not appear to be fixed in this release, the new "refresh puzzles button" serves as a handy additional method to make the bug symptoms disappear. That's much better than clearing app data! |
This still happens... |
This is still happening. If I remove the app data in Android, it will start showing new puzzles but as soon as I get back to the game after leaving, it'll start repeating the same old ones always from the same puzzle. |
Still happening on iOS. App version 6.3.0 |
@veloce Could you reopen this? There are very regular reports about this on lichess forums too, see this search: https://lichess.org/forum/search?text=puzzle+app Edit: After searching some new issues I saw that this was also acknowledged in #876, so nvm |
I am also experiencing the same issues described here. It has made the puzzles experience bad enough for me that I do them much less :( |
The same puzzles: https://lichess.org/training/61982 and https://lichess.org/training/61987 |
|
This problem persists in a severe way, beyond logging out, clearing data, and reinstalling.
The app shows me the wrong score and it doesn't update as I complete puzzles. It sends me through the same puzzles in the same order, so I don't get new puzzles and my results rarely appear on the website.
Every once in a while, I'll see some scores reflected or get a new puzzle, but I can't see a rhyme or reason to it.
Puzzles work fine on the website.
The text was updated successfully, but these errors were encountered: