I made a script to calculate a user's karma
Just save this as karma.py and run it with Python 3.6 or higher.
import requests
import math
INSTANCE_URL = "https://feddit.de"
TARGET_USER = "ENTER_YOUR_USERNAME_HERE"
LIMIT_PER_PAGE = 50
res = requests.get(f"{INSTANCE_URL}/api/v3/user?username={TARGET_USER}&limit={LIMIT_PER_PAGE}").json()
totalPostScore = 0
totalCommentScore = 0
page = 1
while len(res["posts"])+len(res["comments"]) > 0:
totalPostScore += sum([ x["counts"]["score"] for x in res["posts"] ])
totalCommentScore += sum([ x["counts"]["score"] for x in res["comments"] ])
page += 1
res = requests.get(f"{INSTANCE_URL}/api/v3/user?username={TARGET_USER}&limit={LIMIT_PER_PAGE}&page={page}").json()
print("Post karma: ", totalPostScore)
print("Comment karma: ", totalCommentScore)
print("Total karma: ", totalPostScore+totalCommentScore)
Nice job and all.. but I really wish this wasn't a thing.
Karma is something that should stay behind at Reddit imho, it just fosters karma whoring instead of actually contributing to a discussion.
That's one of the things I liked best when I joined Lemmy, that there wasn't a visible karma counter on people's profiles.
Anyway.. rant mode off.
Still, neat job!
but then how would you sell your 500k karma account to some shady company? /s
Agreed 100%
I get the reasoning, I've read the discussions. Still I like the karma thing, not for showing it to other people, but to give me an overview over what I've been doing so far. It's kinda an activity meter for me, and a bit of feedback on how my posts are doing.
I am on Stackoverflow, and obviously I was on Reddit. While I was there, I never actually looked at other people's karma, but my karma motivated me to be more engaged.
I can see the appeal.. I just don't like it myself. It's so easy for it to become a measuring stick or something you actually try to boost.
For me the number of posts and comments on the profile is good enough :)
There are plenty of other ways to track how your posts are doing though, if by that you mean relevance, exposure, engagements, etc. On this, it was at least one aspect of positives that Twitter offered in terms of their Analytics feature. I guess we don't have the tools for something like that here but it might be a nice middle ground.
For me, the entire concept of Karma from a Reddit fashion is that it completely blows up any notion of genuine discussions. It too often lead to people fishing for approval vs saying what they actually believe. Too many like minded people also following and posting after each other, ganging up on opinions opposite their own, hive mind issues everywhere. I tried to comment something similar about this on Kbin and immediately got down-voted for it, so it was pretty clear to me there are people who don't want to be told their karma was not a real measure of their value, nor of their contributions to the site.
Yea, as said by @nlm@beehaw.org it's best that reddit is just left behind ... we don't need to recreate everything it did, and there's plenty of scope to try and create new things here on the fediverse not seen before or, at least to make lemmy better in lots of ways that make sense.
Alternative web-based PWA front ends, community discovery, contributing to the source code of the platform itself, moderation tooling (a big one), making donations easier for users, devs and admins (big one too), documentation and tutorials for new users or those looking to use the API (bet you're well placed for that one) ... etc.
Yeah, this was just a little exercise while getting into the API. I'm looking into making a simple mod bot, if I actually get around to it.
Regarding Karma: I don't care at all for other peoples' karma. Never have, also on Reddit and Stackexchange.
But I do like to kinda see my progress. With Karma I can see how well I am doing in the community in general. And it motivates me to put out more decent content.
I don't really get Karma farming though. Karma is inherently meaningless, so there is no reason to farm it.
I'm with you (and I hope my post didn't come of as too harsh ... you're making tools for users which is awesome!).
And you're right to point out that this is just for personal use. And karma is useful for letting you know how your communities feel about you. For me, I scroll through my posts in my profile page and just scan the scores to get a feel for whether I've pissed anyone off gotten some traction for some reason.
If someone is posting more often than me, I can see how your tool would be useful.
Still, I feel there are questions to be asked about whether it's healthy, but that's me ... you do you!!
Also ... kbin actually has a karma feature like on reddit. You might find it useful.
I am spending too much time on Lemmy as is :)
What I really found interesting, also compared to Reddit, is how few of my posts/comments actually have downvotes.
Yea ... it's interesting isn't it ... what happens when the "engagement rage" is taken out of the room.
Politeness and consideration ... helluva drug!
For instance, I disliked the push to bring a reddit feature to here, mostly out of a feeling that there might be a bit too much "lets keep this is much like reddit as possible" developing, just enough to think about downvoting this post for a second, but realised it would have way too mean and that a post expressing my disagreement was plenty while you were doing genuinely interesting and useful work.
On reddit, sadly, I probably would have downvoted, moved on and not thought twice.
I think, part of the equation might also be that the split between upvotes and downvotes is shown directly. On Reddit, if someone has a comment score of 1, you never know if that's because nobody cares enough to vote at all, or because there's a 50:50 split of up- and downvotes.
So if you downvote here, it is instantly visible and not as anonymous.
Also, if a post has a score of -5 on Reddit, you'd assume that everyone hates that post. But here you'd see that actually 50 people upvoted it and 55 downvoted it.
Just for numerical context: Out of my 15 posts only 3 have any downvotes at all. Out of my 332 comments, only 27 have any downvotes.
Compared with 15/15 posts with upvotes and 240/332 comments that have upvotes.
Please don't.
I think you left this line behind by accident:
l = Lemmy(INSTANCE_URL)You are right, I removed it ~20min before you posted, though. So I guess, the change wasn't synced to your instance yet. Interesting, that the syncing can take that long.
Yeah, I still see the line now. I am not sure if this was a one-off, maybe the edit occurred when I rebooted the instance for a moment and the edit fell through the cracks... Or there might be an actual issue federating edits.
Isn't that the same as "post_score" and "comment_score" in https://feddit.de/api/v3/user?username=Grishaix
That would sound plausible, yes, but apparently it is not. For me, my post_score was ~20% less than the score of all my posts summed up, and my comment_score was ~95% less. I actually opened a bug report for that here: https://github.com/LemmyNet/lemmy/issues/3393
I just ran the script on your user. The API returns a post_score of 125, while the calculated score is 121. For the comment_score the API returned 83, while the calculated score is 397.
Thanks for the clarification.
But to be honest, I'm still not quite sure if we really need this whole "internet points" thing.
Yeah, it's only for each user to run if they want to. I like to. I don't care about others' score. It's just for me to know.
I've read the discussions on that topic, and I agree that it should not be publicly visible, at least not next to the posts. If at all maybe in the user profile. But honestly I don't care about that aspect.
I just want to know where I'm standing.
I've earned ~10% of my Reddit karma in just-over-two-weeks that I have been here, even though there are far fewer people here than on Reddit, so that's concerning ;)
But that's what I wanted to know and I go that info.
It's an alright way to see if someone is commenting on good-faith or not. Anytime I saw someone saying highly controversial things, I'd check their account to see if they were just downvote collecting or actually held the view. It's harder to do that now; my account says I'm at like...-30 something, on a single comment that went beyond a few people's sensibilities. I could have put an /s on it but that defeats the purpose.
But anyone that looks at my profile now to make the "good faith" check will see me at -30, despite other contributions.
It is dumb how people worry about the number, but it does have other uses besides just a popularity indicator.
Seems like this is kbin Vs Lemmy difference, we at kbin get to see people's "reputation" (yes including Lemmy users ...with caveats) from this thread it seems Lemmy doesn't easily expose the same.
That said the reputation system is kbin is currently broken as upvotes don't count - it's a known bug which will no doubt be corrected soon.
I was under the impression that was how it was supposed to function; boosts were the upvote and downvote is the downvote, but boosts were weighted more heavily? Or something? Lmao it seems unnecessarily complicated, just stick with up or down.
Ernest switched the upvote from being boosts to favourites shortly before the Reddit exodus, he did this to better align with Lemmy.
Boosts and Favourites are both wider fediverse things - using them keeps compatibility there.
The switch is implemented in hot and top sorting (as you note I think a upvote/favourite counts 50% of a boost), but it sounds like reputation wasn't fixed at the same time, for now reputation is boosts less downvotes which Ernest himself has acknowledged doesn't make much sense.
It is still being discussed in the issues log!
https://codeberg.org/Kbin/kbin-core/issues/80
I'm getting back into Python for unrelated reasons, and last I was using it, JSON wasn't on my radar yet.
I'm curious about the
.json()method here, which seems to be exposingpostset al. for further manipulation without parsing. Is this really as simple as it appears?Yes, it totally is that easy. At first I used an API wrapper library, but then I checked out the source and there is really no need for it since requests already handles basically everything.
.json()takes the response body of the request and runs it throughjson.decode()and thus spits out a nice Python dict/list structure.It is absurdly simple and powerful.
“An user” sounds weird
But it’s entirely grammatically correctnot correct, just high school grammar level correct, english has exceptions to the "a/an" rule, thanks @HausYet i still want to say “a user”
Fuck you, English
eeeehhhhhhhhhhhhh. the a/an rule is based on the first sound (phone?) of the word, not the spelling. hence "an hour", for example, where the H is silent, but "a heist" where it's voiced.