Welcome to the Lounge

#realJSOP22-Feb-21 8:27

22-Feb-21 8:27

That's the problem. I'm fine with people not finding an article worth their own time, but I can't make it better if I don't know why. That's the whole point of my complaint, so it most definitely *is* a thing.

".45 ACP - because shooting twice is just silly" - JSOP, 2010
-----
You can never have too much ammo - unless you're swimming, or on fire. - JSOP, 2010
-----
When you pry the gun from my cold dead hands, be careful - the barrel will be very hot. - JSOP, 2013

20212a22-Feb-21 8:40

20212a

22-Feb-21 8:40

#realJSOP wrote:
but I can't make it better if I don't know why.

Pretty sure in these cases there is no why.

honey the codewitch22-Feb-21 9:03

honey the codewitch

22-Feb-21 9:03

I wrote this a little earlier to someone else, but it bears repeating here.

Recently I submitted an article and codebase that I had worked quite hard on, and tested (i thought)

I only got a single vote, and it was one star.

That does tank my overall article rep and it makes me wonder if that impacts ranking and thus search engine results, so for reasons there are real world implications of having a downvote troll - someone that religiously downvotes you particularly if you write niche content that is unpopular but nevertheless important for those who need it. Which I do a lot of.

But more than that - and this goes to JSOP's point - that article contained code that I then tried to use in a commercial product on a platform I thought I had tested it for, and it failed right out of the gate. First time i tried it in the wild. I didn't have time to investigate so I used another offering in the alternative, and there it sits until i get free time to attack it again.

What if that one star was someone who ran into the same problem I did with my code?

Hundreds of people downloaded it. That one comment could have saved everyone who used it and the person that wrote it a lot of hassle.

Maybe if they didn't have the one star option they'd be more inclined to write a scathing comment. I'd still much prefer that to a non-descriptive "general failure" message a one star vote sends.

Real programmers use butterflies

20212a22-Feb-21 9:22

20212a

22-Feb-21 9:22

honey the codewitch wrote:
I'd still much prefer that to a non-descriptive "general failure" message a one star vote sends.

I agree. But I'm not going to care if one person keeps downvoting me.

honey the codewitch22-Feb-21 10:23

honey the codewitch

22-Feb-21 10:23

20212 wrote:
if one person keeps downvoting me.

I'd care in that case. But in that case, I care because I don't like being trolled. In that case it's not about the votes. It's about being harassed. That's something I'd pursue with staff, and I'm not afraid to say so. Especially since they're probably doing it to others too.

And someone might say "it's just a website, get a life!"

What about the other person and their life? Sitting their harassing strangers for daring to contribute. At the other end of that website is a real person engaging in activity designed to undercut the site and harass its users, and that nonsense doesn't have a place in a constructive environment because it runs directly at odds with the spirit of sharing and mentorship that has been cultivated here.

So yeah, I'd feel like a Bad User(TM) for not reporting it. If they're doing it to troll they should get banned. Whether how they do it is downvoting or commenting.

Real programmers use butterflies

20212a22-Feb-21 11:34

20212a

22-Feb-21 11:34

honey the codewitch wrote:
harassing

Seriously? You're too soft. Big Grin | :-D

Daniel Pfeffer22-Feb-21 8:04

Daniel Pfeffer

22-Feb-21 8:04

We want to know who comes second to Granny Weatherwax? Smile | :)

(Also known as OriginalGriff...)

Freedom is the freedom to say that two plus two make four. If that is granted, all else follows.
-- 6079 Smith W.

Matthew Dennis22-Feb-21 6:38

Matthew Dennis

22-Feb-21 6:38

Don't worry about the occasional 1 vote. Chris has applied some complex statistical analysis on the votes and excludes outliers from the calculation of the overall rating.
So if you have 20 +5 votes and 1 1 vote, your rating will still be 5.

If you have 10 1 votes, then there might actually be something wrong with the article, at least to some group of users that don't have the required context or experience to appreciate your brilliance.

"Time flies like an arrow. Fruit flies like a banana."

#realJSOP22-Feb-21 6:42

22-Feb-21 6:42

I don't care about the points deduction, I just want to know *why* it was down voted.

Does the code not work? Was my thinking wrong in terms of the code I wrote? Did I not explain the code sufficiently? Did I offend your sensibilities? Was the article not long/short enough? Are you annoyed because there's no downloadable file?

Matthew Dennis22-Feb-21 6:46

Matthew Dennis

22-Feb-21 6:46

In many cases its the same reason that pigeons crap on statues. It's just what they do. No thought at all.

"Time flies like an arrow. Fruit flies like a banana."

#realJSOP22-Feb-21 6:56

22-Feb-21 6:56

Well, it matters to me because this is what I do for a living. If I'd actually "doing it wrong" (as far as the actual code is concerned), I'd like to know.

Doing it "different from the way you'd do it" is not a reasonable complaint, because given a specific task, ten different programmers will come up with at least eight different ways to accomplish that task, and all of those ways would work.

MarkTJohnson22-Feb-21 6:59

MarkTJohnson

22-Feb-21 6:59

#realJSOP wrote:
Doing it "different from the way you'd do it" is not a reasonable complaint, because given a specific task, ten different programmers will come up with at least eight different ways to accomplish that task, and all of those ways would work.

Which I am always grateful for. If there WAS only ONE way to do stuff many of us would be unnecessary.

I’ve given up trying to be calm. However, I am open to feeling slightly less agitated.

Matthew Dennis22-Feb-21 7:01

Matthew Dennis

22-Feb-21 7:01

Actually, given 10 programmers, they will come up with at least 16 ways to do something.

"Time flies like an arrow. Fruit flies like a banana."

Chris Maunder22-Feb-21 7:16

22-Feb-21 7:16

#realJSOP wrote:
Did I offend your sensibilities?

You snuck that in there as a trick question to ensure we were actually reading your post.

cheers
Chris Maunder

#realJSOP22-Feb-21 7:22

22-Feb-21 7:22

That's one of the questions that I don't care about, but included it in the interest of covering all possible bases. Smile | :)

Chris Maunder23-Feb-21 4:03

23-Feb-21 4:03

So this got me thinking.

What if we replace the voting system with the reactions system (I think this is actually one of your suggestions)

The reactions would be

thumbs-up
Great-read
Techno-mazing
Saved-my-arse
bad-code / poor-form
technically-incorrect
hard-to-read

So the feedback must be chosen with a little thought, and what feedback provides guidance. A reaction of the first 4 will translate to a vote of 5, the last 3 to a vote of 1. A score can be calculated, the reactions provide readers with a sense of what people like, and the whole thing uses our current systems (give or take a small connecting wire)

cheers
Chris Maunder

#realJSOP23-Feb-21 5:30

23-Feb-21 5:30

Would more than one reaction be selectable at a time? And would "hard to read" really be applicable (unless you explained what "hard to read" means)? We have a lot of non-English speakers here, and I can see that one going sideways on us.

We also still need in-site notification of new article formum messages (pretty please Smile | :)

Chris Maunder22-Feb-21 7:15

22-Feb-21 7:15

This is, as always, a tricky one for sure.

First: we filter out downvotes from most members if their vote is too far off the general consensus. Drive-by's shouldn't actually affect your score after more than 10 people have voted. However, if a high ranking member downvotes you, their vote, as a trusted member, sticks.

Making votes public is something we've toyed with forever (and we have the code ready to rock and roll) but we've always felt, based on past behaviour, that the tit-for-tat voting wars would just get silly.

So: we could switch to a model where you only get the good votes and no one, not even high ranking members, can downvote anyone's article. That's very easy to do (flip a switch, more or less) but what do we do about poor articles? How do we surface the good articles?

One possibility is that you show the articles with the most "thumbs up"s first, but this doesn't reward quality, it rewards an article for being old (which is the last thing we want).

So taking all that into account, and wanting something that

1. Says "thank you" to members
2. Helps sort the wheat from the chaff
3. Doesn't favour old over new
4. Provides trusted authors with more say
5. Minimises drive by's from new members

we've ended up with our rating system that has exponential voting weighting based on level, auto-filters out outliers, provides a relative (/5) score, and can identify the good and the bad.

The biggest issue (I think) is that a downvote is considered judgement. An upvote is a thanks and a pat on the back and a great dopamine hit. We want the latter but not the former.

So: options?

cheers
Chris Maunder

#realJSOP22-Feb-21 7:27

22-Feb-21 7:27

Chris Maunder wrote:
That's very easy to do (flip a switch, more or less) but what do we do about poor articles? How do we surface the good articles?

Remove the 1-5 voting entirely and have a series of checkboxes, something like:

I found this article useful
I downloaded this code
I used (or will use) this code
I recommend this article to others

And come up with a score based on the answers, and the reputation of the member answering them

Chris Maunder22-Feb-21 7:36

22-Feb-21 7:36

There's a few issues with this.

1. Anything other than a single-click will drastically reduce the number of votes, thus reducing the effectiveness
2. These are all positive votes - there's no way of identifying poor content
3. This provides an absolute number of ticks, but no way to create a time-invariant relative scoring.

A way around that is to simply give in to The Google and just score an article based on # pageviews and time read. We get an average time spent per article and rate them that way.

cheers
Chris Maunder

#realJSOP22-Feb-21 8:10

22-Feb-21 8:10

it's not a matter of identifying poor content, it's the idea of identifying the most useful content. Usefulness can be weighted similar to how you currently do voting (awarding rep points the same way). It should all work out pretty much the same way.

In the end, users are more likely to look at the articles most often marked useful (just like they'd look at articles with a higher score). The only difference is that you won't have me complaining about drive-by down-voting. Smile | :)

W∴ Balboos, GHB22-Feb-21 8:12

W∴ Balboos, GHB

22-Feb-21 8:12

Chris Maunder wrote:
So taking all that into account, and wanting something that

1. Says "thank you" to members
2. Helps sort the wheat from the chaff
3. Doesn't favour old over new
4. Provides trusted authors with more say
5. Minimises drive by's from new members

Chris and the other might powers on-high, you're overthinking this ratings thing.

All that needs to be done is to ask me rate them.

At some point, this reply was obligatory - so why not here?

Ravings en masse^
"The difference between genius and stupidity is that genius has its limits." - Albert Einstein
"If you are searching for perfection in others, then you seek disappointment. If you seek perfection in yourself, then you will find failure." - Balboos HaGadol Mar 2010

Greg Utas22-Feb-21 9:34

Greg Utas

22-Feb-21 9:34

I would suggest no votes without a comment, or at least no 1 through 4 votes without a comment. If I'm not giving it a 5, then I think it can be improved or disagree with it, and I should be willing to say how. And if I provide a useless comment, like "You know less about this than the back end of a donkey", it should be possible to flag my comment so that it, and my vote, can be removed.

Robust Services Core | Software Techniques for Lemmings | Articles
The fox knows many things, but the hedgehog knows one big thing.

Chris Maunder22-Feb-21 9:42

22-Feb-21 9:42

We had this and removed it. See My blog post[^]. It actually made tings worse.

cheers
Chris Maunder