It's not much, but if I'm having a crap day and something I say it getting downvoted, I feel a little better when it ticks up a notch, knowing at least one person isn't trying to just bomb my post. It's basically the tiniest little gift I like to give to people, heh.

I also think that, if enough people put that upvote in, the community could shift how it reacts to things.

Just my two cents

haha I get it. And that's also very kind of you.

I do it too sometimes, trying not to lose perspective of the true purpose of the conversation taking place of course; a single upvote in a bombarded comment won't cause detriment to the conversation but it will provide some motivation for the participant to keep trying to participate (assuming good faith on the poster, as well).

For example in StackOverflow, if I notice a -1 post but the post has some value, I'll upvote it even if it wasn't particularly useful for my issue, in order to give it a chance to be reevaluated without a negative bias of that -1

I find it a bit funny/curios on why we do such things. I've always known I'm not the only one doing that, it's one of those unsaid things that happen on the net, part of its culture.

I think you're too concerned about what strangers think and do around the internet. Most people will "bomb" your subjective side of the argument (the snarky, mean, remarks), and not the underlying idea. I've found out that conveying your arguments while avoiding personal, subjective, attacks is a much better way of getting your point across, since people are more receptive if they don't get the impression they need to be defensive.

Basically, if you're being an asshole, I don't feel a need to save you from yourself.

There aren't many (if any) formal guidelines for comment voting on HN, so this is me driving at what I want out of the community. I don't speak for everyone. For my part, I want civility, and to encourage people to contribute their opinions, popular or not.

The polite responses to misinformed ideas are often some of the most educational, at least for me, and so I'd like an environment where anxiety about the chance of being wrong is low for everyone.

I also don't want any of this to sound like an indictment of hacker news. I keep coming back here because I think the community is really top notch, and the signal to noise ratio is high. Commenting systems and forums have had these problems for ages, and they're hard problems. I just wanted to add my voice to the crowd.

I've also encountered many cases that confused me a bit and at the time didn't quite understand why those people/commenters were being downvoted.

So I've spent a long time just reading HN without commenting, trying to first understand the culture, the news are just as important as the people curating them IMHO. After sometime I've begun to understand the different mindsets and dynamics; there are many discernible reasons to be downvoted by the community, there's for example, being perceived as a shill or a strawman, and etc. And correct me if I'm wrong but there's also the secret/magic algorithms making the mod's jobs easier, I don't think those algorithms are perfect and may make mistakes sometimes, that's understandable.

But there are still enough cases that I suspect ended up being downvoted just because disagreement motive or maybe even for irrational emotion (very rarely do strong emotions come with reason) motive. And there's also the case of Trigger-happy-downvoting individuals, all these latter cases generate such damage that I have * sometimes * even found myself a bit suspicious of the worth of the HN community as a whole.

But I stressed "sometimes" for a reason...

...One has to understand that communities change, that's a fact, and as YC has had success and obtained more and more media attention (and therefore HN as well), different kind of audiences have joined the flock (me included, since a few years ago).

My guess is that HN's community is still adapting to the new audience, must be tough since there must be thousands of real people along with hundreds of fake accounts (some with nefarious purposes) and bots.

In the end, you'll find it more useful to acquire a thick skin and a "deal with it" attitude towards gratuitous negativity (down-voters included), but not an impermeable shell because sometimes there's some useful criticism behind rude negativity, just don't take it personal and master the art of spinning the conversation towards a constructive direction.

Side note: >> anonymously down-voting / reporting someone's opinion just because it's not yours is just as mean-spirited as calling someone names.

Amazingly well phrased, I had to point that one out :)

This is the most important lesson I was never formally taught about discourse. If you can mange to be Kind & Thick-skinned you will be capable of communicating with a far wider group of people than those who are missing one component of that pair.

One could argue for another voting just for agreement/disagreement that wouldn't impact the positioning and colour of the comments, because seeing HN's opinions is also interesting. It would however add another layer of complexity.

As long as people upvote "I learnt something I agree with," and not "I learnt something I disagree with," the effect is the same, the amplification is just tuned a bit. Having a different standard for upvotes and downvotes doesn't do much. If you get less upvotes, it is effectively as if you got more downvotes.

So yeah, I generally agree with PG in that

    > it's ok to use the up and down arrows to express agreement

So my current take on this whole vote up/down to convey agreement/dissagreement discussion would be not voting at all unless you find it pertinent.

Also IMHO, in other types of reasoning one should be a bit faster to downvote, for example 'rudeness', or worse, 'hate messages', for those ones I downvote immediately, or flag and sometimes even educate (if the person seems to be just confused, and I make one attempt only, if the person is not interested in changing their mind, I skip).

Other types of reasoning deserve a different approach, for example; Interesting comments I upvote even if I don't agree. Well redacted but uninteresting comments I leave untouched (I don't vote up nor down).

And there are many other considerations for different scenarios and different types of reasoning, I mostly prefer not to waste time, but I liked this post about kindness and find the topic of community very engaging and relevant to me. Maybe because I'm starting to build a community myself.

I admit that I'm a bit faster to upvote, for anything that deserves positive feedback; attention, approval, agreement, endorsement, celebration, applause, kudos, kindness, etc. And I'm OK with that.

I've stopped up/down voting all together for similar reasons.

For example, upvoting when a subject matter expert gives some context, or the person offers a comment that you found interesting. Conversely downvoting when there is a "me too"/contentless comment.

Personally I most commonly find myself downvoting what I would describe as either rhetoric or offensive.

It is. I've decided to just leave the comments (reactions) as they are and signal upvote by commenting. If the post/comment is of no interest, no comment. Voting in some ways has distorted the real intent of the site: "talking about things hackers like".

Imagine of the rating system did away with votes and instead measured the quality of the discourse on something interesting.

I suspect HN's up/down system follows the pattern of simplicity by design and for good reason; the reason being the flexibility to welcome a cognitive diverse audience.

Apparently HN's decline in audience quality over time was expected from the start¹; quality and popularity seem to be inversely correlated, but the voting privilege threshold is a good way to maintain culture values to some degree.

Correct me if I'm wrong, but given the fact this community was conceived as "an experiment"¹ to explore and empower the hacker community, I suspect one of the points has always been that the "best" ideas float to the top. In this case, the "best" ideas are always relative to the values of the community as a whole; the sum of criteria of the engaged participants, a very diverse audience.

That diverse audience can have multiple motives to vote up or down, according to each individual mindset; it can be agree/disagree, constructive/non-constructive, as you proposed, yes, but it also can be interesting/uninteresting, like/dislike, kind/unkind. Or many other criteria that are not necessarily antonyms; amazing/disgusting, enlightening/TLDR, etc.

So this singular up/down "un-opinionated" system's simplicity allows a cognition diversity one-size-fits-all dynamic, any change in UX would have to be planned for a justifiable purpose and very carefully executed.

¹ https://news.ycombinator.com/newswelcome.html

The difference is often very small and just adds not really necessary options which makes making a choice harder.

How about checking if a user downvoted more than e.g. average number of comments on some post, resulting in his downvotes carrying very little weight if it is true. Unless it's being done already.

I basically stopped commenting (well, except now -- doh) here on HN because of how frequently I saw toxic responses and punishment down-voting. There are a noisy subset of users here who either (a) can't understand or respect generalizations, especially based on older/real-life experience, or (b) demand "proof" for any claim, or (c) will punish any kind of not-narrowly-PC statement. And then the blatantly rude. Folks who say something in a comment they'd never dare to say to your face.

Therefore I've given up on caring and now much prefer Reddit's community. Less pinheads there, greater topic variety, greater sense of fun, can collapse threads, can delete comments much later (if you discover one that attracts a hate mob). I still enjoy reading HN, mostly the off-site articles linked to. But not so much the comment threads here. Not anymore.

Nit-picking and pedantry is toxic in real life and on message boards, however it seems to be ever-present in communication with fellow software engineers. In software engineering, if you miss an = sign or misplace a { or use a single wrong keyword in a 100K line program, the program can fail spectacularly. Debugging a broken program often involves nit-picking syntax--it's an essential tool in a software engineer's arsenal.

This is not universally true for written argument or rhetoric. Being clear is good. Using correct grammar is good. The details count. But attacking a clear, articulate argument by nit-picking grammar or word usage is pointless and does not substantially add to the conversation. Despite this, this behavior seems to exist as a cancer everywhere on the Internet where you encounter written argument or persuasion, including here.

When I'm wearing the Software Engineer hat, yes, the entire world can fail if a single nit is imperfect. But when we wear the Human Conversing or Persuading hat I expect and indeed enjoy seeing and using a more fuzzy logic. I expect more buffer and assume-I'm-not-an-idiot-kid-please-get-off-my-lawn-been-there-done-that-have-the-tee-shirt-thanks.

Compilers and CPUs are the world's most productive utter idiots. The best engineers adjust their thoughts to see them and emulate them, but sometimes this backfires when they forget to exit that mode.

In fact, those people are very useful and most of them have helped me from time to time, so I'm grateful for them.

Side note: >> assume-I'm-not-an-idiot-kid-please-get-off-my-lawn-been-there-done-that-have-the-tee-shirt-thanks

Thank you for that, I'll use it sometime, not sure why or when, but it's so funny I must! :D

maybe turn it into a C poem and wear as a T-shirt?

I once had the following printed on a T-shirt and wore to a Meetup.com event here in Colorado, and out of the crowd an early Bell Labs alumnus (C/UNIX era) emerged and tried to recruit me:

    while (!(success = try()));

    void whatever(params) {
        assert assume_im_not_an_idiot_kid_please_get_off_my_lawn_been_there_done_that_have_the_shirt_thanks(params);
        ...
    }

And I should warn you! I WILL make a T-shirt and give you credit.

and/or from the earnings on my book series The Dread Space Pirate Richard. (srsly)

or... savings from my day job. a good strategy has several fallbacks. :-P

I've just promoted your book up in the long to-read list.

The kindly part is so important; their attitude shows that they don't have an open mind at the moment; dismissing kindly leaves a moral door open so that in the infrequent (but not rare) event that they'll have an open mind later in the day (or later in a lifetime), they may remember your reasoning and revisit the logic in their minds and agree with you, sometimes they acknowledge you merit for it and sometimes they don't, but when they do, bingo, you just made an intellectual ally, but that works only if you were kind and treated them with respect.

Pick your battles; If you insist on proving yourself right and antagonize people (remember even trolls are people) you make gratuitous enemies and waste resources in general; time, energy, focus, etc.

Trust me, I've turned foes into allies several times in my short lifetime. And more importantly, I've saved myself from wasting time and headaches. Kindness works miracles and saves time.

Be kind, rewind (your mind?).

And don't forget to read the welcomming page and the guidelines:

https://news.ycombinator.com/newswelcome.html

https://news.ycombinator.com/newsguidelines.html

That's cool but I'd like to have better control, so I'm designing a custom HN curator myself. And I've noticed some other people have done the very same thing themselves.

So I would not complain because HN needs criticism. I'd rather suggest a technical solution through UX design, for the false alarm flagging mostly.

I'd be less annoyed by a downvote from pg, because I probably deserved it, than a downvote from someone completely random, who I might not care about.

In some ways, I also wish downvoting cost you a point like it does on stackoverflow.

For the longest time I was confused at seeing down-voted comments because I mentally assumed that downvotes mean "this is a troll comment / this is some kind of abusive comment". Not so.

Also (and I've said this before so apologies for repeating myself!), if downvotes do mean 'disagreement', could we get rid of the graying out of down-voted comments, or maybe just make it an account preference? I'm interested in reading viewpoints that I might not agree with and graying out the comment makes it really hard to do so. It also seems to fit better with the model of "downvote = this comment shouldn't be in the conversation".

If a comment is disagreeable but worthwhile, my preference is to give a substantive response. If a comment is value-subtracting, even if I agree with the sentiment, I'll downvote it (and may take the time to explain the downvote if I think my explanation will be helpful.)

The main attributes I consider to make a comment value-subtracting are:

- substantial factual errors (not just a nitpick, but core to the point)

- gratuitous negativity or unnecessary incivility

- waste of space ("me too", memes or jokes except if they're edw519 caliber, excessive references to irrelevant topics such as mentioning someone else's religion several times in a thread that has nothing to do with that)

- extremely poor reasoning or communication, to the degree that trying to engage substantively becomes difficult

It should be common sense but I suspect not everyone with the ability to downvote has such considerations and the HN guidelines¹ offer very little comment on that.

I'm guessing that mods think that once you've risen above the downvote reputation threshold you must be qualified and should have a good criteria on how to use this newly given privilege.

I'm not trying to tell anyone how to do their job, but maybe it would be useful to add some guidelines for that too (Voting, specifically), I think I like your guidelines.

¹ https://news.ycombinator.com/newsguidelines.html

Eg when I don't know how you perceive the comments in the vicinity of the discussion, it is harder for me to understand your comment's intentions and finer points.

Of course this doesn't apply to peripherical features such as the colour of the header bar.

I admire PGs work too, but don't you think this criteria would introduce to the community a herd-like mindset of following perceived moral leaders instead of evaluating ideas for their own standing?

If someone important likes what I said, I'm going to think that I made a good comment. If they dislike it, then I'm going to reconsider what I said a bit more deeply. Neither of these would change my mindset from a simple up/down vote. I'd want a comment or discussion to take place first.

In the case of your example, if someone important likes what you said and you feel rewarded for that, you're more likely to say that (or something similar) again, regardless of the objective merit of the statement.

Say your comment was "Java is stupid". it only gets 3 upvotes, but XYZ likes it. You feel validated, so next time the subject comes up you say it again. Commentary was then shaped by a personality rather than the merit of the commentary.

1st of all. This is not facebook. Upvotes and downvotes are not likes/dislikes.

That said, here's another way to explain what I meant: Commenting is not about pleasing readers (is also not about displeasing, so be kind, like the post recommends). If "someone important" likes or dislikes what you said, should not have such an impact that would make you consider changing your mind, your mind should be changed by feedback in the form of comprehensive information, that kind of feedback is much more meaningful than someone else's emotion (or even disagreement by itself without feedback provided).

There can be many reason/motives to upvote/downvote, but emotion is not reason, acting on emotion alone is just compulsive.

I don't think PG would ever downvote you if he dislikes what you said, I'd like to think he's more sophisticated than that; he would either have an interest (motive) on muting the info you introduce to the conversation (but that's sketchy), or even better he would provide feedback (reason) on why he disagrees with you. (But I have absolutely no idea on how would PG make decisions, I'm just being rhetorical here and using PG as an example of "someone important")

So my point is:

You should not make comments expecting PGs approval (or any other figure you consider "someone important", which is what I stated as a "perceived moral leader"), because that's the very definition of a "herd-like mindset of following perceived moral leaders", if HN introduces tools that allow that mindset, the community's value will suffer. Being inspired by leaders is cool, but allowing them to make your mind just by liking or disliking what you said is very dangerous (and immature).

You should participate in a conversation with confidence, and to obtain confidence you need to be well informed and prepared to test your ideas.

EDIT: Rephrased a few things, but didn't change substance I hope.

I don't agree with your "argument from authority" though - even pg can be wrong sometimes...