It's a bit funny to notice how the folks on HN generally haven't studied
the site they are using. Sure, there is source code available for the
forum through arclanguage.org, but it does not include all of the secret
sauce that PG is actually running on HN.
If you believe that you have unlimited voting on HN, you are mistaken.
Some votes and flags don't count, or better said, count less. There are
thresholds in place to prevent excessive voting, up or down, and
excessive flagging. How you vote/flag is weighted in a number of ways.
For example, if you flag submissions that are heavily up-voted by
others, your flags might count less, but if you flag submissions that
are heavily flagged by others (till [dead]), your flags might count
more.
Can I "prove" the above with actual code running on HN? --NO. The most I
can do is point out posts where PG has (vaguely) explained how things
work, and of course, things may have changed since he posted his
explanations. Also, there's probably tons of other secret sauce that he
has very intentionally never mentioned.
The meaning of "up" and "down" votes has never been defined. Some might
use votes to indicate agreement or disagreement. Others might use votes
to indicate appreciation, contribution, or other personally defined
metrics.
I'll flag spam submissions and posts, but I'll only down-vote when
someone is clearly and intentionally being an ass. Though it might seem
odd compared to other forums, I often up-vote people I disagree with (or
who disagree with me -- same thing) because I appreciate the time they
took to share their views and opinions with me and everyone else. None
of us have a monopoly on knowledge, so finding other ways to look at
something is always beneficial. Agreeing or disagreeing with another
point of view is less important than appreciating and learning from
other ways of looking at things.
How you vote/flag is weighted in a number of ways.
I've certainly noticed that my own karma isn't strictly related to the number of upvotes/downvotes I get in a certain time period, (a day for example). Moreover is seems that posts that are heavily upvoted may not reflect on my karma for a day or two - in fact my karma seemed to be increasing due to a post I made about a week ago, even though I hadn't posted in the meantime.
I always figured that was some sort of anti-spam fuzzing (similar to what Reddit does) in calculating karma, but it's even better (IMO) if what you wrote is correct.
I don't believe Daniel Hillis or Bran Ferren work at Google; they are the principles of Applied Minds, a consulting/R&D firm. Applied Minds spun out Metaweb which Google bought in 2010, I wonder if this patent was part of the acquisition?
I found that the most significant factor in diminished reliability was simply to let people have infinite moderation powers all the time.
Interesting observation since, IMO, unlimited "moderation powers" has been the biggest problem with Digg and (as of late) Reddit. I'm hopeful that HN can avoid this fate (no downvotes until the karma threshold is reached seems to help somewhat).
Personally I see HN's fate as worse than that of Digg and Reddit. HN's karma rating system reinforces an exclusive community. If someone who 'goes against the grain' tries to add diversity to the HN community, they will never get any karma, which means they will never have any power.
IMHO, the worst part about HN is the extreme homogeneity, and the system is designed to reinforce this. The recent discussion about display points is an example of this. Removing points is said to "reduce arguments" ... which is really just another way of "reducing disagreement".
I disagree - I don't fit in perfectly with the typical HN demographic (I'm black and I'm not a programmer), and I've made comments that disagree with consensus, and so long as I've been able to back up those comments, I've gotten upvotes.
Contrast that with Reddit where I can link to all kinds of supporting studies but if it's contrary to the hivemind's opinion, I'll get downvoted to oblivion while someone with a differing opinion based on something he heard from his best friend's sister's boyfriend's brother's girlfriend will get an upvote for an unsupported opinion that aligns with the group.
Of course that's just anecdotal experience on my part, so perhaps you're correct and this comment should be downvoted :)
Agree to this,
Just experienced Karma retaliation a couple of weeks ago myself. After commenting an apposing view on a comment of a high Karma HN person, all my later comments were downvoted in a consistent manner.
It was just to coincidental to attribute all the down votes to sudden lower quality of comments...
I currently rate 1756 karma and I contribute nothing of value to news.yc.
Getting karma here happens with little effort. I have trouble losing karma! It generally only happens when a comment really strikes a nerve[1] like in the suicide thread. My top level comment went pretty far negative (surprises me to see it back to 0 points). My nested comment with nearly the same sentiment as the top level currently sits at +29.
I never check a user's krama on HN prior to giving his opinion any weight. You have less karma than I do. Does that change anything whatever about this discussion? In fact you have a bunch of replies telling you how you get it wrong, and your dissenting comment has done just fine.
Unlimited Points leads to strange abuses. Creating scarcity minimized abuse, and increased the chance that people would use them up. Otherwise 1% of users create 99% of the moderations, and it only takes 1 bad apple to ruin the pie. YMMV!
But the top comments on Reddit and YC for the same news story are as good or better than the top comments on /. for the same story. That would make me feel like the unlimited moderation powers don't do a lot. For example, at the time I am writing this visiting the UN story on Hackers has "Quick, someone log in with all of them, and announce World Peace!" among other comments immediately visible with a +3 score.
> just seeing the Slashdot mention in such a document is awesome.
I think this is the right attitude towards patents, if they do what they're supposed to: advance the state of the art, be inventive/non-obvious/clever/original and so on. (IMO) cmdrtaco is saying that he's happy that an advance in the state of the art builds on his inventiveness - google is standing on the shoulders of him. If only software patents really were what they're supposed to be.
I wish patents were a celebration of inventiveness, as they are in other fields, instead of often obvious and consequentially hated in ours. Inventiveness is the highest ideal of our field; and patents, really, should be about that.
A secondary problem is that the description isn't intelligible to him - who invented prior art for it. One of the requirements of a patent is that someone skilled in the art is able to construct the invention. If the originator of slashdot, who is cited as prior art isn't "skilled in the art", who is? Software patents have truly lost their way.
Patents are required to include prior art. If they are mentioning it as prior art that means they are explicitly claiming that it is not the thing being patented, but that it builds on it in a non-obvious way.
I mean this informationally; I'm not a software patent fan, but let us damn them for what they are, not for what we think they are.
>The invention provides an evaluation system for reliably evaluating large amounts of content. The evaluation system is managed by a primary authority that designates one or more contributing authorities by delegating to each a specific quantity of authority. Each contributing authority may in turn designate and delegate authority to one or more additional contributing authorities, subject to the restriction that the total quantity of authority delegated does not exceed the quantity of authority the contributing authority was itself delegated. Each contributing authority, and optionally the primary authority itself, may evaluate one or more portions of content by associating a rating with each evaluated portion of content. A composite rating for a particular portion of content may then be determined based upon the ratings associated with the portion of content. Preferably, the ratings are combined in a manner that affords a higher priority to the ratings provided by contributing authorities to which a greater quantity of authority was delegated.
Isn't that more or less exactly how the Slashdot moderation system was initially seeded?
This is more or less how the Slashdot tagging system works, but more advanced than how the moderation system works. But still, abstracts in patent filings are generally oversimplifications.
I think one could argue that the Google system foretold by your tagging and moderation system and thus should fail the novelty test.
To your question about bonuses though, when I worked there they were aggressive at getting things filed and just filing was a bonus (didn't have to issue, if it issued that was another (bigger) bonus). Someone once pointed out the not so strange correlation between patent filings and needing a bit of extra money for Christmas shopping :-) So not to worry the Googlers in question were well compensated, enough for many many beers.
That correlation doesn't sound plausible, given the long and unpredictable lead times in the process.
I did an IDF for something June 2010 or so, the patent application was eventually filed in April 2011, and I just got the bonus this week. (And to top it off, I'd left Google in the meanwhile, so getting that bonus at all was a bit of a surprise).
If you believe that you have unlimited voting on HN, you are mistaken. Some votes and flags don't count, or better said, count less. There are thresholds in place to prevent excessive voting, up or down, and excessive flagging. How you vote/flag is weighted in a number of ways. For example, if you flag submissions that are heavily up-voted by others, your flags might count less, but if you flag submissions that are heavily flagged by others (till [dead]), your flags might count more.
Can I "prove" the above with actual code running on HN? --NO. The most I can do is point out posts where PG has (vaguely) explained how things work, and of course, things may have changed since he posted his explanations. Also, there's probably tons of other secret sauce that he has very intentionally never mentioned.
The meaning of "up" and "down" votes has never been defined. Some might use votes to indicate agreement or disagreement. Others might use votes to indicate appreciation, contribution, or other personally defined metrics.
I'll flag spam submissions and posts, but I'll only down-vote when someone is clearly and intentionally being an ass. Though it might seem odd compared to other forums, I often up-vote people I disagree with (or who disagree with me -- same thing) because I appreciate the time they took to share their views and opinions with me and everyone else. None of us have a monopoly on knowledge, so finding other ways to look at something is always beneficial. Agreeing or disagreeing with another point of view is less important than appreciating and learning from other ways of looking at things.