I'm sure many out there consider the moderation system to be something akin to sewage in a modern town: Perhaps necessary, but stinky, and best left to others who make it happen. For those of you who couldn't care less about this journal, feel free to comment on the uselessness of it and discourage me from further filling your inbox with such drivel. For those interested in maintaining your custom, state of the art cess pool, complete with fans, luminescent paint, and alligators, read on.

Personally, I still get a little excited whenever I get modpoints, and whenever I notice a significant code change has occurred, I head over to the Slash CVS and try to pick over the pertinent changes. It's sometimes amusing listening to people speculate about these changes in conspiratorial tones when Slash is all open secrets (Since when did that site go down? Misconfigured squid perhaps? We all know how much trouble /. had with Pound when it started using that to load balance.).

The recent changes did a lot of things. Yes, it gives you modpoints more often if you metamoderate. I've seen someone claiming this is a lie in his sig. Well, here's the thing: When you M2, you have a good chance of getting one mod token. It takes 40 tokens to actually get modpoints (Slash/DB/Static/MySQL/MySQL.pm - sub convert_tokens_to_points) (also, (sql/mysql/defaults.sql - sub maxtokens) which can modify this minimum). As such, if you're not getting mod tokens any other way, it'll take you nearly a month to work your way back to M1.

And why wouldn't you be getting any tokens any other way? Probably because you did something I did right when the system came in: Bad moderation. Now the system keeps track of how 'well' you've moderated in the past, based on M2. The higher this percentage, the higher the odds you'll randomly be assigned tokens. More exactly, the higher your percentage among your peers, the higher the odds. From the comments CmdrTaco has made in his journal, it would seem most people maintain around a 90%-99% approval rating with their M1.

So if you were M2'd badly right when this system was introduced, you probably noticed these shiny new "Metamoderation Results" messages in your inbox. They reported your present approval rating was sitting around "85%" or some such low number. And then, well, you were just voted out of moderating. It looks like democracy is tyrany.

Now, the groovy thing about all of this, is that if you read comments in the code (Slash/DB/Static/MySQL/MySQL.pm - sub _csq_bonuses), you'll notice there's actually a fair bit of explanation for the logic. In fact, there's some detailed thinking and response to user feedback. Do you remember when Slashdot posted an article criticising and praising its own moderation system? They quote that article above the sections which implement specific suggestions. Unfortunately, I'm not sure this knowledge is widespread.

This poses an interesting question. If changes are made to reward 'good' moderation practices, but these changes aren't publicised, will they be effective? It seems the system is designed to have maximum feedback, so most of the benefit of good moderation is that you're asked to moderate again more quickly -- often with a small or nil karma bonus. But I speculate there are some motivated moderators who may already be trying to game the system. People who feel they can moderate fairly, and want to, but can't seem to figure out what to do to get the opportunity to do their civic duty. So, if good people know about these rewards, perhaps moderation will improve. So for friends and trolls alike, let's see if this helps:

• Meta-modding a comment which has been multiply moderated only positively will adversely affect you and have little effect on the actual moderator
• Moderating comments positively shortly after they have been posted will give you a token bonus; Moderating a comment from 1 or 2 (either negatively or positively) will give you a token bonus
• Moderating a comment posted late in a discussion will give you a token bonus
• Moderating replies (non top level comments) will give you a token bonus
• Tokens continue to accumulate even as you mod, so you may mod again very soon now if you do a good job
• Letting your tokens expire is one of the worst things you can do if you want to moderate again soon
• Having a low M2 fairness rating is just about the worst thing you can do if you want to moderate

Personally, I think this system is improving dramatically. The key is to ultimately eliminate the huge amount of redundant posts in high traffic discussions. To do this, you reward people for moderating down marginal redundant comments, and you don't reward users who try to post as fast (but unoriginal) comments as possible. There may never be an effective way to prevent Slashbotting, though. If it's what a large portion of the audience wants to read, why not? Much to the joy of the trolls, I imagine.

Just an odd thing I noted earlier today. warez.slashdot.org used to be a kind joke. It resolved to 127.0.0.1. No longer. It takes one to Slashdot's homepage. Was this instituted recently when Pound was added for load balancing, or was this done a long time ago, and I just didn't notice?

Speaking of which, Pound seems to be doing a crappy job, or someone got cheap and removed several of Slashdot's servers. I'm finding more and more often I have to go to https://slashdot.org (subscriber plum) just to get Slashdot to load midday. Once more, when I'm using http, I've frequently found 500 errors popping up, mostly when loading messages.pl. What gives?

On a related note, Jamie has shown him/herself to be a class act on multiple occasions, including this one. Slashdot gets a lot of flamage thrown at its admin, much of which it doesn't deserve, and some of which it does. Most of the time, it looks like Jamie responds to this feedback and does so calmly and in a business-like manner. Even when throwing my own flames at Slashdot, I was quite taken with the soft handed response I received. Once more, I found it a little surprising that, after what happened to the censorware project, we have both Michael Sims and Jamie on the same team. That takes some tolerance.

There's a fair chance that you've already seen this video before I did. After all, Fark and PA already covered it, both in a fairly derisive manner. Just in case you missed it, there's a fairly rotund teenager attempting some staff techniques, or something. The video itself isn't terrible, and one definitely can't fault the kid for enthusiasm. But the video isn't the matter here.

Just like Andy Kaufman, this appears to tbe theatre which is more centered on the reactions of the audience than on the original material. And my initial reaction may show I'm a moron with no taste: He doesn't look that terrible. Sure, in the latter half of the video as he starts kicking out his feet a bit much and stumbling a bit more, he looks a little uncoordinated. But he realistically doesn't look that bad in my mind, and he's certainly putting some effort in, lack of proper training with a bo aside.

Others' commentary — others who are apparently my peers — seem decidedly more acid. Comments range from criticizing the fellow for his being overweight, to criticizing Star Wars in general, to just plain being an ass calling him whatever names necessary to exhaust a troll's vocabulary. I know sociologists say somethings or another about exclusion rituals being necessary for community building, but this sort of thing continually turns my stomach when I see it. And I've seen it a lot in the past couple years in online communities which are, agan, composed of my peers. Maybe I'm just a dork for not realising how ridiculous this gentleman actually looks and am equallty suitable for such ridicule.

But I sure hope not. I hope that what I'm seeing is the worst of what society has to offer, raised up in a ha-ha-only-serious fashion designed to make us take a closer look at our own reprehensible behavior. I hope Something Awful causes that skin-crawling in many other people. I hope this is a phase and not some kind of generation-defining trend. Criticize the apparent uber geek cruelly, brutally, for the purpose of humor, for the purpose of tying yourselves together into...what? Is this just an abberation caused by the fact that trolls are so easily allowed by anonymity or is it a microcosm?

This is rambly and disjointed. You'll pardon me. I should be studying for finals.

Just in case I'm not the only one who needed to know why asterisks were appearing everywhere next to people's name in the comments: They mark who is a subscriber. It'd be nice to, I dunno, have a real gold star instead of a random asterisk which no one cared to send us all a little message about. Probably already a feature request at the CVS. If not, maybe someone should add it? Maybe me. Anyhow, it appears that the asterisk is determined by the current subscription status of the poster, so if you subscribe now all of your old posts will be asterisked...ed.

As stated at http://cvs.slashcode.com/cgi-bin/cvsweb/slash/themes/slashcode/templates/dispComment;misc;light:

Revision 1.18 / (download) - annotate - [select for diffs] , Tue Apr 22 20:03:09 2003 UTC (4 days, 7 hours ago) by jamie
CVS Tags: useperl, T_2_3_0_89
Changes since 1.17: +2 -1 lines
Diff to previous 1.17
Added comments.subscriber_bonus, the asterisk after subscriber posters'
UIDs in comment headers (aka the "Gold Star").

Someone is going to have to point out that Fark marks their subscribers with a large yellow banner featuring a gold star and the words "Total Fark". Just like how people had to point out to George Lucas that there was this really popular Messiah tale told long before he made "The Phantom Menace".

EDIT 28 April 2003
Might help if that CVS feature request link actually worked.

Slashdot's karma system has some interesting properties to it. Here's one that's particularly near to my heart: moderating will drop your karma much more often than it will raise it. As ones moderations at are meta-moderated, the results affect the moderator's karma. When you are meta-moderated down, you will lose a point of karma unless your karma is already below "badkarma" which defaults to -10. Iff your karma is below "Good" will a "Fair" or "Funny" M2 result in a karma increase, and, then, the odds of it happen decrease as you approach "Good".

So I've been doing a lot more reading, M1, and M2 and little posting. If one reads regularly and insures that "comments.pl" or "article.pl" are almost always the 'last page visited', it takes just about a week to get moderator access. (And this is perhaps part of the reason old-timers frequently say they don't get M1 anymore. Journal pages don't count unless you change your threshhold for the comments, at which point you don't see the journal entry any more.) Therefore, if one is a regular Slashdot visitor, but not a regular commenter, he will lose karma at relatively swift rate, even if he's at 50...er, I mean...the high end of "Excellent".

I've got about a 15% rate of being rated as Unfair moderating, which is something I can deal with. In my opinion, agressive M1 and M2 is exactly how one defines the community one is in, if he wishes to make a change in it. And I wouldn't mind changing some small things with Slashdot. I'm pretty quick to mod down people who post information that is redundant -- already posted by someone else, especially if it's common knowledge such as google caches or references to well known information. Or, worse, people who reiterate a blithe viewpoint already posted in the swarm of 100 or so comments pre-existing on the story. The reason stories have 400 comments, frequently, is because 200 people were too lazy to read through and see if someone already said what they wanted to. I check post dates and insure that there's at least an hour or two between the first and second redundant posting, something I imagine most of the people doing the M2 don't notice. I don't think any of my Redundant mods have been considered Fair. I also frequently mark postings which are intended to be Funny, but aren't, as "Troll" or "Offtopic". Again, M2 hits my karma a little.

There's, of course, the note that you can always moderate using only "Underrated" or "Overrated" and bypass M2 altogether. Doing so, I feel, is dumb. If you don't like someone's post, it's a good idea to tell them why. Further, it intentionally avoids the M2 system which is designed to let you know how the rest of the community views the situation. Not that it's hard to fix your karma. Of course, I've used both at appropriate times. Overrated when someone made a comment which was "Interesting", but completely wrong, as pointed out by one of the children of the comment (with refrences, no less). I've voted a post as "Underrated" which wasn't particularly interesting, but was modded as "Redundant" when it wasn't.

I've found, like I did a long time ago, that it's pretty easy to get modded up. So every now and again I need to post at 1 and get modded up to 5 to fix that problem of my ever-dropping karma. At least, if I want to keep that karma bonus on the occasion it's needed in a 200 comment article and I actually want to have a discussion. Posting well-researched or informed comments is still a nearly sure-fire way to get modded up. Posting an uninformed, cocksure opinion early in an article that experesses a popular opinion, unfortunately, still works as well, and takes a lot less work.

So when finals are over, I'll be posting a little more often, primarily to fix my karma. Cuz, you know, my karma sucks because I moderate.

This goes to linux because prior to actually heavily using linux (college) I had only seen a few random mentions of the 'meta' key in my life. And I have yet to see any Windows/DOS/OS/2 program that had something that responded to a 'Meta' key. Perhaps that's because these are operating systems which have had a strong habit of tying themselves to the present day hardware. Or are a little more user-interface centered. But I ramble.

My point:

Today I finally received a full understanding what a Meta key is. I've been forced to use XEmacs on occasion because some professor found it to be the editor from the gods (or at least said editor prefixed with an 'X') and wanted me to realize why it kicked gvim's ass. It didn't take long for me to ask what a "M-g" meant [the shortcut key for 'go to line' -- I hate using the mouse if I can avoid it, especially when coding]. A friend told me "just tap ESC, then hit g". I said "How stupid," did it, and it worked. Further, since then I've frequently seen mention of an AltGr 'chord'. PuTTY in particular takes note of the thing in one of its options, and I use PuTTY enough now (perhaps unwise given its security history) that not knowing what an option does slightly bugs me. However, I started not to care so much about all of these 'chords'. It was a lost cause. Searches for keyboards only found me standard 101/102 key keyboards, occasionally finding me an AZERTY or Dvorak keyboard, but little else.

But then I got in the habit of using the Linux machines around here, and ESC just didn't work for the Meta key. Or it was difficult enough to time it right so that it did work that I couldn't do it that way. It was a small pain recoding all of the desktop shortcut keys to being Ctrl-Foo instead of Meta-Foo. So, I did a little more searching, and learned that my right Alt (and only my right Alt) key did that job.

And thanks to that E2 node on the 'space cadet keyboard' (above) I finally understand where all of this nonsense came from. And, oddly enough, it links back to those LISP machines I've heard so much folklore about. Stories abound about these things, like easy symbollic debugging and code correction during runtime -- things which were difficult/impossible in debug and remain difficult in gdb (at least hand patching an executable in situ). Simultaneously this is somewhat magical and somewhat annoyingly archaic. The only keyboards I've ever seen with an AltGr key that's labelled as such are attached to AIX machines in the server room.

Incidentally, if someone could provide a link to a picture of said keyboard layout, I'd love to have a look. You'd be amazed how hard it is just to find a decent picture of a properly labelled AZERTY keyboard.

What follows is an overly obsessive analyzation of a minor part of Slashdot. I really am this geeky, but I'm not really lacking as much perspective as the time I took on this might make it appear.

Trolls seem to be popping up lately, and here I am with no mod points, so the best I can do is post to my journal. Observe what several found Funny.

Starting Score: 1 point
Moderation +4
80% Funny
20% Overrated
Extra 'Funny' Modifier -1 (Edit)
Foe Modifier -1 (Edit)
Total Score: 3

I've seen this same thing a couple dozen times before of course, normally posted by someone with a monicker such as "Anal Cox". Unfortunately, the article didn't have much of jack to do with Microsoft or its world dominance. Once more, the comment was posted in reply to another post currently scored -1. So the question is: Who was so dimwitted as to mod this guy up? And who are those other 3 dimwits? Do they have a club?

But before I get too ranty, my point -- I'm seeing a decided drop in the quality of moderation lately. And it seems to be caused a lot by unsophisticated moderation. Take this post by the same creature. It's a troll, as clearly and elegantly explained by an AC response. However, note the relative score of each. With my modifiers, they both result as 2, but that's because AC gets +1, long post gets +1.

This makes me long for a system where I can give half credit to moderation of people not on my friends list. Because, as observed above, even with my modifiers, that troll is still above my default +2 threshhold. I don't want to take the overboard step of making foes -2 because that makes a registered (1), high karma (+1) troll who gets modded up once (+1), drop below (1+1+1-2 == 1 < 2) my threshhold. And even trolls have something useful to say, sometimes. The key is that they only get modded up when they should be. Pardon me as I try to become even more subjective.

Example 2, a Mr. Samir Gupta. A fair bit more clever troll, who actually managed to convince one of my friends to befriend him. He provided this 'Insightful' comment which got him on my foes list. Yes, that comment is completely ludicrous for all sorts of obvious reasons, but he sure sounds convincing when he's BS'ing it. Check his history and you'll find a comment with a wonderful response detailing his long experience at trolling.

But the moderation system seems to blatantly fail in some cases. The mondane, popular comment. The clever troll. The confident, misinformed IANAL. In all of these cases, I would beg to be able to do metamoderation specific to myself so that I could increase the effect of intelligent moderators on my reading. And, yes, I realize the technical difficulties as well as the danger of creating a moderation that insures I only see opinions which I agree with. But there's been at least a few people on my friends list mainly because I disagreed with them, but still found their points well put together.

I would love to properly submit this bug at the apporpriate SourceForge list, but I can't be bothered. Why? They require you to be a registered user to submit, and I'm not. I just wasted 10 minutes confirming this bug, and now they want me to spend another 10 to 20 validating an account just so I can submit a bug that likely won't be fixed unless I take 30 minutes to write and submit a patch as well? Blarg.

With those compalains out of the way, my issues is as follows: With my current comment preferences, a friend of a friend (the little green pill) gets (+1). This has mostly served me well. Oddly, though, someone found it fit to mark me as a friend. No less, I had already flagged this someone as a friend. Therefore, I'm a friend of a friend. Yes, I'm listed on that page. Why this is so I can't fathom. I can buy the claim that being a friend of oneself is a feature. After all, there are enough geek narcisists out there. But being a friend of friend via this method doesn't seem helpful. However, in itself, this isn't realy a bug.

What I do find to be buggy is that when I view my own comments, I get the +1 bonus for being a friend of a friend (!), but I don't get the green pill next to my name indicating that this is so. Testing this in both "Light" and normal modes resulted in this behavior. Setting my FoF modifier to 0 caused my comments to reduce to the appropriate level.

A small annoyance mostly. But I like finding bugs. And I like reporting them. (I rather fixing them, but I'm not exactly a perl master.) Heck, I generally like hearing myself talk. Way to encourage my bad habits, Slashdot. Darn you!

On a nearly completely tangential note, now I apparently have a fan who is also a foe. I'm wondering if he has a foe of a friend modifier. If so, I wager he's seeing this same queer behavior.

1. One who can't make humour appreciated by his peers
2. One who gets excited when his Slashdot karma transitions to a new adjective

I still don't see what makes this quite so interesting. Do not question the moderators, for they work in strange and mysterious ways.

On a semi-related subject, it seems that one of the surest ways to insure that one gets modded up is to make meta-commentary about Slashdot at the end of a cynical comment. This seems to especially be the case if one refers to the general Slashdot populace as pimply-faced teenagers generally younger than 20 without a whit of useful sense.

I'd like to see real statistics on that. Because either a lot of pimply-faced teenagers are incredibly open-minded, welcoming harsh criticism of themselves as 'Insightful'; these similar pimply-faced teenagers are on an even higher level than that and manipulate comments with moderation to display different ideologies; or we just plain have a lot of old, curmudgeonly geeks reading Slashdot who don't quite realize they're not alone.

Regardless, it appears that among the commenters on Slashdot, there's a fair amount of knowledge and experience, often missed by the very knowledgable and experienced. Interesting articles sometimes get posted which I honestly don't even know where to start a write-up on, and then after a few hours, I go back at 5 and learn something about ground-loops.

So I moderated for the first time and luckily found a fair amount of Informative and Insightful comments that had been heretofore neglected. Things with research and links to real web pages with relevant content. Amazing.

I was disappointed with the lack of moderation options, however. 'Troll' and 'Flaimbait' are about the same, but meanwhile there is no 'Uninformed' or 'Incorrect' moderation topic. One is left looking at either 'Offtopic' or 'Overrated'.

As other people have stated, Overrated and Underrated are closer to M2 than M1, but they 'fit' when other moderation reasons don't. The solution is to remove 'Underrated' and 'Overrated' and add more moderation reasons, starting with 'Uninformed' and 'Not Funny'.

Edit 26 Nov 2002

Apparently a partial swing in the right direction has occurred. According to CmdrTaco, Funny M2s are now given either "Funny" or "Unfunny" (as opposed to the normal "Fair" or "Unfair"). But this still leaves M1 with the kludge of "Troll" or "Overrated" for a comment marked as Funny which is bland. Perhaps I'm just too damn negative.

My friends list balloons, and I felt obligated to explain why. Obviously this is a sign of a larger problem.

Wasting time. First I read Slashdot's stories because they passed the time that would otherwise have been dedicated to homework. Soon I began reading the comments when I had run out of unread stories of interest. Working a long shift with little work and a computer helped hone the unhealthy habit. Having found it possible to spend too much time reading Slashdot, I found it necessary to read it efficiently. As such, I created myself an account primarily so I could view the "sparse" mode and view comments at a threshold of 2 by default.

Highly commented stories I've mostly avoided. I've tried reading such things at a threshold of 1 (300 comments) somewhat randomly, and found valuable information in posts rated at 2 or 3. Several times I've tried reading such things at 5 (38 comments) only to find 12 'Funny' posts that are dreadful, 10 'Insightful' posts which are underresearched demagogeury, 6 comments consisting of sentences searching for predicates, linking to previous articles on similar subjects, 3 links to the same google-cached page, and 7 or so comments of value to me.

More recently, I've discovered I could actually get value out of those 500 comment stories with tweaking of my comments. Although adding a "Funny" [-2] definitely aided matters, perhaps the most beneficial thing I discovered was the use of friends and foes.

In short, if I find a comment especially insightful, I've found a friend. If a comment is particularly base and highly scored, my foe list lengthens.

Trolling annoys me. At +4, I find a "Micro$oft sucks" comment veiled in larger words than normal but still with the dollar-'s': '5 (Interesting)'. We've all wondered what Bill could fit up there. But I don't want to read your ruminations on it just now. 'Foe' [-1]. The shortness of my foes list hopefully indicates a mostly intelligent moderation population.

College has taught me to respect good research. But simple, fast statements get modded up quickly and research takes time. Unfortunate because I'm often left dumbfounded by otherwise seemingly fascinating articles. Background material comes in the comments, but frequently rated at 1 or 2, after all of the "Your-anus" jokes have reached '3 (Funny)' and one '4 (Interesting)', once again, due to a Microsoft reference. If I have the time to read a story at 0 and see someone who posts clearly thoughtful and researched commentary, I have a new 'Friend'.

If only I could give +1 to AC, but keep Funny from reaching 5, I'd have a lock.