Here's our chance: Ask Discourse Why Jeff Broke Bad
-
How the /. "Ask [person] a question thing" works:
- One of the shitty /. editors writes an article saying who that guy is and that they accept questions.
- The commenters write one question per comment
- Eventually, the corporate evildoers who ruin /. select a few (up to 10 or so) questions, none of them are usually critical
- The corporate overlords ask the guy the questions
- One of the shitty /. editors puts the questions and answers into a new article
-
Just like fan interviews used to work in magazines 50 years ago! Progress, how does it fucking work?
-
Welcome back
Filed Under: We almost missed you!
-
I'm not sure what I'm reading either. Is Jeff actually responding anywhere in that jumbled soup ?
A good explanation of how it works:
-
@Lorne_Kates said:
@Gaska said:
Clicked one of the links posted here. It took me to a page with single comment. There's a back button on that page that I expected to take me to the page with all comments. It took me back to this forum.
My earlier comments about boycotting Slashdot? It was because of a UI design.
Also, yes.
I assumed it was because their content went to shit, which is why I no longer visit /.
-
@anonymous234 said:
"Time for you to migrate off Discourse".
If only Alex would follow that advice.Have you looked at the migration category? There's a team working on a test migration to NodeBB.
-
How the /. "Ask [person] a question thing" works:
Pity. I'd much prefer reading a real interview in which the interviewer is able to press him to answer questions he'd rather avoid.
-
Using Rep to grade reproducibility.
Yes, that's exactly the point (I'm the AC who posted the question).
Ioannidis's paper is titled "Why Most Published Research Findings Are False" (well worth a read).
His point, which is amply confirmed in my own experience, is that a huge proportion (>> 50%) of published results are plain wrong (in biology, anyway - I've no data about other fields). This is massively wasteful of research resources - not only are the original results bogus, but then others waste more resources trying & failing to replicate them.
There are many reasons for this, but the biggest is lack of accountability - there is not enough incentive to make researchers careful about what they publish.
If we had a online reputation mechanism whereby professional reputations were a function of the (ir)reproducibility of published results, that would go a long way toward closing the feedback loop - good scientists who publish reproducible results would get good (career-enhancing) reputations, and vice-versa.
Today this happens all too slowly and indirectly, leading to all kinds of perverse incentives.
A well-designed online reputation mechanism could improve things.But.... we have consensus...
I mean... isn't that what it's for.
Peer review? Isn't working?
-
Why did you feel it necessary to ban the entire population of the WTDWTF Discourse install from meta.d? After all, these are the people who, over a period of 18 months, have picked up more bugs and inconsistencies in your software than the whole of your team of paid developers and testers. The same people who are now looking to migrate off Discourse.
Ooooohhhhh.... someone here is getting saucy!
-
Peer review? Isn't working?
I guess that's saying that peer review catches some things before publication, but it's not the same as trying to actually reproduce the claimed results. That doesn't happen until after the research is published.
-
Wait....
Reproducing results isn't a part of peer review?
Wow....
-
Ok, so here's my contribution.
How do you live with the pedants? How do you feel about the fact that while important questions go unanswered people are harvesting points simply by taking the word "thanks" off the end of posts? Does it worry you at all that the kind of people most attracted to your site are not interested in actually answering questions?
(My answer) Instead of leaving a system in place and letting the community do the best with it.
Jeff is concerned with social engineering by constructing a system that controls behavior.
Is it a suprise that, in turn, he produces a culture where the rules are more important that the actual purpose.
-
TBH, I'm not entirely sure; I've never been involved in academic research nor written anything intended for publication, so I've never been involved with the process. But I have never been under the impression that the review was that thorough. (Then of course there are the journals where peer review is merely a rubber stamp, if it exists at all.)
-
When rules and law go from "protecting rights" to "controlling the outcomes and behavior"....
-
From what I've read, it varies widely. If the researcher provides data and sufficiently describes a methodology, reviewers might run some calculations or something. But I think that's rare.
-
If the researcher provides data and sufficiently describes a methodology, reviewers might run some calculations or something
This is supposedly how you agree/disagree when doing peer review, do some of your own analysis and point out where you think they should clarify something or why you think they were dumb. That kind of thing. Peer review (not done it myself, but have listened to bitching about doing it from those that do) includes more than just the paper to be published (at least in the technical stuff I heard bitching about).
-
If the researcher provides data and sufficiently describes a methodology, reviewers might run some calculations or something
So... if your work is good, we're going to verify it thoroughly, but if it's bad, fuck it, just publish away?
-
My personal experience (CS papers): If someone publishes his source code and it's not too hard to build, they try to reproduce the results. Otherwise they just check whether they're believable.
Then, of course, there's that other kind of reviewer who rejects papers without even reading them. Those bastards deserve to be shot.
-
A few more points of interest about the /. posting and moderation system:
The posts of normal registered users start with a score of 1. Registered users have a "karma" rating, which a function of good vs. bad mods their posts have received from the community; Funny mods are excluded from this calculation. Users with a high karma rating have a karma bonus added to the initial score for their post, increasing it to 2 by default, although they can opt to turn off this karma bonus for any specific post. Users with low karma ratings have initial post scores of 0, and users with very low karma ratings have initial post scores of -1. Also, any user (whether registered or not) can post anonymously, in which case their post appears as a post by Anonymous Coward and begins with an initial score of 0.
As you said, the minimum rating is -1 and the maximum is +5. A post which is already at -1 cannot be given any more negative mods, and similarly a post at +5 cannot be moderated up further. The moderation type (Funny, Insightful, Troll, etc.) displayed next to the score is based on the type of moderation that it's received the most, with the exception of Overrated and Underrated mods, which change the post's score without affecting its average moderation type.
Funny mods do not contribute toward a user's karma; this makes it a bit risky to post jokes which won't be widely appreciated (or might be considered Flamebait). The other up- and down-mods offset each other in both the post score and contribution to the user's karma rating, so ordinarily there's a limit on how much negative karma you can receive for a given post, until it reaches the minimum score of -1 and cannot receive any more negative mods. However, since Funny mods raise the post score without affecting your karma, the post score can be increased back up above the minimum (without giving any karma back) and it can then receive another negative mod. Occasionally, (and partly to combat this), mods will give Insightful instead of Funny when a post is particularly good and they want to award karma for it.
When reading the discussion, registered users have a large degree of control over which posts they see. By default, posts which are rated 0 or greater are shown; posts which are rated -1 are hidden. This can be changed to hide posts below any specified point threshold; many users like to read at -1 so posts are never really deleted or Jeffed (and in fact they can be moderated back up again even if they've been moderated down to -1). Registered users can change or eliminate the karma bonus (+1 by default), or attribute extra positive/negative point modifiers to posts based on the average moderation (Insightful, Funny, Troll, etc.). This allows you to raise or lower their score to ensure that they will be above or below your reading threshold. You can also "Friend" or "Foe" other registered users, and can set point modifiers to increase or lower the scores of their posts as well.
-
someone here is getting saucy!
Maybe but this type of question also makes us look like some needy ex-girlfriend/boyfriend.
-
It depends. There is a short list of high quality sofware engineering journals 25-30 of them, that do thorough reviews and where the editors care enough to not let any crap through. Then there is a much longer list where the quality control is not that strict.
The same goes for conferences.
A good rsearch paper should describe the methodlogy they have applied to the extent that a reader (or a reviewer) can assess how reproducible the research is. And when you then assess the results, you do it in the light of how rigorous the methodology was applied. But you never have time to actually reproduce the results in the review process - you just assess the credibility. Besides, the reproduction is a nice paper in itself, so why waste it on a review.
That said, I maybe recommend "accept with major revisions" to at most one article in five, and "reject" for the rest when I review.
Filed under sorry for the accalias. I'm on different
-
Then, of course, there's that other kind of reviewer who rejects papers without even reading them. Those bastards deserve to be shot.
I take it you've never tried to submit something to JofUR?
-
So, I got my question(s) posted.
-
I don't see it that way.
I see it like, you're helping an old lady load her car with her groceries, and you put the milk next to the meat and she slaps you.
The expected behavior was, "That's bad to do that, can you move it."
So you say, "Why did you slap me?"
-
I think the answer is very simple.
Programming and Design are two separate disciplines.
As soon as we start to recognize this, we'll be better off.
Given what I gather from Jeff, I'd say he has very limited design talent, and has not hired any. He's relying on a full staff of programming talent.
-
From what I've read, it varies widely. If the researcher provides data and sufficiently describes a methodology, reviewers might run some calculations or something. But I think that's rare.
I think the typical assumption is that experiments never really get re-run until some other researcher wants to expand on the original research, at which point he finds out the experiment was critically flawed.
-
Maybe but this type of question also makes us look like some needy ex-girlfriend/boyfriend.
I've given up on that. A lot of people here just do not understand this whole "dignity" concept.
-
Given what I gather from Jeff, I'd say he has very limited design talent, and has not hired any. He's relying on a full staff of programming talent.
Discourse is awful at both.
-
He's relying on a full staff of yes men.
No, I don't think much of their programming talent. Good programmers would turn round and point out that what they are doing is senseless.Jeff … has very limited … talent,
-
For what it's worth, it looks like some sympathetic and reasonable /. users have modpoints. Posts which aren't blatantly trolling are getting moderated up:
-
For what it's worth, it looks like some sympathetic and reasonable /. users have modpoints. Posts which aren't blatantly trolling are getting moderated up:
Doesn't matter. You can moderate them up to +297 and they still won't send them to Jeff. The only way Jeff will see them is if he reads Slashdot.
-
Anything posted by "Anonymous" is gonna get blackholed, I can guarantee that right now.
-
They've attributed questions to Anonymous Coward before. Check a few of the recent "_____ answers your questions" topics on http://interviews.slashdot.org/.
E.g. (one of the dumber questions they actually used):
And here's a slightly more serious, less softball question from Anonymous Coward to Kim Dotcom:
-
He's relying on a full staff of programming talent.
He might need to pick new staff then.
-
Every time I see "this question is closed as "not constructive", I'd like to give StackOverflow a taste of their own medicine. For example, StackOverflow exec's would be having an board meeting over the phone, and all of a sudden the phone clicks off and a pre-recorded voice says, "This meeting has been closed as primarily opinion-based and not constructive".
-
I'm building a bike shed. What color should I paint it?
I wonder how many of the Discourse-bashing posts are us, and how many are other Discohaters.
-
-
@loopback0 said:
I wonder how many of the Discourse-bashing posts are us
My guess, nearly all.
I would like to think there are other Discourse users who perhaps don't hate it, but who have serious questions about bugs not getting fixed. I mean, we can't be the only users experiencing
; that 
's annoying as heck, and AFAICT, no effort's been put into fixing it.
-
That would find things on /. and be fully aware that Jeff is the one behind Discourse? I mean sure there are PROBABLY people annoyed by it, but that are likely to show up there?
-
They've attributed questions to Anonymous Coward before. Check a few of the recent "_____ answers your questions" topics on http://interviews.slashdot.org/.
E.g. (one of the dumber questions they actually used):
http://i.imgur.com/ex0or1F.gif
http://i.imgur.com/nXvUz2R.gif
-
I think (sadly), I would have to predict that @blakeyrat posting a invective-free post nicely explaining to someone why their opinion about git is wrong is more likely than any of the Discourse-bashing posts being from non-WTDWTFers
-
Why images when we have all this awesome "DiscoMardownBBcdeHTML" stuff:
[quote=Anonymous Coward at Slashdot]
Actually, please allow me to post the full story here since I suppose most slashdot users don't know about it.Once upon a time there was a website called The Daily WTF, it was pretty popular. And they needed to replace their old forum with a newer one, so the owner of the website chose Discourse, because apparently he was friends with Jeff Atwood.
Now, the users in that forum tend to troll each other a lot, and they love to find bugs in crappy software (it's the whole reason for the website). They found a severe XSS vulnerability within 24 hours, and a boatload of bugs shortly after (did you know Discourse has no QA testing?). People weren't happy with the "infiniscroll", the general website slowness, the inconsistent DiscoMardownBBcdeHTML syntax, etc. They started to complain.
The Discourse team came to the forum to answer questions and monitor the "meta/bugs" category (which was collecting several bugs per day). They had some frictions with the community since Jeff Atwood's idea of "civilized discussion" is clearly different than TDWTF's (plus some members in particular love to post inflamatory comments). This went on for some time, then they left.
But the forum was still slow and crashed every other day, and people still wanted to report bugs, so they went to meta.discourse.org, the official forum and bug tracker (Bugzilla, Jira? nope, Discourse). But as I said, Jeff has his own ideas of civilized discourse, which include things like silently deleting your posts for no clear reason, so people were still unhappy. Some TDWTF forum members decided to troll him a bit, doing things like everyone using the same avatar, but nothing particularly bad (IMO). This again went on for some time.
Then disaster happened: the admin of TDWTF forums went to meta.discourse to report that two buttons were in different order in the mobile and desktop views, but he made the mistake of illustrating the desktop view with a mobile screenshot (browser set to desktop mode). Jeff replied "not a bug, desktop view on mobile is not supported". The first admin replied that this had nothing to do with the bug, you can easily reproduce it in a desktop browser. ...and in response, Jeff banned every member of TDWTF, with the only messages "sorry, you are no longer welcome here", and another Discourse developer self-banned from TDWTF with the message "Time for you to migrate off Discourse".
[/quote]
-
Hey, I'm trying really hard here to look for positives.
-
Maybe but this type of question also makes us look like some needy ex-girlfriend/boyfriend.
Why oh why won't
return my calls, after all the pizzas I've had delivered to him?!?
-
Everything is darkness and death.
-
-
A lot of people here just do not understand this whole "dignity" concept.
Hey can I get some me dignity please? I'll do anything for it...
-
This post is deleted!
-
"Leaving to be with family" translates to "giving them ten minutes alone with a pistol is no longer considered acceptable"
FTFY
-
