by [anonymous]
1 min read

11

I noticed enough comments about topic recently that I wanted to run a quick poll about topics and discussion categories on Less Wrong.

Since being told others opinions might cause an anchoring bias, please take the poll here before reading further:

https://docs.google.com/spreadsheet/viewform?formkey=dEVJSTBTdk92UjhTWWp6U1dYSDltclE6MQ

 

My personal opinion at the moment (last chance to avoid having me influence your poll results) is that I want to reduce the number of people who feel that articles posted on Less Wrong are off topic and limit topical complaints. I don't want to do this because I think the complainers are wrong, because if they don't want to see something, they are correct to indicate it. But I feel that there are ways of handling this at a higher level that might be more effective than just up or down voting based on topic alone, and I feel like it would help increase the quality of discussion here to have better topical division.

I think the biggest problem with the current system is this definition for Discussion.

"This part of the site is for the discussion of topics not yet ready or not suitable for normal top-level posts."

 

The fact that Discussion is "not yet ready or not suitable" indicates to me that as a category it's not cutting apart topics well. "I'm making a draft about SIA and SSA that I want to put in Main" and "Someone made a new scientific paper about the OPERA Neutrino anomaly again." and "Is Twilight Sparkle a Rationalist?" Don't seem to mesh well under the same general discussion category, even though I would probably read the community discussion on any of them.

That said, I am open to arguments to establish why the Status Quo is worthwhile as well. And maybe the poll will result in a bunch of people voting for only Main and Discussion.

Please feel free to discuss anything relating to topics or discussion categories or the poll below, and thank you for answering the poll!

New Comment
29 comments, sorted by Click to highlight new comments since:

Not that I'm complaining, but I do wonder how long it's going to be before someone posts a poll asking people what kind of polls they prefer.

The status quo works fine for me. I think it would be a mistake to over-structure things. It isn't (yet) a firehose.

Honestly, I'd rather have Main and Discussion together in one spot and things promoted to the front page from there. My reasoning for this is a couple of things - A. I rarely check main. If I want to read something technical, I'm going to read the sequences anyway. B. Division upon division in places to post probably would lead to atrophy and death of certain sections. For example, I wouldn't expect many people to check a "Pre-Main" thread, largely due to bystander effect.

When I answered the poll I think I said I was unsure in favor of new divisions, but I think I've talked myself out of that. The description of Discussion really just needs to be "everything else".

ETA: What might be helpful is for it to become more of a norm that people use tags like [LINK] in the title when posting in discussion.You could have tags like [OFF TOPIC] or [PRE-MAIN] too

I don't check main because it's harder to check main. You have to go to lesswrong.com, then click main, then click new (if you don't just wanted the promoted ones, which you could access from the main page most of the time anyway).

I don't think everything should go in Discussion because I think there's some value in the 10x Karma that main articles get (and the associated higher level they are judged at). But I think clicking "main" on the main page should take you to all top level posts, not just promoted ones. Promoted articles are already promoted on lesswrong.com, they don't need to be promoted twice.

Yeah, I agree completely. That's another reason I don't check it.

ETA: And there isn't a list with just the titles like there is for Discussion. Makes Discussion a lot easier to browse.

This link gives a list of all posts, Main and Discussion:

http://lesswrong.com/r/all/recentposts

It's very handy.

Wow, thanks. That's great.

How do you reach that without getting linked directly?

I got to it by looking at the form of other URLs and guessing. Indeed, Google informs me that that URL has never been seen on the net before.

Well I'm going to bookmark it to be the site I always go to when I go to LW...

ETA: Click on the Recent Posts on the sidebar. It's apparently clickable.

N.B. That's only posts in Main.

It's only the posts of whichever section you are on when you click it. There's http://lesswrong.com/recentposts for the main section and http://lesswrong.com/r/discussion/recentposts for the discussion section.

Not a bad idea. I only check main once in a while. I use discussion to discuss things I am interested in, main carries too much gravitas to be interesting in that regard.

[-]Jack150

There isn't enough content on Less Wrong to justify further division.

Changing the definition of Discussion would be fine as it was originally put in place in part to be a permanent place for open thread type discussions. Maybe it is an unfair neurotypical assumption but I sort of figure everyone can just figure out what is appropriate for discussion by seeing what gets posted there instead of trying to interpret a definition. Which is to say, I don't think it is a pressing matter but don't see a reason not to have a more expansive definition.

Posts that a typical Less Wrong visitor would like to read are welcome and low quality posts are unwelcome. Topic is only relevant insofar as it predicts whether visitors will want to read it, and it is not nearly as good a predictor of that as writing quality is.

This poll seems to be "Would you like to turn LessWrong into a forum." which I find to be an interesting question. While I've never been involved in forums, the biggest difference I can see is that LW relies on articles to start discussion rather than simply OPs.

I think the LW user interface in general is pretty terrible and ends up causing a lot of problems. The issue you mention is one of many that I think could use some fixing. The question is, who actually is going to fix it, should it turn out that others agree?

I think the LW user interface in general is pretty terrible and ends up causing a lot of problems.

No kidding ! I still have no idea how to efficiently navigate and search long threads. There doesn't seem to be a way.

Trn is the best method I know of for organizing long discussions online. The bright green borders for posts added since the last time you refreshed a page are but a faint glimmer of the glories of trn.

Less Wrong would also be much more usable if it had some sort of advanced search. It would be great to be able to search by some combination of posted by, in reply to, boolean string (text with AND, OR, and NOT connectors) [1], links (the hypertext behind the link), and date range. The more posts and comments we accumulate, the more important advanced search becomes.

At this point, I don't have any way to find old material unless I remember a good bit about it, and it can be hard to find a whole sub-thread, especially if part of it is concealed by a low karma comment or too much depth of comments.

[1] I'm probably missing out on some of the good features in boolean search, but what I describe is at least a start.

LW is an open source project that, in theory, anyone can contribute to. The project is maintained pro bono by Tricycle as a contribution to SIAI. Tricycle maintains a feature tracker somewhere with requests for features/bugfixes, and also accepts code contributions.

On the other hand, it's been a year and a half since I tried, but when I did, I found the LW code pretty impenetrable -- artifacts of its legacy as a Reddit fork.

Small correction: The poll uses the term "Cryogenics" which is the study of creating low temperatures rather than the term "Cryonics" which is what was intended. Not super important, just saying.

[-][anonymous]10

Thanks! I have fixed that. In case anyone wants to know the specific time point, I fixed it after the first 35 results.

"Is Twilight Sparkle a Rationalist?"

I'd upvote an article on this subject (assuming it was well-written, of course). I have no idea what topic such an article would belong to, though, which leads me to believe that we do need to keep some sort of a "general interest" topic around.

I'd rather just have fewer off-topic posts altogether. Reorganising sections won't do that.

Seconding RobertLumley's suggestion. It would make more sense to me if all new posts went to Discussion and some of them got promoted to Main from there. For topic divisions I'd rather use tagging, because it doesn't force readers to check ten different sections of the site for updates, and also because you can have multiple tags on one article.

This already happens to an extent: posts are being moved between Main and Discussion depending on quality and topic. Of late, Main evaporated into a section for selected posts only, everything else goes in Discussion.

I'm happy with the current array of posts we get. I like the occasional xkcd or smbc at about the rate we get them.

"not yet ready or not suitable" seems awfully restrictive. I assumed discussion was basically an area where you can throw stuff up without necessarrilly intending to put anything in main.

edit: I posted this before I read the rest of the comments. As some others have said, the definition of discussion doesn't seem to mesh with what is actually there. If it's current use is fine, the definition shouldn't implicitly prohibit posting something with no intention of making an article for main.

"not yet ready or not suitable" seems awfully restrictive. I assumed discussion was basically an area where you can throw stuff up without necessarrilly intending to put anything in main.

It is. That's the "not suitable" part.