Cool, good changes. Comments:
https://www.lesserwrong.com/sequences/SqFbMbtxGybdS2gRs (click on last post in listing, notice that it has a “next post” link, which takes you to a post in the middle of the listing)
Re 1: Eventually the goal is to have the section at the top keep track of the sequences you are currently in the process of reading, with an easily-available button to remove the relevant sequence from your reading queue, and then to have another section that has some form of personalized recommendations that change relatively frequently and help you discover old content. Which I expect will make it less likely that people will want to hide that section.
I am generally not too keen on hiding the recommendations section completely, since I do think it being visible to everyone, and everyone knowing that it's visible to everyone, establishes a lot of common knowledge and expectations of what kind of content is considered canon (similarly to how the lack of a separate curated section resulted in a lack of common knowledge and made people feel a lot less excited about getting into curated).
FWIW, I don't really parse the sequences section on the front page as "common knowledge about the canon", and my reaction to the change was "oh, good, the large block of noise above the actual content got smaller." Perhaps this is different for people who haven't read them already, but seeing a list of things I've read already that never goes away is really boring and I have long since tuned it out -- my eyeballs skate right over it.
(To be clear, I do value the idea that there should be a canon, and I think content-wise you've probably made good choices about the canon... but I also think that having a block of static stuff that sits above the dynamic "place where people are actually doing things" will just tend to cause people to tune-out the stuff the way I have. Honestly if you're familiar with "ad blindness", it's basically like that.)
Yeah, I definitely agree that the sequences block of content was too large for logged-in users. Which is why the goal is eventually to have dynamic content at the top that is both interesting to the user, and that highlights the canon of the community (I.e. that’s the direction we are going in with this update. With us making it significantly smaller in height, and planning to change it every week. Eventually it would only show you content that you haven’t read yet, and maybe sometimes (but marked as such) a reminder of a sequence you read a long long time ago and might want to read again).
You don’t think I should be able to hide the recommendations sections if I don’t want to see it? That seems like a rather controlling attitude to take toward individual users’ preferences… :(
(And the “currently in the process of reading” thing isn’t a solution either. I am not in the process of reading any of these things. I’ve either already read them or I don’t want to read them. I also don’t want to receive recommendations, or to be helped to discover old content! I just want to hide that part of the UI, because it is simply not relevant to my interests.)
Edit: And I definitely don’t want personalized reading recommendations. That’s just egregious. :\
I think you overinterpreted “personalized”. With that I mostly meant “showing you content you haven’t read yet”.
Fair enough. Well, but then I am confused—how do you square that with “ personalized recommendations that change relatively frequently and help you discover old content”? In any case that sounds like a thing I just don’t want.
Look, the fact is, you can design that feature however you like—but the details of your design matter a great deal less, and impact my experience of the site a great deal less (i.e., less negatively), if I can just hide the darn thing!
Please, seriously consider whether foisting something like this on users, when they don’t want it, is good design.
(Of course, ultimately, this is not a problem for me, in practice, both because I can simply use an ad-blocker to block that element, and because I use GW rather than going to lesserwrong.com directly. So if your position on this is along the lines of “Said Achmiz, and other people who share his preferences, should use ad-blockers / go to greaterwrong.com / etc.”—well, fair enough!)
If we show you a sequence or a post that we think you haven't read yet, and you don't click on it for a significant amount of time (not sure exactly how long, would want to experiment with that), then it seems reasonable to conclude that you've either already read it or don't want to read it, so we would show you the next best thing we think you haven't read yet (and we would probably also have a button that says "I've already read this" somewhere there, to allow you to force the content to cycle).
I do also think that for people who have read enough of the sequences and the other core content on the page, it should be possible to hide the thing, and maybe replace it with a small badge next to your username that reminds you that you are one of the users who has read the Sequences and the Codex, etc. And then to make the section reappear again when we add a new piece of content to the canon (e.g. Inadequate Equilibria).
I am more hesitant to allow people who haven't read the canon to dismiss the canon section, since I do think it makes sense to nudge users towards reading the core content if they haven't done so in one form or another. Presenting users with the explicit option to hide that section sets a tone for the page that it is encouraged to just not engage with that part of the site (if you haven't done so in the past), which strikes me as a pretty slippery slope to go down.
(Obviously there are problems with determining whether someone has read the core content, which is definitely something I don't yet have a perfect solution for. We do know which pages any user visited while logged in, even from the old LessWrong, but people might have read an ebook version of the content, or read the content while logged out, etc. So that's a problem that we will still need to figure out.)
There is a broader vision for the page here, that I haven't yet really gotten around to writing down, which has to do with the fact that 90% of the best content for most users, is older than 6 months. And that content discovery and scholarship should become a core aspect of the site. I expect that in the long-run we will focus quite a bit of our efforts on allowing users to discover old content, and help them learn the content in the canon, and generally focus on a bunch of scholarship-based features.
Seconding Elo; this “badge” thing sounds unfortunate and I don’t want it. Private “achievements” also seem bad (for different reasons).
Edit: I do agree with this:
90% of the best content for most users, is older than 6 months. And that content discovery and scholarship should become a core aspect of the site. I expect that in the long-run we will focus quite a bit of our efforts on allowing users to discover old content, and help them learn the content in the canon, and generally focus on a bunch of scholarship-based features.
But I think that gamification / badges / achievements / nagging / obtrusive, unremovable UI elements / etc., are just not the way to go about this.
I don't want that badge system. I don't want a badge. It makes an "us and them". The most I'd take is "been here a while" which could be taken as a mishonour.
Is this a concept that the community can discuss in a separate post?
Sure, seems fine to open a separate meta-post about this, though in this case I wasn't actually thinking of a public badge, but instead a private badge on your own profile (I guess achievement would fit it better, but I was just thinking of the visual style of a badge).
I do actually feel pretty positive towards public badges (and it's been requested by Lukeprog, so discussing that seems good to me. Though it looks like possibly implementing that is still a while off, so unsure how actionable that discussion would be.)
On the new front page, the things that should say "Show Excerpt" say "Show Except" instead. (Is there a better place for this sort of minor bug report? The Intercom widget is no longer showing.)
I really appreciate this update and I'm strongly in favour of all of the changes here. I'll just mention one small issue, users may be confused that the interface is different when they log in. I still think that the changes are clearly net good, but I wonder if there is an easy way to mitigate this.
It seems relatively common to me that a site changes if the user logs in, so I think the confusion isn't too bad. But in the long-run, the top of the frontpage will probably be in significant parts filled with more personalized recommendations, which will make it feel more natural that it changes when you log in.
We finally got around to a long-postponed update to the frontpage, where logged in users see content that is hopefully more relevant to them.
Some notes: