Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

In response to comment by Jiro on Feedback on LW 2.0
Comment author: gjm 12 October 2017 04:30:59PM 1 point [-]

It is fundamentally designed around a bad idea.

That seems rather extreme. What specific bad idea do you mean?

(Context here suggests that it's something like "the idea that typographical choices for LW2 should match those for the web as a whole", but even if LW2's design makes that assumption and even if it's a bad assumption it doesn't seem fundamental enough to justify your last paragraph.)

In response to comment by gjm on Feedback on LW 2.0
Comment author: Jiro 12 October 2017 08:45:16PM *  0 points [-]

Context here suggests that it's something like "the idea that typographical choices for LW2 should match those for the web as a whole"

The idea that the study of typographical choices for the web is a mature science whose (nontrivial) recommendations can all be taken at face value.

In response to comment by Viliam on Feedback on LW 2.0
Comment author: adamzerner 02 October 2017 12:52:22AM *  1 point [-]

The font seems to be bigger, and the spaces between lines also seem bigger. "On the web, the op­ti­mal size is 15–25 pixels.". "line spacing 120–145% of the point size"

Looks like LW 2.0 is using a 20px font size, and 25px line height, which is in range of what is recommended.

The font is blurry, and gray instead of black.

NN Group supports what you're saying, although the text on LW 2.0 looks plenty dark to me.

The comments don't have boxes, so it is difficult to see where one comment ends and another comment starts, or which comment replies to which one.

Agreed.

Comment author: Jiro 11 October 2017 07:35:54PM *  0 points [-]

Looks like LW 2.0 is using a 20px font size, and 25px line height, which is in range of what is recommended.

Is "what was recommended" similar to "mistakes were made"? It blames it on someone else, while leaving the "someone else" unnamed.

Existing recommendations about text size (and particularly, about not fitting too much text on a line) do not consider that Lesswrong has a different usage pattern than most sites. There are references dating back to 1971, but I can't figure out if any scientific studies were actually conducted at the time to determine this, and at any rate, printed text is not the web.

Also, beware of using some recommendation just because it's easy to measure.

This is basically breaking the site in order to fit "recommendations". LW 2.0 is bad, and everyone involved should feel bad. It is fundamentally designed around a bad idea.

In response to comment by gjm on Feedback on LW 2.0
Comment author: Lumifer 03 October 2017 03:52:30PM 0 points [-]

If you scroll down the front page until you see the "Recent comments" section, there is a small link to "see all comments" underneath the section heading.

In response to comment by Lumifer on Feedback on LW 2.0
Comment author: Jiro 03 October 2017 07:12:24PM 1 point [-]

Not only is that obscure, it's shows the comments as abbreviated and doesn't let you reply to them. It's not so much as a list of comments as it is a list of things that you can use as comments if you take a couple of extra steps.

In response to Feedback on LW 2.0
Comment author: Jiro 03 October 2017 07:03:07PM *  1 point [-]

Trying the site right now from work using Chrome, Firefox, and IE 11:

  • Firefox fails to load images for the magnifying glass used for "search" at the top of the page, and the "expand _ less" and "expand _ more" arrows. Otherwise it mostly works.
  • Clicking the capitalized "LOG IN" on the home page does nothing on IE.
  • On IE (but not Firefox) going to Codex briefly puts up the actual page, then it disappears and switches to a different page "Sorry, we couldn't find what you were looking for." The location bar still shows https://www.lesserwrong.com/codex. This page has a mixed case "Log in" at the top.
  • Going to a featured post doesn't make it actually disappear, but it starts as a properly formatted post (title centered, uppercase LOGIN) then switches to one improperly formatted with the title on the left and the mixed case "Log in". It does not show comments, instead endlessly throbbing the o o o at the bottom of the page.
  • Clicking the mixed case "Log in" produces the normal login box, which lets me type in a username and password and click "SIGN IN", except that it is too far to the right (going off the page if I don't stretch it) and I can't actually click the "SIGN IN". When I hover over it I get a slashed circle and clicking it produces no response.
  • Neither Firefox nor IE produces the "hi, welcome to lesswrong 2.0" at the bottom right of the page, or shows a red number in the conversations icon there.
  • Chrome has no problems.

Only Firefox is restricting any cookies (and I already unrestricted the one I need to log in).

In response to comment by turchin on Feedback on LW 2.0
Comment author: gjm 03 October 2017 02:56:24PM 0 points [-]

It may be worth noting that "6 downvotes" need not mean that 6 people downvoted you. LW2 has "weighted voting" which means that the number of points your upvotes/downvotes change the victim's karma by depends on your own karma level. So maybe you were downvoted twice by weight-3 users, or three times by weight-2 users; in any case, losing 6 points probably corresponds to <6 downvotes.

In response to comment by gjm on Feedback on LW 2.0
Comment author: Jiro 03 October 2017 06:34:32PM 5 points [-]

Weighted karma is a system that heavily violates user expectations and is a bad idea for that reason alone.

Comment author: Jiro 22 September 2017 08:29:39PM 1 point [-]

I am unable to use this open beta because of the problem I describe here.

Comment author: Benito 20 September 2017 07:55:24PM *  0 points [-]

Hi! I just checked on Firefox, and the login dialog box opened for me. If you still have this issue, next time you try to log in (open beta will happen by 4pm today) please ping us in the intercom (bottom right-hand corner of the lesserwrong page), and let us know what browser version you're using.

If your intercom doesn't work, let me know here.

Comment author: Jiro 21 September 2017 05:14:08PM 0 points [-]

It seems to have been a cookie problem so I got it working.

However, I ended up with two logins here. One I never used much, and the other is this one. Naturally, lesserwrong decided that the one that it was going to associate with my email address is the one that I never used much.

I'd like to get "Jiro" on lesserwrong, but I can't, since changing password is a per-email thing and it changes the password of the other login. Could you please fix this?

Comment author: Raemon 20 September 2017 07:00:56PM 0 points [-]

Not 100% I understand your description, but currently the expected behavior when you attempt to login (if not already a part of the beta) is nothing happening when you click "submit" (although in the browser console there'll be an error message)

This is simply because we haven't gotten to that yet, but it should be something we make sure to fix before the open-beta launch later today so people have a clear sense of whether it's working.

Comment author: Jiro 20 September 2017 07:35:05PM 0 points [-]

And the expected behavior when using IE or Firefox is that you can't even get to the login screen? I find that unlikely.

Comment author: SaidAchmiz 19 September 2017 08:59:28PM 1 point [-]

A private beta has been ongoing, clearly.

Comment author: Jiro 20 September 2017 04:02:41PM 0 points [-]

That can't explain it, unless the private beta is accessed by going somewhere other than lesserwrong.com. The site isn't going to know that someone is a participant in the private beta until they've logged in. And the problems I described happen prior to logging in.

Comment author: Elo 19 September 2017 07:16:01PM 0 points [-]

I believe that is on purpose. Login is not open yet.

Comment author: Jiro 19 September 2017 08:34:04PM 0 points [-]

People are clearly posting things there that postdate the DB import, so they must be logging in. Also, that doesn't explain it working better on Chrome than on other browsers.

View more: Next