I also do not like this. HPMOR.com was a much better HPMOR reading experience than HPMOR-on-LW is. There’s hardly any comparison. This change replaces a better experience with a worse one.
Is this a subtle way of discouraging people from reading HPMOR? (Perhaps a good idea, come to think of it…)
FWIW, I also thought HPMOR.com was a better reading experience for HPMOR than current LW is. LW is optimized for a bunch of stuff that HPMOR.com isn’t optimized for. I hope we can get the old site back up.
While I’m feeling a bit miffed that I’m getting a vibe that people at mad at the LessWrong team for a thing that’s, like, totally not our fault, I do wanna go on record as “I actively prefer the LessWrong version.”
It scores at least as well for me on core-reading-experience. I don’t have any issue reading the text. I like that the https://www.lesswrong.com/hpmor page has some pretty pictures, has a progress bar so you can see all your reading progress at once, and you can easily see the spots you haven’t read yet (both in the simple bar at the top and in the table-of-contents-ish thing on the rest of the page), and
(FYI it’s not fully true you need to click back to see the list of posts – if you hover over the book title at the top of a post page, you see a dropdown list of the posts in that book. It’s true you don’t get the entire list of posts but if you’re navigating to different books I think seeing the higher-order structure is kinda useful anyway)
And also obviously it’s bad to not have authors notes. (Even if/when hpmor.com goes live again it’s plausible we should import those too and make a final sequence that’s just those)
I am not mad at the LessWrong team. The reason I framed the title as an accusation was because I figured it was likely since I was sent to your website that you were responsible in some way, or at least were aware of what was going on. I now understand I was mistaken.
As for “improvements” if/when hpmor.com comes back up, I would like to note that I am against them, for the same reasons described in the post. I don’t think it’s obvious at all that some change to the old site would not be bad, at least from the perspective of people who prefer the old site.
I meant the improvements for lesswrong.com’s general functionality, not hpmor.com (not sure if we were on the same page about that). Like, one of lesswrong’s main jobs is to be good at longform reading, and if there are ways it can be straightforwardly improved at that we should do so.
LW is optimized for a bunch of stuff that HPMOR.com isn’t optimized for.
Yes, exactly. LW.com is certainly better at being LW.com than HPMOR.com would be! Single-purpose tools are often better for that purpose than things designed for other purposes entirely…
I hope we can get the old site back up.
This probably won’t be needed, but just in case it is—I’d be happy to host it.
I’ve pulled down a static mirror from archive.org and modified a couple pieces which depended on server-side implementation to use a javascript version, most notably the chapter dropdown. In the unlikely case it’s useful, ping me.
I disagree that the reading experience is notably worse, let alone much worse. I don’t see any mistakes that would make the LW version difficult to read. Things like font size, line length, line height, contrast, etc. all seem to follow best practices.
That said, I am bummed by the switch for nostalgic reasons, but only very slightly. I don’t think I’d pay more than $10 to keep the old version.
I don’t see any mistakes that would make the LW version difficult to read.
On LW, the body text font is not fully opaque (color: #000, opacity: 0.87 in CSS), which (given how text rendering works in browsers) makes it harder to read than HPMOR.com’s text (color: #222; opacity: 1.0)
To navigate to a specific chapter (not just “the next chapter” or “the previous chapter”), on HPMOR.com you had only to select the chapter from a drop-down menu; but on LW, you have to click on the “HPMOR” link at the top, then scroll down
This “HPMOR” link often doesn’t even appear (a bug?), leaving you with no apparent way to navigate to the contents
HPMOR.com had a widget that let you toggle between a narrow and a full-width body text column, for readers who preferred different layouts; LW has no such thing
HPMOR.com had a widget that let you toggle between light and dark mode; on LW, the dark mode toggle is hidden inside a submenu (I bet many people don’t even know it exists!)
HPMOR.com had a Fan Art page which was linked from the top navigation bar; no such thing exists on the LW version (is the fan art archived somewhere other than archive.org?)
The “More Info” page on HPMOR.com, which had a bunch of interesting links and info, seems also to be missing from the LW version
(These deficiencies are, in my experience, quite typical of such content migrations.)
P.S. I said that the change “replaces a better experience with a worse one”, which is a much more general claim than any particulars of typography. But notice that you assumed that I was talking about only the narrow typographic sort of problem! This, in my experience, is also typical. Understandable—but typical.
IMO those issues are all very minor, even when summed.
color: #000, opacity: 0.87 on a #fff background works out to a hex of #212121, which has a contrast ratio of 16.10. This is well above the WCAG’s minimum requirement of 7 for body text and is rated as five stars and “super” by https://coolors.co.
I’d imagine that there isn’t a large user need where they start off on one chapter and need to quickly and easily navigate to a new chapter that isn’t the next chapter. When a user has a specific chapter they want to read I think they’re probably coming from elsewhere. But even if they weren’t, clicking the “HPMOR” link and then scrolling down isn’t too difficult.
My impression is that other sites that are also used primarily for reading also don’t provide this option. Medium is an example and from what I gather is considered to provide users with a great reading experience. To me, this points pretty strongly towards it being at most a minor issue.
Same as the above. Plus I bet that users who care a lot about dark mode have browser extensions installed that they can reach for when needed.
I could see an argument for the lack of fan art being more than “very minor” but personally I’m very bearish on it.
IMO those issues are all very minor, even when summed.
Is that relevant? Imagine that we were discussing the replacement of a ramp with stairs. This has a very minor effect on my experience—is that enough to conclude the change was benign?
color: #000, opacity: 0.87 on a #fff background works out to a hex of #212121, which has a contrast ratio of 16.10.
You absolutely did not understand what I wrote, which had not the slightest connection to any contrast ratios. Please reread.
I’d imagine that there isn’t a large user need where they start off on one chapter and need to quickly and easily navigate to a new chapter that isn’t the next chapter.
Why would you imagine this? I have done this regularly when using HPMOR.com.
Medium is an example and from what I gather is considered to provide users with a great reading experience.
Considered by whom? Medium offers one of the worse reading rexperiences on the web.
I bet that users who care a lot about dark mode have browser extensions installed that they can reach for when needed.
This is not even remotely true. Such browser extensions do not (and, given the way that browsers work, cannot) work well (or, increasingly, at all). In my experience, most people who “care a lot about dark mode” in fact just do not use websites that don’t have a dark mode. (Also, dark mode is, at this point, common enough, and the implementation techniques well enough known, that users simply expect it from a well-designed website.)
I could see an argument for the lack of fan art being more than “very minor” but personally I’m very bearish on it.
This boils down to “I, personally, don’t care about this, therefore it doesn’t matter”, which does not work as a rebuttal of a claim that not having the thing in question is a worse user experience than having it.
I also do not like this. HPMOR.com was a much better HPMOR reading experience than HPMOR-on-LW is. There’s hardly any comparison. This change replaces a better experience with a worse one.
Is this a subtle way of discouraging people from reading HPMOR? (Perhaps a good idea, come to think of it…)
FWIW, I also thought HPMOR.com was a better reading experience for HPMOR than current LW is. LW is optimized for a bunch of stuff that HPMOR.com isn’t optimized for. I hope we can get the old site back up.
While I’m feeling a bit miffed that I’m getting a vibe that people at mad at the LessWrong team for a thing that’s, like, totally not our fault, I do wanna go on record as “I actively prefer the LessWrong version.”
It scores at least as well for me on core-reading-experience. I don’t have any issue reading the text. I like that the https://www.lesswrong.com/hpmor page has some pretty pictures, has a progress bar so you can see all your reading progress at once, and you can easily see the spots you haven’t read yet (both in the simple bar at the top and in the table-of-contents-ish thing on the rest of the page), and
(FYI it’s not fully true you need to click back to see the list of posts – if you hover over the book title at the top of a post page, you see a dropdown list of the posts in that book. It’s true you don’t get the entire list of posts but if you’re navigating to different books I think seeing the higher-order structure is kinda useful anyway)
There are a couple areas that seem like they could be improved (it does seem reasonable to make it one-click to see the full library page, maybe if you click the book title at the top it should take you to https://www.lesswrong.com/hpmor, to the relevant anchor-title, instead of https://www.lesswrong.com/s/EBuZhwCrYuJGp7ax4.
And also obviously it’s bad to not have authors notes. (Even if/when hpmor.com goes live again it’s plausible we should import those too and make a final sequence that’s just those)
I am not mad at the LessWrong team. The reason I framed the title as an accusation was because I figured it was likely since I was sent to your website that you were responsible in some way, or at least were aware of what was going on. I now understand I was mistaken.
As for “improvements” if/when hpmor.com comes back up, I would like to note that I am against them, for the same reasons described in the post. I don’t think it’s obvious at all that some change to the old site would not be bad, at least from the perspective of people who prefer the old site.
I meant the improvements for lesswrong.com’s general functionality, not hpmor.com (not sure if we were on the same page about that). Like, one of lesswrong’s main jobs is to be good at longform reading, and if there are ways it can be straightforwardly improved at that we should do so.
We were not on the same page. I thought you were suggesting changes to the new re-hosted version of hpmor.com. Thanks for clarifying.
Yes, exactly. LW.com is certainly better at being LW.com than HPMOR.com would be! Single-purpose tools are often better for that purpose than things designed for other purposes entirely…
This probably won’t be needed, but just in case it is—I’d be happy to host it.
I’ve pulled down a static mirror from archive.org and modified a couple pieces which depended on server-side implementation to use a javascript version, most notably the chapter dropdown. In the unlikely case it’s useful, ping me.
I disagree that the reading experience is notably worse, let alone much worse. I don’t see any mistakes that would make the LW version difficult to read. Things like font size, line length, line height, contrast, etc. all seem to follow best practices.
That said, I am bummed by the switch for nostalgic reasons, but only very slightly. I don’t think I’d pay more than $10 to keep the old version.
On LW, the body text font is not fully opaque (
color: #000, opacity: 0.87in CSS), which (given how text rendering works in browsers) makes it harder to read than HPMOR.com’s text (color: #222; opacity: 1.0)ETA: A demonstration of why this is bad
To navigate to a specific chapter (not just “the next chapter” or “the previous chapter”), on HPMOR.com you had only to select the chapter from a drop-down menu; but on LW, you have to click on the “HPMOR” link at the top, then scroll down
This “HPMOR” link often doesn’t even appear (a bug?), leaving you with no apparent way to navigate to the contents
HPMOR.com had a widget that let you toggle between a narrow and a full-width body text column, for readers who preferred different layouts; LW has no such thing
HPMOR.com had a widget that let you toggle between light and dark mode; on LW, the dark mode toggle is hidden inside a submenu (I bet many people don’t even know it exists!)
HPMOR.com had a Fan Art page which was linked from the top navigation bar; no such thing exists on the LW version (is the fan art archived somewhere other than archive.org?)
The “More Info” page on HPMOR.com, which had a bunch of interesting links and info, seems also to be missing from the LW version
(These deficiencies are, in my experience, quite typical of such content migrations.)
P.S. I said that the change “replaces a better experience with a worse one”, which is a much more general claim than any particulars of typography. But notice that you assumed that I was talking about only the narrow typographic sort of problem! This, in my experience, is also typical. Understandable—but typical.
IMO those issues are all very minor, even when summed.
color: #000, opacity: 0.87on a#fffbackground works out to a hex of#212121, which has a contrast ratio of 16.10. This is well above the WCAG’s minimum requirement of 7 for body text and is rated as five stars and “super” by https://coolors.co.I’d imagine that there isn’t a large user need where they start off on one chapter and need to quickly and easily navigate to a new chapter that isn’t the next chapter. When a user has a specific chapter they want to read I think they’re probably coming from elsewhere. But even if they weren’t, clicking the “HPMOR” link and then scrolling down isn’t too difficult.
My impression is that other sites that are also used primarily for reading also don’t provide this option. Medium is an example and from what I gather is considered to provide users with a great reading experience. To me, this points pretty strongly towards it being at most a minor issue.
Same as the above. Plus I bet that users who care a lot about dark mode have browser extensions installed that they can reach for when needed.
I could see an argument for the lack of fan art being more than “very minor” but personally I’m very bearish on it.
Same as the above.
Is that relevant? Imagine that we were discussing the replacement of a ramp with stairs. This has a very minor effect on my experience—is that enough to conclude the change was benign?
You absolutely did not understand what I wrote, which had not the slightest connection to any contrast ratios. Please reread.
Why would you imagine this? I have done this regularly when using HPMOR.com.
Considered by whom? Medium offers one of the worse reading rexperiences on the web.
This is not even remotely true. Such browser extensions do not (and, given the way that browsers work, cannot) work well (or, increasingly, at all). In my experience, most people who “care a lot about dark mode” in fact just do not use websites that don’t have a dark mode. (Also, dark mode is, at this point, common enough, and the implementation techniques well enough known, that users simply expect it from a well-designed website.)
This boils down to “I, personally, don’t care about this, therefore it doesn’t matter”, which does not work as a rebuttal of a claim that not having the thing in question is a worse user experience than having it.