LW is mostly pure-text with no images except for occasional graphs. Why is that so? Are the reasons technical (due to reddit code), cultural (it’s better without images), or historical (it’s always been so)?
Ah, a vote for “it’s better this way”. Why do you prefer pure text? Is it because of the danger of being overrun with cat pictures and blinking gif smileys?
Let’s take that particular image. It covers a huge block that could have been filled by text otherwise and conveys relatively little information accurately. It distrupts my reading completely for a little while and getting back to the nice flow takes cognitive effort.
This moment I’m reading on my phone and the image fills the whole screen.
It is because text can be copy-pasted and composed easily since browsers mostly allow selecting any text (this is more difficult in win apps).
Whereas images cannot be copy pasted as simple (mostly you have to find the URL and copy paste that) and images cannot be composed easily at all (you at least need some pic editor which often doesn’t allow simple copy-paste).
This is the old problem that there is no graphical language. A problem that has evadad GUI designers since the beginning.
Um. In Firefox, right-click on the image, select Copy Image. Looks pretty simple to me. Pretty sure it works the same way in Chrome as well.
This is the old problem that there is no graphical language.
I think you’re missing the point of images. Their advantage is precisely that they are holistic, a gestalt—you’re supposed to take them in whole and not decompose them into elements.
Sure, if you want to construct a sequential narrative out of symbols, images are the wrong medium.
Um. In Firefox, right-click on the image, select Copy Image.
And how do you insert it into a comment?
I think you’re missing the point of images. Their advantage is precisely that they are holistic, a gestalt—you’re supposed to take them in whole and not decompose them into elements.
I’d go with laziness and lack of overt demand. I know that people love graphs and images, but I don’t especially feel the need when writing something, and it’s additional work (one has to make the image somehow, name it, upload it somewhere, create special image syntax, make sure it’s not too big that it’ll spill out of the narrow column allotted articles etc). I can barely bring myself to include images for my own little statistical essays, though I’ve noticed that my more popular essays seem to include more images.
I haven’t tried authoring an article myself, but a quick look now seems to indicate that you can’t upload images, only link to them. This means images must be hosted on third parties, meaning you have to upload it there and if not directly under your control, it’s vulnerable to link rot. It seems like this would be inconvenient.
You can upload images to the LessWrong wiki, and then link them from comments or posts. It’s a bit roundabout, but the feature is there. The question is then, should it be made easier?
That’s very common in online forums (for the server load reasons) but doesn’t seem to stop some forums from being fairly image-heavy. It’s not like there is a shortage of free image-hosting sites.
Yes, I understand the inconvenience argument, but the lack of images at LW is pretty stark.
Do you think more people should include graphics in their posts? Do you think more people should include graphics in their comments? Do you think the image-heavy forums you mention get some benefit from being image-heavy that we would do well to pursue?
I’ll observe that I read your comments on this thread as implicitly recommending more images.
This is of course just my reading, but I figured I’d mention it anyway if you are hesitant to make a recommendation for fear of tearing that fence down in ignorance, on the off chance that I’m not entirely unique here.
I understand where you are coming from (asking why this house is not blue is often perceived as implying that this house should be blue) -- but do you think there’s any way to at least tone down this implication without putting in an explicit disclaimer?
do you think there’s any way to at least tone down this implication without putting in an explicit disclaimer?
Well, if that were my goal, one thing I would try to avoid is getting into a dynamic where I ask people why they avoid X, and then when they provide some reasons I reply with counterarguments.
Also, when articulating possible reasons for avoiding X, I would take some care with the emotional connotations of my wording. This is of course difficult, but one easy way to better approximate it is to describe both the pro-X and anti-X positions using the same kind of language, rather than describing just one and leaving the other unmarked.
More generally, assymetry in how I handle the pro-X and anti-X cases will tend to get read as suggesting partiality; if I want to express impartiality, I would cultivate symmetry.
That said, it’s probably easier to just express my preferences as preferences.
avoid is getting into a dynamic where I ask people why they avoid X, and then when they provide some reasons I reply with counterarguments
I think it’s fine. Reasons that people provide might be strong or might be weak—it’s OK to tap on them to see if they would fall down. I would do the same thing to comments which (potentially) said “Yay images, we need more of them!”.
In general, I would prefer not to anchor the expectations of the thread participants, but not at the price of interference with figuring out of what does the territory actually look like.
describe both the pro-X and anti-X positions using the same kind of language
I didn’t (and still don’t) have a position to describe. Summarizing arguments pro and con seemed premature. This really was just a simple open question without a hidden agenda.
There’s a good chance this is not a “fence”, deliberately designed by some agent with us in mind, but a fallen tree that ended up there by accident/laziness.
There’s a design choice on the part of LessWrong against avatar images. Text is supposed to speak for itself and not be judged by it’s author. Avatar imaging would increase author recognition.
I think I agree with that. I do read author names, but I read them after I read the text usually. I frequently find myself mildly surprised that I’ve just upvoted someone I usually downvote, or vice versa.
Most people are much better at remembering faces than at remembering names. Hacker News also has a lot more people and therefore you will interact with the same person less often.
I am not implying that it should, but to answer your question, because limits on accepted forms of expression are not necessarily a good thing. Not necessarily a bad thing, either.
People already mentioned some pros (e.g. graphs and such help cross the inferential distance) and cons (e.g. images break the mental flow of some people).
I’d note that the short help for comments does not list the Markdown syntax for embedding images in comments, and even the “more comment formatting help” page is not especially clear. That LessWrong cultural encourages folk to write comments before writing Main or Discussion articles makes that fairly relevant.
Some people embed graphics in their articles, and this is seen by many as a good thing. I suspect it’s just individuals choosing not to bother with images.
Reading this comment… I suddenly feel very odd about the fact that I failed to include images in my Neuroscience basics for LessWrongians post, in spite of in a couple places saying “an image might be useful here.” Though the lack of images was partly due to me having trouble finding good ones, so I won’t change it at the moment.
I find it harder to engage in System 2 when there are images around. Heck, even math glyphs usually trip me up. That’s not to say graphics can’t do more good than harm (for example, charts and diagrams can help cross inferential distance quickly, and may serve as useful intuition pumps) but I imagine that more images would mean more reliance on intuition and less on logic, hence less capacity for taking things to analytical extremes. So it could be harmful (given the nature of the site) to introduce more images.
I like my flow. I don’t have anything against images if they are arranged in a way that doesn’t distrupt reading. I’m not sure if lw platform allows for that.
Personally I find the absence of images mostly positive, and apparently helpful for staying in System 2 mode. This is a place where we analyze counter-intuitive propositions, so the absence of images may be critical to the optimal functioning of the community.
That’s not to say images don’t have cognitive advantages (the information content can be processed more quickly than text, e.g.) but they can be distracting, and might actually tend to lead one to be less analytical and concise in the long run. Notice how an image-based meme may seem credible or hard to shoot down even when it represents a strawman argument (or even no argument at all). There’s a reason Chick tracts are a thing.
LW is mostly pure-text with no images except for occasional graphs. Why is that so? Are the reasons technical (due to reddit code), cultural (it’s better without images), or historical (it’s always been so)?
I think most people are unaware that they can include images in comments.
A state of affairs which I hope continues.
Ah, a vote for “it’s better this way”. Why do you prefer pure text? Is it because of the danger of being overrun with cat pictures and blinking gif smileys?
Let’s take that particular image. It covers a huge block that could have been filled by text otherwise and conveys relatively little information accurately. It distrupts my reading completely for a little while and getting back to the nice flow takes cognitive effort.
This moment I’m reading on my phone and the image fills the whole screen.
It is because text can be copy-pasted and composed easily since browsers mostly allow selecting any text (this is more difficult in win apps).
Whereas images cannot be copy pasted as simple (mostly you have to find the URL and copy paste that) and images cannot be composed easily at all (you at least need some pic editor which often doesn’t allow simple copy-paste).
This is the old problem that there is no graphical language. A problem that has evadad GUI designers since the beginning.
Um. In Firefox, right-click on the image, select Copy Image. Looks pretty simple to me. Pretty sure it works the same way in Chrome as well.
I think you’re missing the point of images. Their advantage is precisely that they are holistic, a gestalt—you’re supposed to take them in whole and not decompose them into elements.
Sure, if you want to construct a sequential narrative out of symbols, images are the wrong medium.
And how do you insert it into a comment?
That may be true of some images but not all.
I’d go with laziness and lack of overt demand. I know that people love graphs and images, but I don’t especially feel the need when writing something, and it’s additional work (one has to make the image somehow, name it, upload it somewhere, create special image syntax, make sure it’s not too big that it’ll spill out of the narrow column allotted articles etc). I can barely bring myself to include images for my own little statistical essays, though I’ve noticed that my more popular essays seem to include more images.
I haven’t tried authoring an article myself, but a quick look now seems to indicate that you can’t upload images, only link to them. This means images must be hosted on third parties, meaning you have to upload it there and if not directly under your control, it’s vulnerable to link rot. It seems like this would be inconvenient.
You can upload images to the LessWrong wiki, and then link them from comments or posts. It’s a bit roundabout, but the feature is there. The question is then, should it be made easier?
I haven’t tried it, but just knowing that it requires logging in to the wiki, I know that it’s way too hard and I’ll probably use imgur instead.
That’s very common in online forums (for the server load reasons) but doesn’t seem to stop some forums from being fairly image-heavy. It’s not like there is a shortage of free image-hosting sites.
Yes, I understand the inconvenience argument, but the lack of images at LW is pretty stark.
Do you think more people should include graphics in their posts?
Do you think more people should include graphics in their comments?
Do you think the image-heavy forums you mention get some benefit from being image-heavy that we would do well to pursue?
I am hesitant to put forward a recommendation. I don’t know yet and approach this as the Chesterton’s Fence.
That’s fair.
I’ll observe that I read your comments on this thread as implicitly recommending more images.
This is of course just my reading, but I figured I’d mention it anyway if you are hesitant to make a recommendation for fear of tearing that fence down in ignorance, on the off chance that I’m not entirely unique here.
I understand where you are coming from (asking why this house is not blue is often perceived as implying that this house should be blue) -- but do you think there’s any way to at least tone down this implication without putting in an explicit disclaimer?
Well, if that were my goal, one thing I would try to avoid is getting into a dynamic where I ask people why they avoid X, and then when they provide some reasons I reply with counterarguments.
Another thing I would try to avoid is not questioning comments which seem to support doing X, for example by pointing out that it’s easy to do, but questioning comments which seem to challenge those comments.
Also, when articulating possible reasons for avoiding X, I would take some care with the emotional connotations of my wording. This is of course difficult, but one easy way to better approximate it is to describe both the pro-X and anti-X positions using the same kind of language, rather than describing just one and leaving the other unmarked.
More generally, assymetry in how I handle the pro-X and anti-X cases will tend to get read as suggesting partiality; if I want to express impartiality, I would cultivate symmetry.
That said, it’s probably easier to just express my preferences as preferences.
I think it’s fine. Reasons that people provide might be strong or might be weak—it’s OK to tap on them to see if they would fall down. I would do the same thing to comments which (potentially) said “Yay images, we need more of them!”.
In general, I would prefer not to anchor the expectations of the thread participants, but not at the price of interference with figuring out of what does the territory actually look like.
I didn’t (and still don’t) have a position to describe. Summarizing arguments pro and con seemed premature. This really was just a simple open question without a hidden agenda.
All right.
You could put a “light” disclaimer, like “I’m curious” or “(not that I’m complaining)”.
Edit (post downvote): (not that I’m saying you should have) :D
I read them this way too.
There’s a good chance this is not a “fence”, deliberately designed by some agent with us in mind, but a fallen tree that ended up there by accident/laziness.
There’s a design choice on the part of LessWrong against avatar images. Text is supposed to speak for itself and not be judged by it’s author. Avatar imaging would increase author recognition.
I think I agree with that. I do read author names, but I read them after I read the text usually. I frequently find myself mildly surprised that I’ve just upvoted someone I usually downvote, or vice versa.
And yet names are visually quite distinct. I find authorship much more obvious here than on HN.
Most people are much better at remembering faces than at remembering names. Hacker News also has a lot more people and therefore you will interact with the same person less often.
Why shouldn’t it be?
I am not implying that it should, but to answer your question, because limits on accepted forms of expression are not necessarily a good thing. Not necessarily a bad thing, either.
People already mentioned some pros (e.g. graphs and such help cross the inferential distance) and cons (e.g. images break the mental flow of some people).
It doesn’t feel like a limit to me, just something that very seldom occurs to me to do because I very seldom have any use for it.
(I sometimes link to images—maybe next time I’ll consider including them directly in the comment.)
I’d note that the short help for comments does not list the Markdown syntax for embedding images in comments, and even the “more comment formatting help” page is not especially clear. That LessWrong cultural encourages folk to write comments before writing Main or Discussion articles makes that fairly relevant.
Some people embed graphics in their articles, and this is seen by many as a good thing. I suspect it’s just individuals choosing not to bother with images.
Reading this comment… I suddenly feel very odd about the fact that I failed to include images in my Neuroscience basics for LessWrongians post, in spite of in a couple places saying “an image might be useful here.” Though the lack of images was partly due to me having trouble finding good ones, so I won’t change it at the moment.
I find it harder to engage in System 2 when there are images around. Heck, even math glyphs usually trip me up. That’s not to say graphics can’t do more good than harm (for example, charts and diagrams can help cross inferential distance quickly, and may serve as useful intuition pumps) but I imagine that more images would mean more reliance on intuition and less on logic, hence less capacity for taking things to analytical extremes. So it could be harmful (given the nature of the site) to introduce more images.
I like my flow. I don’t have anything against images if they are arranged in a way that doesn’t distrupt reading. I’m not sure if lw platform allows for that.
Personally I find the absence of images mostly positive, and apparently helpful for staying in System 2 mode. This is a place where we analyze counter-intuitive propositions, so the absence of images may be critical to the optimal functioning of the community.
That’s not to say images don’t have cognitive advantages (the information content can be processed more quickly than text, e.g.) but they can be distracting, and might actually tend to lead one to be less analytical and concise in the long run. Notice how an image-based meme may seem credible or hard to shoot down even when it represents a strawman argument (or even no argument at all). There’s a reason Chick tracts are a thing.