5 ms·

A few months ago I had a discussion here with someone regarding good/bad design, what was personal preference and what could be considered objective judgment.

A few months ago I had a discussion here with someone regarding good/bad design, what was personal preference and what could be considered objective judgment. They mentioned something about a design philosophy in which there should never be more than one button, or if there were two then one should be the main button that people really should click on. It seems there is an effort to raise a whole school of new designers on the idea that there isn't that much of a difference between software users and chicken in a farm.

contravariant 5y ago
Ironically if there's one thing utterly useless as a way of conveying information it would be a singular button.
marginalia_nu 5y ago
I think there's a middle ground. Providing clear options and deliberate choices can be done through not having a thousand options and buttons. Wikipedia is a clear example of not taking this into consideration. There's a hundred links mixed with the text, and buttons and doodads everywhere around it, which makes it very clear to actually think about what you want to do. If you compare Encarta '96 with Wikipedia: <https://memex.marginalia.nu/pics/links/encarta.png https://memex.marginalia.nu/pics/links/encarta.png> <https://memex.marginalia.nu/pics/links/wikipedia.png https://memex.marginalia.nu/pics/links/wikipedia.png> It's clearly very different. The Encarta design is a lot more quiet. If you're reading, then you are reading. There are sparse inline links, but nothing near the shotgun linkedness of Wikipedia. I've experimented with stripping down Wikipedia, producing this: <https://encyclopedia.marginalia.nu/wiki/Plato https://encyclopedia.marginalia.nu/wiki/Plato> It's definitely several steps too far (but taking things too far is the best way to get the feel for an idea), but I find that much easier to read, because the choices are moved out of the text and into the bottom of the text. That is, you read first, and decide what to do next once you are finished reading.
- pferde 5y ago
  All I see on the Marginalia Plato article is a site that forces an uncomfortably narrow article width on me, while eschewing almost all the advantages of hypertext in the name of some ill-conceived notion of "quieter reading". I'm not saying it's all bad, though - it's something that could be an option offered to the unfortunate people with attention deficit disorder. The forced text width, however, is inexcusable.
  - marginalia_nu 5y ago
    Yeah I think that's probably from my own habit of never using full screen windows. Though I'm not sure how to deal with the fact that it's very hard to read long lines of text. Most printed books, newspapers, and so forth use that width (max 60ch) specifically because it's hard to follow lines longer than that. I chose to follow book guidelines specifically because that format has been largely stable for some thousand years and is at this point the gold standard for legibility. Usually when things stagnate for that long, it's not because nobody has thought to try something different in all that time. Maybe multiple columns could be an option, I guess. Not sure how to do that well with varying paragraph sizes, though.
    - morsch 5y ago
      On Firefox mobile, it's more like 11 characters wide. Pretty much unusable.
      - marginalia_nu 5y ago
        Hmm, strange. It doesn't have these proportions? <https://memex.marginalia.nu/pics/encyclopedia-mobile.png https://memex.marginalia.nu/pics/encyclopedia-mobile.png> I sadly can't diagnose since my phone isn't powerful enough to run Firefox. Looks as intended on Firefox desktop with emulation mode, though.
        morsch 5y ago
        Huh. It looks fine now. Did you change anything?
- maest 5y ago
  Not much to add except wanted to say that I enjoyed becoming aware of the impact Wikipedia's design decision had on me, as a reader. Seeing the encarta/marginalia version vs the wikipedia one really drives home the point. It's interesting how the design had a strong impact on my experience, and yet I never really noticed it before. I feel a bit like a fish suddenly aware of the water they're in.
  - mrob 5y ago
    I was confused by the claim that the Encarta version is quieter until I read that it was because of the inline links. I thought the Wikipedia version was quieter because it doesn't have a distracting background image. I think this confusion is evidence that inline links are actually not distracting. I don't feel I have to make a decision whenever I see one because I know they'll still be there later.
    - marginalia_nu 5y ago
      Maybe it's individual variations? I struggle immensely reading text with inline links. I immediately start skimming and my attention is strongly drawn to the linked words.
      - dahfizz 5y ago
        Maybe you can add custom CSS rules that make links the same color? I also much prefer the wikipedia interface to the alternatives.
        marginalia_nu 5y ago
        This pretty much is a custom CSS. I was also more than a bit annoyed by the huge downloads and 30 second page loads.
- wongarsu 5y ago
  The Encarta version is much better if you want to learn about Plato, while Wikipedia makes it much easier to learn about thing related to Plato. The notion that the relation to other things is at least as important as the thing itself permeates Wikipedia. Many smaller articles, like e.g. Differentiable Programming [1] probably link to more topics than the article itself has words. I think a "stripped down" Wikipedia is an experiment worth trying, but it would serve a subtly different purpose, with some use cases being left behind and some being better served. 1: https://en.wikipedia.org/wiki/Differentiable_programming https://en.wikipedia.org/wiki/Differentiable_programming
  - marginalia_nu 5y ago
    I think the big problem with Wikipedia's design is that it caters to too many audiences with disparate needs using the same interface, both reading and navigating and cross-referencing and editing. I think perhaps having multiple views for different needs would be a way forward.
    - tablespoon 5y ago
      > I think the big problem with Wikipedia's design is that it caters to too many audiences with disparate needs using the same interface, both reading and navigating and cross-referencing and editing. > I think perhaps having multiple views for different needs would be a way forward. I think what this thread is missing is that Encarta and Wikipedia are vastly different products. If Wikipedia had multiple views for different needs it would literally die, because it's community edited. If its UI encouraged staying withing a narrow lane, then it would have even more problems with getting readers to cross over to being editors.
  - spookybones 5y ago
    I read something once that when you come across a link, your mind subconsciously hesitates whether or not to open it and that these thoughts detract from your focus on the material at hand. I often wonder if it would be more beneficial for websites to transition to footnotes with all pertinent links at the bottom of the page.
- MagnumOpus 5y ago
  If you want to strip away the MediaWiki adornments and editing/interactivity tools from Wikipedia, you just have to go to the mobile page: https://en.m.wikipedia.org/wiki/Plato https://en.m.wikipedia.org/wiki/Plato It is still consumption-oriented and stripped down, but retains far more non-textual information than your experiment.
martin_a 5y ago
> one should be the main button that people really should click on I don't think there's something wrong with CTAs. You can misuse/abuse this technology as you can with every, but providing a structured way to achieve a desired result is not bad per se. It helps build muscle/"brain" memory and can help you perform your task better/quicker.
scetron 5y ago
There was an item posted here this week with the same exact conversation. It's honestly a bit eye opening with the way UX is going. I am all for sensible defaults, but being able to tweak things here and there should really be an option, and not one seen as a failure by the designers...
zwkrt 5y ago
Agreed, but the reason this happens is not in a bubble. "User interface design" is not a natural category of design. People in the field will point to "The Design of Everyday Things" as some sort of progenitor of the field, but if you read it, it is kind of just a pop-psychology book. The truth is (IMO) that UX is more closely related to marketing than it is to technical design work. The entire point of having a UX designer is to get marketing speak and advertising techniques right into the heart of the engineering process. If engineers just built tools in the way that was the most sensible to them, they would have too many options and give the user far too much control. Ostensibly the reason given why this is bad is that it is 'confusing' for the user, but the reality is that it allows the end user to use the product in ways that do not maximize the profit. Even worse, users might use the product in a way that does not reflect the spirit of The Brand. It is telling that UX revolves around creating user stories, because this is exactly how advertising is done; “Martha, a woman in her mid thirties, is trying to balance home life with her need to efficiently walk her dog…”
musesum 5y ago
> there should never be more than one button ... which you have to look at, because the position and function changes. There is no muscle memory with dynamic UI. I recall a visually impaired woman who couldn't hail an Uber. I could tell she was very fast with the old UI. But, the button had shifted. She was going to call her Husband to drive over from Oakland to pick her up in San Francisco. Instead, I showed her where the button was.
lariati 5y ago
I think it is mostly a trade off. Using modern design philosophy you won't risk getting a really terrible UI but the trade off is you will not get something new and innovative. I am sure there are all kinds of issues with the scale of complexity also. It would seem nearly impossible for a designer to come up with a good innovative UI if they are not an expert in using the system themselves so the chicken in a farm ends up being the least bad option.

contravariant 5y ago

Ironically if there's one thing utterly useless as a way of conveying information it would be a singular button.

marginalia_nu 5y ago

I think there's a middle ground. Providing clear options and deliberate choices can be done through not having a thousand options and buttons. Wikipedia is a clear example of not taking this into consideration. There's a hundred links mixed with the text, and buttons and doodads everywhere around it, which makes it very clear to actually think about what you want to do. If you compare Encarta '96 with Wikipedia: <https://memex.marginalia.nu/pics/links/encarta.png https://memex.marginalia.nu/pics/links/encarta.png> <https://memex.marginalia.nu/pics/links/wikipedia.png https://memex.marginalia.nu/pics/links/wikipedia.png> It's clearly very different. The Encarta design is a lot more quiet. If you're reading, then you are reading. There are sparse inline links, but nothing near the shotgun linkedness of Wikipedia. I've experimented with stripping down Wikipedia, producing this: <https://encyclopedia.marginalia.nu/wiki/Plato https://encyclopedia.marginalia.nu/wiki/Plato> It's definitely several steps too far (but taking things too far is the best way to get the feel for an idea), but I find that much easier to read, because the choices are moved out of the text and into the bottom of the text. That is, you read first, and decide what to do next once you are finished reading.

pferde 5y ago
All I see on the Marginalia Plato article is a site that forces an uncomfortably narrow article width on me, while eschewing almost all the advantages of hypertext in the name of some ill-conceived notion of "quieter reading". I'm not saying it's all bad, though - it's something that could be an option offered to the unfortunate people with attention deficit disorder. The forced text width, however, is inexcusable.
- marginalia_nu 5y ago
  Yeah I think that's probably from my own habit of never using full screen windows. Though I'm not sure how to deal with the fact that it's very hard to read long lines of text. Most printed books, newspapers, and so forth use that width (max 60ch) specifically because it's hard to follow lines longer than that. I chose to follow book guidelines specifically because that format has been largely stable for some thousand years and is at this point the gold standard for legibility. Usually when things stagnate for that long, it's not because nobody has thought to try something different in all that time. Maybe multiple columns could be an option, I guess. Not sure how to do that well with varying paragraph sizes, though.
  - morsch 5y ago
    On Firefox mobile, it's more like 11 characters wide. Pretty much unusable.
    - marginalia_nu 5y ago
      Hmm, strange. It doesn't have these proportions? <https://memex.marginalia.nu/pics/encyclopedia-mobile.png https://memex.marginalia.nu/pics/encyclopedia-mobile.png> I sadly can't diagnose since my phone isn't powerful enough to run Firefox. Looks as intended on Firefox desktop with emulation mode, though.
      - morsch 5y ago
        Huh. It looks fine now. Did you change anything?
maest 5y ago
Not much to add except wanted to say that I enjoyed becoming aware of the impact Wikipedia's design decision had on me, as a reader. Seeing the encarta/marginalia version vs the wikipedia one really drives home the point. It's interesting how the design had a strong impact on my experience, and yet I never really noticed it before. I feel a bit like a fish suddenly aware of the water they're in.
- mrob 5y ago
  I was confused by the claim that the Encarta version is quieter until I read that it was because of the inline links. I thought the Wikipedia version was quieter because it doesn't have a distracting background image. I think this confusion is evidence that inline links are actually not distracting. I don't feel I have to make a decision whenever I see one because I know they'll still be there later.
  - marginalia_nu 5y ago
    Maybe it's individual variations? I struggle immensely reading text with inline links. I immediately start skimming and my attention is strongly drawn to the linked words.
    - dahfizz 5y ago
      Maybe you can add custom CSS rules that make links the same color? I also much prefer the wikipedia interface to the alternatives.
      - marginalia_nu 5y ago
        This pretty much is a custom CSS. I was also more than a bit annoyed by the huge downloads and 30 second page loads.
wongarsu 5y ago
The Encarta version is much better if you want to learn about Plato, while Wikipedia makes it much easier to learn about thing related to Plato. The notion that the relation to other things is at least as important as the thing itself permeates Wikipedia. Many smaller articles, like e.g. Differentiable Programming [1] probably link to more topics than the article itself has words. I think a "stripped down" Wikipedia is an experiment worth trying, but it would serve a subtly different purpose, with some use cases being left behind and some being better served. 1: https://en.wikipedia.org/wiki/Differentiable_programming https://en.wikipedia.org/wiki/Differentiable_programming
- marginalia_nu 5y ago
  I think the big problem with Wikipedia's design is that it caters to too many audiences with disparate needs using the same interface, both reading and navigating and cross-referencing and editing. I think perhaps having multiple views for different needs would be a way forward.
  - tablespoon 5y ago
    > I think the big problem with Wikipedia's design is that it caters to too many audiences with disparate needs using the same interface, both reading and navigating and cross-referencing and editing. > I think perhaps having multiple views for different needs would be a way forward. I think what this thread is missing is that Encarta and Wikipedia are vastly different products. If Wikipedia had multiple views for different needs it would literally die, because it's community edited. If its UI encouraged staying withing a narrow lane, then it would have even more problems with getting readers to cross over to being editors.
- spookybones 5y ago
  I read something once that when you come across a link, your mind subconsciously hesitates whether or not to open it and that these thoughts detract from your focus on the material at hand. I often wonder if it would be more beneficial for websites to transition to footnotes with all pertinent links at the bottom of the page.
MagnumOpus 5y ago
If you want to strip away the MediaWiki adornments and editing/interactivity tools from Wikipedia, you just have to go to the mobile page: https://en.m.wikipedia.org/wiki/Plato https://en.m.wikipedia.org/wiki/Plato It is still consumption-oriented and stripped down, but retains far more non-textual information than your experiment.

martin_a 5y ago

> one should be the main button that people really should click on I don't think there's something wrong with CTAs. You can misuse/abuse this technology as you can with every, but providing a structured way to achieve a desired result is not bad per se. It helps build muscle/"brain" memory and can help you perform your task better/quicker.

scetron 5y ago

There was an item posted here this week with the same exact conversation. It's honestly a bit eye opening with the way UX is going. I am all for sensible defaults, but being able to tweak things here and there should really be an option, and not one seen as a failure by the designers...

zwkrt 5y ago

Agreed, but the reason this happens is not in a bubble. "User interface design" is not a natural category of design. People in the field will point to "The Design of Everyday Things" as some sort of progenitor of the field, but if you read it, it is kind of just a pop-psychology book. The truth is (IMO) that UX is more closely related to marketing than it is to technical design work. The entire point of having a UX designer is to get marketing speak and advertising techniques right into the heart of the engineering process. If engineers just built tools in the way that was the most sensible to them, they would have too many options and give the user far too much control. Ostensibly the reason given why this is bad is that it is 'confusing' for the user, but the reality is that it allows the end user to use the product in ways that do not maximize the profit. Even worse, users might use the product in a way that does not reflect the spirit of The Brand. It is telling that UX revolves around creating user stories, because this is exactly how advertising is done; “Martha, a woman in her mid thirties, is trying to balance home life with her need to efficiently walk her dog…”

musesum 5y ago

> there should never be more than one button ... which you have to look at, because the position and function changes. There is no muscle memory with dynamic UI. I recall a visually impaired woman who couldn't hail an Uber. I could tell she was very fast with the old UI. But, the button had shifted. She was going to call her Husband to drive over from Oakland to pick her up in San Francisco. Instead, I showed her where the button was.

lariati 5y ago

I think it is mostly a trade off. Using modern design philosophy you won't risk getting a really terrible UI but the trade off is you will not get something new and innovative. I am sure there are all kinds of issues with the scale of complexity also. It would seem nearly impossible for a designer to come up with a good innovative UI if they are not an expert in using the system themselves so the chicken in a farm ends up being the least bad option.