5 ms·

"Don't use right-pointing carets on line items" "Avoid them to stay consistent with the platform and in order to not have the user guess as to what the meaning

by ge0rg 15y ago

"Don't use right-pointing carets on line items" "Avoid them to stay consistent with the platform and in order to not have the user guess as to what the meaning of those carets may be." Instead, let the user guess if an entry is just an entry or opens another screen full of interesting things.

babebridou 15y ago
I use the following rule of thumb: "if you're going to display a new screen from touching an item with a label, append '...' to the label".
- mef 15y ago
  So you're effectively reproducing the caret by appending text to the label. Why wouldn't Android just use a caret or some other indicator by default? It's small UI details like this that can mean the difference between a sexy interface and a clunky one.
  - babebridou 15y ago
    I agree. I'm not defending my decision to do it that way, it comes down to a habit from my Swing days, a good one then, and probably a bad one now that I'm working to correct. I'm just stating the plain fact that I have my way of doing it, Apple have their way of doing it, and Android has no way of doing it. I believe the right way to be something else, though: Never display a list that mixes data and actions. Even better: never display a list at all. Lists are bad for the user. Scrolling is bad. That's my honest belief.
    - dquigley 15y ago
      "I believe the right way to be something else, though: Never display a list that mixes data and actions. Even better: never display a list at all. Lists are bad for the user. Scrolling is bad." - Interesting, I'd love to know if there is more of a reason for this, specifically why scrolling is bad? Either way, it's an interesting idea, and I'm tempted to agree with you. I would love to see the open source community, Android included, expand on what is done on this one site, and continue testing, discussing, and demoing theses types of user experience and interface questions, so we can all benefit from it.
      - babebridou 15y ago
        I've tried to reason my belief. It's hard. The only really earthly reason why I don't like scrolling is that (don't laugh) touch interfaces hurt my fingers. I actually feel pain whenever I touch the iPad or Xoom's screen, or my macbook pro's touchpad. I gave my magicmouse to my sister no more than a couple hours after I had purchased it, because I knew I couldn't sustain the mix of friction & electrical pain for more than a minute at a time. I honestly don't know what came first, the pain or the repulsion for scrolling in general, but the association is fierce. Then I started reasoning on why people would prefer a world without scrolling. The first thing that came to mind is that scrolling does not exist a lot in real life, apart from a few instances: using binoculars, using a magnifier, going forward/backward on some sort of static security camera looking at traffic or something like that. For everything else, we naturally use paging. Paging is strange: using two or three fingers, you can page by any amount you see fit, and every time, the entry you're looking for is on display at the exact same place on your eye. Items can be specified with their page number (324), then a 2D position (fifth line, seventh word). At some point you can even use muscle memory to reach the correct page while you position your eye for receiving the info. Scrolling also has an interesting attribute: it's entirely a computer concept. You have a buffer of information and you increment the offset. But as an end user, using only scrolling, you can't easily perform things such as dichotomy - everyone knows dichotomy even if they don't know the name of it. Go forward a bunch, go backwards half a bunch, go forward half of a half of a bunch, find the entry. Another thing that I don't like with scrolling is how it messes up with the "5-7 modulo 2" rule of thumb for decision making (1). It's not that there are more elements than needed, it's that they get pushed up & down when you scroll, so they never appear to be a fixed point in space where you can go in a subset of 3-9 destinations. As a result, you can't select a subset from 3-9 subsets, and you can't select an item from 3-9 items, you actually have to scroll all the way and parse all elements in order to make an informed decision from, say, 15 or 150 possible ones. All these questions (and more (2)) itched me since the day I realized that there could be a computer world without scrolling and maybe we could start dreaming/having nightmares about it... Anyway this is a belief and I'm generally a man of reason - it's tough to put it into words what you know you definitely can't prove :) (1) : http://en.wikipedia.org/wiki/The_Magical_Number_Seven,_Plus_or_Minus_Two http://en.wikipedia.org/wiki/The_Magical_Number_Seven,_Plus_... (2) : mouse wheel fatigue when browsing websites or expanding trees, why isn't a newsfeed or a time related search result on a logarithmic scale so I can see it all, always, since the start of the century, why did we mix UI and content on webpages when we could simply remove the navigation panes...
- roel_v 15y ago
  Yeah, that's more or less what the Windows UI guidelines say too.
demallien 15y ago
Yup, that line jumped out at me as well. Another one was complaining about people putting labels on the back button. Heaven forbid that you should actually inform your users as to what is going to happen when they press a button who's action depends on context. I can understand that they don't want people simply copying the interface from another platform, but complaining when people steal ideas from another platform when your own platform is deficient? That's just stupid.
- Kudos 15y ago
  There's a built in back button in Android, sticking another one on the interface looks completely wrong. It is invariably cheap iOS ports or webkit wrappers tuned for iOS that do this.
  - ge0rg 15y ago
    As the guidelines outline, the (physical) Back button has a different meaning from the Up button in the top left corner of your app.
- myko 15y ago
  The platform isn't deficient. Android users are familiar with the way the ActionBar works - any design that explicitly labels their back button is deficient and wasteful (on Android).
pja 15y ago
Instead, let the user guess if an entry is just an entry or opens another screen full of interesting things. If it's just an entry, then it will either have a checkbox or a text box whilst inactive list members which do nothing are grey instead of white (at least on Gingerbread, I don't have an ICS device to hand). So there's no ambiguity. You can argue that a caret or ellipsis might convey the affordance of expanding the item by selecting it and I wouldn't necessarily disagree, but the Android UI as designed is at least unambiguous.
king_jester 15y ago
The example shows a ListView of items where each entry responds to a click and does something. In that context, you should never ever use indicators on the rows of the ListView as the list selector highlight when clicking an item signifies that an action will happen (if you want a list with no click actions, disable the list selector so nothing happens when a row is touched).

by ge0rg 15y ago

babebridou 15y ago

I use the following rule of thumb: "if you're going to display a new screen from touching an item with a label, append '...' to the label".

mef 15y ago
So you're effectively reproducing the caret by appending text to the label. Why wouldn't Android just use a caret or some other indicator by default? It's small UI details like this that can mean the difference between a sexy interface and a clunky one.
- babebridou 15y ago
  I agree. I'm not defending my decision to do it that way, it comes down to a habit from my Swing days, a good one then, and probably a bad one now that I'm working to correct. I'm just stating the plain fact that I have my way of doing it, Apple have their way of doing it, and Android has no way of doing it. I believe the right way to be something else, though: Never display a list that mixes data and actions. Even better: never display a list at all. Lists are bad for the user. Scrolling is bad. That's my honest belief.
  - dquigley 15y ago
    "I believe the right way to be something else, though: Never display a list that mixes data and actions. Even better: never display a list at all. Lists are bad for the user. Scrolling is bad." - Interesting, I'd love to know if there is more of a reason for this, specifically why scrolling is bad? Either way, it's an interesting idea, and I'm tempted to agree with you. I would love to see the open source community, Android included, expand on what is done on this one site, and continue testing, discussing, and demoing theses types of user experience and interface questions, so we can all benefit from it.
    - babebridou 15y ago
      I've tried to reason my belief. It's hard. The only really earthly reason why I don't like scrolling is that (don't laugh) touch interfaces hurt my fingers. I actually feel pain whenever I touch the iPad or Xoom's screen, or my macbook pro's touchpad. I gave my magicmouse to my sister no more than a couple hours after I had purchased it, because I knew I couldn't sustain the mix of friction & electrical pain for more than a minute at a time. I honestly don't know what came first, the pain or the repulsion for scrolling in general, but the association is fierce. Then I started reasoning on why people would prefer a world without scrolling. The first thing that came to mind is that scrolling does not exist a lot in real life, apart from a few instances: using binoculars, using a magnifier, going forward/backward on some sort of static security camera looking at traffic or something like that. For everything else, we naturally use paging. Paging is strange: using two or three fingers, you can page by any amount you see fit, and every time, the entry you're looking for is on display at the exact same place on your eye. Items can be specified with their page number (324), then a 2D position (fifth line, seventh word). At some point you can even use muscle memory to reach the correct page while you position your eye for receiving the info. Scrolling also has an interesting attribute: it's entirely a computer concept. You have a buffer of information and you increment the offset. But as an end user, using only scrolling, you can't easily perform things such as dichotomy - everyone knows dichotomy even if they don't know the name of it. Go forward a bunch, go backwards half a bunch, go forward half of a half of a bunch, find the entry. Another thing that I don't like with scrolling is how it messes up with the "5-7 modulo 2" rule of thumb for decision making (1). It's not that there are more elements than needed, it's that they get pushed up & down when you scroll, so they never appear to be a fixed point in space where you can go in a subset of 3-9 destinations. As a result, you can't select a subset from 3-9 subsets, and you can't select an item from 3-9 items, you actually have to scroll all the way and parse all elements in order to make an informed decision from, say, 15 or 150 possible ones. All these questions (and more (2)) itched me since the day I realized that there could be a computer world without scrolling and maybe we could start dreaming/having nightmares about it... Anyway this is a belief and I'm generally a man of reason - it's tough to put it into words what you know you definitely can't prove :) (1) : http://en.wikipedia.org/wiki/The_Magical_Number_Seven,_Plus_or_Minus_Two http://en.wikipedia.org/wiki/The_Magical_Number_Seven,_Plus_... (2) : mouse wheel fatigue when browsing websites or expanding trees, why isn't a newsfeed or a time related search result on a logarithmic scale so I can see it all, always, since the start of the century, why did we mix UI and content on webpages when we could simply remove the navigation panes...
roel_v 15y ago
Yeah, that's more or less what the Windows UI guidelines say too.

demallien 15y ago

Yup, that line jumped out at me as well. Another one was complaining about people putting labels on the back button. Heaven forbid that you should actually inform your users as to what is going to happen when they press a button who's action depends on context. I can understand that they don't want people simply copying the interface from another platform, but complaining when people steal ideas from another platform when your own platform is deficient? That's just stupid.

Kudos 15y ago
There's a built in back button in Android, sticking another one on the interface looks completely wrong. It is invariably cheap iOS ports or webkit wrappers tuned for iOS that do this.
- ge0rg 15y ago
  As the guidelines outline, the (physical) Back button has a different meaning from the Up button in the top left corner of your app.
myko 15y ago
The platform isn't deficient. Android users are familiar with the way the ActionBar works - any design that explicitly labels their back button is deficient and wasteful (on Android).

pja 15y ago

Instead, let the user guess if an entry is just an entry or opens another screen full of interesting things. If it's just an entry, then it will either have a checkbox or a text box whilst inactive list members which do nothing are grey instead of white (at least on Gingerbread, I don't have an ICS device to hand). So there's no ambiguity. You can argue that a caret or ellipsis might convey the affordance of expanding the item by selecting it and I wouldn't necessarily disagree, but the Android UI as designed is at least unambiguous.

king_jester 15y ago

The example shows a ListView of items where each entry responds to a click and does something. In that context, you should never ever use indicators on the rows of the ListView as the list selector highlight when clicking an item signifies that an action will happen (if you want a list with no click actions, disable the list selector so nothing happens when a row is touched).