Closed Klona13 closed 1 year ago
I don't get this, how do I show which sorting direction is active if using buttons?
Something like this next to the dropdown menu (but more consistent looking duh)
I still don't understand; can you show a mockup what you have on your mind?
Ascending Descending
OK I see. There are a couple of problems with this design.
"Newest first" in the mockup should be "Date", "Last update" or something.
The intent of the button is not really clear in my opinion. Not many people will understand this without stopping and thinking what's this about.
As always with the toggle button, it's not clear if it displays the current value or the action performed. In other words if the button label spells A-Z ↓, it's not clear if it means that A-Z ↓ is currently active, or if it means that clicking the button will set it to A-Z ↓.
To change the sort direction I need to perform two searches instead of just 1 (click the button twice, not once).
It's not immediately clear how this applies to certain values. What does A-Z ↓ mean: that the earliest, or the newest levels will be first? That the highest, or lowest rated levels will be at the top? This is why icon packs usually distribute icons that represent volumetric values with bars, and numeric values with 0-9. At the same time changing the icon depending on the dropdown value is a huge no-no.
I struggle with understanding how "descending" and "ascending" applies to my search results, even with icon representations. I always have to stop and think for a moment, or I just resort things until the results look like I want. This is poor experience; it requires less mental effort to just select "Newest first" and press "search" instead of composing your search out of two elements "Last update" then "descending". I think that this is why all online shops I know (allegro.pl, ebay.com, amazon.de, rakuten.co.jp) have the kind of dropdown that we currently have.
Based on these thoughts I believe this is a step back.
How about this?
The purpose of this is to prevent the list from becoming too long (12 if we go by your idea and 6 with mine). It's hard to wrap the logic around my idea as you've said as it would require a bit more thinking.
It's sacrificing a bit of user experience for a more fleshed out look, but really after the first 2-3 times of using it people will know how to use it. And it would encourage people to see what it does... kind of like exploration. :)) wdyt
I think that no user wants to learn how to use a particular website. This approach takes more space, but allows for quick visual scanning. Common prefixes make the scanning just as quick as if it were single items, perhaps even slightly faster.
That's the point of it. Visual scanning takes longer than just pressing a button to get what you want. Especially since they're being renamed from something like: Largest size or Size (large first) To just: Size And then you get the button that's almost self explanatory. It's not completely vague as to what it actually does. Name (Alphabetically first) Release (Latest first) Update (Last updated) Rating (Highest first) Download count (Highest first) Size (Biggest first)
Putting repetitive information like this is very ugly.
But it's not just pressing a button to get what you want. You need to select what you want to sort on. For this, you need a visual scan anyway. In fact, you need to open the dropdown, because all the items are hidden by default, and only then do the scanning. Then you need to click your way to get the Ascending or Descending order. You need to pause for a moment to understand what this means. And you probably won't even look at the button but instead look at the results, to compare one item to another and see if it's going down or up. It does take longer. And this is why Amazon and other big players are not doing it this way.
In my opinion fewer, more common widgets are less ugly than this multi-state custom button that's supposed to accomplish a pretty common thing (sorting) and breaks POLA; plus my other points from the previous post still stand.
Ok we can go with your way and we'll see how it will turn out. But bear in mind that the "big players" don't have 12 items on the dropdown list:
Yes, that is another point – do all of those sorting options make sense? In my opinion only these make sense:
Name | Why |
---|---|
Release: newest first | default, to play new levels |
Release: oldest first | to play all levels |
Average rating | (best to worst) to play popular levels, first criterion |
Downloads: popular first | to play popular levels, second criterion |
Downloads: unpopular first | to play hipster levels, first criterion |
Least reviews | to play hipster levels, second criterion |
Things I think are not useful:
Can we go by
It's fun to see which levels are the biggest, some times people pick based on that. I agree with the rest.
We should get rid of "Least reviews" too cause this page achieves that: https://trcustoms.org/reviews/level_suggestions
With this, we have 8!
Would it be possible to fit levels with "not enough ratings" between slightly negative and mixed? So this doesn't happen:
Hmm, these two also have a similar issue when a level has no download available:
I can't even think of how to solve that...
Actually, "Biggest size" is fine cause the "no download" levels display at the last pages... @rr- So we may able to solve "Smallest size" if this is possible: when that option is selected, levels with "no download" will be shown at the end (last pages) instead.
(Up and Down arrows buttons) Sort by: Name Release Update Rating Downloads Size
The up and down arrows are for ascending and descending. The default sorting type is Descending to show the biggest or latest number.