You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Same thing as #205, but round 2 😉 . The same reasoning applies:
Zulip's avatars should be displayed as squares (with rounded corners), not as circles.
This is based on a lot of consideration -- circular avatars are problematic because many images that we'd like to use in avatars (e.g. bot logos) don't actually look good in circular form, whereas just about anything looks good in a rounded square.
One example: our own bot avatars look quite bad when cropped to a circle, as they now are on the user-profile screen (AccountDetailsScreen).
Relatedly: we shouldn't have had to have a round 2 of this, and we should avoid a round 3.
One root cause of round 2 is that in round 1 (in #208) we left the "circle" case lying around in the code, even though we had determined that we had no intention of actually using it. That's a bad pattern for a number of reasons -- most of all, it makes the code more complex and confusing to understand -- and this is another of them.
The text was updated successfully, but these errors were encountered:
This isn't used, and makes the code more complex. We always
effectively use the default of "rounded", never the alternative of
"square"; so just hard-wire the "rounded" behavior.
Back in 805d35d, fixing zulip#3398, we took out the old "circle" option
here, which produced quite bad results when used. The "square" option
could be OK in the right design context; but we don't actually use it,
and haven't since 57f45f4 a couple of years ago. If we want that
functionality in the future, we can add it back in the relevant places
then.
Same thing as #205, but round 2 😉 . The same reasoning applies:
One example: our own bot avatars look quite bad when cropped to a circle, as they now are on the user-profile screen (
AccountDetailsScreen
).Relatedly: we shouldn't have had to have a round 2 of this, and we should avoid a round 3.
One root cause of round 2 is that in round 1 (in #208) we left the "circle" case lying around in the code, even though we had determined that we had no intention of actually using it. That's a bad pattern for a number of reasons -- most of all, it makes the code more complex and confusing to understand -- and this is another of them.
The text was updated successfully, but these errors were encountered: