Usability issues (UX) in redesigned UI (2024.12.0)

Understood entirely. And I’m not attacking BW. I understand fully that there are trade-offs and business reasons for everything that forum users wouldn’t be aware of.

IMO everybody should chill and give it a few weeks. Don’t roll back or investigate alternatives – BW is still the best. Try to get used to it and soon enough it will improve.

1 Like

I get that. And I agree. But I reacted to “just asking fanboys” - because that didn’t sound right either.

@edelstone Honest question: what would you have said to an auto-fill button like that (and ignore my editing - I don’t have the original):

?

@Nail1684 I personally agree with others here that the entire row should be the autofill target with view/edit offloaded into the ellipses menu. Lacking that functionality, I would tell the user to go with the wider layout in settings > appearance – since there already exists a workaround for your character visibility complaint, there is no need to shrink the autofill target to compensate.

1 Like

Actually I don’t disagree now. That’s why I made that suggestion:

But I must say, I’m not in the mood to discuss the inconsistency-issue again (see aboooooove), so I’ll leave it at that. :sweat_smile:

PS:

Yeah, that option didn’t exist two months ago. :woozy_face:

1 Like

Since @Nail1684 is campaigning for his idea, I will make a plug for my own proposal.

The main concept is to enlarge the autofill button size by leveraging the nonfunctional space currently used to display the item icon (e.g., website favicon, or card logo), and to place the autofill button on the left (this satisfies the principle of placing important actions first, while also preventing the autofill button form shifting locations depending on what quick-copy icons have been displayed on the right side of the item).

In the draft concept illustrated below, the autofill button encompasses only the item’s icon, as well as a pictograph (or word) added to remind the user that the button action is autofill.

image

For items that do not have any URI match (or if the autofill action should not be available for any reason), the item icon can still be displayed, but without the button border or autofill pictograph. This will keep the clickable item name in a consistent location always.

The same UI button could also be made available in search results, where clicking the left button would result in a forced autofill (autofilling when there is no URI match). To remind the user that autofilling without a URI match can be risky, the button border and/or background could have a different color in this context, or a different pictograph could be used (even something extreme, such as a :warning: symbol).

1 Like

That’s what is meant here, in case someone hasn’t seen it already (I only noticed this today):

It’s still the same principle. No need to correct the problem twice. Increase layout options or shrink fill target, not both.

I think I don’t agree completely. If they didn’t shrink the target then everyone would be forced to a wider view. ?!

Sure, everyone concerned about character truncation. Other people are concerned about the size of the “fill” button, clearly. So which one should win? Hence, proper testing required so they can actually see how people use the product.

1 Like

Well, the only valid reason for needing a “fill” button is for websites that have a large number of matching credentials, and in such cases, it is essential to be able to see enough characters to be able to tell the credentials apart. So the Venn diagram between the set of users concerned about character truncation and the set of users who are concerned about the size of the “fill” button may have significant overlap.

Agree 100%!

1 Like

Thanks, and sorry for my misunderstanding. However, I do seem to recall that previously, when clicking on an address field on a form and then opening the Chrome extension, Bitwarden usually revealed the records in my Identities. Clicking on one then filled the address fields. Now, the extension opens and initially presents Autofill suggestions followed by Favorites. So, to get to my Identities, I have to click the Vault Filter, click on Type, click Identity, click the stacked dots adjacent to the desired Identity, and then click Autofill. Again, please correct me if I’m wrong.

I could accept your proposed change as a compromise, although can you explain the practice of adding (TOTP, Passkey) to the site name? That extra length is the root cause of the problem you are seeing. I try to keep the site name as succinct as possible.

1 Like

This idea may be unpopular or unworkable for reasons I haven’t considered; you guys tell me. Bring the Tab view back except call it Fill instead. Make a huge Fill button on the left side of the row that autofills while clicking the row itself opens the row for viewing consistent with the Vault view, web app, and mobile app. This would address the loss of a large autofill target and the undesired display of all vault items while continuing with the new design principals. I think a bonus benefit with this is that the “Autofill suggestions” section on the Vault would no longer be needed freeing up real estate for an expansion of the new filtering feature. I’d love to have a filter option that allows quick identification of items with passkeys, TOTP, etc

That is my compensation for the lack of “tags” and to be able to search my vault for e.g. all items with a passkey, TOTP, …

(PS: Ah, and therefore, that part of the vault item title is nothing I necessarily have to “see”… it is mostly enough, that I can search for those “tags”)

Though that is certainly a good idea, to keep the titles short and succint, I wouldn’t say that it is the “root cause of the problem”. Personally, I think now, that the root cause of the problem is more, that now a new auto-fill button and the text of the item (title & username) compete with each other about the given space.

Therefore my suggestion, to “merge” large parts of the item with the button (maybe again…but maybe not again, because the old design didn’t look like a button, but my idea would be, to make it a button - and obviously that also means that it would look like a button).

It would be similar to the previous behaviour (of an auto-fill-item) - though when it is a button I think the previous inconsistency would still be at least reduced - and there would no longer be that competition about the given space between a “separate auto-fill-button” and the item’s identifying texts…

(actually, I see more problems with the current button… I wrote about that in previous posts… far above :sweat_smile:)

a .2 version of the extension, plus other tasks in progress and user-created documentation of options have improved several of the most obvious problems with the new UI. That’s good.

The biggest flaws impacting my experience come from one other area that doesn’t have any work-arounds. The new UI concept merging the Tab and Vault pages into one page is the root cause of so many of the other complaints.

These two pages in the old UI serve two completely different purposes. Tab was for Fill suggestions related to the website of the current browser tab. Vault could be used like a bookmark list of favorites for launching with (optional) auto-fill, and for other workflows that were not related to a current browser tab. A user is only ever dealing with one of those things at a time, so there cases where both are needed are the same dialog is extremely small.

Some users apparently only use one of them, and some apparently only use the other one. Other users use both and knew when to click between the two for the different purposes. Merging the two pages into one page is disruptive for all 3 groups of users, leading to the issues of “why are all my sites / usernames showing on the dialog” and “why can’t I find the favorite I was looking for without scrolling?”

Size of buttons, size of fonts, default action after a click are all important details, but the fundamental UX flaw is putting two different workflows, two different purposes together on the same dialog page.

Everything else can probably be fixed eventually, or worked around. Splitting Tab and Vault can’t be worked around. These need to be two separate pages, with an option for which is displayed by default or always remember the last one used.

3 Likes

Reading some of the earlier posts, it seems all of my main points have already been said. So instead, I will provide some additional insight into how I see things in the hopes that it will be beneficial in some way.

Optimizing the UI is an admirable goal, one that I firmly believe in. However, first and foremost, UI’s should be optimized for your users workflows. Then other aspects of the UI can be optimized within the framework of the workflows. Doing things the other way around simply trades one issue for another. This results in “the cure being worse than the disease”, as you end up with something that is very easy to use, but aggravating for many with the extra work involved to complete each task.

I understand that some users need the UI to be easy to operate without having to learn by reading or watching anything. As such, there is no one-size-fits-all solution. So you will either need to pick one type of user and code the UI for that specific use case, while alienating the rest, or you need to provide UI options to allow each individual to decide what works best for them.

Whenever my development team implements changes to existing features, including the user interface, we always implement a setting that defaults to continue working the way it always has. Anyone wishing a different experience, now has the option to tweak things. This prevents a lot of confusion and upset from the existing user base, and has served us well over the years. We only break this rule in extreme circumstances.

If you want to simplify the UI setup, while providing various options, then how about implementing a setting to specify what type of user they are. Do they want a simplified UI, the most optimal workflows, middle of the road or Jedi master who can customize their own UI configuration?

I have seen some browser extensions bring up a webpage after installation. Perhaps you should utilize this in future when making big changes. It would be very useful for informing or training the users on big changes. Another idea would be to use it to implement a setup wizard. This wizard could ask a few targeted questions to setup the browser extension to the individuals needs. Such as asking them what their use case is (basic password management, use of security keys, etc.), show UI pictures and have them pick which they like the most, etc.

On a side note, I am struggling with arthritis from heavy keyboard and mouse usage. So while a few extra clicks to access a feature may not seem a lot to you, they can have a huge impact to my quality of life. Especially for password and form filling as I use that throughout my day, 365 days a year. Even a few clicks per task, can really add up. Currently, a good day for me is when I’m done and don’t have more pain than when I started. Your awesome password manager is one of my most used tools, so any workflow optimizations are a huge help to me.

Thanks for adding a setting back in to allow us to more easily access the copy username and copy password feature. That is a huge help to me and my fingers. The biggest thing I am still missing is more keyboard shortcuts, especially for form filling. Keyboard shortcuts make it really easy for people such as me, to access features via voice commands. Attempting to control the mouse be a voice can be very frustrating and time-consuming, especially for smaller buttons such as your new form fill button.

I just double checked, and it seems the keyboard shortcut, that I was so desperately needing, has been completed. Thank you so much! I think we need the better feedback mechanism, unless I missed something. It have been nice to receive a notification when a feature request I voted on/reply to was completed. :wink:

2 Likes

Before I misunderstand you: are you saying you think (with the old extension) there were 1) users who only used the “Tab” tab and 2) users who only used the “Vault” tab and 3) users who used both tabs and this third group were also those people who had no problem with having the two separate tabs?

I mostly like the new design, but the “Fill” button on each login needs to be a little larger. It’s easy to miss, especially for those of use used to the old way, where we just click on the item. And second, why not give us a second Fill button on the next page that we get, after we do click on the item? As is, I have to go back a page to get to the Fill button again.

2 Likes

The previous behavior was that whether you clicked on an address field or not (or whether an address form was even present in the webpage or not), opening the browser extension would display identities in the “Identities” section on the “Tab” page (assuming that you had enabled the option “Show identities on Tab page”).

The new behavior is supposed to be that Identities are displayed in the “Autofill Suggestions” section on the browser extension’s “Vault” page, if and only if Bitwarden detects that the webpage contains a form for inputting identity data (also, the option “Show identities as autofill suggestions on vault view” must be enabled for this to work). The problem is that Bitwarden’s new algorithm for detecting the presence identity form fields is extremely unreliable, with a failure rate estimated at 90%.

Therefore, the new behavior for Identity (and Card) autofilling is causing problems primarily because the implementation is buggy. Nonetheless, even if the bugs are fixed, and Bitwarden’s form field detection algorithm starts working perfectly (something that still doesn’t work 100% when it comes to detection of login fields, even after 8 years of development!), there will still some problems caused by the design of the new UI — in particular, Identities displayed in the “Autofill Suggestions” area will be sorted in descending order of most recent use, and there will be no visible demarcation between the list of Login items and the list of Identity items within the “Autofill Suggestions” section.

1 Like