Microsoft's Copilot button is a shining example of the form factor trap.

Who remembers when Siri came out and voice assistants were the future of computing? Samsung built a physical Bixby key into their phones (and went through massive pains to prevent users from disabling it). Microsoft had a dedicated Cortana button ship pinned to Windows 8's dock. Google Assistant was a default Android homescreen.

1/

https://www.theverge.com/2024/1/4/24023809/microsoft-copilot-key-keyboard-windows-laptops-pcs

#uxDesign #ProductManagement

Microsoft’s new Copilot key is the first big change to Windows keyboards in 30 years

Microsoft is bringing a Copilot key to new Windows-powered laptops and PCs. It’s the first big change to the Windows keyboard layout in nearly 30 years.

The Verge

Apple, Microsoft, Google, and Samsung assured us they were building the future.

Nobody wanted the future they built. After a couple of years, Bixby, Cortana, and Assistant are gone from the spotlight. But not to worry - Microsoft has a new thing it wants to stick into your face! This time it's a physical keyboard button!

2/

LLMs are just the new VUI. A technology that makes *some* things easier, in *some* cases, but is being sprayed on everything because it's the Hot New Thing, and rushing out features is easier than doing real user research and modeling use cases from observed needs.

3/

And when they need features but don't have a conception of how they will solve user problems, low product maturity teams turn to form factors. It's trivial to generate a dozen "ideas" when they are all some variation of "implement a design pattern from somewhere else that we don't have yet."

4/

Add a button. Add a chatbot. Add a dashboard.

Those are all form factors. Which is fine. But when your idea is "the everything button" or "the everything dashboard" what I hear is "we don't know or care what users need, just put everything in there and let them sort it out."

And users will sort it out. They sorted it out with Bixby, through 3rd party patches to disable it. And they'll sort out the Copilot button too.

5/5

.@PavelASamsonov There's an underlying worldview here that's a similar motive to other prevailing trends in our society, and that is the idea that someone powerful enough can make the truth be whatever they want to through sheer strength of will and wielding said power. They start with the assumption that the use of power will make users behave the way they want us to rather than starting with the opposite question, "What is the reality here, and how do we make something that conforms to that truth?" But see, that gives a little bit of power to the user, and they don't want to admit that.

It has a lot in common with fascism, that.

@PavelASamsonov aka to a hammer, everything is a nail. Truly elegant product design is a subtle art.

@PavelASamsonov

Well said. The UX philosophy of “Solution first, ask questions later.”

@PavelASamsonov after having tried to set up gifted devices at christmas for my relatives, I came to realize that much open source software will by now have a better usability; not because of the expertly designed UIs but because there are less features+advertisement+banners thingies.
@simulo I'm torn on this because on the one hand, yes, but on the other hand much OSS follows the design principles of "this is the way I like it" which often results in a very difficult UI for people whose mental model doesn't match the creator's
@PavelASamsonov totally agree. I am not praising the state of UX in open source software, it is just that all the "enshittification" of basic infrastructure (like operating systems) makes them difficult to use (except you opt in in all their data collection and services).

@simulo @PavelASamsonov #OpenSource software is 1 giant step forward for privacy and calm (no upselling/nudging) and 1 giant step backward for #UX

My hope is that the state of both will continue to improve.

@scottjenson @simulo @PavelASamsonov For many open source projects, #UX is not at all a "giant step backward".

I'd say that #OpenSource generally prioritizes getting things working, functional, and tested before working on #UX, but many OSS projects have matured to the point where UX is a major focus -- and often better than commercial alternatives.

See current #Drupal, #NextCloud, #HomeAssistant, #ForgeJo, #BitWarden, #Mastodon, and dozens of other examples...

@freelock @simulo @PavelASamsonov Of course, there are positive examples of #UX in #opensource It's just that most projects with good UX are large and funded. They 'get it'. There just aren't that many unfortunately.

My point isn't that #FOSS is bad, I'm very much here to participate and help. It's only that it has lots of room to improve. How can we take the projects that get it and transfer it to smaller, less mature teams?

@freelock @scottjenson @simulo @PavelASamsonov I agree with many of those, but Nextcloud definitely should not serve as a positive example for "good UX" in #FOSS software.
@preya @freelock @simulo @PavelASamsonov care to elaborate? (happy to have a DM if that's easier)

@scottjenson @preya @freelock @simulo @PavelASamsonov I'd like to know, too 😃 @jancborchardt probably would also want to know.

I use Nextcloud everyday (work/personal), all the core apps. It's simple and functional. Can it be improved? Everything can be improved. But the only way it can be improved is if people report issues on Github, so the team is aware, can triage, and see if it's an actual problem or a niche need of one user. If skills permit, contribute fixes with PRs. It's open source.

@preya Sure there are rough edges getting everything going in NextCloud. But I love that I can save a file and then immediately send a link to somebody without them needing to create an account to download it.

And with OnlyOffice (admittedly not open source), document collaboration beats the pants off Google Office, which for me has become crippleware.

The Memories app in Nextcloud has a really great UX!

@viktor @scottjenson @simulo @PavelASamsonov @jancborchardt

@freelock @viktor @scottjenson @simulo @PavelASamsonov @jancborchardt Hmm, I don't know, it seems we've had vastly different experiences. Neither OnlyOffice nor Collabora are usable for me in a collaborative way. The lack of speed and massive delay on mouse/keyboard inputs together with the visual bugs in the UI (mostly z-index problems) make them unusable for me. As much as I hate Google, their product is just on another level. Comparing these two is really not helpful.
@freelock @viktor @scottjenson @simulo @PavelASamsonov @jancborchardt I do agree on the Memories app though. It's definitely one of the nicer ones in the Nextcloud App Store.

@preya @freelock @scottjenson @simulo @PavelASamsonov @jancborchardt That sounds like a config/hardware issue. Our Nextcloud Office runs rather smoothly.

Google has the luxury of managing the entire stack and optimize it for their code. But when self-hosting, you're it. You're your own Google. You have to do everything. You have to optimize your stack but also make sure to get proper hardware for your needs. Nextcloud Office needs a dedicated server, the built-in CODE option isn't scalable.

@preya @viktor @scottjenson @simulo @PavelASamsonov @jancborchardt I literally cannot open any Google Office doc without going through a hoopla where it forces me to log in, tells me I can't access it because I refused to pay for our "forever free" office account, and then it finally lets me switch to a different Gmail account where I can access it.

Never mind that we do pay Google for cloud services, YouTube premium, phones, lots of stuff, just not their office service that NextCloud replaces

@viktor @scottjenson @freelock @simulo @PavelASamsonov @jancborchardt We're not talking about a "can be improved" situation. No one expects Nextcloud to be perfect and bug free. But the state of the project has degraded constantly over the last years. Reporting issues has become useless. There's thousands of open issues and no one is triaging them or answering to the community. Nextcloud GmbHs internal priorities and processes are completely intransparent to the community.
@preya @scottjenson @freelock @simulo @PavelASamsonov @jancborchardt There are over 15,000 closed issues on GitHub for the core server alone. If nobody was fixing anything, there wouldn't be that many closed issues. Maybe issues you care about are taking longer than you would like, but everything is prioritized and eventually gets fixed. Lots of work, limited resources. But serious bugs get fixed quickly. That's why there was a 28.0.1 release a few weeks after 28.0.0.
@viktor @preya @scottjenson @freelock @simulo @PavelASamsonov regarding transparency: The planning boards of all of our teams are public at https://github.com/orgs/nextcloud/projects
Additionally we have Talk channels for contributors, you can find the one of the design team on https://nextcloud.com/design for example.
Not sure what is "completely intransparent" here?
Build software better, together

GitHub is where people build software. More than 150 million people use GitHub to discover, fork, and contribute to over 420 million projects.

GitHub
@scottjenson @freelock @simulo @PavelASamsonov This is hard to convey in 500 characters (or even multiple toots) and obviously is very subjective. But one of the greatest UX fails in #Nextcloud is the whole sharing topic. There's actually at least three different and independent sharing functions (regular share, groupfolders and circles) which function completely independently. So as soon as you allow more than one of those within your team, all hell breaks loose.
@scottjenson @freelock @simulo @PavelASamsonov People don't understand the difference between a group folder and a "share". The UX around this is horrible. There should be dialogs, warnings and help pages explaining the differences, as soon as you have multiple of those mechanisms in place. It's the number one thing that brings most of my customers to hate their Nextcloud after a couple of months/years of use.