[somehnacct3757]

The thread is about removing the label, not the whole search box.

Which is a sort of silly thing to complain about, but Windows 11 has a lot of taskbar feature regressions. You can't even pin the thing to the left or right edge anymore. So I understand why users are frustrated.

This all represents a philosophical shift in Windows ("the taskbar is Microsoft's, not yours") and it will be a painful adjustment for power users.

[323]

> You can't even pin the thing to the left or right edge anymore.

What if only 0.05% of users do that? Is it really worth all that entropy in code and test matrix for such a low usage feature?

[csydas]

It's the wrong way to Fram the question.

If I have a program that emits a painful audio bit every time the user does X and only 0.05% of users disable it by going to a configuration value they need to search for through dozens of (my own) SEO'd results, that doesn't mean that only 0.05% of users want it disabled, it just means that only 0.05% of users were able to figure out how to stop my program from doing X.

I work on a product where our power user base is known for requesting registry options to change/disable feature x/y, and our market research has consistently shown a registry value doesn't enter most people's minds; to further this point, our user base is specifically IT persons. Even our Linux user base (actual day to day linux users) don't consider asking if we have a config file to edit somewhere in any overwhelming way.

The point I take from this data (which I know is anecdata for you) is that most users assume software "out of the box" is what you see is what you get. Making these changes and options and not including a way to remove them is not a good way of tracking interest since the sample size that is actually willing to take the time to complain and post compared to those that just look at other software entirely are far different. You're dealing with a complete unknown in the Power User space; regular users are even less inclined to see how to disable it and hiding it behind registry edits is actually a dissuasion.

[323]

MIcrosoft is well known, and one of the first which did extensive user-testing of UIs, video tapped, experts analyzing through one-way mirrors, questioning, ...

I doubt that they removed the side taskbar option without very extensive testing and discussions. In fact, it's a well repeated criticism of Microsoft, that they have TOO MANY options, not too few:

https://www.joelonsoftware.com/2006/11/21/choices-headaches/

[thesuperbigfrog]

>> I doubt that they removed the side taskbar option without very extensive testing and discussions.

And yet Windows 11 adoption remains low.

The nonsense UI changes to the taskbar are the primary reason I am still using Windows 10.

The issue with having so many features is that people use those features and become reliant upon them. Removing those features means that the product is not as good as its predecessor and thus fewer people move to the new version.

[323]

> And yet Windows 11 adoption remains low.

I suspect the TPM 2.0 requirement is the bigger reason. Let's face it, most Windows users are unaware of what exact version they are running. If it prompted to auto-update to Windows 11 they would click yes.

[thesuperbigfrog]

>> I suspect the TPM 2.0 requirement is the bigger reason. Let's face it, most Windows users are unaware of what exact version they are running. If it prompted to auto-update to Windows 11 they would click yes.

I am sure that may be true for some users, but certainly not for all.

My computer meets TPM 2.0 requirements, but I turn down every Windows 11 upgrade offer.

The business I work for buys new PCs to replace old ones, but the company OS image is still Windows 10.

[csydas]

Well, except the article and the argument aren't what we were talking about.

You were talking about % of users using an option and the cost of using it, to which I argued that it's quite difficult even with perfect telemetry to know whether people are not using an option because they don't want to use it or they just cannot find it. Telemetry data like this just tells you N% of users do X; you don't know why just from the fact that they do it. The user base for my product does X which is contrary to the intended design all the time, and we don't know why until we get to one on one discussions as to why they tried it that way, and even that is very difficult data to get.

Your article here is not really related, it's about choices in UI that are confusing (and also cannot be enabled/disabled from the UI), and in fact I even understand this point to undermine the first; even if Microsoft continues to do large amounts of user testing, are they doing anything meaningful with the data if they continue to have too many choices?

I don't think there's a lot of sense to speculate on why Microsoft made a choice unless they publish a claim as to why they made a choice; we can point to data and anecdata endlessly to support positions but it doesn't actually make a case for the affirmative in either way.

What I am to say is that even with user testing, countless options, and so on, at best you can say that on a vanilla installation, N% of a sample group did X. With extrapolating statistical math, you can assume with a margin of error that if N% of users from Z sample set did X, then Y% of users will likely do the same with some margin of error.

From anecdata I can know from other persons who call themselves Windows power users and/or technical persons, I know that the common response to most Windows changes is not that it's good or bad, but that you just learn to deal with it after awhile. Personally I take this as more negative, that is it's not active adoption it's begrudging acceptance, but that's a personal take.

[323]

> Your article here is not really related, it's about choices in UI that are confusing

But remove any of the shutdown options, and someone who uses it will complain saying that it's the perfect one for them.

> know that the common response to most Windows changes is not that it's good or bad, but that you just learn to deal with it after awhile.

So what's the alternative, not changing nothing ever? We would still have the Windows 3.1 interface (no taskbar) according to this logic, I'm sure the transition from the 3.1 model to the Windows 95 model was extremely disrupting to Windows 3.1 users who had muscle memory for it. Yet in retrospect it was the correct choice. Should Windows 95 kept as an alternative the Windows 3.1 interface? Forever? Should it still be present in Windows 11 just because there are 10 users who loved it?

I got your point about hidden features and removing them, I sometimes discover wonderful features in software I've been using for years, but at the same time, I accept that progress also means losing things sometimes. You get 10 new features or improvements, you lose 2. Other say "no, I can't lose nothing, I will chose the existing 2 features versus the 10 new ones". That's ok too, they can keep using Windows 10.

[jdsully]

That is 20 years and 2 CEO’s ago. Companies change.

[somehnacct3757]

That's not how Microsoft historically operates. Programs I wrote 20 years ago still work on Windows using ancient apis that they carry, because they care about backwards support. Whatever obscure msc or exe you ran in the 90s is still there, still runs, and still configures the OS. To this day you can check weird boxes deep in your disk config to put buggy behavior back into Windows, because it fixes a bug a single customer had decades ago.

My theory is simpler: it was too hard to add new features to the taskbar (like wide search bars and now text fields) when you had to think about a tall and a wide configuration. So they cut the tall way. Now they can cram all sorts of features into your taskbar.

It represents a shift in MS philosophy, closer to how Apple mandates progress in their OS.

[323]

But backward compatibility it's not the same thing as OS customization.

I also run software last compiled in 1997, but Windows 11 looks a whole lot different than Windows 95 where that software was compiled.