Linked by Thom Holwerda on Wed 14th Sep 2011 22:20 UTC
Windows This is mandatory listening and watching material for understanding the design methodology and ideas behind the Metro interface in Windows 8 (and thus, Windows Phone 7). All this sounds great in theory, and Jensen Harris, one of the minds behind Metro, is clearly passionate about it - and I love people who are passionate about their work. It's just that to me, the Metro UI doesn't seem to work very well for actual work. I want window management! I'm taking all this into account for an article on Metro in the Developer Preview. Stay tuned.
Thread beginning with comment 489698
To view parent comment, click here.
To read all comments associated with this story, please click here.
galvanash
Member since:
2006-01-25

So you think that general-purpose computing has no future and that people should have to learn several interfaces over the course of their life to get stuff done, part of which you consider too complicated yourself ?


Wow. Where did I say that? I think general purpose computing has PLENTY of future, in fact I think it is growing rapidly and the number of people who will actually use a computers in a complex content creation or scientific capacity is just going to keep going up.

At the same time, I recognize that is not the only way computers are used. They are also becoming more prominent as information consumption and entertainment devices. For many, many people this is primarily what they are used for. A UI that is primarily tailored towards this use case seems to make lots of sense to me. I don't see why people see this as a bad thing...

I agree that there's a problem with current GUIs, but I don't think that removing complexity altogether is the solution. Complexity is needed sometimes. What must disappear is unneeded complexity, or complication, and this is a very different problem.


And that is exactly what I see Windows 8 doing. In fact that is what distinguishes it from iOS - it doesn't throw the baby out with the bath water. iOS was a great effort at simplification, and obviously it has worked to some degree - consumers "get" it and like it. But as you say, sometimes complexity is needed - and iOS is too oversimplified. Windows 8 / Metro seems to have a bit more meat on the bone so to speak - there is more to it and functionally I think the UI paradigms it introduces have a lot more potential for effective, even powerful interfaces. But it also has a fallback mechanism. You don't lose the ability to have complex interaction with the machine - it is simply pushed back into a different mode of operation (i.e. classic desktop).

It is 2011. We STILL use CLI interfaces. We will STILL use them 10 years from now, probably more. They are not going away anytime soon. Why? Because they are the best way to interact with a computer for certain tasks. Everyone said GUIs would kill them off, and everyone was wrong. But that doesn't mean GUIs are bad, they are just different and are better for some things than CLIs.

Metro is trying to do the same thing in a sense. Im not saying it is perfect or anything, but it is trying to make a UI suitable for normal people - something they will naturally like to use rather than something they have to work hard at learning to use. It will NOT kill off traditional window oriented UIs for a long time - but that doesn't mean it is bad...

Electric switches can be used to make plane cockpits, that are too complex for untrained peoples. Does it mean that we should get rid of them in every consumer device ?


No. But it does mean we shouldn't model consumer devices after airplane cockpits...

Edited 2011-09-15 13:54 UTC

Reply Parent Score: 2

Neolander Member since:
2010-03-08

"So you think that general-purpose computing has no future and that people should have to learn several interfaces over the course of their life to get stuff done, part of which you consider too complicated yourself ?"

Wow. Where did I say that? I think general purpose computing has PLENTY of future, in fact I think it is growing rapidly and the number of people who will actually use a computers in a complex content creation or scientific capacity is just going to keep going up.

At the same time, I recognize that is not the only way computers are used. They are also becoming more prominent as information consumption and entertainment devices. For many, many people this is primarily what they are used for. A UI that is primarily tailored towards this use case seems to make lots of sense to me. I don't see why people see this as a bad thing...

I wanted to reply to the post where you both said that Metro was the way to go because it was accessible and that the traditional desktop would be necessary for serious content for a long time. Then the tediousness of cellphone browsing made me click a "reply" button a bit to early. Many apologies.

My problem with this philosophy is that it segregates users into two relatively large groups : those who do "complex" content creation (which can be something as commonplace as writing a CV in Word), and those who just use their computer as a consumer device.

This means two parallel codebases to maintain, two unrelated sets of usage patterns to master, two different computer handling philosophies... Even if OSs try to do both, it will always feel schizophrenic and uncomfortable and be harder to learn. Kind of like desktop Linux : you have pretty nice GUIs for some common functionality, but must dive in an unrelated interface based on CLI and text config files for anything advanced.

I see this dual-interface strategy as an attack towards the concept of a general-purpose personal computing device, presenting a unified interface that's easy to grasp for the newcomer and the consumer, but keeping around the ability to manipulate more complex stuff.

"I agree that there's a problem with current GUIs, but I don't think that removing complexity altogether is the solution. Complexity is needed sometimes. What must disappear is unneeded complexity, or complication, and this is a very different problem."

And that is exactly what I see Windows 8 doing. In fact that is what distinguishes it from iOS - it doesn't throw the baby out with the bath water. iOS was a great effort at simplification, and obviously it has worked to some degree - consumers "get" it and like it. But as you say, sometimes complexity is needed - and iOS is too oversimplified. Windows 8 / Metro seems to have a bit more meat on the bone so to speak - there is more to it and functionally I think the UI paradigms it introduces have a lot more potential for effective, even powerful interfaces.

I don't know... Some of the Metro design elements showcased on the article's video seem a bit counter-productive to me. The "feel alive" and "as little chrome as possible" parts in particular leave me wondering. But let's give them the benefit of doubt for now...

But it also has a fallback mechanism. You don't lose the ability to have complex interaction with the machine - it is simply pushed back into a different mode of operation (i.e. classic desktop).

Do we need a different mode of operation ? Can't we clean up the classic desktop until it becomes friendly enough for everyone ? Not a pure content consumption interface, not a mess of power-user features either, just a well-organized and clean generic information processing, I/O, and storage tool, mirroring the genericity of computer hardware.

The beauty of computers are that they can be programmed to do everything. Why equip them with specialized interfaces that restrict this power, even though stacking up enough interfaces side by side would result in something relatively general purpose ?

It is 2011. We STILL use CLI interfaces. We will STILL use them 10 years from now, probably more. They are not going away anytime soon. Why? Because they are the best way to interact with a computer for certain tasks. Everyone said GUIs would kill them off, and everyone was wrong. But that doesn't mean GUIs are bad, they are just different and are better for some things than CLIs.

Well, a pure CLI interface like bash feels, in my debatable opinion, quite out of place in a GUI environment, and the CLI/GUI synergy is one of the main reasons why desktop Linux is not for everyone. The CLI interfaces which I enjoy most today (search engines, IDEs, computational software...) are actually CLI/GUI hybrids, which have managed to keep the speed, purity and power of CLI interfaces, and combine it with the discoverability, information density, and fun factor of GUIs.

Metro is trying to do the same thing in a sense. Im not saying it is perfect or anything, but it is trying to make a UI suitable for normal people - something they will naturally like to use rather than something they have to work hard at learning to use. It will NOT kill off traditional window oriented UIs for a long time - but that doesn't mean it is bad...

It is not generally bad, but I believe that like iOS it is limited in purpose and range of applications. That could prove to be a significant problem for something like desktop Windows, which aims at being a general-purpose OS.

"Electric switches can be used to make plane cockpits, that are too complex for untrained peoples. Does it mean that we should get rid of them in every consumer device ?"

No. But it does mean we shouldn't model consumer devices after airplane cockpits...

Point taken !

Edited 2011-09-15 20:51 UTC

Reply Parent Score: 2

galvanash Member since:
2006-01-25

My problem with this philosophy is that it segregates users into two relatively large groups : those who do "complex" content creation (which can be something as commonplace as writing a CV in Word), and those who just use their computer as a consumer device.

SNIP...

Even if OSs try to do both, it will always feel schizophrenic and uncomfortable and be harder to learn.

Kind of like desktop Linux : you have pretty nice GUIs for some common functionality, but must dive in an unrelated interface based on CLI and text config files for anything advanced.


It depends... Again, I'm not completely sold on Metro - I think it is a good start and this is the right direction and all, but it is lacking in some ways. That said, Windows still has a CLI. When you buy a book on how to use Windows you probably won't find much if any coverage of the command line or its usage. Same with OSX. But it is still there, because it is sometimes really, really useful. Its mere existence is not a barrier to learning the basic (even advanced) usage of either OS (unlike Linux), but there can be many benefits if you want to delve into it and some people do. It was not always like this though - back during the Windows 3.1/95 days saying "you don't need to know DOS at all" was simply not true. Things evolve...

I see the relationship between Metro and the traditional GUI the same way. It will be a long while until anyone can say "you really don't need to mess with the classic desktop", but it will get there eventually. People using Windows say "you really don't need to know the command line" all the time - its mostly true... but its still there and some people cannot live without it.

Do we need a different mode of operation ? Can't we clean up the classic desktop until it becomes friendly enough for everyone? Not a pure content consumption interface, not a mess of power-user features either, just a well-organized and clean generic information processing, I/O, and storage tool, mirroring the genericity of computer hardware.


Well we have been trying for 25 years... And almost every operating system in existence has had to give way to having at least 2 interface modes (GUI + CLI) - even Apple gave up on GUI only... Its not ideal, but it is practical.

The beauty of computers are that they can be programmed to do everything. Why equip them with specialized interfaces that restrict this power, even though stacking up enough interfaces side by side would result in something relatively general purpose ?


I have yet to see a UI that is simple enough for consumer use and powerful enough for expert use... If someone builds one Ill be the first to applaud it - but until then we have to make do.

Well, a pure CLI interface like bash feels, in my debatable opinion, quite out of place in a GUI environment, and the CLI/GUI synergy is one of the main reasons why desktop Linux is not for everyone.


See my previous commment... Windows and OSX and virtually everything else still has CLIs too. If you don't need it and never use it then it amounts to one icon (which you can delete) buried in your start menu - ironically much the same way that the classic desktop is buried in Metro.

It is not generally bad, but I believe that like iOS it is limited in purpose and range of applications. That could prove to be a significant problem for something like desktop Windows, which aims at being a general-purpose OS.


No doubt. I think it has legs though, but I'm just a guy with an opinion. Well see I guess.

Reply Parent Score: 2