Linked by Thom Holwerda on Tue 6th Sep 2005 12:56 UTC, submitted by Anonymous coward
Graphics, User Interfaces This opinion piece gives 8 reasons as to why HCI (Human-Computer Interaction) is in its stone age. It talks of screen corners, visual attention, the spatial paradigm (oh my...) and much more.
Order by: Score:
I don't quite get it
by MikeGA on Tue 6th Sep 2005 13:27 UTC
MikeGA
Member since:
2005-07-22

I don't quite see what this guy is getting at. He seems to be on a bit of a rant about how he dislikes modern GUIs, yet there seems to be no real useful suggestions in there.

Reply Score: 2

RE: I don't quite get it
by QuantumG on Tue 6th Sep 2005 13:34 UTC in reply to "I don't quite get it"
QuantumG Member since:
2005-07-06

That's because he's bitching. He's not trying to make constructive conversation, he's trying to motivate you to think what would be better by pointing out the obvious failings of the modern GUI. Speaking of which, I *hate* one-knob faucets.

Reply Score: 3

Four corners
by Anonymous on Tue 6th Sep 2005 13:35 UTC
Anonymous
Member since:
---

SymphonyOS, that now is on Alpha4 stage, is built upon the concept of the four corners of the desktop doing different work for the OS (computer view, program view, task view, search view).
Read the design laws on the website
www.symphonyos.com and take a look to the screenshots.

Reply Score: 0

to paraphrase:
by Anonymous on Tue 6th Sep 2005 13:37 UTC
Anonymous
Member since:
---

He seems to be saying, "Don't you hate how computers are simple yet complex when they should be complex yet simple? After all these years, the morons at Microsoft/Apple haven't even invented a computer that will wipe my bottom after going potty. Why doesn't everyone just recognize the wisdom of my untested theories that I've read off of other people's weblogs? Everyone is a moron but me." And that's kind-of it.

Reply Score: 5

RE: to paraphrase:
by Thom_Holwerda on Tue 6th Sep 2005 13:42 UTC in reply to "to paraphrase:"
Thom_Holwerda Member since:
2005-06-29

Erm, this guy has some valid points-- it isn't just a mindless rant.

Especially the screen corner thing is perfectly valid; this is derived from Fitts' Law-- the only constant in these matters. Read more on Fitts' Law here: http://en.wikipedia.org/wiki/Fitts_Law .

The visual attention section is another very valid point.

Reply Score: 5

RE[2]: to paraphrase:
by QuantumG on Tue 6th Sep 2005 13:47 UTC in reply to "RE: to paraphrase:"
QuantumG Member since:
2005-07-06

Ya know, as I was reading his "valid" point about visual attention I was scrolling his blog with my mouse wheel, which requires no visual attention because my muscle memory knows how to find my mouse without looking, locate the mouse wheel and invoke it without me having to take my eyes off the text. It also happens to work on every application I run.

Reply Score: 1

RE[3]: to paraphrase:
by Thom_Holwerda on Tue 6th Sep 2005 13:52 UTC in reply to "RE[2]: to paraphrase:"
Thom_Holwerda Member since:
2005-06-29

Ya know, as I was reading his "valid" point about visual attention I was scrolling his blog with my mouse wheel, which requires no visual attention because my muscle memory knows how to find my mouse without looking, locate the mouse wheel and invoke it without me having to take my eyes off the text.

Yeah, but first you had to locate your browser launcher (requires eye/head movements and attention shifts), then you had to locate your OSNews bookmark (requires eye/head movements and attention shifts), and when you finally got to the article, you had to put your mouse pointer inside the content of the window (requires eye/head movements and attention shifts).

But yeah, once you got up to that point you can scroll without eye/head movements and attention shifts. Good for you ;) .

Reply Score: 5

RE[4]: to paraphrase:
by whenney on Tue 6th Sep 2005 15:53 UTC in reply to "RE[3]: to paraphrase:"
whenney Member since:
2005-07-06


Yeah, but first you had to locate your browser launcher (requires eye/head movements and attention shifts), then you had to locate your OSNews bookmark (requires eye/head movements and attention shifts), and when you finally got to the article, you had to put your mouse pointer inside the content of the window (requires eye/head movements and attention shifts).


Unless he was using QuickSilver, in which case it would just be "Cmd-Space O S" to get to the OS news website from anyhere, even if you were starting from a different virtual desktop. No eye/head movements at all :-)

I guess the point is that many of the things that he complains are lacking have been implemented, just not on the systems he happens to have used.

Another example of a well-designed UI is emacs. Since everything can be achieved via key presses, there is no need to ever divert your visual attention (well, OK, sometimes you have to look down at the minibuffer but that is always in the same place). Unfortunately, it is rather difficult to navigate the new-look OSNews site inside emacs. For some reason, the "Reply" links don't show up in w3m. Any chance of fixing that Thom?

Reply Score: 1

RE[4]: to paraphrase:
by ma_d on Tue 6th Sep 2005 17:59 UTC in reply to "RE[3]: to paraphrase:"
ma_d Member since:
2005-06-29

alt+f2
firefox
ctl+l
osnews.com
tab [ad infinitum]
enter
page-down

Look ma', no mouse ;) .

Reply Score: 2

RE[3]: to paraphrase:
by StephenBeDoper on Tue 6th Sep 2005 21:04 UTC in reply to "RE[2]: to paraphrase:"
StephenBeDoper Member since:
2005-07-06

Without having read the article, I would like to point out that the existence of a counter example does not invalidate his point. HCI especially is far too general for there to be concrete rules and formulas that govern every possible situation.

For example: I could design a UI that adhered to Fitt's Law perfectly, but is still awful from a usability point of view. Would that disprove Fitt's Law? No, it would merely be a bad application of it.

Calling these things "laws" is somewhat misleading, there aren't really any laws in the typical sense when it comes to design; but there are general guidelines that can be applied to help make design decisions in specific situations.

Fitt's Law may tell you that UI elements in certain locations can be reached faster/more accurately, but it says SFA about what particular UI elements should go in those locations.

Reply Score: 2

RE[2]: to paraphrase:
by ma_d on Tue 6th Sep 2005 17:58 UTC in reply to "RE: to paraphrase:"
ma_d Member since:
2005-06-29

The argument over screen corners has nothing to do with fitt's law. It has to do with whether you intentionally hit screen corners or often accidentally hit screen corners; since you aren't clicking the corner intention is non-derivable. And if you attach a time for the derivation than you've got people complaining about a time they waste.
So you're limited to instantly reversable actions in association with screen corners. Which is what OS X does, and I think it's off by default anyway.
You can do more, but you'd better keep it off by default or else your environment is handicapped for new users: "I just let go of the mouse [the mouse moves via cord pull] and a web browser opened..."

The guys makes note of issues we're all aware of. He doesn't bother to mention fixes for them.
I wouldn't call it drivel, but it's not much more than that..

Reply Score: 1

RE[3]: to paraphrase:
by cheriot on Wed 7th Sep 2005 03:32 UTC in reply to "RE[2]: to paraphrase:"
cheriot Member since:
2005-08-30

"The argument over screen corners has nothing to do with fitt's law."
Incorrect. Since the mouse can not move beyond the corner, they essentially have an infinate length/hieght, which gives a higher value to buttons placed in the corner.

"since you aren't clicking the corner intention is non-derivable"
I agree completely. Using the placement of a mouse without a click to derive intention seems really dumb. If you want the action then a click will be faster. If you don't click then you don't want the action and won't get it. Unfortunately, menu designers disagree so we're stuck trying to read the menu and keep the mouse from wondering to the wrong spot and changing the menu.

Reply Score: 1

RE[2]: to paraphrase:
by Anonymous on Wed 7th Sep 2005 10:13 UTC in reply to "to paraphrase:"
Anonymous Member since:
---

aga ;)

Reply Score: 0

Good, but too simplistic...
by nathan_c on Tue 6th Sep 2005 13:41 UTC
nathan_c
Member since:
2005-07-12

This post is actually quite good, and the general opinion of "hey, the UI hasn't changed in years and could be a lot better" is really true. Unfortunately though, there are numerous problems to overcome before implementing a better UI, and one of them is change. Computers and now mainstream and changes must be evolutionary in nature and revolutionary in nature. People are creatures of habit and don't want to change (if you disagree with this, think of the top five things you decided to change on New Years... has it happened?). Here are my thoughts on his post:

1. Corners are the most accessible part of the screen. However, I have OSX and have Expose keyed to the corners. My wife (a Windows user) hates it because she hits of the corners by accident and unintentionally fires Expose. Why? Because corners mena nothing to her. Most people will find corner actions extremely unfriendly for this reason. They are stuck in a habit and no amount of tell them how wonderful it is will stop them from griping.

2. So is it possible to design a system that's suits both beginners and professionals?

Yes, it is, but the author thinks that the GUI should be expanded. However, by nature, GUI offer very little power compared to the command line and someone who knows how to use it. Unlike the modern GUI, the command line (in the unix world at least) has been making leaps and bounds in terms of power. The possibilities are endless. A GUI that integrates key components of a command line would be a vastly superior tool than current GUIs.

3. I bet you my bunny the former Soviet union could have designed a better operating system GUI than any of the software vendors of today.

Maybe, maybe not. In an ideal world, having a best-of-breed OS with one GUI and the best preferences would work. Well - it would work for the people who designed it. However, people work differently from each other, and choice is a huge part of human productivity. Just like no tool fills all roles, no tool fits all humans. The key here is to provide the best preferences for most people, and then give the others the extensibility to easily change and mold it to their own needs. His example of the kitchen is very valid, but you have to remember, not all kitchens are alike - and spatial awareness only work when you are familiar with it. Forcing familiarity by limiting customization just alienates those for which the defaults don't work.

So, my opinion is that the GUI needs lots of work, but it is naive to think that it will change rapidly because of people's fear of change. It is wishful thinking to think that one GUI with no customization will provide better productivity because humans themselves are different. Also, there's no getting around it. The CLI is still an extremely flexible interface and I would have a very hard time thinking of any GUI paradigm that could replace its power. A hybrid approach between GUI and CLI is probably a better and more realistic solution.

my 2 cents...

Reply Score: 5

rcsteiner Member since:
2005-07-12

Some examples from the OS/2 WorkPlace Shell include:

* The ability to extend any folder context menu via drag and drop (including the desktop context menu), which is very nice for making quick context-sensitive menus of commonly used programs.

* The ability to place a series of programs and data files in a single folder and mark the folder as a work group. This groups those elements logically, and the desktop will automatically open or close all of those programs and documents when the folder itself is opened or closed. Very useful for programming environments when you might want to open an IDE, a couple of shell windows, and various help files and other assistants.

* The ability to create, manipulate, and destroy desktop folders and other objects from scripts on the command line (in OS/2's case, it's done via Rexx).

Those features are all over ten years old.

Also, w.r.t. the command line: it hasn't been exactly standing still in the DOS, Windows, or OS/2 worlds either, at least if you've ever used 4DOS, 4NT, 4OS2, or any of the other JP Software shells available.

The fact that Microsoft hasn't seen fit to extend that to its standard shells is more a testament to Microsoft's stagnation than anything else.

Reply Score: 1

rcsteiner Member since:
2005-07-12

I never hit Submit???????

Please delete the first instance of my reply. Thanks!

Reply Score: 1

rcsteiner Member since:
2005-07-12

Some examples from the OS/2 WorkPlace Shell include:

* The ability to extend any folder context menu via drag and drop (including the desktop context menu), which is very nice for making quick context-sensitive menus of commonly used programs.

* The ability to place a series of programs and data files in a single folder and mark the folder as a work group. This groups those elements logically, and the desktop will automatically open or close all of those programs and documents when the folder itself is opened or closed. Very useful for programming environments when you might want to open an IDE, a couple of shell windows, and various help files and other assistants.

* The ability to create, manipulate, and destroy desktop folders and other objects from scripts on the command line (in OS/2's case, it's done via Rexx).

* the ability to create third-party extensions to existing desktop functions (objects) which directly inherit characteristics from the parent class on which they are based, and which in turn can be parents of additional third-party classes with inheritable attributes. This allows Joe Programmer to create an FTP class based on a folder class, and then turn around and modify the original folder class in such a way as to impact both the original desktop folders and his FTP folder derivative.

Those features are all over ten years old.

Also, w.r.t. the command line: it hasn't been exactly standing still in the DOS, Windows, or OS/2 worlds either, at least if you've ever used 4DOS, 4NT, 4OS2, or any of the other JP Software shells available.

The fact that Microsoft hasn't seen fit to extend that to its standard shells is more a testament to Microsoft's stagnation than anything else.

Reply Score: 0

nathan_c Member since:
2005-07-12

You are right ... huge enhancements to the GUI can be made and should be - and it's a testament to people's unwillingness to change (or companys that are unwilling to implement) that OS/2 has had many great usability features for years, but the major desktop makers ignore these only to add cool new features like transluscency (?) for title bars and such. What about BeOS and its shell features (live queries, etc)? Also rather canned - well at least some of them are being picked up.

And the command line - yes, it's definitely progessing in other worlds (other than unix) too. However, MS hasn't extened it because of their stagnation and deminuation of the CLI - but that's just the point. MS has claimed that the GUI is the supreme user interface, when in reality, it's not. MS could do tons of things to their GUI to make it more powerful (and they should - corners would be a nice place to start :-)), but in the end, power users will still rely on a good CLI for tasks that require the most flexibility and extensibility. That's why I think an integrated environment between CLI and GUI will be the ultimate solution. We'll have to see if anyone will implement such a thing or how it would work...

Reply Score: 1

rcsteiner Member since:
2005-07-12

Overlapping windows are nice for expert users who need to view multiple things at once, but a typical user only needs to view the documents they work on and perhaps two programs at one time.

Why not create a GUI which has the following simple features:

(1) A Games page (fullscreen with pretty buttons)

(2) An Applications page (fullscreen with pretty buttons)

(3) A Documents page (fullscreen with pretty buttons)

(4) The ability to split the screen to view two programs at the same time (vertical or horizontal split can be toggled)

That gets rid of a lot of the typical desktop UI complexity like icons and folders, but allows most office or home users to do what they want.

Even though I like to think of myself as an advanced user, I send to set up my Win95 OSR2 boxes in this way using QuikMenu 4 as the shell. Good for LAN parties. In order to get to things like the filemanager or the control panel, one has to get to the password-protected Admin menu. :-)

Reply Score: 1

hobgoblin Member since:
2005-07-06

hmm, quikmenu 4 reminds me of something im using here, powerpro.

www.ppro.org

Reply Score: 1

rcsteiner Member since:
2005-07-12

The site for QUikMenu 4 is here:

http://www.neosoftware.com/qm4.html

Oldtimers might recognize the name -- QuikMenu III was a very nice GUI menu system for DOS:

http://www.neosoftware.com/qm.html

Reply Score: 1

hobgoblin Member since:
2005-07-06

err, i was just saying that powerpro and quikmenu looked like they had many of the same capabilitys.

atleast that was my impressing after a quick readthru of the page linked to in the orignial post.

i was not saying that they where the exact same program, atleast that was not my intent.

Reply Score: 1

RE: Good, but too simplistic...
by DHofmann on Wed 7th Sep 2005 00:42 UTC in reply to "Good, but too simplistic..."
DHofmann Member since:
2005-08-19

However, by nature, GUI offer very little power compared to the command line and someone who knows how to use it.

It doesn't have to be this way. It's just easier to write a command line app with all the features you want than to design a good GUI that has the same features.

Reply Score: 1

This guy isn't an expert -
by e2mtt on Tue 6th Sep 2005 13:41 UTC
e2mtt
Member since:
2005-08-08

I agree with the bit using the screen corners... a long-time pet peeve of mine, and you can't even find 3rd-party apps that do anything about it.

About not needing or allowing customization is just wrong though...(beside somewhat contradicting item #2) People with different needs on their PC NEED different configurations to do their things best. No other industries' experts, anywhere in the whole world, are so set against custom solutions. From personal items to industry, today's competitive edge is all about customizing products and services to fit the user/consumer. Many HCI "experts" seriously need to get off their high horses and stop degrading customizable. (However, good defaults are essential!)

One last item... People who bash Microsoft with the "click Start to shut down" mockery aren't clever at all. So maybe they wish MS used some other name for the "start" menu, but I really can't think of a better place to put the shutdown/logoff buttons then where they are, considering how the rest of the system is set up. After all, I don't think we need a entire dedicated "shutdown" menu.

Reply Score: 1

v RE: This guy isn't an expert -
by Anonymous on Tue 6th Sep 2005 13:53 UTC in reply to "This guy isn't an expert -"
Thom_Holwerda Member since:
2005-06-29

Ookaze,

Any OS or desktop environment has its oddities. How about the Gnome Places menu? It contains a link to 'Find files'. Since when is a search dialog a place? Since when is a file a place?

Or what about the dock in OS X? Apps with a link insie the dock do not get a 'taskbar' entry, yet apps that aren't linked to from the dock have... Right, also makes a lot of sense.

I could name dozens, if not hundreds more of these oddities. They exist *everywhere*.

It seems to me *you* are way too involved in the anti-MS way.

Reply Score: 5

Anonymous Member since:
---

How about the Gnome Places menu ? It contains a link to 'Find files'. Since when is a search dialog a place? Since when is a file a place?

Let me answer : 'Find files ...' is not a place, but it sure is place related. And a file is a place since, well, first UNIX ? In Unix, everything is a file, even directories (places). Even Windows uses zip files as places, navigating through them (this confuses newbies BTW). Some companies (MS ?) put everything and the kitchen sink in a file (Office documents) like it's a place. Well, with Vista (and already with BeOS or OSX ?), even 'search criteria' are places !!!
So no, I never thought of 'Find files ...' as not being well place in 'Places' menu. Find another flaw.

I could name dozens, if not hundreds more of these oddities. They exist *everywhere*.

If they are of the same nature as the one for Gnome, I foresee most are actually not oddities.

It seems to me *you* are way too involved in the anti-MS way

Perhaps that's true, but as long as it allows me to see a better design than a shutdown in a start menu, it must be a good thing then. So thanks.

Reply Score: 1

Thom_Holwerda Member since:
2005-06-29

'Find files ...' is not a place, but it sure is place related.

Well, you have to start the shutdown sequence, so in that sense shutdown is most definitely related to start.

Better argument, please.

Reply Score: 5

Anonymous Member since:
---

'Find files ...' is not a place, but it sure is place related.
Well, you have to start the shutdown sequence, so in that sense shutdown is most definitely related to start.
Better argument, please.


Why a better argument ? I never said 'shutdown' was not at its place in the Start menu (I could have though), I just said there are other far better places to put it in, answering someone who could not think of a better place (talk about narrow thinking).
So the usability of the places menu is very good IMHO, and ...
What I think of the Start menu, related to usability ? You give a big evidence. By your words, we can put in the Start menu, everything that can be started. Which means nearly anything ... oh wait ! Starting preferences ? What nonsense is that ? Anyway, that explains why the Windows Start menu is such a mess, and virtually unusable once you have some apps.

Reply Score: 0

RE[2]: This guy isn't an expert -
by e2mtt on Tue 6th Sep 2005 14:14 UTC in reply to "RE: This guy isn't an expert -"
e2mtt Member since:
2005-08-08

@ Ookaze

...brain eaten by the MS way? Yours may be eaten by anti-MS animosity.

My reasons: I don't use the shutdown button often at all... maybe once a week? I am sure that my system gets restarted by an Update, newly installed program, or the reset button :-) more often then the standard shutdown menu. Therefore, I don't think shutdown needs a desktop or main taskbar icon. Placing it one click deep in a main menu, rather then using precious desktop space, seems about right to me.

Reply Score: 1

Anonymous Member since:
---

...brain eaten by the MS way? Yours may be eaten by anti-MS animosity.

Which seems to be better, as I can at least think of a better place for a shutdown button, than buried in a Start menu.

My reasons: I don't use the shutdown button often at all... maybe once a week? I am sure that my system gets restarted by an Update, newly installed program, or the reset button :-) more often then the standard shutdown menu.

So you were right. The fact is that I also rarely shutdown my session (once a month ?), but I have put the icon on one bar anyway. I rarely use the menu (thanks to sessions and spatial navigation), everything is there and I want the shutdown to be just there too. Using the menus has become tedious to me. I've done the same on my wife's desktop, but that's mostly because the shutdown button is associated with the lock screen button.

Therefore, I don't think shutdown needs a desktop or main taskbar icon. Placing it one click deep in a main menu, rather then using precious desktop space, seems about right to me.

Perhaps it does not need it, but for usability it is far better, and it is a possible location. And it does not steal precious desktop space. It is at least more useful than most icons on the taskbar. Most newbie user grasp instantly how to shutdown the computer with such an icon, it is the same icon than on their stereo. I can't say the same about Windows shutdown sequence ... That's why people shut their computers down with the power button back in the Win9x days, it was more obvious !

Reply Score: 0

rcsteiner Member since:
2005-07-12

...which strikes me as a perfectly logical place to put it.

Reply Score: 1

Thom_Holwerda Member since:
2005-06-29

Ideally, there shouldn't be a 'shutdown' option at all anywhere in a GUI. Any computer should shutdown (and bootup) instantly when the power button is pressed. Just like any other appliance. And yes, computers are just that-- appliances. Complicated- yes, but to most people not more important than say a TV or stereo set.

Computers are severely lacking in that area. Sleep functions alleviate it a bit, but is by no means enough. Instant on and off, that's the future.

Reply Score: 5

e2mtt Member since:
2005-08-08

Ideally, there shouldn't be a 'shutdown' option at all anywhere in a GUI...

Good Point.

Of course I would put it in somewhere, or at least make it possible to create a shortcut/link that would shutdown, but you are right.

Why not just be able to push the power button off?

Reply Score: 1

Bnonn Member since:
2005-09-02

I just tried hitting the power button on my Ubuntu box running a fairly standard Intel board, and it had the effect of executing shutdown -h now.

Windows does the same thing. It's called advanced power management. Just because you can hit the power button, though, doesn't mean you don't want to be able to select to shut down via software. What if you can't easily reach the power button? What if you want to restart or hibernate? There are good reasons for having a shutdown menu.

Reply Score: 1

StephenBeDoper Member since:
2005-07-06

If you're running XP (or, I believe, 2k) on a motherboard with decently modern power management support, you should be able to select what action Windows takes when you press the power button (shutdown, hibernate, suspend, etc).

Reply Score: 1

rcsteiner Member since:
2005-07-12

One could almost say "infinitely more complex", since a stereo tends to have a very fixed set of functions that it performs, and it tends to have all of its software embedded in firmware (if it has any at all).

A personal computer, on the other hand, can run an array of software that is impossible to predict -- take any 100 home computer users and compare their machines, and the chances are very good that no two machines will be the same in terms of the software that is installed (and the programs which are the same between them will almost certainly differ somewhat in configuration).

As long as personal computers have both dynamic and permanent memory (i.e., RAM and disk), I suspect you'll see some requirement for an explicit "shutdown" while cached information is copied to permanent storage.

I suppose the shutdown function *could* be incorporated into the hardware to hide it from the end user, but the function would still be there...

Reply Score: 1

ohbrilliance Member since:
2005-07-07

"Ideally, there shouldn't be a 'shutdown' option at all anywhere in a GUI. Any computer should shutdown (and bootup) instantly when the power button is pressed."

You're assuming the computer is within close proximity.
And besides, instant shutdown is hardly friendly if you didn't intend to shut down. At least software shutdown will prompt you for confirmation and actions on unsaved files.

I've experienced Windows going into a prolonged but irretrievable shutdown upon hitting the power button, and it completely threw my workflow as it closed multiple applications and documents.

So, one man's obvious ideal is not necessarily that of all.

Reply Score: 1

v more of a rant ...
by Anonymous on Tue 6th Sep 2005 13:42 UTC
RE: more of a rant ...
by ma_d on Tue 6th Sep 2005 18:02 UTC in reply to "more of a rant ..."
ma_d Member since:
2005-06-29

A mouse gestures. One of the very few things you can gain muscle memory for on a mouse (other than the obvious pushing of buttons).
Good point. Your point was about 8 times more informative than the article.

Reply Score: 1

re: more of a rant ...
by Anonymous on Tue 6th Sep 2005 13:48 UTC
Anonymous
Member since:
---

brightside (http://live.gnome.org/PowerUserTools) lets you configure the corners under Gnome. So now you can scratch 1 too ;)

Reply Score: 1

Ewww
by edwdig on Tue 6th Sep 2005 13:50 UTC
edwdig
Member since:
2005-08-22

I can't stand HCI freaks. By freaks, I don't mean, say, people complaining about Linux UI's, but people who obsess over screen corners and things like that. They rant and rave about current UIs sucking, and give incredibly vague guidelines to "improve" things without thinking about the overall picture.

Responding to the points individually:

1. Screen corners may be easy to reach, but they are completely detached from everything. Generally speaking, you are looking in the center of the screen, which makes the corners the hardest things to see. We'd have to significantly change our current UI's to indicate the corners would be useful, but the article gives no suggestions on doing that. Nor does it give any examples of what features should be activated by the corners. No, "check email in the screen corner" doesn't even make sense.

2. This is just an opinion the author is stating to get across a silly image in your head. He's not saying anything to justify it, or suggesting anything to change it. Pointless.

3. This one is a bunch of little points.

As for every little detail requiring you to pay attention to it, that's a cross of two things. One is the sheer number of things you can do requiring screen space, limiting the area you can give to each button. The other is the lack of tactile feedback - maybe if the mouse could somehow offer resistance when you moved the cursor over a button, you could pull off something, but I don't see that one as likely.

Having to finding your position in text every time you do something that obviously would require the text to move. Current OS's already do their best to minimize the need. If I resize my Mozilla window, it will do its best to keep the same section of text onscreen. If I hit page down, the last lines of the screen will get moved to the top so that I have some overlap to find my place with.

Space bar scrolling - Is he unaware of what the Page Up and Page Down keys do? Unless you're using a crappy laptop keyboard (read: most laptop keyboards), th Page up and Page Down keys are some of the easiest keys on the keyboard to find without looking or groping around. I'm also amazed that he thinks scrolling is a more important feature than typing text when you're in a text entry field.

4. There's not really a point here, just an observation.

5. You gotta love someone ranting how all UI's suck, but this would be so much better complaining that we don't need choice.

6. You can't rely on spatial for two reasons. First is the lack of tactile feedback. Knowing the forks are on the left doesn't prevent you from looking for them unless you can also feel a difference between the left and right side of the draw they're in. You also need to be able to feel that what you're touching is actually a fork. The other reason spatial doesn't really work is due to limited screen space. Things constantly overlap each other, or change due to context. That doesn't happen in your kitchen. The fork draw doesn't become the pencil draw when you start writing a letter. The two never overlap, which is why you can recognize them spatially.

7. Those are really just marketing terms used to sell stuff to people afraid of computers. From people actually designing applications, you'll hear terms such as "document oriented" or "task oriented".

8. There is no example of how he thinks it should work instead. If you've got a decent email app that supports HTML mail, you could do the photo operations, copy the image to the clipboard, and paste it in the email. The app would automatically take care of the attachment part he seems to despise.

Reply Score: 3

RE: Ewww
by QuantumG on Tue 6th Sep 2005 13:57 UTC in reply to "Ewww"
QuantumG Member since:
2005-07-06

As SymphonyOS has shown, UI designers can actually produce something now and then. If only every usability expert got out a scripting language like Python and started coding instead of bitching. I really don't get his comments about single knob faucets. Is he actually trying to suggest they are a good idea? I hate them.

Reply Score: 2

RE[2]: Ewww
by edwdig on Tue 6th Sep 2005 20:51 UTC in reply to "RE: Ewww"
edwdig Member since:
2005-08-22

What's bad about single knob faucets? I really don't see a disadvantage. Nothing drives me more crazy tho than dual knob showers... damn near impossible to keep the water temperature reasonable.

Reply Score: 1

Try Litestep or DesktopX !
by e2mtt on Tue 6th Sep 2005 14:00 UTC
e2mtt
Member since:
2005-08-08

These kind of articles come along all the time...

I now throw out a challenge to all budding Usability experts... put together a desktop using Litestep or DesktopX or one of the many shells/desktop replacements out there, either by figuring it out themselves or collaborating with one of the very talented skinners out there, and post it for us to try out. I know Litestep supports corner activations. (An example of someone doing this is the Developer ObjectBar skin by Scott MacDonald)

If someone comes up with a good one, I would use it and even buy it, but most custom shells/skins you download are even worse for usability then stock Windows.

Fell free to integrate it with other apps too, custom written or available such as Approcket, I look forward to trying it out.

Of course many Mac users combine Expose with the freeware Quicksilver app, and feel they have found interface perfection already.

Reply Score: 1

OSS Offers Best Vehicle
by Anonymous on Tue 6th Sep 2005 14:05 UTC
Anonymous
Member since:
---

Nice rant (not).

Okay, so the GUI environment we use today is stale. So is the filesystem. We all know that.

Here's my argument: in all the years of OSS and Linux, these problems haven't been solved. Yes, you can find some examples of fixes to the GUI here and there, like the aforementioned Brightside. But nothing revolutionary. Every bit of OSS that's really worth using is another implementation of what's already out there. Heck, I had a four-corners thingy for Win 3.1 over 10 years ago, and I had one for OS/2 Warp as well.

So where's the OSS beef? Why haven't we, as a community, produced the ground-breaking next-gen anything? OSS provides the perfect vehicle for it.

Until the OSS community produces something fresh and new, the bulk of computer users are going to ignore Linux on the desktop.

Reply Score: 1

RE: OSS Offers Best Vehicle
by ma_d on Tue 6th Sep 2005 17:54 UTC in reply to "OSS Offers Best Vehicle"
ma_d Member since:
2005-06-29

Stale being used as a synonym for tried and true...

Reply Score: 1

RE[2]: OSS Offers Best Vehicle
by rcsteiner on Tue 6th Sep 2005 19:29 UTC in reply to "RE: OSS Offers Best Vehicle"
rcsteiner Member since:
2005-07-12

Just because folks are happy with the way things are now doesn't mean new things can't be added as options.

If we were stuck only using tried and true tools, everyone would be using things like vim and bash, and that would be a UI nightmare for those of us brought up with CUA interfaces and 4DOS. :-)

Reply Score: 1

Four corners on tablet
by Anonymous on Tue 6th Sep 2005 14:15 UTC
Anonymous
Member since:
---

Four corners of the screen although once very useful, is pretty useless for machines with more than one monitor or (most especially) tablet pcs.

Reply Score: 1

Down With Corners!
by Anonymous on Tue 6th Sep 2005 14:22 UTC
Anonymous
Member since:
---

Personally, I don't really use Expose on my Mac - I'm certainly not keen on corners of the screen triggering things.

It's an interesting article, let down by the slightly ranty nature and the fact that, once again, the theory really needs to be backed up with practice. That's the fundamental problem with articles like this - it's okay to an extent, but really it needs working prototypes to be of any use to others. Someone needs to go past the theory and ranting and actually *do* something. Instead, the very people the author seems to bitch about, like Apple, are the very guys doing the actual work and producing usable interfaces.

The other issue is a personal one, I've read many "future of UI" articles over the years, and some of the UIs really work against my personal preferences. Interfaces involving speech and audio really, really get on my nerves - as do those that have minimal customisation.

Reply Score: 0

OT: Language Nazi
by Anonymous on Tue 6th Sep 2005 14:30 UTC
Anonymous
Member since:
---

The term is "opinion piece" not "opinional piece". There is no such word "opinional" in the English Language.

You just can't go around makening up your own wordzels.

Reply Score: 1

RE: OT: Language Nazi
by Thom_Holwerda on Tue 6th Sep 2005 14:36 UTC in reply to "OT: Language Nazi"
Thom_Holwerda Member since:
2005-06-29

I'm not changing it, due to your post. You can do this politely, or don't do it at all, please. If you would've kindly asked, I would have changed it.

Btw, English is not my native language. I'll talk to you once *you* can run a newssite in a to you foreign language.

Reply Score: 5

Consider this a more polite request.
by rcsteiner on Tue 6th Sep 2005 15:18 UTC in reply to "RE: OT: Language Nazi"
rcsteiner Member since:
2005-07-12

Please change the word "opinional" to something more appropriate, like the suggested "opinion".

The use of poor grammar and vocabulary in OSNews articles and headlines reflects badly on the site as a whole.

Thanks!

Reply Score: 3

RE[2]: OT: Language Nazi
by Anonymous on Tue 6th Sep 2005 19:39 UTC in reply to "RE: OT: Language Nazi"
Anonymous Member since:
---

>>Btw, English is not my native language.<<

Obviously, given the fact that you cannot differntiate rudeness from a joke! =)

I'm a native New Yorker, so most of what I say and how I say it seems rude, except to other New Yorkers. It's the curse we live with being superior to the rest of youse guys.

I was not trying to be rude, or insult you. I was trying to point it out in an amusing manner. Please accept my most humble apology.

Reply Score: 0

RE[2]: OT: Language Nazi
by Anonymous on Tue 6th Sep 2005 15:54 UTC in reply to "OT: Language Nazi"
Anonymous Member since:
---

Well.. language is not static. I've come up with many new words in danish and so have many others. Language always changes. What was wrong yesterday is right tomorrow.

So to Thom_Holwerda > Congratulations ;) You've invented a new word ;)

And to the anonymous language nazi > Go home and lie down, and take another beer.. you clearly need it :p

dylansmrjones
kristian AT herkild DOT dk

Reply Score: 0

RE: OT: Language Nazi
by Anonymous on Tue 6th Sep 2005 20:23 UTC in reply to "OT: Language Nazi"
Anonymous Member since:
---

This is a bit off-topic. My apologies. But I wanted to share an amusing, but enlightening experience I had recently which relates.

I was reading an article written in English on a web site based in India. The author had used the term "loose" to mean "lose" which is one of my language related pet peeves. I emailed the author, politely pointing out the misuse of the word. He emailed me back, politely pointing out that in India, "loose" is the spelling in common use. He refrained from going so far as to point out that the spelling "loose" actually makes more phonetic sense than "lose", but his response reminded me that it really does.

English (for better or worse) has become the language that different parts of the world use to communicate with each other. This is convenient for us native English speakers because the rest of the world has to learn our native language rather than the other way around.

The flip side is that English does not belong to us, alone, anymore. Like OSS, it belongs to the world.

So these days I suppress my language nazi tendencies. ;-)

Reply Score: 0

Some points are right
by ma_d on Tue 6th Sep 2005 14:32 UTC
ma_d
Member since:
2005-06-29

4.) This is blatantly wrong. If it were the same thing it'd be more confusing and annoying.
5.) Preferences are the only way to accomplish the task of "being easy and growing with the user." Sane defaults and rediculously powerful, but slightly hidden, user preferences keep the beginner's timidity down and the power users happy to change things.
6.) That'd be true, if I couldn't move across the screen within 1cm of mouse movement. Typing into a terminal is more likely to create muscle memory for things.
7.) That's a low blow ;) .
8.) I agree with this one totally. There's just still not a completely done clipboard system. Windows comes close, I think you can actually paste an image into outlook there?

Reply Score: 1

email photo task
by rightWingNutJob on Tue 6th Sep 2005 15:10 UTC
rightWingNutJob
Member since:
2005-07-07

The task about rotating a photo, shrinking it, and sending it to someone is quite easy if you are using a Mac. Assuming you are looking at the picture you want to send in iPhoto:
1. Click the photo.
2. Click the rotate button.
3. Click the email this photo button.
4. Select the size you want from the dialouge.
5. Click the compose button in the dialouge.
6. Type the email address (or the name, if you have them in your address book)
7. Type your optional message.
8. Click the send button.

Or does he want a magic button that knows how the rotate the photo correctly, knows what size you want, and knows who you want to send it to without asking?

Reply Score: 1

RE: email photo task
by nathan_c on Tue 6th Sep 2005 21:13 UTC in reply to "email photo task"
nathan_c Member since:
2005-07-12

Or does he want a magic button that knows how the rotate the photo correctly, knows what size you want, and knows who you want to send it to without asking?

From his example at least, it seems like he wants voice activated commands. :-) Hmm, that's very far from being part of a GUI (more like a AUI - audible user interface. However, I would love it if speech recognition was at the point in modern mainstream operating systems to be really useful and *integrated* - ie, why am I having to type these comments - it's 2005 already).

Reply Score: 1

i just
by present_arms on Tue 6th Sep 2005 15:24 UTC
present_arms
Member since:
2005-07-09

hit the power button on my sons dell running XP and it logged off and computer duly turned off,,, i did the same on my linux box (i did have to tell linux to do this if power button was hit though) so it seems u can just turn the computer off safely with the power switch

Reply Score: 1

v ooooh i need a subjezxct so here it is
by Anonymous on Tue 6th Sep 2005 15:39 UTC
This is not HCI...
by Anonymous on Tue 6th Sep 2005 15:54 UTC
Anonymous
Member since:
---

HCI is an academic discipline. About 20-30 years ago the advice of "usability gurus" and what was going on in HCI were quite similar; that's not true anymore. The most important development in HCI was they stopped timing people doing tasks and looking for ways to shave microseconds off those tasks and started doing something useful. They started looking at workflows and context and how people actually use their computers. And to cut a long story short, the consensus of modern HCI research is, ironically, that customisation is fundamental to human-computer interaction. The user needs to be able to integrate their computer use with their workflow. That is why professional applications are highly customisable.

The other major result was that computer users are highly sophisticated, but only within their niche. Business users might be highly capable in some areas of Excel but know nothing of basic file management. Being able to customise your workspace to your domain-specific task is an important part of any human task and, it turns out, it's no different with computers. There's a ton of research on this. Meanwhile, so-called "usability experts" go on and on about "simplicity" (i.e., crippling the interface) based on rumours they've heard about unpublished tests Apple did in the 80's. It's ridiculous.

Reply Score: 5

RE: This is not HCI...
by hobgoblin on Tue 6th Sep 2005 16:17 UTC in reply to "This is not HCI..."
hobgoblin Member since:
2005-07-06

so basicly we should have a system where your desktop enviroment is basicly as rewriteable as a webpage. and where you have the ability to take your work enviroment with you on some storage media so that you can sit down on any machine and still have your enviroment rather then having to figure out the other users enviroment.

think about this: a sysadmin gets a call about a faulty office desktop. when he arrives he pops in his own enviroment, one customized for diagnostics. the moment its loaded he can do a visual check for most common fault. this just by looking at what the desktop config is displaying. this would allow him to speed up the diagnistics work.

so if nothing shows up there is one of two posibilitys. its a uncommon fault, or the user have done something strange to his personal setup.

Reply Score: 2

RE: This is not HCI...
by Anonymous on Tue 6th Sep 2005 20:10 UTC in reply to "This is not HCI..."
Anonymous Member since:
---

"Being able to customise your workspace to your domain-specific task is an important part of any human task and, it turns out, it's no different with computers. There's a ton of research on this."

Which is were?

"The most important development in HCI was they stopped timing people doing tasks and looking for ways to shave microseconds off those tasks and started doing something useful."

You're confusing ergonomics and human factors with HCI.

"About 20-30 years ago the advice of "usability gurus" and what was going on in HCI were quite similar; that's not true anymore."

"Drawing programs: Much of the current technology was demonstrated in Sutherland's 1963 Sketchpad system. The use of a mouse for graphics was demonstrated in NLS (1965).

Text Editing: In 1962 at the Stanford Research Lab, Engelbart proposed, and later implemented, a word processor with automatic word wrap, search and replace, user-definable macros, scrolling text, and commands to move, copy, and delete characters, words, or blocks of text. "

More like the foundation for the modern GUI was being laid.

Reply Score: 0

RE[2]: This is not HCI...
by Anonymous on Tue 6th Sep 2005 23:52 UTC in reply to "RE: This is not HCI..."
Anonymous Member since:
---

You're confusing ergonomics and human factors with HCI.

No, I'm not. Google on Computer-Supported Cooperative Work (CSCW), workplace studies, workflow, distributed cognition, etc, or look at a recent collection or journal in the field. This kind of work has been going on in HCI for years.

Reply Score: 0

RE: This is not HCI...
by John Nilsson on Wed 7th Sep 2005 05:35 UTC in reply to "This is not HCI..."
John Nilsson Member since:
2005-07-06

Well shure, customization is a good thing. In most cars I can adjust the steering wheel and seat position. Thats customization.

While beeing able to choose which side of the car to put the steering wheel, wheter to drive by stick and have all pedal funktions accecible by hands isn't customization that's application desing, and better left to professionals. As a user I simply pick the product that suits me.

Reply Score: 1

This is the stone age of computers
by ple_mono on Tue 6th Sep 2005 16:04 UTC
ple_mono
Member since:
2005-07-26

We kind of just (some 50 years back or so) found out that we can make a wheel spin and acctually use it to get around. This is how i see my computer using mouse/keyboard paired up in front my monitor. These perhaphial devices do what they are supposed to do, but this way sure feels clumsy.
I often wish i could use BOTH my hands as a tool directly on the screen, like two mouse-pointers at the same time with diffrent focus, performing two tasks at once. Man i would kick ass in Age of empires!!
So I vote prepare UI design for such "mulithreaded" handling of events, because what I know of, no such modern UI interface exists.
Why not combine this with Speech regognition commanding and gesture recognition (like xstroke) on both "pointers". Man, I would be a golden god ;)
I still think the mouse and keyboard serves a purpose though...

Reply Score: 1

Author watched to much Star Trek?
by Anonymous on Tue 6th Sep 2005 16:35 UTC
Anonymous
Member since:
---

My take the author sees problems that don't exist.As long as the apps work as they are supposed to do than that's it.All the rest becomes quickly ant-f8cking for nothing.

While ergonomically designs are appreciated one must not forget to get all working properly.Furthermore where's the emphasize on code quality and vulnerabillity acessment?Increasingly yet another piece of code is piled on another piece of code.When are they starting to clean the pile instead of focussing mainly on features and fame?

Reply Score: 0

Stay off the shibby!
by mannyv on Tue 6th Sep 2005 17:40 UTC
mannyv
Member since:
2005-07-18

Dude, but that bong down and take a walk!

Reply Score: 1

One knob to avoid scalding
by Anonymous on Tue 6th Sep 2005 17:50 UTC
Anonymous
Member since:
---

If the design professor Henry Petroski (in his book Small things considered) is to beleived, the one-knob faucet wasn't even invented for handicapped persons, but for all sorts of users to avoid being scalded. (See also this site: http://www.honoraryunsubscribe.com/alfred_moen.html)
I would call the blog just another lot af hot air, without any real substance.
Goran J, Varberg, Sweden

Reply Score: 1

Choices
by jonas on Tue 6th Sep 2005 17:58 UTC
jonas
Member since:
2005-07-08

As someone who uses mostly GTK applications inside the KDE environment with XFWM4 as my window manager, I don't like it when people go off on tirades about choices.

Choices should be there, but non-obtrusive. Defaults are fine for regular users, but power users need access to change those defaults. "Advanced" tabs and buttons don't work; regular users will peruse all options if they are given to them, and give up if they are too numerous.

But, there's nothing that turns me off more than having to dig through code when there is already a perfectly fine text configuration file for something somewhere. I'm reminded of fluxbox in particular; where its a 10 line source code patch to reverse desktop scrolling and add scroll-shading, but no option for (relatively popular) behavior tweaking exists.

I like in particular KDE's window manager approach. There's no GUI option to change the window manager, but it swaps out easily enough by setting an environment variable.

When I hear "choices are bad", I think of my time in windows, where I am forced to find and click the title bar to move windows and the resize handle to resize them instead of my beloved alt+right/left mouse button, where I'm forced to not have rollup or virtual desktop scrolling (or virtual desktops) because for some people it might be confusing.

Reply Score: 1

by Mystilleef on Tue 6th Sep 2005 18:43 UTC
Mystilleef
Member since:
2005-06-29

After reading the article, it seems GNOME is doing a very good job already.

Reply Score: 1

Keyboard and mouse need a redesign
by w-ber on Tue 6th Sep 2005 19:54 UTC
w-ber
Member since:
2005-08-21

What I'd like to see most on my keyboard is analog controls: things like knobs and sliders. Why does the whole thing need to consist of digital buttons? Imagine what it would be like to be able to use a real knob to adjust the volume in the mixer or to fine-tune a colour in image editor!

And the mouse needs force-feedback. That's one area currently unused and forgotten in user interfaces. Make the mouse twitch just a little on top of a button, resist ever so slightly when entering or leaving a window. I think, and I have no way of testing this, of course, that this would help tremendiously in user interfaces if implemented properly. (I'm not talking about breaking the user's wrist, but only tiny movements.)

Reply Score: 1

Oh dear.
by profiled on Tue 6th Sep 2005 20:13 UTC
profiled
Member since:
2005-08-30

Someone else pointed this out but Stevie Wonder and Ray Charles are *BLIND* not deaf.

So emailing him a picture rotated 25 degrees, or a blank image would have about the same effect.

Reply Score: 1

Eugenia?
by Anonymous on Tue 6th Sep 2005 20:28 UTC
Anonymous
Member since:
---

I would be interested in hearing our resident UI expert comment upon this article.

Personally, when I got to the part where he was advocating the spatial metaphor, I realized that there was little chance of our agreeing on anything. ;-)

Reply Score: 0

blah
by Anonymous on Tue 6th Sep 2005 22:27 UTC
Anonymous
Member since:
---

1. Already known. Forgot to mention screen sides, as well, and the point directly under the mouse.

2. ok?

3. (didn't read)

4. Yes. Thank god someone gets it--open/save dialogs [i]need[i] to go.

5. ...

6. Already known. Great article on the subject: http://arstechnica.com/articles/paedia/finder.ars/2 .

7. ok?

8. Lack of integration and decent UIs.

Symphony OS: This is a good example of one step forward, two steps back. Actually, it's more like 0.25 steps forward, 100 steps back. The use of screen corners is certainly unique, but the overall lack of elegance and quality of the rest really screws it over. Some things I don't like:
1) Complete context switch every time you activate one of the corner dongles.

2) The use of the desktop for user interaction, which gets obscured by windows and is a real pain to use.

3) Application management is horribly inelegant, and brute-force.

4) Doesn't tie together well. Not a lot of interface/application reuse.

-bytecoder

Reply Score: 0

I haven't read the article or comments but
by Phil on Tue 6th Sep 2005 23:05 UTC
Phil
Member since:
2005-07-06

With that out the way: The problem with HCI is simple, it's the people who do it.

I don't want to generalize here, but they are all self-important egotistical wannabe academics who think that if they do enough research it makes their project important, even if all they are doing is ripping off something that a designer invented 5 years ago simply because it seemed like a good idea.

I shall now prove my point with this idiot: http://www.ecs.soton.ac.uk/~mc/, and her dumb project: http://mspace.fm/.

And if anyone needs more, I'll attach some of her lecture notes...

Bitter, me?

Reply Score: 1

Anonymous Member since:
---

"I don't want to generalize here, but they are all self-important egotistical wannabe academics who think that if they do enough research it makes their project important, even if all they are doing is ripping off something that a designer invented 5 years ago simply because it seemed like a good idea."

But you will, and you'll note that a lot of the people who contributed to the modern GUI have academic backgrounds, even if they're employer was commercial.

Think of HCI much like computers having two sides. The CS side that basically develops the theory, and then there's everyone else who turns that theory into results.

Reply Score: 0

gestures, attention
by Anonymous on Tue 6th Sep 2005 23:33 UTC
Anonymous
Member since:
---

One thing he mentioned is the suckiness of removing attention from what you are reading to resize the window or whatever, then having to find your spot in the text again. I work around this by highlighting the sentence I am on before resizing the window or text or whatever. Makes it a lot easier to find my spot again when I'm done. *the more you know* ;)

Gestures rule for letting you perform tasks without shifting attention. KHotkeys (part of KDE) lets you use gestures in any application you please. Unfortunately you have to go through and create them all yourself, perhaps not the friendliest thing for new users. On the other hand, if you create them, it's pretty safe that you know what each of them does. They are a lot easier to remember if you design them yourself.. Anyway, the capability is there.

Also have to give points to KDE for the top level Macintosh like menus. They aren't on by default, but they are there, ready and waiting for anyone who wants to get a little use out of Fitts law. As long as you aren't on a dual screen setup, it's worth a shot.

Reply Score: 0

RE: gestures, attention
by Anonymous on Tue 6th Sep 2005 23:48 UTC in reply to "gestures, attention"
Anonymous Member since:
---

That reminds me, I forgot to talk about menubars: I don't like them. Global menu bars/windows (NeXT) are way too imposing for such an uncommonly used item, and local menubars are inelegant with SDI. Simply putting the commonly used functions on the top level window (think palm apps), along with factoring out functionality, should allow you to place the rest in context menus. Add in a document-centered, modular interface and window tabbing and you've got yourself a winner.

-bytecoder

Reply Score: 0

removal of save/load dialogs?
by hobgoblin on Tue 6th Sep 2005 23:43 UTC
hobgoblin
Member since:
2005-07-06

only way i can see that happen is if you create new files not by firing up the app you use to make them but have to select a "new file" dialog inside the folder where you want to store it.

but what about when you want to transfer files from say a email onto the drive directly? as there are no dialogs you cant click on it and select save. should you use drag and drop to move it onto a folder window? and should the file then be deleted from said mail or should it just be copied?

ok, that can be controled by holding a key as you do it. just make sure its consistent. if you drag and drop something inside windows the default action changes depending on the target. if its on a diffrent drive then suddenly the default action is copy, not move!

and the perfect level of integration would be if i could call up say the photoshop toolbox to edit a image embedded inside any other document, email or whatever without having to copy it to the clipboard or similar.

basicly the diffrent functions of the current apps should embedd themselfs inside a generic window. kinda like what you can do with konqueror and kpart plugins. in fact kparts are a interesting concept ;)

Reply Score: 1

RE: removal of save/load dialogs?
by Anonymous on Tue 6th Sep 2005 23:55 UTC in reply to "removal of save/load dialogs?"
Anonymous Member since:
---

only way i can see that happen is if you create new files not by firing up the app you use to make them but have to select a "new file" dialog inside the folder where you want to store it.
Close. A 'new' popup menu on the main window would be more convenient.


but what about when you want to transfer files from say a email onto the drive directly? as there are no dialogs you cant click on it and select save. should you use drag and drop to move it onto a folder window? and should the file then be deleted from said mail or should it just be copied?

Since attachments are part of the email, and since emails are (literally) just documents, you would separate it the same way as dragging and dropping an image from a document to a folder. Non-displayable "attachements" could show proxy icons.


ok, that can be controled by holding a key as you do it. just make sure its consistent. if you drag and drop something inside windows the default action changes depending on the target. if its on a diffrent drive then suddenly the default action is copy, not move!

It doesn't have to be consistent, so long as it corresponds with what the user thinks should happen. This generally requires usability studies.


and the perfect level of integration would be if i could call up say the photoshop toolbox to edit a image embedded inside any other document, email or whatever without having to copy it to the clipboard or similar.

Why would you need "photoshop"? Instead, a viewer/editor for each document type would be provided for each document type, and toolsets would be distributed instead of stand-alone apps.

-bytecoder

Reply Score: 0

e2mtt Member since:
2005-08-08

@ -bytecoder

Nice ideas. There are so many possible ways to change things and make them better, that I know we probably will never see. I wish this kind of thing was easier to develop, because I would love to see complete new experimental shells (on top of regular Linux/Mac/Windows so there would be apps to run) that would allow us to experiment with better ways to interface to computers within the existing KB/mouse/screen parameters.

Reply Score: 1

Anonymous Member since:
---

@ bytecoder:

exactly this kind of task or document centric integration was the underlying concept of opendoc (the better, more advanced cousin of MS 'do less with more' OLE), mostly (but iirc not solely) developed by apple in the 90ies, marketed together w/ system 7.5, and shortly after buried and forgotten: there weren't enough people, i.e. application (sic!) developers, to jump on the train... scalability, modularity and interoperability à la unix shell in a GUI? nah, whuddyathinky'ah, we don't cater for such extravagancies...

much left to (re)discover. ;-)

Reply Score: 0

hobgoblin Member since:
2005-07-06

well i got bitten by the drag and drop copy/move in windows, thats why i talk about it. basicly i was used to dragging between folders on diffrent drives and getting the copy indicator as default. therefor, when i just dragged the content of one folder into a diffrent folder for some testing on the same drive i didnt spot the fact that the default action suddenly was move, not copy. so when i cleaned up my experiment i found that i had deleted the only copy of said files!

who the fuck expects the system to do move within a drive and copy onto a diffrent drive? make one the default (preferably copy in my view) and the other optional, and stick to it!

yet another reason to stick with the copy and paste actions on menus or keyboard combos. atleast then one is sure what action thats done ;)

as for wrapping files in binary files that allow you to access basic actions to them, not sure. there are realy two problems with that, viruses and the distribution rights. basicly with a system like that any file can contain a virus. including every picture thats used on a webpage. the distribution problem is realy a minor one as its only realy a problem if the idea of proprietary software is maintained like it is today.

Reply Score: 1

I can sum up the article in one word
by mark on Wed 7th Sep 2005 01:34 UTC
mark
Member since:
2005-07-06

Simplify.

Good job of stretching that concept into 8 paragraphs though.

Reply Score: 2

In defense of the author.
by Anonymous on Wed 7th Sep 2005 01:52 UTC
Anonymous
Member since:
---

1) S/He is complaining about "companies" not making a decent HCI. I suppose this means mainstream products. No use in telling about less used, concept products which work ok. This is not what s/he's talking about.

2) Mainstream products really suck. S/he _is_ right about this. And, after all these years, they should not suck at all, for there's been enough time to improve the user experience. Maybe "ease-of-use" is a secondary criterion for companies; they care more about "ease-of-understand" and "familiar". We don't buy unfamiliar things we don't understand. That's the whole point of mimicking Windows. So, "pretty looks" always will win over "functional controls" in a company's heart.

3) Spatial works. Don't come with "lack of tactile feedback" or other "creative" excuses. It works and very well, btw. The great DOS programs of yore were so great because they were terribly fast, so fast a user would get into "flow", a state where your hands disconnect from the mind and manipulation almost become a second nature, like driving. A good example of this was XTree Gold.

4) Point 8 is ok by me, too. Humans possess an incredible internal structure to deal with language. It's idiotic to restrain ourselves to pointing, even if this approach is better for certain tasks. As an example, see those mouse-based password entry java applets: we just cope with them because of the added security, but they are so worse that writing the password!

Lºb (formerly Lee Nooks)

Reply Score: 0

RE: In defense of the author.
by segedunum on Wed 7th Sep 2005 12:16 UTC in reply to "In defense of the author."
segedunum Member since:
2005-07-06

Spatial works. Don't come with "lack of tactile feedback" or other "creative" excuses. It works and very well, btw.

It most certainly does not, and here's why. In the article the author talks about spatial viewing in terms of being in a kitchen, looking in drawers and seeing forks, spoons etc. in there. Fair enough.

In the computing world there just isn't any of that because we have a totally unlimited, infinite amount of objects we can store, and not only that, but multiple objects fit into more than one category depending on the situations and contexts you put them in. We also have that problem in the real world as well when organising and filing (which is why computers get bloody well used in the first place!!), so why the author thinks this is the fountain of all knowledge is anybody's guess. A more accurate comparison would have been the difficulties of filing lots of different papers and diferent types of data versus how computers can be used to solve that age-old problem.

A kitchen is an absolutely woeful comparison (probably sounds good because the author thinks real-world examples give him/her credibility) because there is almost unlimited space to organise the objects in there (well as large as the room, but much bigger than a monitor). The objects stored within that kitchen are finite and well defined (discrete), and once we know where to put one category of objects in a one particular drawer we're sorted. Life is just not that simple, which is why people have been very interested in the problems computers (and especially PCs) have been solving for them for over 25 years and why we have things like relational databases. What happens when an object could be stored in any one of four drawers?! Yes, things need to get better in the computing world. For example, we need to have much better, more fluid organisation of objects and get away from things like folders and rigid organisation (things like drawers!). But, things are the way they currently are for a reason.

It's wonderful when there is a finite amount of funtionality and data to organise, but in the computing world there is not, simply because that is the very reason why people use it! You're not constrained by real world limitations like filing cabinets and drawers which are not exactly a perfect fit for the problems people want to solve.

This author, whoever he or she is, and the rest of you who think things could ever possibly work like this in the computing world need to get yourselves back into nursery and find out about life, how to organise things and the sorts of problems you face (or people used to face) in the world of work. This is definitely one of the daftest blogs/articles masquerading as something sensible I have ever seen.

Reply Score: 2

largely misses the bigger picture
by Anonymous on Wed 7th Sep 2005 04:07 UTC
Anonymous
Member since:
---

while there might even be some valid points in this rant, it mostly re-cites statements which are at best stale news to anybody who has been in the field of HCI or interface design for longer than your average introductory lesson. and while it is nit-picking on the surface, it doesn't even scratch the deeper problems. take e.g. #4, the complaint about multiple representations of the file system. superficially even valid, imho, but: the problem is not the representation but the whole concept of work being based on hierarchically organized files. huh? 'file' is simple and everybody knows how...? go ahead, explain the concept of files in hierarchical directories and all their everyday implications to an intelligent(!) but in this respect clueless layperson, good luck! i grew up with it, i am perfectly able to deal with it -- and i still think it's an offence to human mental capabilities and thought structures.

mind you, this does not imply, that developers should not take HCI aka interface/interaction design and its issues much more seriously, incl. nit-picking (if it doesn't miss the bigger picture): a smooth and satisfying workflow (sic!), gaming experience, or whatever *does* depend on details.

Reply Score: 0

re_re Member since:
2005-07-06

>go ahead, explain the concept of files in hierarchical directories and all their everyday implications to an intelligent(!) but in this respect clueless layperson, good luck!<

anybody with orgational skills should have no problem understanding heirarchial directories, it's all logic and organization..... if they can navigate a well organized file cabinet, they can navigate heirarchial directories on a computer.

I have never..... (no not once) had trouble explaining this to anybody.... even my father who knows literally nothing about computers.

simply explain it as a a library card catalog or as a file cabinet with categories and sub categories, it's not difficult.

the heirachial directory structure stays.... it is the most logical and easy to use.

Reply Score: 1

Anonymous Member since:
---

"simply explain it as a a library card catalog or as a file cabinet with categories and sub categories, it's not difficult.

the heirachial directory structure stays.... it is the most logical and easy to use."

It however breaks down when working with large amounts of data, that fit in more than one catagory. Plus not everyone works that way. Note the difference between someone who keeps everything neat and organized, verses one who just put's things in piles around them. IMHO data organization and cataloging are what computers are for. All people really need to worry about (directly or indirectly) is that they can put data into a data store, and they can get it out. The rest is details for the computer.

Reply Score: 0

ma_d Member since:
2005-06-29

Symbolic links.
Your file structures aren't for data organization, if you are organizing data you aren't a typical user... You will of course need something more powerful to organize things.

Computers are for data organization and cataloging, that's one use. And there are thousands of programs written to do it with thousands of diverse kinds of data.
No one actually thinks that search is horrible, the problem is when people start depending on it. And then things change around. Say you implement your virtual folders. You write your search algorithm. You test it. You release it. Mass adoption occurs. Someone discovers something it does wrong, woops. You fix it. You then break 6 millions people's virtual folders while fixing 6 millions people's virtual folders. No big deal, if they knew where there stuff was to begin with...
That's a problem, it's not unsurpassable. But it's one of many that comes with reliance on an search/sort as opposed to knowing.
The next problem is changing meta information.

"The rest is details for the computer."
Wrong. The rest is details for the programmer. And therein lies the problem; you're still organizing it using a flawed human. Except now the flawed human is some guy who did it for 18 million people using a logic flow. You don't know what his logic flow is either...

I think I'd rather see applications do a better job of organizing the files they save...

Reply Score: 1

re_re Member since:
2005-07-06

I understand what you are saying but I don't necesseraly believe that it is better for humanity to make things easier (dumbed down, not faster). I think it is better to make people learn then to make things easy for them (within reason).... I'm not saying they should have to learn the Unix command reference, but that they should have to posess basic orgazational skills and if they don't.... they are not fit for the job market (within corresponding fields) plain and simple. I wouldn't want to employ someone with no organizational skills.

Making things overly easy is not necesseraly good for efficiency because if an undereducated employee runs into a problem they are screwed... I want somebody with problem solving skills.

Reply Score: 1

typo:
by re_re on Wed 7th Sep 2005 06:46 UTC
re_re
Member since:
2005-07-06

organizational skills

Reply Score: 1

screen corners, gui in general, folders
by Anonymous on Wed 7th Sep 2005 08:18 UTC
Anonymous
Member since:
---

Hitting the corners doesn't seem to be a good idea since screens get larger. So the way to the corners increases as well. Unless you have a *very* high mouse velocity you are running into trouble unless you have a XXL-mousepad.
For me GUI is best for doing things you don't do regularly. If I do something over and over again I will remember what keys to press instead.

I can't understand why some people actually think folders are a bad idea. I know many people who can't keep anything organized in the real world but they have no problems organizing their files.

One thing I have never seen mentioned and what really bothers me is why you can't have a gui that - if you miss an icon by a few pixels - checks if your click has any effect (clicking the desktop usually has not) and if it doesn't have an effect associates the click with the icon/app/whatever closest to the point you clicked.
You know, I really hate that feeling:
'God damn machine! Just because of some dirt on my mousepad I missed that app by 1mm. But still you should know what I meant!'

The perfect user interface would imho be smart enough to guess what you mean if eg one of your friends could guess it (which would of course include some time for adaptation).

Reply Score: 0

I know Mouse Pad Personally
by Anonymous on Wed 7th Sep 2005 09:41 UTC
Anonymous
Member since:
---

"Unless you have a *very* high mouse velocity you are running into trouble unless you have a XXL-mousepad."

Its called brushing and 10 y.olds can do it.

The thing missing about screen corners is... Multiple Screens. ( Yea, Im a photoshop guy ).

but wait.. the guy hits a note of absoulte genius.

"Any five-year-old earth child has probably already figured out that the screen corners are the easiest points to hit - the only locations hittable without looking."

What he thinks is that His idea about screen corners is what is important, but everybody missed that ITS A FIVE YEAR OLD OPERATING A COMPUTER that's important. That way, there is almost no training. Why do you need training?

If computer GUIs were simple, then five year olds could operate them easily, and also stressed out Ph.D. in Forensic microbiology could also.

I once configured a new XP installation for someone who was using Windows 95. A week after she got it, she had learned it very well, asking no questions except one. She asked 'Can you get show desktop to work?". I put one in her programs menu, one on the desktop, and one in the quick launch bar.

IF I COULD I WOULD HAVE ASSIGNED A SCREEN CORNER FOR IT. Ah, Apple was listening, and made something like that in Expose. Kudos Apple, now get the rest of the decades of HUI work you threw out.

Reply Score: 0

Working with OS GUIs since 1987, I ...
by Anonymous on Wed 7th Sep 2005 13:28 UTC
Anonymous
Member since:
---

... feel the need to spent my 2 cents here.

1.) I use two of the display corners. Main menu and clock. The other corners are needed for apps/content. And no automatic app starts when my mouse hits a corner, please. I'd go mad.
Corners are currently no issue imho.

2.) "(No t33n-N30, the answer isn't »Pr3f3r3nc3Zz!!!!!!!! 1337-H4XX0R5!!!«.)"
Yes it is. If you're no beginner anymore, preferences is exactly what you need.

5.) Ahh ... thank you. I prefer e.g. to have several different methods of copying files. Sometimes the shell's the fastest, sometimes some norton commander like tool, sometimes even the friggin explorer view.

6.) "Spatial navigation is in our nature."
Forget metaphors. When I switched from Amiga to Win95, I cried tears of joy because of one thing: Non-Spatial navigation was possible!
For *me*, it's faster, more efficient, more concise and way less complex and irritating.

8. Won't work. What you'd need to say would be something like "I've taken some picture about 20 days ago, remember? I'd like to send it as mail. No, don't activate the printer, I meant as email. Show me the one with the church. The picture, not the mail! The Chuch!Picture of the church! From 20 days ago! No, not that one. Next ... no, I mean the one inside. No, not the icon inside the church, the inside view of the church. No, not the view from the church over the town! Ah f***, where's the mouse ..."

Reply Score: 0

Screen corners
by Rodrigo on Thu 8th Sep 2005 08:43 UTC
Rodrigo
Member since:
2005-07-06

I don't know about this guy's computer but I just have 4 here. I can't see how much stuff I would be able to associate with them.

I configured my Mac OS to run exposé when I hit the lower left corner but I gotta confess I get annoyed when I move my mouse there non-intentionally...I tried other corners but with the same effect.

Reply Score: 1

So much nothing.
by deathshadow on Thu 8th Sep 2005 12:52 UTC
deathshadow
Member since:
2005-07-12

Let's analyze these one at ta time.

1> Screen corner controls tend to piss me off, because much like 'sticky-keys' they always trip when I DON'T WANT THEM. First damned thing I turn off in any OS that supports it.

2> This one has lots of words but doesn't actually SAY anything.

3> 4> 5> wow, was anyone able to make any sense at all of what is said in these? Gibberish ranting about so much nothing.

6> Well, I hate spatial navigation too (I'm a tree type of guy... It's laid out like a tree, show me a damned TREE!)

7> Piss poor mix of George Carlin and Bill Engvall... again meaningless

8> Poor baby... Not sure what he's arguing for but certainly seems worked up over nothing.

But again, no two people ever see eye to eye on anything...

Reply Score: 1

Screen corners argument overused
by Anonymous on Thu 8th Sep 2005 13:10 UTC
Anonymous
Member since:
---

The screen corners argument is overused and screen corners are just not all that efficient.
Compared to dragging the mouse anywhere else on screen yes, arguably they are, but compared to keyboard shortcuts they're not.
Especially on something like OS X where mouse acceleration is just too slow but that's a different problem..

Another problem I have with the screen corners idea is large monitors. On my 24" monitor I have to turn my head to see a corner. And add 1920x1200 res to that and even with decent mouse acceleration getting to corners is a pain.

Much easier to hit F9-F11 for me then it is to bind Expose to screen corners.

Reply Score: 0