Just because you try to catch a mistake is no excuse for permitting it in the first place when it's so easy to avoid.
Besides which, 1) users routinely ignore dialogs and mindlessly click through them w/o reading them, because they're horribly abused and overused, and 2) the entire notion of needing to deliberately save files is rather broken anyway. Why not save data, including copious undo buffers, all the time? Is it really all that difficult?
Naw, I bitch about how hard the entirety -- not just the mere GUI -- of the *nix interface is to use. And this is largely justified because it was never designed with usability in mind, nor did it follow any design processes likely to result in superior UI.
I absolutely do not claim that Windows or MacOS are the epitome of good interface. The holy grail is likely a 'Do What I Mean' UI that's so far out of reach that it's a waste of time considering it.
However, there is a method by which we can improve interfaces, and it is decidedly not being used nearly as much as it ought to be, especially in the Unix realm.
Frankly, I'd throw away MacOS or Windows in a heartbeat if confronted with something decidedly superior. I owe them no loyalty, and I'd be instantly upset that they failed to deliver the best UI available. (i.e. that they failed to, at the very least, maintain parity)
Are you really claiming that the OS X CLI is so carefully designed for, and integrated with, the GUI that they are essentially inseperable, each supporting the other?
I don't think so.
Drag and drop is nice, and yet isn't even 1/1000th as far as I'd want to take things. I still can't select a file with the mouse, then type in a command in the CLI to move it, SANS drag and drop to an open terminal.
Can you type in a CLI command, then right click on it to get a list of options you can select with the mouse?
Can you type in a URL and have a browser window automatically come up?
If they were closely tied together, you could. Even the concept of a terminal would likely be lost... one line would probably be sufficient.
I mean Unix -- not XWindows. (though I don't care for that either) Unix, the underlying operating system, is what I am saying (non-trollishly, if anyone will listen) sucks.
I would furthermore not use bash, or rm, or cp, etc. With respect to these, consider two factors.
One, CLIs such as bash are not integrated with the GUI. While I have no problem with permitting users to use them seperately, and even offering the GUI alone as the default, they need to work together. This is absolutely not simply having multiple terminals open at once. Rather, permitting each form of input to be used at any point, interchangably, in concert with the other. E.g. You could select a group of files with the mouse, then type in a command to rename them according to some rule (avoiding having to do them individually). Then drag the still-selected files into another folder with the mouse (avoiding having to type a path), then use a pipe to send them out via email, then double click, or drag and drop them into a program to edit them with.
Two, the GUI and CLI commands must be reconciled. On the Mac, files are deleted by moving them to the Trash, then emptying it. The rm command does NOT do this! Yet consistency is essential to any good UI... it's all that keeps a scroll bar from shutting off the computer. The methods by which commands work, what they're called, etc. have to all be consistant across both UI methods. This is made a little easier with my proposal for integration above, as a command akin to (but obviously not) 'ls' would merely open a folder window; not provide a textual listing in a terminal, unless someone were perversely using a CLI alone.
CLIs haven't improved substantially for several decades -- there's no excuse for that, and no sense in continuing to use outdated ones today.
OTOH, I'm all for having an OS built from scratch be POSIX compliant... but only after the standalone design work is finished.
I'm upset regarding tendancies towards affording importance towards paths. The layout of files. The filesystem (and related issues, e.g. metadata). The security and multiuser models.
All of this crap seems to have been adopted for no reason other than because it is what is used in the Unix world. That is not good enough! It has to be worthwhile ON ITS OWN. But there is a gigantic amount of room for improvement! I sincerely doubt that Apple even considered ways in which to surpass the hoary old Unix baggage. Nevertheless, they should've. Providing a superior experience is all that matters.
Stability of Unix != Unix. Compatability with Unix != Unix.
And as for Aqua, it may look nice, but it's not well-thought out. There've been all sorts of UI bugs all year, and no end in sight.
This isn't really true. I need to hit the sack, but check out what Raskin, and Tog have had to say about the subject. Note that the decision for one button was quite arbitrary, and not done with any solid evidence that it was a good call.
OTOH, paths as a rule aren't that great. On the Mac, for most of the time (God only knows about MFS -- I don't want to check) paths aren't even used as GUIDs. There's a GUID number for that. There's little reason to know paths.
A more general purpose 'find' box might be better. I've long advocated integrating a CLI and GUI together so that you could type in certain commands and see them carried out in _windows_. Naturally it'd need a whole new shell, but that's expected. Certainly it'd be nice to have a window of downloaded files, and be able to type "Select *.mp3" or some such and then be able to drag those selected files w/ the mouse to a different folder w/o typing in it's path.
As for the mouse buttons, this is merely contingency planning. If you develop a UI and do not anticipate and plan for users, you have fucked up. Add features that make a two, or three, or five button mouse beneficial. But make sure that the computer can always work with a one button mouse, or even no mouse at all. (in extreme situations) What's POSSIBLE isn't the same as what's OPTIMAL.
The UI should not impair color-blind users either. This doesn't mean you can't have red, or green, or yellow, or blue... rather that the color alone shouldn't be essential, even if it's commonly used.
This is commonplace stuff.
Windows, incidentially, is NOT INTENDED to be used with two button mice. MS's UI guidelines dictate that all commands on the 2d button need to be accessible from the menubar. (Apple does likewise) Hell, you don't need a mouse at all.
Re: Control panels, I've never run into this. (barring the recent decision to put certain clock settings in the international panel in OS X, but I hate OS X, and that's just another item on the list)
Windows suffers from inconsistant layout, too many poorly explained options, poorly arranged (esp. unresizable) panels, etc. Consider Phone and Modem options being seperated from the Network panel....
1) NeXT had technologies years ahead of its time? Should I include the 3.5 NeXT floppy that I have that has to be a SCSI device in that list?
Be was good, but Be's true strength was a willingness to start from a clean slate. Apple's last real attempt to do so was the Mac. MS never has. NeXT never did either.
3) Cancel is typically 'Escape' or 'Command-Period' Nicer software, e.g. BBEdit would display shortcut keys if you depressed the Command button. (which is definately a necessary feature)
You surely must admit that the consistant button placement in such dialogs is good though. I'm always having to hunt for the 'Yes/OK' button on Windows, because the damn thing is on the left, and the number of buttons shifts it around.
I'm not sure that I could agree with 'the commercial support of Windows' given that it's really pervasive all across the world, but I agree with your general point.
The tricky issue, as I see it though -- and this isn't mere trolling, but an opinion arrived at after a lot of consideration, chiefly with regards to usability issues -- the tricky issue is that Unix is a total piece of crap. Being the best Unix is like having the fastest Trabant.
If you are _developing_ an OS based on what is essentially marketing hype, that is that it is Unix, you are likely not going to do too well.
It is better to leave aside the 'name', and instead devote your resources to developing a killer UI, a solid back end to support it, and having no baggage at all. Not even the baggage that Unix brings with it. The end result is infinitely more important than the name attached to it. There was no reason to adopt Unix, given that it doesn't have a monopoly (heh) on stability, which is pretty much the only redeeming feature. (as far as it goes with regards to what Apple's doing; they don't want people looking at _all_ their source)
The Mac could've been a GUI built on top of the Apple II, CP/M, or MS-DOS. If the mere name factor had been desirable, it would've. Certainly there were no end of GUIs that were.
Instead they built the damn thing from virtually first principles to accomplish a goal that had nothing to do with the marketing of the underlying technology, and it worked out really well. Frankly, it's the only way to go.
(too bad Be didn't come along years earlier -- they also realized the value of a fresh start)
Congratulations! You discovered why it is that usability testing with real, live users is important.
I'm not being sarcastic either -- I really mean it. I wish more people would actually look to see how people use their software, and make improvements in response to that.
First, Windows lets clicks on widgets take effect on windows that are not in the foreground. Thus you could (and I frequently do, still firmly believing that the Mac method of clicking to change focus and foreground ONLY is superior) accidently hit a close button that was positioned near something important in a different window.
Secondly, Windows has the close button smack-dab next to the minimize and maximize/restore widgets. Those two are not destructive, but it's again quite easy to accidently hit the wrong one. On the Mac, when a second titlebar widget was introduced, they put the damn thing as far away as it could get.
Why OS X is adopting the Windows approach to both of these things, and compromising their UI, I'll be damned if I know.
Actually, the traditional method of renaming a file on the Mac is to select it and to either a) hit return, or b) hold the cursor over the name field briefly. Then you rename as desired, and hit return when you're done.
Honestly, there ISN'T a rename command... I wouldn't mind providing one that accomplished the same thing for people not expecting this long, long established method to work, but I wouldn't make it manditory as you suggest. To do so would be to throw away nearly over a decade and a half of good history... more so than Apple, in their madness has already done so.
(Oh, and it's Control-Click, or Right-Click)
Personally, I don't have a real problem with the old way, particularly given that yours would make it somewhat annoying to _reposition_ the icons within the same window. (Windows tends to forget icon placement, but the spacial Finder is an essential mainstay of the Mac, and I wouldn't want to lose it for no good reason)
The complaint re: clicking has been around since the days of Hypercard... I don't think that too much has changed.
What's my incentive in accepting such a situation? It does work both ways, and a compromise is called for. Neither absolute position is acceptable.
Each side has its own advantages too... organizations like NYT tend to be organized in promoting their interests. But the public are more powerful and can more readily cause such activities to become illegal.
Actually, agreeing to prepare a contract may indeed be agreeing to an actual contract. (making the drawing up of papers a mere formality, which evidences the terms _already_ mutally agreed upon)
No, there is a very old common law rule called the Statute of Frauds that can be used to make some verbal contracts unenforcable if they last a certain amount of time, or deal with certain subjects. And most states have other regulations along these lines too.
Although I'm certainly no MA lawyer (or any kind of lawyer).
In most day-to-day contracts, no one cares if you sign your name in cursive. As long as you make some kind of mark that definately identifies you as a party to it, that's enough. Even printed names are sufficient, so email headers ought to work fine.
The signature requirement is not a formality to closing a contract, though it can act as one. Rather, it is intended to ensure that people aren't accidently made parties or not made parties. This objection is only valid if he claims someone else was trying to sell his house, and pretending to be him.
The only tricky bit is whether or not property contracts require certain formalities that this doesn't comply with. Sounds like there aren't any, though.
There was certainly a lot of stuff going on in it... and the movie looks like it will be interesting enough to get my $5. (matinee showing)
But _man_, the clips in the trailer just seemed to be thrown together without any real order behind them. The previous trailer was a lot better in terms of editing.
Re:Look, BS. Never thought you'd find *that* on /.
on
Penguin2Apple
·
· Score: 1
I'm not entirely certain that this is true...
For a lot of tasks, I think it can be safely assumed that users would never notice the limits imposed by such a security model (which, agreed, could be turned off).
I mean, how often do people run executables from within Outlook that are intentionally supposed to have permission to delete all of your files? The main trick, as I see it, is to somehow protect and identify commands given from the console as opposed to those which programs attempt to run on their own. (e.g. allow a user to open files with this app, don't allow the app to make the request w/o user intervention)
I'm not saying it's easy to develop this. But dammit, Apple made a name for itself by doing _hard_ things that had big payoffs.
Anyway, there's some details on such a system called EROS on the net. (which is experimental, but then Mach wasn't too far out of the labs when NeXT picked it up) There may be others as well.
I do agree with your general contention though. I'm just all for making those mechanisms invisible.
Re:Look, BS. Never thought you'd find *that* on /.
on
Penguin2Apple
·
· Score: 1
Well, let me clarify in that there are definately times when multiuser systems are useful! I have no problems with them becoming more common... the MS document describing a sort of distributed computing environment (minus the monopolistic evil) would rock. But anyone who sits down at a computer and logs into their account should never notice that they're on a multiuser system unless they have to admin it for some reason.
Nope -- I'm legitimately concerned. Check out my history with regard to Apple and UI here on/.
I am ONLY interested in good UI. I don't care who provides it, but I'm sharply critical of anyone who doesn't. Apple, which had so satisfied me in the past has utterly disappointed me.
I'd go back in a heartbeat if only they had something that lived up to their legacy.
Re:Look, BS. Never thought you'd find *that* on /.
on
Penguin2Apple
·
· Score: 1
The Unix multiuser model is ancient and loathesome. It is beneficial to NO ONE save for the admins of multiuser machines, who represent a miniscule minority of all users in aggregate. A system that confers upon ordinary users, or single users who do not share their machines, the benefits of a single user model (less complexity, more freedom) is far superior, and, sadly, still waiting to be developed. Apple _COULD'VE_ worked on it, but they dropped the ball.
Naturally you'd want to tie it into a security model that was superior to that of Unix as well. In this day and age, with the prevalence of worms and trojans, it's inexcusable to permit such malware to wipe out a user's entire space. That the remainder of the system is fine is small comfort. WinNT has a more granular, but still not amazing security system. I'm more interested in a pervasive capabilities security model myself. One in which a user is allowed to delete files, but certain of his programs, e.g. malware that naturally lacks permission to do so, cannot.
Regarding the second item, they are not contradictory. The Mac was designed as, above all else, a humanistic computer. A computer intended to work with human beings as it found them, with as little imposition upon them as possible. The degree to which this can be achieved would naturally increase over time. It hasn't. Instead we're beginning to see backsliding, as technical concerns override usability concerns. This should _never_ happen. If compromises must be made, it is a necessary evil, and should be rectified as quickly as possible. This process of rectification is advancement in UI design. Apple isn't even doing good design (again, I'm not _just_ talking about GUIs and skins here) anymore.
As noted above, clunk, clunk. I have a NeXT Cube, I've used it quite a lot, the fact that it has an overblown reputation doesn't prevent it from being as lousy as it is.
The graphics model was intolerably slow (check out Doom on black hardware sometime;), encumbered by licensing issues, and incapable of easily achieving certain effects taken for granted today. Furthermore, it was never accellerated with hardware, and any unique features it might have held, e.g. a vector-based UI, were never exploited. (indeed, this is still true -- Aqua is raster-based) Widgets made to look 3d through a tromp l'oeil effect are of dubious worth, (look at what an eyesore Win95 et al are!) but also were present since System 1 on the Mac.
Miller columns, incidentally, have been around since the 70's -- they were used in conjunction with SmallTalk at PARC.
As for dev tools, I'm not arguing about that. I'm not claiming that they're bad -- but users who would never ever need or want them likely constitute 99.44% of the computer using population. While I'd like to see more people program, the last thing I would ever do would be to design an UI that was geared for such a tiny minority... it ensures that only they would ever use it, and not even that, for programmers follow users around. (and are users themselves a lot of the time)
Well, Rhapsody never really got off the ground because they were attempting to maintain a certain degree of compatability with the old MacOS, and it was crippling the project.
OS X is natively NOT compatable with MacOS. Instead it runs a (mediocre) emulator. This is a perfectly valid option, I'm totally in support of it, and I think Copland could've gotten a lot further if they'd explored this. Pink, sadly, died a political death, not a technological one.
As for NextStep, clunk, clunk. That is the sound of my tapping the black cube by my desk! NextStep was certainly more showy than System 6, but it was not particularly more useful. Indeed, the Unix legacy really ruined it from my perspective.
It is not that issue specifically, but rather that is one of many issues that bothers me. There are, indeed, possible alternatives which at the very least minimize the usability problem created by that, however, but OS X is slavishly reimplementing the same model used for over thirty years. I sincerely doubt that anyone at Apple has even considered whether or not there's a better method.
I realize that Windows is no better in this regard, but that doesn't mean that I'm not critical of it either.
As for what I want in a UI... I've been toying with it a bit on and off. At my old job I used to design them, always paying strict attention to the needs of users, and doing testing to ensure that I was on the right track. Apple's got loads of excellent work archived away... unfortunately, they use virtually none of it.
Between Linux and OS X -- yes, I'd agree that OS X is better. But honestly, I think that that's about _all_ that it's better than, rather in the same way that a Yugo is far superior to a Trabant, and neither is more appealing than walking.
Unfortunately, the UI in OS X, and by this I mean the interface as it runs to the core of the machine, including multiuser model, security, filesystem, etc. and not merely GUI, is inherently flawed and, I think, impossible to fix.
The UI concepts that were the heart and soul of the Mac since '84 should never have been cast aside, and the development of progressively better UI should never have been allowed to stagnate since the early 90's.
OS X is little more than a reheated version of NextStep, an OS that flopped dramatically. Was it always said to have a great UI? Sure... but again, only as far as Unix went. (e.g. NetInfo is vastly better than the equivalent tools on Unix, and a total POS compared to what most Mac and Windows users are used to these days)
If Apple had been willing to make hard decisions, such as dropping compatability with the Mac save through an emulator -- just as it is _now_ -- I think that they could've done better on their own.
I've been migrating from MacOS to Windows, but I fear it could be a very long time before anyone really develops a good UI again.
Slashdot Mirror
Just because you try to catch a mistake is no excuse for permitting it in the first place when it's so easy to avoid.
Besides which, 1) users routinely ignore dialogs and mindlessly click through them w/o reading them, because they're horribly abused and overused, and 2) the entire notion of needing to deliberately save files is rather broken anyway. Why not save data, including copious undo buffers, all the time? Is it really all that difficult?
Naw, I bitch about how hard the entirety -- not just the mere GUI -- of the *nix interface is to use. And this is largely justified because it was never designed with usability in mind, nor did it follow any design processes likely to result in superior UI.
I absolutely do not claim that Windows or MacOS are the epitome of good interface. The holy grail is likely a 'Do What I Mean' UI that's so far out of reach that it's a waste of time considering it.
However, there is a method by which we can improve interfaces, and it is decidedly not being used nearly as much as it ought to be, especially in the Unix realm.
Frankly, I'd throw away MacOS or Windows in a heartbeat if confronted with something decidedly superior. I owe them no loyalty, and I'd be instantly upset that they failed to deliver the best UI available. (i.e. that they failed to, at the very least, maintain parity)
Are you really claiming that the OS X CLI is so carefully designed for, and integrated with, the GUI that they are essentially inseperable, each supporting the other?
I don't think so.
Drag and drop is nice, and yet isn't even 1/1000th as far as I'd want to take things. I still can't select a file with the mouse, then type in a command in the CLI to move it, SANS drag and drop to an open terminal.
Can you type in a CLI command, then right click on it to get a list of options you can select with the mouse?
Can you type in a URL and have a browser window automatically come up?
If they were closely tied together, you could. Even the concept of a terminal would likely be lost... one line would probably be sufficient.
You misunderstand me.
I mean Unix -- not XWindows. (though I don't care for that either) Unix, the underlying operating system, is what I am saying (non-trollishly, if anyone will listen) sucks.
I would furthermore not use bash, or rm, or cp, etc. With respect to these, consider two factors.
One, CLIs such as bash are not integrated with the GUI. While I have no problem with permitting users to use them seperately, and even offering the GUI alone as the default, they need to work together. This is absolutely not simply having multiple terminals open at once. Rather, permitting each form of input to be used at any point, interchangably, in concert with the other. E.g. You could select a group of files with the mouse, then type in a command to rename them according to some rule (avoiding having to do them individually). Then drag the still-selected files into another folder with the mouse (avoiding having to type a path), then use a pipe to send them out via email, then double click, or drag and drop them into a program to edit them with.
Two, the GUI and CLI commands must be reconciled. On the Mac, files are deleted by moving them to the Trash, then emptying it. The rm command does NOT do this! Yet consistency is essential to any good UI... it's all that keeps a scroll bar from shutting off the computer. The methods by which commands work, what they're called, etc. have to all be consistant across both UI methods. This is made a little easier with my proposal for integration above, as a command akin to (but obviously not) 'ls' would merely open a folder window; not provide a textual listing in a terminal, unless someone were perversely using a CLI alone.
CLIs haven't improved substantially for several decades -- there's no excuse for that, and no sense in continuing to use outdated ones today.
OTOH, I'm all for having an OS built from scratch be POSIX compliant... but only after the standalone design work is finished.
I'm upset regarding tendancies towards affording importance towards paths. The layout of files. The filesystem (and related issues, e.g. metadata). The security and multiuser models.
All of this crap seems to have been adopted for no reason other than because it is what is used in the Unix world. That is not good enough! It has to be worthwhile ON ITS OWN. But there is a gigantic amount of room for improvement! I sincerely doubt that Apple even considered ways in which to surpass the hoary old Unix baggage. Nevertheless, they should've. Providing a superior experience is all that matters.
Stability of Unix != Unix.
Compatability with Unix != Unix.
And as for Aqua, it may look nice, but it's not well-thought out. There've been all sorts of UI bugs all year, and no end in sight.
This isn't really true. I need to hit the sack, but check out what Raskin, and Tog have had to say about the subject. Note that the decision for one button was quite arbitrary, and not done with any solid evidence that it was a good call.
OTOH, paths as a rule aren't that great. On the Mac, for most of the time (God only knows about MFS -- I don't want to check) paths aren't even used as GUIDs. There's a GUID number for that. There's little reason to know paths.
A more general purpose 'find' box might be better. I've long advocated integrating a CLI and GUI together so that you could type in certain commands and see them carried out in _windows_. Naturally it'd need a whole new shell, but that's expected. Certainly it'd be nice to have a window of downloaded files, and be able to type "Select *.mp3" or some such and then be able to drag those selected files w/ the mouse to a different folder w/o typing in it's path.
As for the mouse buttons, this is merely contingency planning. If you develop a UI and do not anticipate and plan for users, you have fucked up. Add features that make a two, or three, or five button mouse beneficial. But make sure that the computer can always work with a one button mouse, or even no mouse at all. (in extreme situations) What's POSSIBLE isn't the same as what's OPTIMAL.
The UI should not impair color-blind users either. This doesn't mean you can't have red, or green, or yellow, or blue... rather that the color alone shouldn't be essential, even if it's commonly used.
This is commonplace stuff.
Windows, incidentially, is NOT INTENDED to be used with two button mice. MS's UI guidelines dictate that all commands on the 2d button need to be accessible from the menubar. (Apple does likewise) Hell, you don't need a mouse at all.
Re: Control panels, I've never run into this. (barring the recent decision to put certain clock settings in the international panel in OS X, but I hate OS X, and that's just another item on the list)
Windows suffers from inconsistant layout, too many poorly explained options, poorly arranged (esp. unresizable) panels, etc. Consider Phone and Modem options being seperated from the Network panel....
1) NeXT had technologies years ahead of its time? Should I include the 3.5 NeXT floppy that I have that has to be a SCSI device in that list?
Be was good, but Be's true strength was a willingness to start from a clean slate. Apple's last real attempt to do so was the Mac. MS never has. NeXT never did either.
3) Cancel is typically 'Escape' or 'Command-Period' Nicer software, e.g. BBEdit would display shortcut keys if you depressed the Command button. (which is definately a necessary feature)
You surely must admit that the consistant button placement in such dialogs is good though. I'm always having to hunt for the 'Yes/OK' button on Windows, because the damn thing is on the left, and the number of buttons shifts it around.
7) Not their fault, but I feel your pain.
I'm not sure that I could agree with 'the commercial support of Windows' given that it's really pervasive all across the world, but I agree with your general point.
The tricky issue, as I see it though -- and this isn't mere trolling, but an opinion arrived at after a lot of consideration, chiefly with regards to usability issues -- the tricky issue is that Unix is a total piece of crap. Being the best Unix is like having the fastest Trabant.
If you are _developing_ an OS based on what is essentially marketing hype, that is that it is Unix, you are likely not going to do too well.
It is better to leave aside the 'name', and instead devote your resources to developing a killer UI, a solid back end to support it, and having no baggage at all. Not even the baggage that Unix brings with it. The end result is infinitely more important than the name attached to it. There was no reason to adopt Unix, given that it doesn't have a monopoly (heh) on stability, which is pretty much the only redeeming feature. (as far as it goes with regards to what Apple's doing; they don't want people looking at _all_ their source)
The Mac could've been a GUI built on top of the Apple II, CP/M, or MS-DOS. If the mere name factor had been desirable, it would've. Certainly there were no end of GUIs that were.
Instead they built the damn thing from virtually first principles to accomplish a goal that had nothing to do with the marketing of the underlying technology, and it worked out really well. Frankly, it's the only way to go.
(too bad Be didn't come along years earlier -- they also realized the value of a fresh start)
I'm not being sarcastic either -- I really mean it. I wish more people would actually look to see how people use their software, and make improvements in response to that.
Well, you ignore two things.
First, Windows lets clicks on widgets take effect on windows that are not in the foreground. Thus you could (and I frequently do, still firmly believing that the Mac method of clicking to change focus and foreground ONLY is superior) accidently hit a close button that was positioned near something important in a different window.
Secondly, Windows has the close button smack-dab next to the minimize and maximize/restore widgets. Those two are not destructive, but it's again quite easy to accidently hit the wrong one. On the Mac, when a second titlebar widget was introduced, they put the damn thing as far away as it could get.
Why OS X is adopting the Windows approach to both of these things, and compromising their UI, I'll be damned if I know.
Actually, the traditional method of renaming a file on the Mac is to select it and to either a) hit return, or b) hold the cursor over the name field briefly. Then you rename as desired, and hit return when you're done.
Honestly, there ISN'T a rename command... I wouldn't mind providing one that accomplished the same thing for people not expecting this long, long established method to work, but I wouldn't make it manditory as you suggest. To do so would be to throw away nearly over a decade and a half of good history... more so than Apple, in their madness has already done so.
(Oh, and it's Control-Click, or Right-Click)
Personally, I don't have a real problem with the old way, particularly given that yours would make it somewhat annoying to _reposition_ the icons within the same window. (Windows tends to forget icon placement, but the spacial Finder is an essential mainstay of the Mac, and I wouldn't want to lose it for no good reason)
The complaint re: clicking has been around since the days of Hypercard... I don't think that too much has changed.
What's my incentive in accepting such a situation? It does work both ways, and a compromise is called for. Neither absolute position is acceptable.
Each side has its own advantages too... organizations like NYT tend to be organized in promoting their interests. But the public are more powerful and can more readily cause such activities to become illegal.
Actually, agreeing to prepare a contract may indeed be agreeing to an actual contract. (making the drawing up of papers a mere formality, which evidences the terms _already_ mutally agreed upon)
;)
Typically, it depends.
No, there is a very old common law rule called the Statute of Frauds that can be used to make some verbal contracts unenforcable if they last a certain amount of time, or deal with certain subjects. And most states have other regulations along these lines too.
Although I'm certainly no MA lawyer (or any kind of lawyer).
In most day-to-day contracts, no one cares if you sign your name in cursive. As long as you make some kind of mark that definately identifies you as a party to it, that's enough. Even printed names are sufficient, so email headers ought to work fine.
The signature requirement is not a formality to closing a contract, though it can act as one. Rather, it is intended to ensure that people aren't accidently made parties or not made parties. This objection is only valid if he claims someone else was trying to sell his house, and pretending to be him.
The only tricky bit is whether or not property contracts require certain formalities that this doesn't comply with. Sounds like there aren't any, though.
It's not a secret. I really am Superman.
There was certainly a lot of stuff going on in it... and the movie looks like it will be interesting enough to get my $5. (matinee showing)
But _man_, the clips in the trailer just seemed to be thrown together without any real order behind them. The previous trailer was a lot better in terms of editing.
I'm not entirely certain that this is true...
For a lot of tasks, I think it can be safely assumed that users would never notice the limits imposed by such a security model (which, agreed, could be turned off).
I mean, how often do people run executables from within Outlook that are intentionally supposed to have permission to delete all of your files? The main trick, as I see it, is to somehow protect and identify commands given from the console as opposed to those which programs attempt to run on their own. (e.g. allow a user to open files with this app, don't allow the app to make the request w/o user intervention)
I'm not saying it's easy to develop this. But dammit, Apple made a name for itself by doing _hard_ things that had big payoffs.
Anyway, there's some details on such a system called EROS on the net. (which is experimental, but then Mach wasn't too far out of the labs when NeXT picked it up) There may be others as well.
I do agree with your general contention though. I'm just all for making those mechanisms invisible.
Well, let me clarify in that there are definately times when multiuser systems are useful! I have no problems with them becoming more common... the MS document describing a sort of distributed computing environment (minus the monopolistic evil) would rock. But anyone who sits down at a computer and logs into their account should never notice that they're on a multiuser system unless they have to admin it for some reason.
This just isn't the case at present.
Nope -- I'm legitimately concerned. Check out my history with regard to Apple and UI here on /.
I am ONLY interested in good UI. I don't care who provides it, but I'm sharply critical of anyone who doesn't. Apple, which had so satisfied me in the past has utterly disappointed me.
I'd go back in a heartbeat if only they had something that lived up to their legacy.
The Unix multiuser model is ancient and loathesome. It is beneficial to NO ONE save for the admins of multiuser machines, who represent a miniscule minority of all users in aggregate. A system that confers upon ordinary users, or single users who do not share their machines, the benefits of a single user model (less complexity, more freedom) is far superior, and, sadly, still waiting to be developed. Apple _COULD'VE_ worked on it, but they dropped the ball.
;), encumbered by licensing issues, and incapable of easily achieving certain effects taken for granted today. Furthermore, it was never accellerated with hardware, and any unique features it might have held, e.g. a vector-based UI, were never exploited. (indeed, this is still true -- Aqua is raster-based) Widgets made to look 3d through a tromp l'oeil effect are of dubious worth, (look at what an eyesore Win95 et al are!) but also were present since System 1 on the Mac.
Naturally you'd want to tie it into a security model that was superior to that of Unix as well. In this day and age, with the prevalence of worms and trojans, it's inexcusable to permit such malware to wipe out a user's entire space. That the remainder of the system is fine is small comfort. WinNT has a more granular, but still not amazing security system. I'm more interested in a pervasive capabilities security model myself. One in which a user is allowed to delete files, but certain of his programs, e.g. malware that naturally lacks permission to do so, cannot.
Regarding the second item, they are not contradictory. The Mac was designed as, above all else, a humanistic computer. A computer intended to work with human beings as it found them, with as little imposition upon them as possible. The degree to which this can be achieved would naturally increase over time. It hasn't. Instead we're beginning to see backsliding, as technical concerns override usability concerns. This should _never_ happen. If compromises must be made, it is a necessary evil, and should be rectified as quickly as possible. This process of rectification is advancement in UI design. Apple isn't even doing good design (again, I'm not _just_ talking about GUIs and skins here) anymore.
As noted above, clunk, clunk. I have a NeXT Cube, I've used it quite a lot, the fact that it has an overblown reputation doesn't prevent it from being as lousy as it is.
The graphics model was intolerably slow (check out Doom on black hardware sometime
Miller columns, incidentally, have been around since the 70's -- they were used in conjunction with SmallTalk at PARC.
As for dev tools, I'm not arguing about that. I'm not claiming that they're bad -- but users who would never ever need or want them likely constitute 99.44% of the computer using population. While I'd like to see more people program, the last thing I would ever do would be to design an UI that was geared for such a tiny minority... it ensures that only they would ever use it, and not even that, for programmers follow users around. (and are users themselves a lot of the time)
Well, Rhapsody never really got off the ground because they were attempting to maintain a certain degree of compatability with the old MacOS, and it was crippling the project.
OS X is natively NOT compatable with MacOS. Instead it runs a (mediocre) emulator. This is a perfectly valid option, I'm totally in support of it, and I think Copland could've gotten a lot further if they'd explored this. Pink, sadly, died a political death, not a technological one.
As for NextStep, clunk, clunk. That is the sound of my tapping the black cube by my desk! NextStep was certainly more showy than System 6, but it was not particularly more useful. Indeed, the Unix legacy really ruined it from my perspective.
It is not that issue specifically, but rather that is one of many issues that bothers me. There are, indeed, possible alternatives which at the very least minimize the usability problem created by that, however, but OS X is slavishly reimplementing the same model used for over thirty years. I sincerely doubt that anyone at Apple has even considered whether or not there's a better method.
I realize that Windows is no better in this regard, but that doesn't mean that I'm not critical of it either.
As for what I want in a UI... I've been toying with it a bit on and off. At my old job I used to design them, always paying strict attention to the needs of users, and doing testing to ensure that I was on the right track. Apple's got loads of excellent work archived away... unfortunately, they use virtually none of it.
Between Linux and OS X -- yes, I'd agree that OS X is better. But honestly, I think that that's about _all_ that it's better than, rather in the same way that a Yugo is far superior to a Trabant, and neither is more appealing than walking.
Unfortunately, the UI in OS X, and by this I mean the interface as it runs to the core of the machine, including multiuser model, security, filesystem, etc. and not merely GUI, is inherently flawed and, I think, impossible to fix.
The UI concepts that were the heart and soul of the Mac since '84 should never have been cast aside, and the development of progressively better UI should never have been allowed to stagnate since the early 90's.
OS X is little more than a reheated version of NextStep, an OS that flopped dramatically. Was it always said to have a great UI? Sure... but again, only as far as Unix went. (e.g. NetInfo is vastly better than the equivalent tools on Unix, and a total POS compared to what most Mac and Windows users are used to these days)
If Apple had been willing to make hard decisions, such as dropping compatability with the Mac save through an emulator -- just as it is _now_ -- I think that they could've done better on their own.
I've been migrating from MacOS to Windows, but I fear it could be a very long time before anyone really develops a good UI again.
A 'Thundertree' huh? Sounds good, I guess, but when I was growing up, I just went over to the municipal thunderdome.