In most cases Windows won't even let you install on older hardware.
What ? Windows will typically install - and be usable with some minor upgrades (typically RAM and, with Vista, a video card) on machines back to around the 7 year old mark (from my vague recollection of an interview I saw, Microsoft target a slightly-above-average 5-year old machine as the "minimum requirements"). This has been true, well, basically forever.
Any Mac comparably aged to a PC too old to "let you install" would certainly fare no better with OS X, and almost certainly worse. Heck, even my <3 year old iBook struggles to run OS X at more than a snail's pace.
Windows is *far* kinder to older hardware than OS X is.
Windows will pre-emptively swap out memory when the machine is idle. Note that it doesn't actually mark the memory as free, so only needs to swap it back in if the machine has been under some memory pressure since (and the memory has had to be used for something else). It does this so that if swapping really does have to occur (eg: because you fire up some huge application), then all the system has to do is mark the memory free, rather than swap it out and then mark it free.
Linux does the same thing (although you can tune it not to). There's not really a problem with it, except for people who get neurotic about a flickering hard disk light:).
The one problem that can happen in Windows is that it grows the buffer cache too aggressively, which can result in some of that pre-emptively swapped out memory being marked free when it probably shouldn't be.
What about the limited terms of patents and copyrights? You could argue that they are also making what is "yours" "ours". I wouldn't agree with that argument though, since you cannot own "intellectual property", you are merely enjoying a few time-limited exclusive rights to your intellectial products. You own the rights, but not the intellectual products.
Firstly, "limited copyright terms" are rapidly becoming folklore. Secondly, I was speaking outside of the concept of "intellectual property", in a more ethereal sense about something that is "yours" in the sense that its something you have created, rather than an object you have some sort of legally-enforced power over.
The GPL depends utterly on copyright to work. While I personally don't agree with almost all aspects of copyright law, the fact remains that the GPL and that it would be pointless and useless without it.
No it isn't. If your software depend on software developed by someone else, it is only fair that the other developer has a say in how you license your combined (derivative) work.
Why is it fair if that GPLed code only makes up, say, 5% of the final product ?
After all, nothing prevents you from writing an equivalent package yourself, where you would have all rights to it. If yu don't like being forced to use the GPL for derivative works, don't use GPL software as a foundation for yours.
How predictable. This argument always appears in an attempt to deflect commentary about the GPL away with the "but no-one makes you use it" straw man.
But it doesn't require that. Unless of course, your "own" code depends on the GPL software. If it does, you simply have to accept that you have to comply with the license of that software.
Clearly you misunderstood what I meant by "GPL-code-derivative". I was referring to code that would be considered "derivative code" by the GPL. I would have thought that was obvious, since my comments would be utterly pointless without that assumption.
I don't see what all the complaining is all about.
I was complaining about the deceptive use of the term "free" as a synonym for "GPLed" and how the deliberate choice of such loaded terminology identifies it as common political propaganda being use to further an idealistic goal.
I personally don't like the GPL, as I think it frequently ends up being an unfair exchange, deceptively "marketed" as generosity (in both directions - although I don't mind the LGPL, it's much more likely to result in an equitable transaction). However, I don't dispute someone's choice to license their code however the hell they want. What I *do* find annoying, is the deceptive use of the word "free" as a synonym for "GPL", the hostility from the cult of GNU towards people who don't agree with their loaded definition of that word and the implication that those people are wrong in some sort of objective manner.
I still assert that the analogy holds. Both sets of restrictions are there to stop people from reducing other people's freedom.
If you follow the view that the additional privileges the GPL bestows are, in fact, worthy of the term "freedoms" rather than just the results from a simple transaction of goods and services one might also get in a traditional contract, then the analogy holds. Personally, I don't agree with that in the slightest and I take offense when people like RMS - and those who subscribe to his views - assert that I am making some sort of moral failure by doing so.
Here's another way of looking at it. When someone kills you, or steals from you, or misrepresents you, you have suffered harm and your "human rights" - as the term is fairly commonly accepted across the civilised world - have been infringed. When you choose to voluntarily use a piece of non-GPLed software instead of a piece of GPLed software, neither of these things have happened. That's the difference. Feeling safe and secure in your person and property is a genuine "freedom
If the tactile feedback really worked, though, they'd be looking at the screen constantly, rather than letting their eyes off the buttons only every 5-10 keypresses, no?
Er, no ? Why would they ? When I know I can write, for example, "friend" by just hitting 374363, and being able to feel exactly where those buttons are is trivial, why would I be looking at it constantly ?
I don't even send a lot of texts - maybe 2-3 a day - and I can write SMSes with only a casual glance. People who do send a lot (some younger colleagues at work) can quite literally write messages dozens of words long without even looking at (or, apparently, concentrating on) the phone. I sincerely doubt they'll be able to do that with an iPhone.
Maybe then I'd also receive fewer SMSes with completely mistaken words that just happened to be on the same keys, as well. I can't quote any examples in English, but I think everyone knows what I mean...
I can think of one - "he" and "if". On my Nokias, you toggle between the alternatives using the '*' key and it's basically an automatic reaction after a while (ie: writing "if" becomes "41*").
This is not a fan-boi trolling post; speaking as a UI designer for a few OSS projects, I am seriously curious about what you find so wrong with the Dock. I understand that people's workflows are different, and that what fits me may not fit you, but please share your concerns with me.
Many (including me) have covered this before, however, some highlights:
* The indication as to what is a running program (vs a launch icon) and what is not is too subtle
* Similar objects look identical (eg: a bunch of minimised terminals or word docs, or a bunch of folder shortcuts) unless you either turn the zoom right up (which produces other problems), keep it big all the time (which wastes screen space), or constantly wipe the mouse across it to get captions.
* It's centre-oriented (I realise this is changeable with appropriate tools, but the point is that's how it arrives out of the box). Additionally, it doesn't extend to the bottom of the screen. That makes it (relatively) hard to hit.
* Further, it means everything on it moves constantly, destroying any hope of leveraging muscle memory.
* The intermingling of icons with distinctly different functions (more apparent in the right side of the Dock where you get minimised windows hanging around with file and folder aliases and.... the Trash). The Dock should have at least 3 - maybe even 4 - different "zones".
* Icons on the Dock behave inconstently (and, worse, _destructively_) with "similar" icons in most other parts of the UI (eg: drag something off and it disappears, rather than copies/moves). Icons on the Dock *should* be aliases that can be manipulated like other aliases.
* The icons on the Dock don't tie back to anything (ie: they're not physical files that can be manipulated).
* You can't drag-and-drop via the Dock to an arbitrary window.
* Context menus are infuriatingly slow, especially for folder aliases.
* Related to this, moving from an arbitrary window in one application to an arbitrary window in another application is tedious (I know Expose has disguised this problem, but I'm talking in the context of the original release).
* Following on, I really don't like the application-centric paradigm (this is more a personal preference thing, I'll grant).
Basically, it's difficult to find anything the Dock does _right_ (other than look cool in demos). The Windows Taskbar is superior in pretty much every functional way[0], as were the Application and Apple Menus it replaced from Classic MacOS (my only adjustment to them would have been to make the Application Menu cascade out a window list for each application on it and to improve drag & drop capabilities).
The Dock is a very confused piece of UI. It tried to be a program launcher, [filesystem] shortcut bar, task switcher and window manager, and consequently ends up sucking at all of them. If you really are a UI designer, steer well clear of the Dock for inspiration. According to pretty much all well understood UI guidelines, it's a disaster. It was quite clearly made primarily to look cool and with usability (and good UI principles) as a distant secondary priority.
[0] Even that annoying collapse-a-bunch-of-windows-into-a-button feature only serves to make that aspect (task switching) of the Taskbar *as bad* as the Dock when it triggers - and like the Dock's centre-orientation, it's "fixable" with tweaking utils (to bump up the threshold).
Every time I've seen it demo'ed pressing "home" brings up the standard 4 options at the bottom of the screen. "Phone" is always lower left. After that Favorites is two over.... and so on.
How do you know how far without any physical feedback ?
As long long as the interface is responsive, "muscle memory" wouldn't seem to be an issue.
You seem to be missing the point I'm trying to make, which is that with decent physical feedback how responsive the UI isn't doesn't matter, because you can confidently be a couple of steps ahead of it.
Please. Watch the demos and notice the "flicking" gestures used to scroll and navigate. Watch how the "scroll" speed matches the velocity at which your fingers move. Gestures are intuitive as hell.
I've watched the demos. It reminds me of the first OS X demos and how "cool" they were. Then we got the UI train wreck that was the Dock.
And texting? Watch someone text sometime. Very few people (even on Treos) are "touch-text'ists", and most are starting intently at the phone while they're doing it. And if you're moving to the iPhone from a RAZR or some other phone that has a standard 12-key-pad, having separate letter "keys" (even virtual ones) would be a godsend over having to hit "7" four times to get an "S", or waiting for the last letter to "enter" up so you can get a "A" after you've entered a "B". Thanks, but no thanks.
I think you need to spend more time watching people under the age of 30 texting, and less time watching technophile, 50-year-old CEOs. Predictive text input systems have been around for 7 - 10 years, "learning" ones for at least 5 years, and anyone remotely familiar with doesn't write SMSes the way you describe. Indeed, I can't think of anyone I know who regularly texts that gives the screen more than a casual glance every 5-10 keypresses.
I wouldn't be surprised in the least if regular texters used to a traditional keypad and predictive input will be at least as fast as people using an on-screen keyboard.
I think you're dramtically over-estimating the benefits of tatile feedback, and ignoring how interface action, responsives, and audible feedback can compensate, or even improve on the experience.
I don't. However, I'm willing to be convinced, which is why I'm waiting before passing judgement.
Aaah, crap. You mean I actually have to *look* at a device I'm using. Good god, how will I cope?
Dunno. Can you touch type ?
And incidentally, I find I'm pretty good at noticing the feedback when my finger hits a screen. I don't waft it in front of the screen, or push it through any more.
I've no doubt. But how do you know which part of the screen you've hit, what was under it at the time and whether or not the UI has responded correctly ?
Do me a favour and save your arguments about how the lack of feedback will kill this device until you've actually tried one.
You must have me confused with someone else, because I made no such argument.
The UI has had an awful lot of work put into it, I doubt it's going to be difficult to use.
I doubt it is either, however, that doesn't mean it will be *efficient* to use, or as easy as it could be using real buttons. Added to which, a) a lot of Apple's recent UI efforts haven't exactly been confidence-inspiring and b) they don't have any experience in this market yet, so even it "being good" is not a given.
And to be honest, even in the worst case scenario where you're right and it is slower to use, I think that's a small price to pay for all the benefits.
I've yet to see a lot of significant benefits over the alternatives.
The larger screen, and much larger buttons for dialing numbers are two features I'd sacrifice an awful lot for.
Your phone must really suck if you feel that the standard number pad is so awkward to use that you think one on a phone-sized touchscreen would be easier.
Indeed. Apple were among the first to get this sort of "correctional" interface behavior right in other spheres (e.g., even early versions of the Mac OS allowed you to "stray" a little when moving from a menu item into its sub-menu; rather than requiring the cursor to remain precisely within the bounds of the selected menu item, the mouse could venture outside a small distance -- so long as it continued generally moving toward the submenu -- without causing the submenu to disappear).
"Corrective" isn't really the way to describe it - that would imply the system restricts the mouse movement for you.
What MacOS did was allow for a certain amount of "looseness" in the mouse movement. It did this by creating a triangle bounded by the mouse pointer (at the time of the click or menu expansion), the top-left of the submenu and the bottom-left of the submenu. As long as the point stayed within the boundaries of that triangle, the submenu stayed open.
Yeah, I'm really going to know on a Treo that I'm on C and not V. Lots of tactile feedback in dozens of identical little buttons.
I was actually thinking of a regular numberpad, however, you'll still get more from "dozens of identical little buttons" than you will from a smooth piece of plastic with "dozens of little buttons" drawn on it.
Assumption.
No. Deduction.
And probably totally ignores all of the other things one does on a "normal" phone like handling voicemail, or diving down in the menu tree to change a setting, or things like multiway conference calls that no one does on their existing phone because they never read the manual and memorized the procedure.
No, it doesn't. Quite the contrary, in fact.
But hey. You've convinced yourself you're not going to like it. Fine. Horses for courses. And one less person ahead of me in line...
Ah, you must be an Apple zealot, someone for whom "maybe it won't be perfect because of these reasonable arguments" translates to "it sucks horribly and will never work" when referring to any Apple product.
Which part of the I'll-wait-and-see position I advocated is leading you to believe I've "convinced myself" ?
If contacts have worked well for you before, then they will work just as well for the iPhone right?
Can't say. Haven't used it. My response based on the mockups on Apple's website is "maybe".
The point at hand was how having a contact list is no replacement for simply typing in numbers, which I argue is untrue and you would seem to also assert.
A contact list is great if it's populated. Not having a way to directly type in number is a pretty major PITA if need to, however. You need both to be well done for a phone to be good.
Personally I have found use of contacts on the phones I've owned to be annoying, and so I don't use them all the time even if I have them loaded. That would thus be the counterargument for why use of contacts in an iPhone would not be of use, which is what I was refuting.
I just don't see what the fuss about the iPhone's contact list is about. Certainly from looking at the mockups, it doesn't seem to offer any opportunity for significant improvement over existing methods.
No, actually I type and know I missed it by observing the screen - that's the reason to touch-type, right? So you can observe the screen as you work? On the iPhone the screen and your typing are as one.
Well, I'd be quite willing to bet your subconscious recognised that you've made an error long before (relatively speaking) your conscious mind sees the error on screen, but I doubt you'll agree since you seem to have made your mind up.
What Apple is trying to do is to make contacts actually usable to well, contact people with. Just because you've not had that experience in the past on a phone does not mean it cannot be done.
I've had that experience with every mobile phone I've ever owned. Even with hundreds of "contacts" in my phone book I can get to one of them in a few seconds and (at worst) half a dozen button presses. For the 5 or so people I call frequently and regularly, it's just a matter of holding down a number on the keyboard or saying their name. Heck, it was true (except for the voice recognition) on the Nokia 5310 I had a decade ago.
Of course, I've never owned anything but Nokias - other phones' phonebook interfaces may well suck as much as you imply, but they certainly don't by definition.
How sure are you? I'd say relative distance from a fixed location has a lot to do with how you type. When your finger is over the keyboard, then goes down to press a key how often are you feeling the other keys around it, vs. just hitting the kay you want? I find as I type this message I make little use of the keys around the key I am pressing, and simply type directly based on position.
Ah, but you know instantly when you've duffed it and missed a key or hit between two keys, don't you, even without looking at the screen ?
Tactile feedback isn't just about knowing when you get it right, it's (probably more importantly) knowing about when you've gotten it *wrong*.
If you want to play semantic games and claim that by compiling BSD code and putting it into a box, it's no longer the same code, I'm sure you and any other like minded people can go amuse yourselves that way.
I didn't, you (apparently) did. You didn't say "compiled", you said "modified".
Most reasonable people (and, for that matter, US Copyright law) see it as clearly still being Foo (albeit a modified Foo), and thus have lost the freedom to modify and enhance that Foo.
If it's just compiled - that is, unmodified - you have lost nothing, because you can just go out and get the BSDLed source for Foo and modify it to your heart's content.
As I have mentioned, the killer app for a phone is break away from anachronistic "dialing" the phone. There is no reason why making a call should not be as simple as stating someone's name.
It already is, from that perspective. Heck, my Nokia 6310i has voice recognition for dialling and it's 5+ years old. However, the problems with the system aren't trivial:
* You need the contact in your phone.
* You either have to tag them with a voice-dial -or- the phone needs a good voice recognition system and you have to remember what you called the contact
* Duplicates (in the case of a voice recognition system) require manual intervention anyway.
If you have a small number of frequently-dialled contacts, the whole "voice dialling" thing works ok. If you don't, it doesn't and going via the "phonebook" is easier.
The lack of tactile input is, IMHO, going to be a major usability problem - but I'm reserving my final judgement until I can actually use one.
Does the mouse on your computer provide tactile feedback when you move the cursor over a button? Trackpads or pointers on laptops?
I'd argue that's a meaningfully different form of UI interaction.
Do any other touchscreens (e.g., in grocery store checkout lines) do so? Do any PDAs with touchscreens provide tactile feedback? I can't think of many, if any, that do, and that doesn't seem to have hindered them.
That's actually a relatively interesting point and I'd be interested to see any studies of error rates and input speeds for traditional numberpads vs touch-screen numberpads. I'd expect the touchscreens to lose, however, *especially* in the context of any sort of multitasking.
How often do you actually use a phone without looking at it? Even when I'm just hitting speed dial buttons I'm usually looking at the phone to double-check that it's calling the right person. Especially relevant: how often do you use advanced features like web surfing or text messaging/email without looking at the phone? Unless you've got a screen reader in there, don't you kind of have to look at it to use those features? Ditto for watching video on a handheld device.
It's not just about using it without looking at it, it's about using it while looking at it, but not really having to concentrate.
The tactile feedback you get from real buttons tell you both where you are in the UI relative to other elements (ie: where your thumb is on the buttons) and when you have successfully complete an action (ie: the "click" when the button connects). This helps you both move your thumb quickly to the right spot by feel (even if you're also looking) and know that you've done something without having to actively concentrate on it. The iPhone will lack both these features and, hence, you'll have to actively watch the UI not only for the feedback that you'd normally respond to with muscle memory, but also correct positioning within the UI you'd normally do from muscle memory. Basically, you'll have to concentrate harder to use it.
I disagree about the fragility of the touch screen. If children's devices (Nintendo DS) can have a touch screen, I don't see why adult devices should be concerned about the fragility of such a thing.
Because a Nintendo DS is ca. US$100 and an iPhone will ba ca. US$500.
I also suspect that since you'll have to do everything to the iPhone through the touch UI, and because it's not going to use a stylus, the iPhone's screen will see more "use".
As for the tactile feedback, I think you're underestimating the UI mechanisms used to use the device. The most pressing activity on a phone is dialing.
Text messages are another that springs instantly to mind. Realisticaly, though, it's basically everything that involves using the phone. You've probably never thought much about the tactile feedback and hence don't realise how much you use it, even if only unconsciously.
For example: you know that to get to a certain UI element you have to do a certain number of button pushes. Say, two down arrow presses, across one and "enter". With real buttons, you get feedback when a button is pushed and your muscle memory can make getting to UI elements you are familiar with an automatic, practically instant procedure (because you don't need to watch it to make sure each action was successful, that you didn't "overshoot", etc). Further, you know where your thumb is relative to the next button you have to push, because you can feel it. With the iPhone, you will have to actively watch the UI as it changes to navigate through it, to know which intermediate step it is at during each phase and to know where your thumb needs to be.
Yes, because they assume "free" and "GPLed" (or equivalent) are synonymous and use the terms interchangeably, despite that not being a definition most people would perceive.
I don't think so. Many if not most people would consider our western democracies "free countries", where you are allowed to do whatever you want except harming people.
They would, but they'd be wrong in a very real sense, as evidenced by the numerous laws involving victimless crimes (pretty much anything surrounding drugs, for example).
However, I understand the point you are trying to make, that we ostensibly have a "free" society due to the restrictions placed on certain activities (eg: murder).
However, it's not analagous, as laws like that are about keeping things that are "yours", "yours", whereas the GPL is about making things that are "theirs", "yours".
These restrictions are analogous to the restrictions placed on software licensed under the GPL. Thus, I would hardly call the use deceptive.
Perhaps if you subscribe to the FSF doublespeak. However, requiring someone to GPL their own code if they wish to distribute some sort of GPL-code-derivative isn't in the same ballpark - heck, not even playing the same game - as laws against, say, murder and theft. The former are assumptions about what you should be able to do with *other people's work*, the latter about the assumptions that can be reasonably made regarding *yourself* and *your property*.
That is, unless you happen to agree with the collossal arrogance of people who consider the "right" to modify and redistribute someone else's work is somehow on par with so-called "human rights" like freedom of speech and self-defense. If that's true I doubt we'll be able to find a common ground.
The word "right", in the context of various "human rights", carries a heavy meaning and using it for such mundane things as licensing of source code significantly cheapens its value.
"Open Source" is a *vastly* more honest term that "free software", because it conveys accurate meaning without relying on loaded/ambiguous definitions of the word "free" which only a relative handful of people even know, let alone subscribe to. Of course, the term "free software" was chosen precisely *because* of that loaded meaning, which in my mind puts it in the same arena as the kind of political propanda that is used to justify - amongst others - the various "War On X" projects.
Suppose you release package Foo under the BSD license. Now a company uses a modified version of that software in their product Bar. As an end user, I buy a Bar. Now I've got a product running a non-free version of Foo.
No, you have a product called FooBar. I.e. a combination of Foo and Bar.
The company won't give me the sources to their version, so I can't modify or enhance it, or fix bugs. That's about as non-free as software gets, and it was clearly facilitated by the BSD license.
The code for Foo remains "free". The code for Bar, does not. The BSDL code remains "free", even if a product that is using it is not.
Slashdot Mirror
It doesn't. Those people aren't helping Apple's business.
I think you need to elaborate here on "helping ... business".
No, this is evil. You don't treat people this way who are helping your business.
How does it compare to, say, people hacking OSX around to make it work on regular PCs ?
In most cases Windows won't even let you install on older hardware.
What ? Windows will typically install - and be usable with some minor upgrades (typically RAM and, with Vista, a video card) on machines back to around the 7 year old mark (from my vague recollection of an interview I saw, Microsoft target a slightly-above-average 5-year old machine as the "minimum requirements"). This has been true, well, basically forever.
Any Mac comparably aged to a PC too old to "let you install" would certainly fare no better with OS X, and almost certainly worse. Heck, even my <3 year old iBook struggles to run OS X at more than a snail's pace.
Windows is *far* kinder to older hardware than OS X is.
Never quite understood that one though.
Windows will pre-emptively swap out memory when the machine is idle. Note that it doesn't actually mark the memory as free, so only needs to swap it back in if the machine has been under some memory pressure since (and the memory has had to be used for something else). It does this so that if swapping really does have to occur (eg: because you fire up some huge application), then all the system has to do is mark the memory free, rather than swap it out and then mark it free.
Linux does the same thing (although you can tune it not to). There's not really a problem with it, except for people who get neurotic about a flickering hard disk light :).
The one problem that can happen in Windows is that it grows the buffer cache too aggressively, which can result in some of that pre-emptively swapped out memory being marked free when it probably shouldn't be.
(End simplified explanation.)
What about the limited terms of patents and copyrights? You could argue that they are also making what is "yours" "ours". I wouldn't agree with that argument though, since you cannot own "intellectual property", you are merely enjoying a few time-limited exclusive rights to your intellectial products. You own the rights, but not the intellectual products.
Firstly, "limited copyright terms" are rapidly becoming folklore. Secondly, I was speaking outside of the concept of "intellectual property", in a more ethereal sense about something that is "yours" in the sense that its something you have created, rather than an object you have some sort of legally-enforced power over.
The GPL depends utterly on copyright to work. While I personally don't agree with almost all aspects of copyright law, the fact remains that the GPL and that it would be pointless and useless without it.
No it isn't. If your software depend on software developed by someone else, it is only fair that the other developer has a say in how you license your combined (derivative) work.
Why is it fair if that GPLed code only makes up, say, 5% of the final product ?
After all, nothing prevents you from writing an equivalent package yourself, where you would have all rights to it. If yu don't like being forced to use the GPL for derivative works, don't use GPL software as a foundation for yours.
How predictable. This argument always appears in an attempt to deflect commentary about the GPL away with the "but no-one makes you use it" straw man.
But it doesn't require that. Unless of course, your "own" code depends on the GPL software. If it does, you simply have to accept that you have to comply with the license of that software.
Clearly you misunderstood what I meant by "GPL-code-derivative". I was referring to code that would be considered "derivative code" by the GPL. I would have thought that was obvious, since my comments would be utterly pointless without that assumption.
I don't see what all the complaining is all about.
I was complaining about the deceptive use of the term "free" as a synonym for "GPLed" and how the deliberate choice of such loaded terminology identifies it as common political propaganda being use to further an idealistic goal.
I personally don't like the GPL, as I think it frequently ends up being an unfair exchange, deceptively "marketed" as generosity (in both directions - although I don't mind the LGPL, it's much more likely to result in an equitable transaction). However, I don't dispute someone's choice to license their code however the hell they want. What I *do* find annoying, is the deceptive use of the word "free" as a synonym for "GPL", the hostility from the cult of GNU towards people who don't agree with their loaded definition of that word and the implication that those people are wrong in some sort of objective manner.
I still assert that the analogy holds. Both sets of restrictions are there to stop people from reducing other people's freedom.
If you follow the view that the additional privileges the GPL bestows are, in fact, worthy of the term "freedoms" rather than just the results from a simple transaction of goods and services one might also get in a traditional contract, then the analogy holds. Personally, I don't agree with that in the slightest and I take offense when people like RMS - and those who subscribe to his views - assert that I am making some sort of moral failure by doing so.
Here's another way of looking at it. When someone kills you, or steals from you, or misrepresents you, you have suffered harm and your "human rights" - as the term is fairly commonly accepted across the civilised world - have been infringed. When you choose to voluntarily use a piece of non-GPLed software instead of a piece of GPLed software, neither of these things have happened. That's the difference. Feeling safe and secure in your person and property is a genuine "freedom
If the tactile feedback really worked, though, they'd be looking at the screen constantly, rather than letting their eyes off the buttons only every 5-10 keypresses, no?
Er, no ? Why would they ? When I know I can write, for example, "friend" by just hitting 374363, and being able to feel exactly where those buttons are is trivial, why would I be looking at it constantly ?
I don't even send a lot of texts - maybe 2-3 a day - and I can write SMSes with only a casual glance. People who do send a lot (some younger colleagues at work) can quite literally write messages dozens of words long without even looking at (or, apparently, concentrating on) the phone. I sincerely doubt they'll be able to do that with an iPhone.
Maybe then I'd also receive fewer SMSes with completely mistaken words that just happened to be on the same keys, as well. I can't quote any examples in English, but I think everyone knows what I mean...
I can think of one - "he" and "if". On my Nokias, you toggle between the alternatives using the '*' key and it's basically an automatic reaction after a while (ie: writing "if" becomes "41*").
You mean apart from the width of the iPhone? Or are you claiming that it grows and shrinks?
No, I'm saying that without looking at it, it will be difficult to tell if you're over one of those buttons or between two of them.
"Two over" means nothing when there's no way of counting how many you've already gone past.
And that ammendment makes a big difference, as my original reply to you was based on my step-daughter's Sidekick III which has a QWERTY keyboard.
Well, not really, because a QWERTY keyboard on a similarly small phone-sized touch screen is going to be worse or, at best, equivalent.
This is not a fan-boi trolling post; speaking as a UI designer for a few OSS projects, I am seriously curious about what you find so wrong with the Dock. I understand that people's workflows are different, and that what fits me may not fit you, but please share your concerns with me.
Many (including me) have covered this before, however, some highlights:
* The indication as to what is a running program (vs a launch icon) and what is not is too subtle .... the Trash). The Dock should have at least 3 - maybe even 4 - different "zones".
* Similar objects look identical (eg: a bunch of minimised terminals or word docs, or a bunch of folder shortcuts) unless you either turn the zoom right up (which produces other problems), keep it big all the time (which wastes screen space), or constantly wipe the mouse across it to get captions.
* It's centre-oriented (I realise this is changeable with appropriate tools, but the point is that's how it arrives out of the box). Additionally, it doesn't extend to the bottom of the screen. That makes it (relatively) hard to hit.
* Further, it means everything on it moves constantly, destroying any hope of leveraging muscle memory.
* The intermingling of icons with distinctly different functions (more apparent in the right side of the Dock where you get minimised windows hanging around with file and folder aliases and
* Icons on the Dock behave inconstently (and, worse, _destructively_) with "similar" icons in most other parts of the UI (eg: drag something off and it disappears, rather than copies/moves). Icons on the Dock *should* be aliases that can be manipulated like other aliases.
* The icons on the Dock don't tie back to anything (ie: they're not physical files that can be manipulated).
* You can't drag-and-drop via the Dock to an arbitrary window.
* Context menus are infuriatingly slow, especially for folder aliases.
* Related to this, moving from an arbitrary window in one application to an arbitrary window in another application is tedious (I know Expose has disguised this problem, but I'm talking in the context of the original release).
* Following on, I really don't like the application-centric paradigm (this is more a personal preference thing, I'll grant).
Basically, it's difficult to find anything the Dock does _right_ (other than look cool in demos). The Windows Taskbar is superior in pretty much every functional way[0], as were the Application and Apple Menus it replaced from Classic MacOS (my only adjustment to them would have been to make the Application Menu cascade out a window list for each application on it and to improve drag & drop capabilities).
The Dock is a very confused piece of UI. It tried to be a program launcher, [filesystem] shortcut bar, task switcher and window manager, and consequently ends up sucking at all of them. If you really are a UI designer, steer well clear of the Dock for inspiration. According to pretty much all well understood UI guidelines, it's a disaster. It was quite clearly made primarily to look cool and with usability (and good UI principles) as a distant secondary priority.
[0] Even that annoying collapse-a-bunch-of-windows-into-a-button feature only serves to make that aspect (task switching) of the Taskbar *as bad* as the Dock when it triggers - and like the Dock's centre-orientation, it's "fixable" with tweaking utils (to bump up the threshold).
Indeed, I can't think of anyone I know who regularly texts that gives the screen more than a casual glance every 5-10 keypresses.
I should amend this to say "except for the poor sods using "smartphones" with QWERTY-style keyboards". Those things are a PITA.
Every time I've seen it demo'ed pressing "home" brings up the standard 4 options at the bottom of the screen. "Phone" is always lower left. After that Favorites is two over.... and so on.
How do you know how far without any physical feedback ?
As long long as the interface is responsive, "muscle memory" wouldn't seem to be an issue.
You seem to be missing the point I'm trying to make, which is that with decent physical feedback how responsive the UI isn't doesn't matter, because you can confidently be a couple of steps ahead of it.
Please. Watch the demos and notice the "flicking" gestures used to scroll and navigate. Watch how the "scroll" speed matches the velocity at which your fingers move. Gestures are intuitive as hell.
I've watched the demos. It reminds me of the first OS X demos and how "cool" they were. Then we got the UI train wreck that was the Dock.
And texting? Watch someone text sometime. Very few people (even on Treos) are "touch-text'ists", and most are starting intently at the phone while they're doing it. And if you're moving to the iPhone from a RAZR or some other phone that has a standard 12-key-pad, having separate letter "keys" (even virtual ones) would be a godsend over having to hit "7" four times to get an "S", or waiting for the last letter to "enter" up so you can get a "A" after you've entered a "B". Thanks, but no thanks.
I think you need to spend more time watching people under the age of 30 texting, and less time watching technophile, 50-year-old CEOs. Predictive text input systems have been around for 7 - 10 years, "learning" ones for at least 5 years, and anyone remotely familiar with doesn't write SMSes the way you describe. Indeed, I can't think of anyone I know who regularly texts that gives the screen more than a casual glance every 5-10 keypresses.
I wouldn't be surprised in the least if regular texters used to a traditional keypad and predictive input will be at least as fast as people using an on-screen keyboard.
I think you're dramtically over-estimating the benefits of tatile feedback, and ignoring how interface action, responsives, and audible feedback can compensate, or even improve on the experience.
I don't. However, I'm willing to be convinced, which is why I'm waiting before passing judgement.
Aaah, crap. You mean I actually have to *look* at a device I'm using. Good god, how will I cope?
Dunno. Can you touch type ?
And incidentally, I find I'm pretty good at noticing the feedback when my finger hits a screen. I don't waft it in front of the screen, or push it through any more.
I've no doubt. But how do you know which part of the screen you've hit, what was under it at the time and whether or not the UI has responded correctly ?
Do me a favour and save your arguments about how the lack of feedback will kill this device until you've actually tried one.
You must have me confused with someone else, because I made no such argument.
The UI has had an awful lot of work put into it, I doubt it's going to be difficult to use.
I doubt it is either, however, that doesn't mean it will be *efficient* to use, or as easy as it could be using real buttons. Added to which, a) a lot of Apple's recent UI efforts haven't exactly been confidence-inspiring and b) they don't have any experience in this market yet, so even it "being good" is not a given.
And to be honest, even in the worst case scenario where you're right and it is slower to use, I think that's a small price to pay for all the benefits.
I've yet to see a lot of significant benefits over the alternatives.
The larger screen, and much larger buttons for dialing numbers are two features I'd sacrifice an awful lot for.
Your phone must really suck if you feel that the standard number pad is so awkward to use that you think one on a phone-sized touchscreen would be easier.
Indeed. Apple were among the first to get this sort of "correctional" interface behavior right in other spheres (e.g., even early versions of the Mac OS allowed you to "stray" a little when moving from a menu item into its sub-menu; rather than requiring the cursor to remain precisely within the bounds of the selected menu item, the mouse could venture outside a small distance -- so long as it continued generally moving toward the submenu -- without causing the submenu to disappear).
"Corrective" isn't really the way to describe it - that would imply the system restricts the mouse movement for you.
What MacOS did was allow for a certain amount of "looseness" in the mouse movement. It did this by creating a triangle bounded by the mouse pointer (at the time of the click or menu expansion), the top-left of the submenu and the bottom-left of the submenu. As long as the point stayed within the boundaries of that triangle, the submenu stayed open.
Yeah, I'm really going to know on a Treo that I'm on C and not V. Lots of tactile feedback in dozens of identical little buttons.
I was actually thinking of a regular numberpad, however, you'll still get more from "dozens of identical little buttons" than you will from a smooth piece of plastic with "dozens of little buttons" drawn on it.
Assumption.
No. Deduction.
And probably totally ignores all of the other things one does on a "normal" phone like handling voicemail, or diving down in the menu tree to change a setting, or things like multiway conference calls that no one does on their existing phone because they never read the manual and memorized the procedure.
No, it doesn't. Quite the contrary, in fact.
But hey. You've convinced yourself you're not going to like it. Fine. Horses for courses. And one less person ahead of me in line...
Ah, you must be an Apple zealot, someone for whom "maybe it won't be perfect because of these reasonable arguments" translates to "it sucks horribly and will never work" when referring to any Apple product.
Which part of the I'll-wait-and-see position I advocated is leading you to believe I've "convinced myself" ?
If contacts have worked well for you before, then they will work just as well for the iPhone right?
Can't say. Haven't used it. My response based on the mockups on Apple's website is "maybe".
The point at hand was how having a contact list is no replacement for simply typing in numbers, which I argue is untrue and you would seem to also assert.
A contact list is great if it's populated. Not having a way to directly type in number is a pretty major PITA if need to, however. You need both to be well done for a phone to be good.
Personally I have found use of contacts on the phones I've owned to be annoying, and so I don't use them all the time even if I have them loaded. That would thus be the counterargument for why use of contacts in an iPhone would not be of use, which is what I was refuting.
I just don't see what the fuss about the iPhone's contact list is about. Certainly from looking at the mockups, it doesn't seem to offer any opportunity for significant improvement over existing methods.
No, actually I type and know I missed it by observing the screen - that's the reason to touch-type, right? So you can observe the screen as you work? On the iPhone the screen and your typing are as one.
Well, I'd be quite willing to bet your subconscious recognised that you've made an error long before (relatively speaking) your conscious mind sees the error on screen, but I doubt you'll agree since you seem to have made your mind up.
What Apple is trying to do is to make contacts actually usable to well, contact people with. Just because you've not had that experience in the past on a phone does not mean it cannot be done.
I've had that experience with every mobile phone I've ever owned. Even with hundreds of "contacts" in my phone book I can get to one of them in a few seconds and (at worst) half a dozen button presses. For the 5 or so people I call frequently and regularly, it's just a matter of holding down a number on the keyboard or saying their name. Heck, it was true (except for the voice recognition) on the Nokia 5310 I had a decade ago.
Of course, I've never owned anything but Nokias - other phones' phonebook interfaces may well suck as much as you imply, but they certainly don't by definition.
How sure are you? I'd say relative distance from a fixed location has a lot to do with how you type. When your finger is over the keyboard, then goes down to press a key how often are you feeling the other keys around it, vs. just hitting the kay you want? I find as I type this message I make little use of the keys around the key I am pressing, and simply type directly based on position.
Ah, but you know instantly when you've duffed it and missed a key or hit between two keys, don't you, even without looking at the screen ?
Tactile feedback isn't just about knowing when you get it right, it's (probably more importantly) knowing about when you've gotten it *wrong*.
If you want to play semantic games and claim that by compiling BSD code and putting it into a box, it's no longer the same code, I'm sure you and any other like minded people can go amuse yourselves that way.
I didn't, you (apparently) did. You didn't say "compiled", you said "modified".
Most reasonable people (and, for that matter, US Copyright law) see it as clearly still being Foo (albeit a modified Foo), and thus have lost the freedom to modify and enhance that Foo.
If it's just compiled - that is, unmodified - you have lost nothing, because you can just go out and get the BSDLed source for Foo and modify it to your heart's content.
As I have mentioned, the killer app for a phone is break away from anachronistic "dialing" the phone. There is no reason why making a call should not be as simple as stating someone's name.
It already is, from that perspective. Heck, my Nokia 6310i has voice recognition for dialling and it's 5+ years old. However, the problems with the system aren't trivial:
* You need the contact in your phone.
* You either have to tag them with a voice-dial -or- the phone needs a good voice recognition system and you have to remember what you called the contact
* Duplicates (in the case of a voice recognition system) require manual intervention anyway.
If you have a small number of frequently-dialled contacts, the whole "voice dialling" thing works ok. If you don't, it doesn't and going via the "phonebook" is easier.
The lack of tactile input is, IMHO, going to be a major usability problem - but I'm reserving my final judgement until I can actually use one.
Does the mouse on your computer provide tactile feedback when you move the cursor over a button? Trackpads or pointers on laptops?
I'd argue that's a meaningfully different form of UI interaction.
Do any other touchscreens (e.g., in grocery store checkout lines) do so? Do any PDAs with touchscreens provide tactile feedback? I can't think of many, if any, that do, and that doesn't seem to have hindered them.
That's actually a relatively interesting point and I'd be interested to see any studies of error rates and input speeds for traditional numberpads vs touch-screen numberpads. I'd expect the touchscreens to lose, however, *especially* in the context of any sort of multitasking.
How often do you actually use a phone without looking at it? Even when I'm just hitting speed dial buttons I'm usually looking at the phone to double-check that it's calling the right person. Especially relevant: how often do you use advanced features like web surfing or text messaging/email without looking at the phone? Unless you've got a screen reader in there, don't you kind of have to look at it to use those features? Ditto for watching video on a handheld device.
It's not just about using it without looking at it, it's about using it while looking at it, but not really having to concentrate.
The tactile feedback you get from real buttons tell you both where you are in the UI relative to other elements (ie: where your thumb is on the buttons) and when you have successfully complete an action (ie: the "click" when the button connects). This helps you both move your thumb quickly to the right spot by feel (even if you're also looking) and know that you've done something without having to actively concentrate on it. The iPhone will lack both these features and, hence, you'll have to actively watch the UI not only for the feedback that you'd normally respond to with muscle memory, but also correct positioning within the UI you'd normally do from muscle memory. Basically, you'll have to concentrate harder to use it.
I disagree about the fragility of the touch screen. If children's devices (Nintendo DS) can have a touch screen, I don't see why adult devices should be concerned about the fragility of such a thing.
Because a Nintendo DS is ca. US$100 and an iPhone will ba ca. US$500.
I also suspect that since you'll have to do everything to the iPhone through the touch UI, and because it's not going to use a stylus, the iPhone's screen will see more "use".
As for the tactile feedback, I think you're underestimating the UI mechanisms used to use the device. The most pressing activity on a phone is dialing.
Text messages are another that springs instantly to mind. Realisticaly, though, it's basically everything that involves using the phone. You've probably never thought much about the tactile feedback and hence don't realise how much you use it, even if only unconsciously.
For example: you know that to get to a certain UI element you have to do a certain number of button pushes. Say, two down arrow presses, across one and "enter". With real buttons, you get feedback when a button is pushed and your muscle memory can make getting to UI elements you are familiar with an automatic, practically instant procedure (because you don't need to watch it to make sure each action was successful, that you didn't "overshoot", etc). Further, you know where your thumb is relative to the next button you have to push, because you can feel it. With the iPhone, you will have to actively watch the UI as it changes to navigate through it, to know which intermediate step it is at during each phase and to know where your thumb needs to be.
Simple: patents. Infringe a patent and the owner of your BSD code is the patent holder.
Broken patent laws are an entirely different discussion.
Is the FSF use of the word free really deceptive?
Yes, because they assume "free" and "GPLed" (or equivalent) are synonymous and use the terms interchangeably, despite that not being a definition most people would perceive.
I don't think so. Many if not most people would consider our western democracies "free countries", where you are allowed to do whatever you want except harming people.
They would, but they'd be wrong in a very real sense, as evidenced by the numerous laws involving victimless crimes (pretty much anything surrounding drugs, for example).
However, I understand the point you are trying to make, that we ostensibly have a "free" society due to the restrictions placed on certain activities (eg: murder).
However, it's not analagous, as laws like that are about keeping things that are "yours", "yours", whereas the GPL is about making things that are "theirs", "yours".
These restrictions are analogous to the restrictions placed on software licensed under the GPL. Thus, I would hardly call the use deceptive.
Perhaps if you subscribe to the FSF doublespeak. However, requiring someone to GPL their own code if they wish to distribute some sort of GPL-code-derivative isn't in the same ballpark - heck, not even playing the same game - as laws against, say, murder and theft. The former are assumptions about what you should be able to do with *other people's work*, the latter about the assumptions that can be reasonably made regarding *yourself* and *your property*.
That is, unless you happen to agree with the collossal arrogance of people who consider the "right" to modify and redistribute someone else's work is somehow on par with so-called "human rights" like freedom of speech and self-defense. If that's true I doubt we'll be able to find a common ground.
The word "right", in the context of various "human rights", carries a heavy meaning and using it for such mundane things as licensing of source code significantly cheapens its value.
"Open Source" is a *vastly* more honest term that "free software", because it conveys accurate meaning without relying on loaded/ambiguous definitions of the word "free" which only a relative handful of people even know, let alone subscribe to. Of course, the term "free software" was chosen precisely *because* of that loaded meaning, which in my mind puts it in the same arena as the kind of political propanda that is used to justify - amongst others - the various "War On X" projects.
Suppose you release package Foo under the BSD license. Now a company uses a modified version of that software in their product Bar. As an end user, I buy a Bar. Now I've got a product running a non-free version of Foo.
No, you have a product called FooBar. I.e. a combination of Foo and Bar.
The company won't give me the sources to their version, so I can't modify or enhance it, or fix bugs. That's about as non-free as software gets, and it was clearly facilitated by the BSD license.
The code for Foo remains "free". The code for Bar, does not. The BSDL code remains "free", even if a product that is using it is not.