Yes, but you don't use the volume knob on your home stereo to control which song is being played EITHER. Apple's pretty much screwed things up -- the knob is unforgivably modal.
'Cos I'm a man of the world. Happily, most of my mp3s are either English or Japanese. But there are the ocassional Russian, Hebrew or Chinese songs in there, and hell -- there are a damn lot of languages in this world of ours. Apple's had big internationalization pushes before. (indeed, that was one of the prime reasons for the dual forked filesystems)
The LCD display is too small, it remains to be seen what the power consumption or usability of the backlight is, the four buttons (five, actually, I suspect) are likely insufficient, and probably rather modal. I dare not imagine how badly they've ginnied up the volume control. Apple's support for ID3 is woefully insufficient on iTunes and on iPod. (so is everyone else's, more's the pity)
I will grant them a half-point on the auditory feedback on the wheel (but no tactile feedback! for shame) and a half-point for displaying multiple languages simultaneously (but no plug ins for future language expansion? tsk tsk)
Oh, well, so basically you spoke generally even though you had a specific meaning in mind that was not difficult to articulate.
Thanks.
You are of course entitled to your opinion that it is wrong. As you might have guessed by my.sig here, I take a slightly dimmer view of copyright. I don't mind it in a general way, but I don't much like what we've got now in many specifics. I understand why there are those who are entirely opposed to the notion, however. I can't say that they're _wrong_ per se; in both cases, I think that we are interested in achieving the same result in the end.
I think that you're deluding yourself, however. If copyright infringements are morally wrong, and you have, or should have, knowledge that P2P filesharing is very predominently used in illegal manners, save by citizens of countries that do not have copyright laws, for (rare) uncopyrighted materials or materials for which indiscriminate sharing is permitted, you ought to be against them. I see your position as being against yellow fever, but unwilling to condemn mosquitoes in the process, even though they're nominally innocent.
I agree with you regarding teaching moral values however. There are many differing moral positions, and mere endorsement by legal authority is unconvincing.
Your second assessment of what copyrights are ("IP" is another misleading term -- the copyrights may be property, but the copyrighted material, e.g. songs, are not. I'd avoid it, so as to keep out of threads like this one we're having) is much more on the ball.
I would add though, that there are three interests to be served, and not just the one you mention. Copyright is a delicate balancing act between 1) the promotion of learning; 2) the promotion of the public domain, and; 3) the interests of the author. The entire, judicially reconized objective of copyright in the US (which has about the only sane system in the world) is to promote the public interest, through promoting to a lesser degree, the private interest of authors. Copyright's not natural; it is not earned or deserved; it is granted by the government, and then only if they want to do so. They may abstain, and authors can suck eggs.
Re:Suggestion for users about the ads...
on
Slashdot Updates
·
· Score: 0, Flamebait
That's Taco's problem -- not mine.
I absolutely _loathe_ advertising. Cannot stand it in the least, and I go through a bit of effort to avoid it. I eagerly await the development of some sort of computerized glasses, so that I can filter it out irl. (kind of the reverse of those ads they have at baseball games now)
Frankly, I've not even got any obligation to look at ads... filters are only slightly more complicated than closing my eyes when they're onscreen, and I don't see you advocating doing away with that.
TV has faced the same thing for decades. They can put on ads, but they can't make people watch 'em.
If/. can't support itself, let it go subscription. And then, because I have no desire to pay for a smaller community of posters who aren't compensated for the only good content around here, I'll take my business elsewhere.
Slashdot's best interests are the furthest thing from my mind, in much the same way that I shed no tears as dot-coms melted down.
You said that "[w]holesale copying of music against the permission of its creators is wrong," but I've got to take issue with this.
It's not wrong.
People copy music wholesale without the permission of the creators ALL THE TIME. Indeed, Disney is known for this. They have two entire movies, their "Fantasia" series, which liberally copy music without permission from the creators. (many of whom were long dead)
And if the copyright scheme in this country were like that of the early Republic, copying music would be perfectly allright, and not a copyright violation at all. A lot later and you'd merely have to wait for the copyright to expire -- which wouldn't take terribly long.
It's about as wrong as installing a picket fence at your house that doesn't comply with zoning regulations, in many cases. Reasonable people are not only perfectly capable of arguing over whether some particular act ought to be infringement, and even whether we ought to have copyrights at all. (which are not mandated)
You don't give children much credit either. They are often pretty capable of calling a spade a spade. (c.f. "The Emperor's New Clothes")
Both are indeed illegal. Yet piracy remains copyright infringement, not stealing. Stealing involves taking something and thus depriving the rightful owner of it.
Copying something does not involve taking, deprivation, or even anything capable of being owned. (copyrights are ownable, content is not)
Except, of course, that Jobs had nothing whatsoever to do with Pong. *Nothing*
Willy Higinbotham invented a sort of ur-pong called "Tennis for Two" in the late 50's. And there was another pong precursor on the Magnavox Odyssey in the very early 70's.
Nolan Bushnell and Al Alcorn developed Pong in the early 70's.
At that time, Jobs, who knew very little about electronics, despite what he claimed ("Infinite Loop" has some scathing comments on this) would've either been a senior in high school, or in college in Oregon, in the process of dropping out and not eating.
Jobs didn't work at Atari until the mid-70's. He was over his head right away, enlisted Woz who was actually responsible for Breakout, and wound up with a design that was indeed so unusual that it had to be redesigned anyway.
Bushnell promised and delivered $5000 to Jobs for it. Jobs promised Woz half -- of $700. (i.e. $350) Woz didn't find out until years later that he had been cheated.
Pardon. Where filename extensions are used by web browsers where there is no actual web server. I'm afraid that although I am a former web designer, I invariably found it more convenient to keep files locally and preview them in a browser, than to work directly off of a server and preview them properly.
Where there was nothing interactive, it worked out okay.
Where things were interactive, there were already craploads of problems (that a smart developer could solve and make some money from) with regards to difficulties in maintaining multiple front ends in varying stages of development on a working, but not live backend, and in simultaneous front and back end development.
It is indeed good that the _real_ web, uses headers instead.... now if we can only convince people to accept that and dump their filename extensions....
I didn't say that it wouldn't fail due to business issues -- I said that it was a better approach to developing better UI from the standpoint of _ONLY_ being concerned with UI.
It's depressing though that Be never had an opportunity to succeed.
I'll note, however, that NeXT, who wrote the OS that OS X is very clearly mostly based upon, ALSO did not succeed, or have any number of significant applications. Yet support of OS X often implicitly includes support of NextStep. Unfounded support, IMO, as a former NextStep user. (best UI on a Unix doesn't translate into best UI overall, or even a good UI)
I mostly try to stick to looking at good UI, and stay away from the business angle, thouhg I'd not want to develop UI that was intended solely for nefarious purposes. (thus I like the idea that MS had re: integrating IE with the OS, but they should nevertheless be broken up)
Oh, OS X is not uniformly bad -- perhaps I should've made that clear. It is just mostly bad. I am actually very fond of sheets. I think they're a damn good idea. There are a few things in OS X that are worthwhile, certainly. But it brings relatively little to the table that is new or a substantial improvement.
Indeed, I don't recall that I've seen rejections of sheets at all, although the utility of dialogs of any stripe has been criticized soundly by many different commentators.
Re: Linux, I simply haven't seen anything that leads me to believe that that's true. It _ought_ to be, and I very much want it to be because I like the ideas behind the Linux community, but I haven't seen it. Would you mind citing some examples of originals and their new and improved derivatives.
Yeah, well, it's a lot easier to find computers to play games on nowadays -- Unix was developed for games, after all;)
Plan9 does have some interesting things... if it were my job to deal with HCI at such fundemental levels, and given that I'm presently radically changing professions, it won't be, I'd install it. It does sound interesting. At present, I am content to read about it as fuel for pontification.
And long filenames in Windows were no fun when you had to see how they really worked... ye olde "MICROS~1.FOO"
And now they're gone, and longer (but insufficiently long, IMO) filenames are here. And people had to take it on the chin during the transitional period, but that isn't forever, and there's a clear advantage.
People developing on Windows should just start using Streams because they're good, and through use will become the standard way of doing things.
We'd still be on six character filenames* in TOPS-10 if legacy issues were ever allowed to _prevail_. Be concerned, but don't let them dictate to you, is all.
Be the first, my anonymous friend -- use the damn feature, it's there for you to do so!
(*Turning the game "Adventure" into "ADVENT" -- bleh)
I wouldn't expect CLI revolutions from the Unix folks, frankly.
I think that Unix has some good ideas behind it:
*I am attracted to the idea of mountpoints, though not confident in them entirely. They mesh well, however, with the Win95 notion of a 'Computer' icon, and I'm fond of that.
*Many small tools working together is a great idea, but insufficiently easy to use at present.
*Configuration files that are universally humanly readable are a good idea, although ASCII is not.
*The power in CLIs is desirable, having it locked up in an emulator of a 60's era teletype is not.
*Multiuser systems are desirable, global filesystem structures are not (my unproven hypothesis) and at any rate, suspending and resuming a user environment a la 'screen,' probably with heavy use of features borrowed from hibernation is really necessary to beef it all up.(some of the MS stuff seen recently re: fungible computing resources is _excellent_ though the security and business issues need to be hashed out -- i'd like being able to log in to two machines side by side, and use the 2nd as an additional monitor, and extra computational power all for the same session)
*ACL security has got to go. Reimplement it on top of something better, like (I'm told) capabilities, if you need it, but totally redo the fundementals.
*Flat filesystems have to go, but this is already being worked on, I hear.
*Links are cool -- now make 'em work across disks. Combine them with forked filesystems.
*etc, etc, etc.
But you'll never hear this coming out of the Unix crowd.... Imitation has its place, and Unix is fundementally an imitation of Multics, but for God's sake, take some initative!
Apple's PC Exchange software, which originally came along in the early 90's, transparently (to Mac users) created invisible files to store metadata and resource forks, letting the files be mere data forks.
It's not pretty if you're looking at it in DOS, or something else that doesn't respect the PCEx encoding, but it works great for a Mac user.
As for streams, that's your own damned fault. Imagine that streams are just another way to handle tarring files together. Now stream all the support files and folders in, say, MS Office into the executable. One icon, no clutter, double click and it just runs. But with a good shell that can optionally explore the streams, you can still dig in and customize on the rare ocassions you need to.
(and could couple that with some markup for formatting so that the stream-contents were presented in a more user-friendly manner than an ordinary Windows folder)
Streams rock; Apple just should've done a better job in marketing them.
Sigh. Did those monitors _work_? I'd kill for a good 20" greyscale. Greyscale monitors don't have discrete physical pixels -- the entire phosphor coating is continuous. This makes them hella good for reading text.
Additionally, monitor glass is leaded, and presents a toxic waste hazard. In a number of places it's illegal to just throw them out. (aside from being morally repugnant everywhere)
Be had the right idea. Start from a clean slate, and design in the backwards compatability in a sane manner.
Linux has over thirty years of history weighing around its neck like an albatross. It needs to be dumped, rebuilt, and made POSIX compliant so that it _can_ run the old stuff, but _isn't_ the old stuff.
So, I'm curious.... Regardless of all that you say, why the hell are you storing file type metadata (and presumably that per-specific-file-basis creator metadata) in the filename?
After all -- you will never change the type code unless you change the type as well, correct? Not so long as you can easily assign a specific file to open with a specific -- DIFFERENT -- application than the default for that type. Would it not be unnecessary to change file.jpeg to file.exe?
Of the few types that seem more interchangable -- HTML is of course structurally often, but not always, the same as TXT -- wouldn't a more MIME-like system that permitted the file to simultaneously be known as _both_, since it _is_ both, be superior?
About the only exception I can think of to this general rule, where filename extensions are important are for the web, and between other information passed to the client from the server, and the browser's ability to use other metadata filetypes, in-file declared type information, and other Unix-like magic numbers, it is wholly unnecessary and ascribable to lazy or shortsighted programming.
At any rate, a three character string is worse than a four character string, and both are worse than an arbitrarily long string that describes the type in _human_readable_ format. Again, MIME strikes me as doing a better job for type most (not all) of the time, than either camp's codes.
Like the way that NTFS or FAT can't support Mac metadata or forked files. Oh, oh wait. My mistake. They _CAN_, you dolt. Hell, we backed up Mac files on one of the Win2K machines at my old job _constantly_. Ran like a champ.
I suppose next you'll have never heard Linus talking about how support for nonnative filesystems, and all that comes with them (specifically including Mac fs'es) is necessary for Linux, so as to be a good neighbor in heterogeneous environments?
The MacOS method is good -- not perfect, but better than virtually anything else out there. (though as a Mac user, I really find BeOS's handling of this fascinating) Other systems would do well to adopt _IT_. You'll never improve a thing if what you have now is good enough. Remember 8.3 filenames? Weren't those fine too?
Dragging disks to the trash to eject them is a _FEATURE_. I swear, I am not kidding. God's truth, it's a feature.
Now sit here beside the fire, my children, and receive the lore of early Mac disk management....
As a cost-savings measure, because Apple had (wisely) chosen to use the brand-new Sony 3.5" floppies with a whopping 400kB of capacity, the Mac had only one drive. (and this was a _big_ floppy for the time, in terms of storage space) Although users could have a second, or even a lot of external, daisy-chained FDDs, they couldn't be assumed to.
So there was a problem: how would a user use two floppies simultaneously? After all, 1) the noun-verb language of the GUI demands that there be a visible target for an icon to be moved. And anyway, 2) many users would want an OS disk, an application disk, and a data disk... maybe a lot.
The solution was this: the volume was slightly divorced from the media!
That is, if you want to copy 'Empty Folder' (because the original OS couldn't create new folders) from disk Fred to disk Barney, and Fred contains a copy of the OS to boot off, you'd do this:
1) Boot up from Fred.
2) Select Fred on the desktop, and use the Eject Disk command in the menu. This ejects the physical disk, but leaves a 'shadowed' copy of the volume on the desktop.
3) Insert Barney, which is then mounted on the desktop.
4) Drag 'Empty Folder' from the shadowed Fred volume to the fully active Barney disk.
5) The OS will at this point, autoeject Barney, leaving a shadowed copy of _its_ volume on the desktop, and ask for Fred to be inserted
6) Insert Fred, and the OS (which obviously couldn't've cached this) copies 'Empty Folder' to memory, then autoejects Fred, and asks for Barney to be inserted
7) Insert Barney, and the OS writes 'Empty Folder' to it, leaving a shadowed copy of Fred, still on the desktop.
Old-time Mac users will be familar with the infamous Disk Swap Tango.
However! What is of note here, is that the Eject Disk command literally ejects the disk, but does not unmount the volume. In order to dismount a volume, you use the entirely seperate Put Away command.
In fact, if you use Put Away on a volume that is active because the disk is physically inserted, the disk is ejected AND the volume is dismounted. Clearly, Put Away should have been a popular command.
Except that, ultimately, the developers making the damn thing found this cumbersome. Even thought the UI people (who are human, after all) were telling them that this was the best way to do it. So one programmer, following the Mac edicts of 'there's more than one way to do it' and 'direct manipulation is superior to abstract manipulation' (i.e. moving things with icons, clicking on close boxes, is better than using the menus to accomplish the same goals) made a shortcut whereby if you dragged an active or inactive disk/volume to the trash, it would be Put Away. (and of course, if the disk was present, ejected)
Although this was immediately picked up on by the HCI people as a bad idea -- because doesn't that imply that the disk is being erased? -- they found that it was, in practice, a damn lot more useful and easy to remember than the above confusion with the menus.
A few years later, of course, hard disks became commonplace, and the need for this behavior was mostly lost. Nowadays in fact, Eject Disk both dismounts _and_ ejects the disk, instead of only the latter.
So it was _never_ a kludge. It was in fact a really good shortcut that wound up becoming more common than the behavior that it was originally intended to be a power user's way of accomplishing! In fact, tests in the mid 90's indicated that changing the Trash into an Eject icon was disconcerting, and so never really pursued at the time, though it had been on the drawing board for ages.
It's not foolishness. Not in the least. I will agree, of course, that a physical eject button wired to the OS so that it is aware that a disk is dismounted is also a good idea. But given the needs in the early/mid 80's, the old behaviors made perfect sense.
Well, 1) many commentators with an eye towards, and knowledge about HCI find OS X to be a step backwards from MacOS. Good UI features in Windows, Unix (there are one or two), BeOS, and other, smaller, OSes are very often not present. OS X is a nice-looking, not especially usable system.
2) Linux will never get anywhere if it _just_ mimics others. Indeed, if OS X had merely mimiced features from those other UIs, it would not really amount to much. What is essential is that the reasoning behind what other developers do is understood, that when you want to create something new you apply a rigorous series of scientific tests, and that you communicate your findings effectively. Copying popular features in a cargo-cult fashion will not result in a good, usable OS.
3) Linux developers must have no sacred cows. The very roots of the system must be vetted, and if found lacking, fixed. Good UI is at the heart of it; it is not a facade. Examine the filesystem, the way that the computer behaves in crises, boot times, the security model, how and whether multiple users should be handled. Ignore those, and you'll be building on sand.
I would think that copyright law is entirely inapplicable for an uncopyable work. Of course, the CDs are eminently copyable, so it's difficult to see any application.
OTOH, it is arguable that they are not copyable _enough_ and thus the publishers are abusing their copyright and are in need of corrective measures.
Stephen King did not try it _right_. This is most commonly known as the Street Performer Protocol, and it involves 1) the work being completed and delivered to an escrow agent who will 2) release it to the public domain if the set cost is met.
I eagerly await hearing news of someone trying it for real.
While I will agree -- fair use doctrine does not create a right -- I disagree vehemently regarding the prevention of copying w/o contracts to that effect.
It is an abuse of copyright, it chills fair use, it chills the public domain requirement for copyrighted materials (i.e. that they must ultimately be copyable), and it violates the novel space shifting doctrine. (as well as in some cases, statutory exceptions, e.g. software backups)
Slashdot Mirror
Yes, but you don't use the volume knob on your home stereo to control which song is being played EITHER. Apple's pretty much screwed things up -- the knob is unforgivably modal.
'Cos I'm a man of the world. Happily, most of my mp3s are either English or Japanese. But there are the ocassional Russian, Hebrew or Chinese songs in there, and hell -- there are a damn lot of languages in this world of ours. Apple's had big internationalization pushes before. (indeed, that was one of the prime reasons for the dual forked filesystems)
No, that's virtually no UI alright.
The LCD display is too small, it remains to be seen what the power consumption or usability of the backlight is, the four buttons (five, actually, I suspect) are likely insufficient, and probably rather modal. I dare not imagine how badly they've ginnied up the volume control. Apple's support for ID3 is woefully insufficient on iTunes and on iPod. (so is everyone else's, more's the pity)
I will grant them a half-point on the auditory feedback on the wheel (but no tactile feedback! for shame) and a half-point for displaying multiple languages simultaneously (but no plug ins for future language expansion? tsk tsk)
I was of course, referring to the works for which there was no copyright or for which the copyright had since expired.
Similarly, Beauty and the Beast -- no royalties. 101 Dalmatians -- royalties.
Certainly, there will be a day when Stravinsky's estate receives no more. At least, I certainly hope so.
Disney's not totally spotless either. There are various allegations floating around... I distinctly remember the ones regarding The Lion King.
Oh, well, so basically you spoke generally even though you had a specific meaning in mind that was not difficult to articulate.
.sig here, I take a slightly dimmer view of copyright. I don't mind it in a general way, but I don't much like what we've got now in many specifics. I understand why there are those who are entirely opposed to the notion, however. I can't say that they're _wrong_ per se; in both cases, I think that we are interested in achieving the same result in the end.
Thanks.
You are of course entitled to your opinion that it is wrong. As you might have guessed by my
I think that you're deluding yourself, however. If copyright infringements are morally wrong, and you have, or should have, knowledge that P2P filesharing is very predominently used in illegal manners, save by citizens of countries that do not have copyright laws, for (rare) uncopyrighted materials or materials for which indiscriminate sharing is permitted, you ought to be against them. I see your position as being against yellow fever, but unwilling to condemn mosquitoes in the process, even though they're nominally innocent.
I agree with you regarding teaching moral values however. There are many differing moral positions, and mere endorsement by legal authority is unconvincing.
Your second assessment of what copyrights are ("IP" is another misleading term -- the copyrights may be property, but the copyrighted material, e.g. songs, are not. I'd avoid it, so as to keep out of threads like this one we're having) is much more on the ball.
I would add though, that there are three interests to be served, and not just the one you mention. Copyright is a delicate balancing act between 1) the promotion of learning; 2) the promotion of the public domain, and; 3) the interests of the author. The entire, judicially reconized objective of copyright in the US (which has about the only sane system in the world) is to promote the public interest, through promoting to a lesser degree, the private interest of authors. Copyright's not natural; it is not earned or deserved; it is granted by the government, and then only if they want to do so. They may abstain, and authors can suck eggs.
That's Taco's problem -- not mine.
/. can't support itself, let it go subscription. And then, because I have no desire to pay for a smaller community of posters who aren't compensated for the only good content around here, I'll take my business elsewhere.
I absolutely _loathe_ advertising. Cannot stand it in the least, and I go through a bit of effort to avoid it. I eagerly await the development of some sort of computerized glasses, so that I can filter it out irl. (kind of the reverse of those ads they have at baseball games now)
Frankly, I've not even got any obligation to look at ads... filters are only slightly more complicated than closing my eyes when they're onscreen, and I don't see you advocating doing away with that.
TV has faced the same thing for decades. They can put on ads, but they can't make people watch 'em.
If
Slashdot's best interests are the furthest thing from my mind, in much the same way that I shed no tears as dot-coms melted down.
You said that "[w]holesale copying of music against the permission of its creators is wrong," but I've got to take issue with this.
It's not wrong.
People copy music wholesale without the permission of the creators ALL THE TIME. Indeed, Disney is known for this. They have two entire movies, their "Fantasia" series, which liberally copy music without permission from the creators. (many of whom were long dead)
And if the copyright scheme in this country were like that of the early Republic, copying music would be perfectly allright, and not a copyright violation at all. A lot later and you'd merely have to wait for the copyright to expire -- which wouldn't take terribly long.
It's about as wrong as installing a picket fence at your house that doesn't comply with zoning regulations, in many cases. Reasonable people are not only perfectly capable of arguing over whether some particular act ought to be infringement, and even whether we ought to have copyrights at all. (which are not mandated)
You don't give children much credit either. They are often pretty capable of calling a spade a spade. (c.f. "The Emperor's New Clothes")
Both are indeed illegal. Yet piracy remains copyright infringement, not stealing. Stealing involves taking something and thus depriving the rightful owner of it.
Copying something does not involve taking, deprivation, or even anything capable of being owned. (copyrights are ownable, content is not)
These are fine differences, but they're there.
Except, of course, that Jobs had nothing whatsoever to do with Pong. *Nothing*
Willy Higinbotham invented a sort of ur-pong called "Tennis for Two" in the late 50's. And there was another pong precursor on the Magnavox Odyssey in the very early 70's.
Nolan Bushnell and Al Alcorn developed Pong in the early 70's.
At that time, Jobs, who knew very little about electronics, despite what he claimed ("Infinite Loop" has some scathing comments on this) would've either been a senior in high school, or in college in Oregon, in the process of dropping out and not eating.
Jobs didn't work at Atari until the mid-70's. He was over his head right away, enlisted Woz who was actually responsible for Breakout, and wound up with a design that was indeed so unusual that it had to be redesigned anyway.
Bushnell promised and delivered $5000 to Jobs for it. Jobs promised Woz half -- of $700. (i.e. $350) Woz didn't find out until years later that he had been cheated.
Pardon. Where filename extensions are used by web browsers where there is no actual web server. I'm afraid that although I am a former web designer, I invariably found it more convenient to keep files locally and preview them in a browser, than to work directly off of a server and preview them properly.
Where there was nothing interactive, it worked out okay.
Where things were interactive, there were already craploads of problems (that a smart developer could solve and make some money from) with regards to difficulties in maintaining multiple front ends in varying stages of development on a working, but not live backend, and in simultaneous front and back end development.
It is indeed good that the _real_ web, uses headers instead.... now if we can only convince people to accept that and dump their filename extensions....
I didn't say that it wouldn't fail due to business issues -- I said that it was a better approach to developing better UI from the standpoint of _ONLY_ being concerned with UI.
It's depressing though that Be never had an opportunity to succeed.
I'll note, however, that NeXT, who wrote the OS that OS X is very clearly mostly based upon, ALSO did not succeed, or have any number of significant applications. Yet support of OS X often implicitly includes support of NextStep. Unfounded support, IMO, as a former NextStep user. (best UI on a Unix doesn't translate into best UI overall, or even a good UI)
I mostly try to stick to looking at good UI, and stay away from the business angle, thouhg I'd not want to develop UI that was intended solely for nefarious purposes. (thus I like the idea that MS had re: integrating IE with the OS, but they should nevertheless be broken up)
Oh, OS X is not uniformly bad -- perhaps I should've made that clear. It is just mostly bad. I am actually very fond of sheets. I think they're a damn good idea. There are a few things in OS X that are worthwhile, certainly. But it brings relatively little to the table that is new or a substantial improvement.
Indeed, I don't recall that I've seen rejections of sheets at all, although the utility of dialogs of any stripe has been criticized soundly by many different commentators.
Re: Linux, I simply haven't seen anything that leads me to believe that that's true. It _ought_ to be, and I very much want it to be because I like the ideas behind the Linux community, but I haven't seen it. Would you mind citing some examples of originals and their new and improved derivatives.
Yeah, well, it's a lot easier to find computers to play games on nowadays -- Unix was developed for games, after all ;)
Plan9 does have some interesting things... if it were my job to deal with HCI at such fundemental levels, and given that I'm presently radically changing professions, it won't be, I'd install it. It does sound interesting. At present, I am content to read about it as fuel for pontification.
And long filenames in Windows were no fun when you had to see how they really worked... ye olde "MICROS~1.FOO"
And now they're gone, and longer (but insufficiently long, IMO) filenames are here. And people had to take it on the chin during the transitional period, but that isn't forever, and there's a clear advantage.
People developing on Windows should just start using Streams because they're good, and through use will become the standard way of doing things.
We'd still be on six character filenames* in TOPS-10 if legacy issues were ever allowed to _prevail_. Be concerned, but don't let them dictate to you, is all.
Be the first, my anonymous friend -- use the damn feature, it's there for you to do so!
(*Turning the game "Adventure" into "ADVENT" -- bleh)
I wouldn't expect CLI revolutions from the Unix folks, frankly.
I think that Unix has some good ideas behind it:
*I am attracted to the idea of mountpoints, though not confident in them entirely. They mesh well, however, with the Win95 notion of a 'Computer' icon, and I'm fond of that.
*Many small tools working together is a great idea, but insufficiently easy to use at present.
*Configuration files that are universally humanly readable are a good idea, although ASCII is not.
*The power in CLIs is desirable, having it locked up in an emulator of a 60's era teletype is not.
*Multiuser systems are desirable, global filesystem structures are not (my unproven hypothesis) and at any rate, suspending and resuming a user environment a la 'screen,' probably with heavy use of features borrowed from hibernation is really necessary to beef it all up.(some of the MS stuff seen recently re: fungible computing resources is _excellent_ though the security and business issues need to be hashed out -- i'd like being able to log in to two machines side by side, and use the 2nd as an additional monitor, and extra computational power all for the same session)
*ACL security has got to go. Reimplement it on top of something better, like (I'm told) capabilities, if you need it, but totally redo the fundementals.
*Flat filesystems have to go, but this is already being worked on, I hear.
*Links are cool -- now make 'em work across disks. Combine them with forked filesystems.
*etc, etc, etc.
But you'll never hear this coming out of the Unix crowd.... Imitation has its place, and Unix is fundementally an imitation of Multics, but for God's sake, take some initative!
FAT can. It's just handled differently.
Apple's PC Exchange software, which originally came along in the early 90's, transparently (to Mac users) created invisible files to store metadata and resource forks, letting the files be mere data forks.
It's not pretty if you're looking at it in DOS, or something else that doesn't respect the PCEx encoding, but it works great for a Mac user.
As for streams, that's your own damned fault. Imagine that streams are just another way to handle tarring files together. Now stream all the support files and folders in, say, MS Office into the executable. One icon, no clutter, double click and it just runs. But with a good shell that can optionally explore the streams, you can still dig in and customize on the rare ocassions you need to.
(and could couple that with some markup for formatting so that the stream-contents were presented in a more user-friendly manner than an ordinary Windows folder)
Streams rock; Apple just should've done a better job in marketing them.
Sigh. Did those monitors _work_? I'd kill for a good 20" greyscale. Greyscale monitors don't have discrete physical pixels -- the entire phosphor coating is continuous. This makes them hella good for reading text.
Additionally, monitor glass is leaded, and presents a toxic waste hazard. In a number of places it's illegal to just throw them out. (aside from being morally repugnant everywhere)
I don't think that it went deep enough.
Be had the right idea. Start from a clean slate, and design in the backwards compatability in a sane manner.
Linux has over thirty years of history weighing around its neck like an albatross. It needs to be dumped, rebuilt, and made POSIX compliant so that it _can_ run the old stuff, but _isn't_ the old stuff.
So, I'm curious.... Regardless of all that you say, why the hell are you storing file type metadata (and presumably that per-specific-file-basis creator metadata) in the filename?
After all -- you will never change the type code unless you change the type as well, correct? Not so long as you can easily assign a specific file to open with a specific -- DIFFERENT -- application than the default for that type. Would it not be unnecessary to change file.jpeg to file.exe?
Of the few types that seem more interchangable -- HTML is of course structurally often, but not always, the same as TXT -- wouldn't a more MIME-like system that permitted the file to simultaneously be known as _both_, since it _is_ both, be superior?
About the only exception I can think of to this general rule, where filename extensions are important are for the web, and between other information passed to the client from the server, and the browser's ability to use other metadata filetypes, in-file declared type information, and other Unix-like magic numbers, it is wholly unnecessary and ascribable to lazy or shortsighted programming.
At any rate, a three character string is worse than a four character string, and both are worse than an arbitrarily long string that describes the type in _human_readable_ format. Again, MIME strikes me as doing a better job for type most (not all) of the time, than either camp's codes.
Like the way that NTFS or FAT can't support Mac metadata or forked files. Oh, oh wait. My mistake. They _CAN_, you dolt. Hell, we backed up Mac files on one of the Win2K machines at my old job _constantly_. Ran like a champ.
I suppose next you'll have never heard Linus talking about how support for nonnative filesystems, and all that comes with them (specifically including Mac fs'es) is necessary for Linux, so as to be a good neighbor in heterogeneous environments?
The MacOS method is good -- not perfect, but better than virtually anything else out there. (though as a Mac user, I really find BeOS's handling of this fascinating) Other systems would do well to adopt _IT_. You'll never improve a thing if what you have now is good enough. Remember 8.3 filenames? Weren't those fine too?
What, are you kidding?
Dragging disks to the trash to eject them is a _FEATURE_. I swear, I am not kidding. God's truth, it's a feature.
Now sit here beside the fire, my children, and receive the lore of early Mac disk management....
As a cost-savings measure, because Apple had (wisely) chosen to use the brand-new Sony 3.5" floppies with a whopping 400kB of capacity, the Mac had only one drive. (and this was a _big_ floppy for the time, in terms of storage space) Although users could have a second, or even a lot of external, daisy-chained FDDs, they couldn't be assumed to.
So there was a problem: how would a user use two floppies simultaneously? After all, 1) the noun-verb language of the GUI demands that there be a visible target for an icon to be moved. And anyway, 2) many users would want an OS disk, an application disk, and a data disk... maybe a lot.
The solution was this: the volume was slightly divorced from the media!
That is, if you want to copy 'Empty Folder' (because the original OS couldn't create new folders) from disk Fred to disk Barney, and Fred contains a copy of the OS to boot off, you'd do this:
1) Boot up from Fred.
2) Select Fred on the desktop, and use the Eject Disk command in the menu. This ejects the physical disk, but leaves a 'shadowed' copy of the volume on the desktop.
3) Insert Barney, which is then mounted on the desktop.
4) Drag 'Empty Folder' from the shadowed Fred volume to the fully active Barney disk.
5) The OS will at this point, autoeject Barney, leaving a shadowed copy of _its_ volume on the desktop, and ask for Fred to be inserted
6) Insert Fred, and the OS (which obviously couldn't've cached this) copies 'Empty Folder' to memory, then autoejects Fred, and asks for Barney to be inserted
7) Insert Barney, and the OS writes 'Empty Folder' to it, leaving a shadowed copy of Fred, still on the desktop.
Old-time Mac users will be familar with the infamous Disk Swap Tango.
However! What is of note here, is that the Eject Disk command literally ejects the disk, but does not unmount the volume. In order to dismount a volume, you use the entirely seperate Put Away command.
In fact, if you use Put Away on a volume that is active because the disk is physically inserted, the disk is ejected AND the volume is dismounted. Clearly, Put Away should have been a popular command.
Except that, ultimately, the developers making the damn thing found this cumbersome. Even thought the UI people (who are human, after all) were telling them that this was the best way to do it. So one programmer, following the Mac edicts of 'there's more than one way to do it' and 'direct manipulation is superior to abstract manipulation' (i.e. moving things with icons, clicking on close boxes, is better than using the menus to accomplish the same goals) made a shortcut whereby if you dragged an active or inactive disk/volume to the trash, it would be Put Away. (and of course, if the disk was present, ejected)
Although this was immediately picked up on by the HCI people as a bad idea -- because doesn't that imply that the disk is being erased? -- they found that it was, in practice, a damn lot more useful and easy to remember than the above confusion with the menus.
A few years later, of course, hard disks became commonplace, and the need for this behavior was mostly lost. Nowadays in fact, Eject Disk both dismounts _and_ ejects the disk, instead of only the latter.
So it was _never_ a kludge. It was in fact a really good shortcut that wound up becoming more common than the behavior that it was originally intended to be a power user's way of accomplishing! In fact, tests in the mid 90's indicated that changing the Trash into an Eject icon was disconcerting, and so never really pursued at the time, though it had been on the drawing board for ages.
It's not foolishness. Not in the least. I will agree, of course, that a physical eject button wired to the OS so that it is aware that a disk is dismounted is also a good idea. But given the needs in the early/mid 80's, the old behaviors made perfect sense.
Well, 1) many commentators with an eye towards, and knowledge about HCI find OS X to be a step backwards from MacOS. Good UI features in Windows, Unix (there are one or two), BeOS, and other, smaller, OSes are very often not present. OS X is a nice-looking, not especially usable system.
2) Linux will never get anywhere if it _just_ mimics others. Indeed, if OS X had merely mimiced features from those other UIs, it would not really amount to much. What is essential is that the reasoning behind what other developers do is understood, that when you want to create something new you apply a rigorous series of scientific tests, and that you communicate your findings effectively. Copying popular features in a cargo-cult fashion will not result in a good, usable OS.
3) Linux developers must have no sacred cows. The very roots of the system must be vetted, and if found lacking, fixed. Good UI is at the heart of it; it is not a facade. Examine the filesystem, the way that the computer behaves in crises, boot times, the security model, how and whether multiple users should be handled. Ignore those, and you'll be building on sand.
I would think that copyright law is entirely inapplicable for an uncopyable work. Of course, the CDs are eminently copyable, so it's difficult to see any application.
OTOH, it is arguable that they are not copyable _enough_ and thus the publishers are abusing their copyright and are in need of corrective measures.
Stephen King did not try it _right_. This is most commonly known as the Street Performer Protocol, and it involves 1) the work being completed and delivered to an escrow agent who will 2) release it to the public domain if the set cost is met.
I eagerly await hearing news of someone trying it for real.
While I will agree -- fair use doctrine does not create a right -- I disagree vehemently regarding the prevention of copying w/o contracts to that effect.
It is an abuse of copyright, it chills fair use, it chills the public domain requirement for copyrighted materials (i.e. that they must ultimately be copyable), and it violates the novel space shifting doctrine. (as well as in some cases, statutory exceptions, e.g. software backups)