This may look like a Good Thing to Joe Sixpack, who only cares about whether he's gonna get some good food on the table, but it may be a potential Bad Thing.
How will this affect regular, plants (non saltwater plants) and their use in the long run? Furthermore, if you remember your history, Rome was famous for salting Carthege's soil in the Punic Wars, thus ruining their land and making it impossible to farm on. Could our precious soil be damaged by mass use of salt water, especially if there is an irrigation error and it gets all over the place? I don't know if it's worth the environmental risks just so a few food corporations can make an extra buck off the unused land. The environment would be best left preserved and allowed to thrive instead of being worn out by man's overuse.
There is a difference between these two terms: a directory is a filesystem construction; while a folder is a user interface construction.
Most of the times, there is a one-to-one mapping between those concepts, but there are exceptions:
In Win95, "Control Panel", "Network Neighbourhood" etc. are folders but not directories
Mac OS has some directories that are only used internally (Network Trash Folder, Temporary Items,...): they are directories, but they are not folders
OS/2 WPS takes it even further: a folder is a subclass of WPFolder (which can be almost anything). Sometimes, the folder object and the associated directory doesn't even have the same name
MS GUI may beat X in some cases, this however isn't due to X running in user space. It is more about X running via TCP/IP, and design differences.
X only uses TCP/IP if the server and the client are not located on the same machine. If they are, X uses UNIX domain sockets (same as named pipes under NT). And using the shared memory extension, you can even work directly (GTK+ and QT use it AFAIK).
In NT 3.51, the GUI was located in user space and the was system dog-slow. In NT 4, they moved the GUI in kernel space for speed gains; at the expense of stability.
Erm, Apple Macintosh has symlinks (actually called aliases) since System 7, which even predates Windows95; and unlike UNIX symlinks they don't break when you move the target. And most Apple users are definitaly *not* geeks.
> Acutally, in Quake 3 all models are the same(unless you use mods or something).
Erm, Slash and Anarki don't make noise when running on most surfaces, which can be a *very big* advantage, even in single-player, and one of the Bones characters is very hard to notice because of its gray color.
It already exists: it's called LessTif (though without the CDE part) and implements both Motif 1.2 and Motif 2.1 (is that version acually in use?). Check www.lesstif.org.
"On my mac, Swing simply doesn't look (even with their Mac PLAF) nor feel like a mac application."
The explanation for this is quite simple: the apple implementation of Java sucks.
SWING for Mac is created by Sun, it has *nothing* to do with MrJ. SWING guidelines especially state that the menubar on mac systems should use the inferior implementation (menu bar attached to window, *belch*).
I'm a bit confused about this statement. I hope you don't mean that its impossible to recreate the same GUI on different platforms because that is exactly what swing does.
This is just the problem: Mac applications have to behave to the Apple UI guidelines; Windows applications have to behave to MS UI guidelines (even though Office and IE give a bad example) and Motif apps have to behave to the CUA guidelines.
Later on you mention that it is impossible to entirely incorporate the look and feel of one platform without violating the look and feel on other platforms. You're right, so what? The goal of pluggable look and feel is not to copy into every detail the look and feel of a specific platform but to be able to customize the applications look and feel without touching the application. A nice application of this is to immitate the look and feel of different platforms to provide users with a familiar looking GUI. A more interesting advantage is that it enables application developers to give a unique look to their applications.
Isn't this just a *disadvantage*? I want my applications to look consistent with the rest of my system, just like many others. Otherwise, you can call Windows 3.1 also cross-platform, as you can use WINE/WABI to run them under UNIX based systems and VirtualPC/SoftWindows to run them under Mac. They also are horribly out of sync with the rest of the system.
It's silly to claim that and nobody is doing that.
NeXTstep, and their implementations on Mac and Windows (WebObjects) exactly does that.
"NeXTstep and MacOS X Server provide an at least as powerful framework as Swing, but is much faster and doesn't have those inconsistencies."
Irrelevant because you need apple hardware to run it. Nice design but not crossplatform (in both our interpretations of this term).
I assume you don't know that in fact there exists an implementation that runs above Windows (called WebObjects). Apps then have exactly the look and feel of Windows applications, cause there the interface is clearly separated from the code, as opposed to SWING or the bunch of X based toolkits. And BTW, NeXTstep runs on Intel boxes.
When using a swing application, I simply have no a real different feel than running Virtual PC. It looks and feels like you're running an emulator (only a very slow one, even VirtualPC running Windoze is faster). On my mac, Swing simply doesn't look (even with their Mac PLAF) nor feel like a mac application.
The biggest problems with Swing are:
You cannot have a cross-platform GUI when you create the widgets in your code. The only possible solution is to completly separate the GUI from the application, so that you can plug in another interface definition depending on the platform (mac, windows, motif, gtk, kde,... all are very incompatible)
Sun's arrogance: they think they can simply wipe all existing user interfaces (which are much better and certainly much more researched) from the planet and impose their own one "Java L&F" which is an ugly copy of windoze 95.
Some appearant inconsistencies:
The shortcut modifier for menu items is command on mac, control on windows, alt on motif.
Menubars are on the top of the screen on mac, top of windows on windows and motif.
Windows are never layered on mac or motif.
The default button in dialogs are at the right side for mac (because you read left-to-right) and left for windows and motif.
"harmful" buttons on mac (like "Don't Save") are clearly separated from "harmless" buttons (like "Save" and "Cancel"); on windows and motif, they are packed together.
Buttons have meaningful names on mac ("Save", "Don't Save", "Print") while Windows, because of limitation of the MessageBox call, uses generic names ("Yes", "No", "OK").
Any user interface which ignores those differences cannot called "cross platform" without redefining the word (which you would more expect from Microsoft).
About the speed: NeXTstep and MacOS X Server provide an at least as powerful framework as Swing, but is much faster and doesn't have those inconsistencies.
Re:The Linux emulation is not perfect!
on
Which BSD?
·
· Score: 1
Using the ipcs and ipcrm commands, you can track down SYSV-IPC resources and kill them. Much better than a reboot;-)
It is indeed a pity that SYSV-IPC has no reference counting like files and the other IPC mechanisms (like pipes, locks and FIFO's)
This is because of a design bug in the MessageBox Windows API function, it only allows for "Ok", "Cancel", "Yes" and "No" buttons and only in some combinations. If you want to use comprehensive dialogs, you would have to write a dialog yourself (create the DLG template, the dialog callback function,...) which is enough work to scare away lazy programmers.
In MacOS (and NeXTstep), it's a HI requirement that buttons have meaningfull names (like: 'Save', 'Don't Save' and 'Cancel'). Also, "harmful" actions (like 'Don't Save') have to be clearly separated from "harmless" actions ('Save' and 'Cancel'). NeXT takes it a bit to the extreme where sometimes very long titles are used.
Currently, I've yet to see the first fully cross-platform toolkit. The problem with the current toolkits is that they emulate one environment on the other, which give the feel you're using an emulator. Some important differences:
MacOS and NeXTstep use a separate menu bar, which means you can have applications without windows; X based toolkits attach the menu bar to every window and Windows uses a strange hack called MDI (a window in a window).
On MacOS/NeXTstep, usefull names are used for buttons while on Windows, it's limited to 'Ok', 'Cancel', 'Yes' and 'No'. This is especially visible in message boxes.
In MacOS/NeXTstep dialogs, the default button is the rightmost one, and 'harmfull' buttons are clearly separated from 'harmless' buttons; in Windows, those buttons are put together and the default one is usually the leftmost one.
For binary menu items, MacOS uses a change in the name (like Show Ruler/Hide Ruler) while Windows typically uses a check mark
Windows interfaces heavily make use of toolbars, MacOS doesn't.
Different vocabulary (e.g. Exit vs Quit)
Different menu layout (e.g. the position of the About item)
...
If you add the OpenLook interface, it gets even worse ('Close' in OpenLook means 'Minimize' in Windows; 'Quit' in OpenLook means 'Close' in Windows...)
On Apple, it easy because there are better standard interfaces (like e.g. SCSIManager). You can easily put a third party SCSI card in your box, attach a CD-ROM drive to it and boot from that drive. It's simply the PC BIOS that is so primitive. If a new and better global specification for BIOSes was created (which hopefully also supported protected mode), we wouldn't have trouble with all those silly device drivers (only *one* generic SCSI driver would be needed).
The problem here is not Win98 but the fact that PC's are such chaos. The BIOS doesn't have serious SCSI support (compare this to the SCSIManager on Mac where you don't need drivers for other SCSI cards) nor even basic CD-ROM support. The reason that PC's are such nightmare to set up is that there are no standard interfaces for most components (only keyboard, up to two hard disks, floppy and very basic video support are standard, all the rest is up to the whims of the card manufacturers).
> 2) its still sun that choose whats going in into os
It's also Linus who chooses what's going into the Linux kernel (just look at KGI) so what's your point? Everyone can also create kernel modules for Solaris, or how do you think that drivers for various add-on cards are created?
> It turns out that as shipped, Mac OS X can't even do filesharing with non-Macintosh machines.
(I assume you mean Mac OS X Server as Mac OS X (client) hasn't shipped yet)
Really? The Mac OS X Server I run at home does have NFS (client *and* server) out of the box, just like e.g. Solaris. Please check your facts before posting false rumours.
This is not true, while it *seems* you only copy files, the Finder automatically writes a boot signature when it sees that you created a valid system folder (it also changes the icon of that folder so that it contains a little mac classic (pre 8.0) or the MacOS logo (8.0+). It would be like if you copy io.sys, msdos.sys and command.com to a floppy, dos/windows would automatically fill in the boot sector.
WindowMaker supports this feature as well: if I right-click in the minimize button of one of my NetScape windows, every NetScape window hides under the NetScape application icon.
GNUstep has the same feature, and is window manager independent. Select "Hide" from the application menu and all windows of that application hide in their application icon.
MacOS, Amiga and NeXTstep/OpenStep/GNUstep/MacOSX don't need layered windows, as MDI is just a dirty work-around for the misplacement of the menu-bar. M$ introduced MDI in Windows 2.0 when they realized they needed a shared menu for multiple document windows and need to be able to display a menu when there are no document windows.
Argh, I don't know how people can actually like MDI, it's just like running MacOS where every other application is always hidden. *sigh*
Oh and btw, in the design specifications of Windows95, M$ wanted to go full SDI (just look at the Windows Explorer) and even stated that in the documentation for Visual C++ 4.
Very easy, especially if you have already used the Slackware distribution (SW uses a more BSD-like filesystem layout than RedHat, SuSE or Debian which use a more SysV like layout). I figured out how to use and configure FreeBSD in a matter of minutes. The only difficulty I find is the configuration of the kernel, here is Linux much easier than FreeBSD (this could be better if the FreeBSD kernel config file were better documented). I personally prefer FreeBSD for a server as it has the concept of immutable and append-only files (immutable files can only be changed while in single-user mode) while I prefer Linux (RedHat) or MacOS X Server for workstations.
It should be possible to create such format. It would be like the "fat" executables on MacOS which can contain the M86K and PPC versions in one file. One file containing the x86, Sparc, MIPS, PPC, m86k... versions:-)
Why hasn't Sun done this?
on
Java for EGCS
·
· Score: 1
I can use dynamic loading of classes in an Objective-C application that is as native as possible (the dynamic classes reside in a shared library, the main classes in the app). Why won't that be able then in a compiled Java app?
Slashdot Mirror
This may look like a Good Thing to Joe Sixpack, who only cares about whether he's gonna get some good food on the table, but it may be a potential Bad Thing.
How will this affect regular, plants (non saltwater plants) and their use in the long run? Furthermore, if you remember your history, Rome was famous for salting Carthege's soil in the Punic Wars, thus ruining their land and making it impossible to farm on. Could our precious soil be damaged by mass use of salt water, especially if there is an irrigation error and it gets all over the place? I don't know if it's worth the environmental risks just so a few food corporations can make an extra buck off the unused land. The environment would be best left preserved and allowed to thrive instead of being worn out by man's overuse.
Most of the times, there is a one-to-one mapping between those concepts, but there are exceptions:
Erm, Apple Macintosh has symlinks (actually called aliases) since System 7, which even predates Windows95; and unlike UNIX symlinks they don't break when you move the target. And most Apple users are definitaly *not* geeks.
> Acutally, in Quake 3 all models are the same(unless you use mods or something).
Erm, Slash and Anarki don't make noise when running on most surfaces, which can be a *very big* advantage, even in single-player, and one of the Bones characters is very hard to notice because of its gray color.
They are still for sale at Black Hole Inc.
It already exists: it's called LessTif (though without the CDE part) and implements both Motif 1.2 and Motif 2.1 (is that version acually in use?). Check www.lesstif.org.
I think it's more a hardware/BIOS problem than a problem with OpenBSD, as even under windoze 95 the same problem happens.
The explanation for this is quite simple: the apple implementation of Java sucks.
SWING for Mac is created by Sun, it has *nothing* to do with MrJ. SWING guidelines especially state that the menubar on mac systems should use the inferior implementation (menu bar attached to window, *belch*).
I'm a bit confused about this statement. I hope you don't mean that its impossible to recreate the same GUI on different platforms because that is exactly what swing does.
This is just the problem: Mac applications have to behave to the Apple UI guidelines; Windows applications have to behave to MS UI guidelines (even though Office and IE give a bad example) and Motif apps have to behave to the CUA guidelines.
Later on you mention that it is impossible to entirely incorporate the look and feel of one platform without violating the look and feel on other platforms. You're right, so what? The goal of pluggable look and feel is not to copy into every detail the look and feel of a specific platform but to be able to customize the applications look and feel without touching the application. A nice application of this is to immitate the look and feel of different platforms to provide users with a familiar looking GUI. A more interesting advantage is that it enables application developers to give a unique look to their applications.
Isn't this just a *disadvantage*? I want my applications to look consistent with the rest of my system, just like many others. Otherwise, you can call Windows 3.1 also cross-platform, as you can use WINE/WABI to run them under UNIX based systems and VirtualPC/SoftWindows to run them under Mac. They also are horribly out of sync with the rest of the system.
It's silly to claim that and nobody is doing that.
NeXTstep, and their implementations on Mac and Windows (WebObjects) exactly does that.
"NeXTstep and MacOS X Server provide an at least as powerful framework as Swing, but is much faster and doesn't have those inconsistencies."
Irrelevant because you need apple hardware to run it. Nice design but not crossplatform (in both our interpretations of this term).
I assume you don't know that in fact there exists an implementation that runs above Windows (called WebObjects). Apps then have exactly the look and feel of Windows applications, cause there the interface is clearly separated from the code, as opposed to SWING or the bunch of X based toolkits. And BTW, NeXTstep runs on Intel boxes.
The biggest problems with Swing are:
Some appearant inconsistencies:
Any user interface which ignores those differences cannot called "cross platform" without redefining the word (which you would more expect from Microsoft).
About the speed: NeXTstep and MacOS X Server provide an at least as powerful framework as Swing, but is much faster and doesn't have those inconsistencies.
It is indeed a pity that SYSV-IPC has no reference counting like files and the other IPC mechanisms (like pipes, locks and FIFO's)
This is because of a design bug in the MessageBox Windows API function, it only allows for "Ok", "Cancel", "Yes" and "No" buttons and only in some combinations. If you want to use comprehensive dialogs, you would have to write a dialog yourself (create the DLG template, the dialog callback function,...) which is enough work to scare away lazy programmers.
In MacOS (and NeXTstep), it's a HI requirement that buttons have meaningfull names (like: 'Save', 'Don't Save' and 'Cancel'). Also, "harmful" actions (like 'Don't Save') have to be clearly separated from "harmless" actions ('Save' and 'Cancel'). NeXT takes it a bit to the extreme where sometimes very long titles are used.
- MacOS and NeXTstep use a separate menu bar, which means you can have applications without windows; X based toolkits attach the menu bar to every window and Windows uses a strange hack called MDI (a window in a window).
- On MacOS/NeXTstep, usefull names are used for buttons while on Windows, it's limited to 'Ok', 'Cancel', 'Yes' and 'No'. This is especially visible in message boxes.
- In MacOS/NeXTstep dialogs, the default button is the rightmost one, and 'harmfull' buttons are clearly separated from 'harmless' buttons; in Windows, those buttons are put together and the default one is usually the leftmost one.
- For binary menu items, MacOS uses a change in the name (like Show Ruler/Hide Ruler) while Windows typically uses a check mark
- Windows interfaces heavily make use of toolbars, MacOS doesn't.
- Different vocabulary (e.g. Exit vs Quit)
- Different menu layout (e.g. the position of the About item)
- ...
If you add the OpenLook interface, it gets even worse ('Close' in OpenLook means 'Minimize' in Windows; 'Quit' in OpenLook means 'Close' in Windows...)Erm...
/proc
/proc, doesn't it?
$ uname -a
SunOS jefke 5.7 Generic_106541-07 sun4u sparc SUNW,Ultra-5_10
$ ls
0 181 22952 23357 23367 23383 23723 25323 275 302 490
1 198 22953 23358 23368 23384 23725 25939 277 304 493
114 2 22954 23359 23369 23388 238 25941 2786 306 499
116 203 22955 23360 23370 23389 24678 26077 2787 308 500
126 214 22956 23361 23371 23394 24682 26082 279 321 505
151 219 22965 23362 23372 23396 24842 26161 281 3600 508
153 228 22966 23363 23373 234 25285 26162 283 3601 509
154 22949 22967 23364 23374 23419 25287 269 298 3603 630
167 22950 22974 23365 23378 23422 25311 271 3 394
179 22951 23356 23366 23379 23468 25313 273 300 406
$
So, it does have a
On Apple, it easy because there are better standard interfaces (like e.g. SCSIManager). You can easily put a third party SCSI card in your box, attach a CD-ROM drive to it and boot from that drive. It's simply the PC BIOS that is so primitive. If a new and better global specification for BIOSes was created (which hopefully also supported protected mode), we wouldn't have trouble with all those silly device drivers (only *one* generic SCSI driver would be needed).
The problem here is not Win98 but the fact that PC's are such chaos. The BIOS doesn't have serious SCSI support (compare this to the SCSIManager on Mac where you don't need drivers for other SCSI cards) nor even basic CD-ROM support. The reason that PC's are such nightmare to set up is that there are no standard interfaces for most components (only keyboard, up to two hard disks, floppy and very basic video support are standard, all the rest is up to the whims of the card manufacturers).
> 2) its still sun that choose whats going in into os
It's also Linus who chooses what's going into the Linux kernel (just look at KGI) so what's your point? Everyone can also create kernel modules for Solaris, or how do you think that drivers for various add-on cards are created?
> It turns out that as shipped, Mac OS X can't even do filesharing with non-Macintosh machines.
(I assume you mean Mac OS X Server as Mac OS X (client) hasn't shipped yet)
Really? The Mac OS X Server I run at home does have NFS (client *and* server) out of the box, just like e.g. Solaris. Please check your facts before posting false rumours.
This is not true, while it *seems* you only copy files, the Finder automatically writes a boot signature when it sees that you created a valid system folder (it also changes the icon of that folder so that it contains a little mac classic (pre 8.0) or the MacOS logo (8.0+). It would be like if you copy io.sys, msdos.sys and command.com to a floppy, dos/windows would automatically fill in the boot sector.
GNUstep has the same feature, and is window manager independent. Select "Hide" from the application menu and all windows of that application hide in their application icon.
Argh, I don't know how people can actually like MDI, it's just like running MacOS where every other application is always hidden. *sigh*
Oh and btw, in the design specifications of Windows95, M$ wanted to go full SDI (just look at the Windows Explorer) and even stated that in the documentation for Visual C++ 4.
Very easy, especially if you have already used the Slackware distribution (SW uses a more BSD-like filesystem layout than RedHat, SuSE or Debian which use a more SysV like layout). I figured out how to use and configure FreeBSD in a matter of minutes. The only difficulty I find is the configuration of the kernel, here is Linux much easier than FreeBSD (this could be better if the FreeBSD kernel config file were better documented). I personally prefer FreeBSD for a server as it has the concept of immutable and append-only files (immutable files can only be changed while in single-user mode) while I prefer Linux (RedHat) or MacOS X Server for workstations.
AFAIK the reason for the delay of LinuxPPC is that it is still beta for the b&w PowerMac G3 (It does work however).
It should be possible to create such format. It would be like the "fat" executables on MacOS which can contain the M86K and PPC versions in one file. One file containing the x86, Sparc, MIPS, PPC, m86k... versions :-)
I can use dynamic loading of classes in an Objective-C application that is as native as possible (the dynamic classes reside in a shared library, the main classes in the app). Why won't that be able then in a compiled Java app?