Hmm, I read that as 'intelligent hamster controlled by MIDI sequencer'... I was wondering, what self-respecting intelligent being will let himself be controlled by a machine-program?
Or, the RAD for OS X, Cocoa on Objective-C. Add garbage collection to that, and it will probably be competitive with VB.
The only problem is that there is effectively only one Objective-C compiler out there (GCC) and only one open Cocoa implementation (GNUstep), and GNUstep has quite some distance to bridge before arriving at OS X 10.3 Cocoa.
Windows binaries are relocatable, if properly designed; otherwise, they still need to have files scattered all over the system.
As for figuring out the absolute path? There are ways of doing this -- various getwd(), getcwd() etc. and realpath(const char *). Unfortunately, none of them figure out the absolute path of an executable... because that information does not seem to be (readily) available. argv[0] is not enough!
Also, libraries are searched for at runtime, not link-time -- you could put the needed libraries anywhere, so long as the path to it is specified in LD_LIBRARY_PATH. (But, specifying '.' in any sort of PATH is dangerous!)
Microsoft(R) Windows(R) -- not Microsoft(TM) Windows(TM)
Re:How does one dispute math as a universal concep
on
The Golden Ratio
·
· Score: 1
One is still one... However, if they do not understand the concept of finite objects, then there might be a problem.
However, since they still percieve objects as objects, they might choose to count timelines rather than an object-time coordinate. Speaking of which, they still need to have a coordinate system to identify a position in space-time...
If there is a sentience that percieves absolutely everything... Hmm, it is possible that it does not understand the concept of finite numbers.
You have yet to encounter extreme-RISC instruction sets, I take it then.
There is the subtract-and-branch-if-negative variant, that is probably much more painful than the x86.
For systems programming, x86 is horrible. But for applications programming, it is OK. Application programmers don't have to fret over the insane layout of a GDT entry, or the virtual-physical address mode, or the A20 line, or the privillege ring, etc. Then again, with only 8 general purpose registers, it could be considered more difficult.
Still, I believe a load-store (RISC-style) system architecture with a large register file (at least 32, of each type, or 64 if registers are orthogonal in use) is easier to program in.
And yet, for the same reason there are some things only megacorporations can do, there are some things only governments can do.
Developing a space program is very expensive. Research is expensive. Actually building it is expensive, in material cost and human cost.
Privatising it will encourage the reduction in cost, but the entry barrier is still there. It's a causal loop really, only those which have lots of resources can run a space program, and they don't have a lot of incentive to reduce costs. (Other than a rapidly dwindling budget (or a threat thereof))
So, you're saying that HTTP authentication is stupid?
http://username:password@www.domain.net/path/fil e is just a way of storing the authentication data in the URL. It would be translated to a HTTP request like
Hmm, what is the origin of this particular piece of flamebait? Seems quite old. Apple System Software and Multifinder died quite some time ago. MacOS cooperative multitasking and no-CLI died with OS 9.
Besides, we all know we can't run EMACS on Amiga... yet.
When dragging removable (data storage) media to trash, the trash icon magically changes itself to an eject icon.
Not as consistent as it could be.
But most of the time, the eject button on the keyboard can be used to eject a CD. The only time you need to manually eject media is if it is not a CD, or if the keyboard is not working.
Perhaps, because they have been sent to Earth, and keeping them on the Rover would be redundant. Thus, they are the perfect candidates for removal.
The problem was apparently inode exhaustion. You can experience this for yourself by running this little shell script:
i=0
while true
do
touch $i
i=$(($i + 1))
done
The Magic SysRq key makes a Request to the Linux System -- Alt + SysRq + (on x86)
For more details, study/usr/src/linux/Documentation/sysrq.txt
Very useful. Far more convenient than pressing some chord that brings up a GUI screen with buttons to press. The framebuffer may be corrupted, for instance, or the mouse non-responsive, or the GUI server frozen, etc.
Slashdot Mirror
emerge -C kde does not remove KDE.
emerge -C kde removes the KDE metapackage, which is a no-content package.
emerge -C kde does not remove the dependencies of the KDE metapackage.
On top of all that, emerge kde is not neccessary to enjoy KDE.
Any Apache authentication module will do... Also, because it isn't filesystem-backed, there is no need to have one system account per SVN account.
SVN supports copy-on-write.
But symbolic links and aliasing was not available last I checked.
Hmm, I read that as 'intelligent hamster controlled by MIDI sequencer'... I was wondering, what self-respecting intelligent being will let himself be controlled by a machine-program?
That's why we have ANSI trigraphs...
But... I don't think they bothered making sure C was programmable in only morse characters.
Mass isn't really destroyed in this case anyway... I don't really see many signs of mass-energy conversion happening.
Might want to be careful with that.
If 'yourbuffer' is in fact a pointer... sizeof(yourbuffer) returns 4.
Most of the time, buffers are statically allocated on the stack, so sizeof(yourbuffer) works as expected.
Just too bad that stack buffers are the easiest to exploit...
Or, the RAD for OS X, Cocoa on Objective-C. Add garbage collection to that, and it will probably be competitive with VB.
The only problem is that there is effectively only one Objective-C compiler out there (GCC) and only one open Cocoa implementation (GNUstep), and GNUstep has quite some distance to bridge before arriving at OS X 10.3 Cocoa.
Windows binaries are relocatable, if properly designed; otherwise, they still need to have files scattered all over the system.
As for figuring out the absolute path? There are ways of doing this -- various getwd(), getcwd() etc. and realpath(const char *). Unfortunately, none of them figure out the absolute path of an executable... because that information does not seem to be (readily) available. argv[0] is not enough!
Also, libraries are searched for at runtime, not link-time -- you could put the needed libraries anywhere, so long as the path to it is specified in LD_LIBRARY_PATH. (But, specifying '.' in any sort of PATH is dangerous!)
eclasses, like ebuilds, are written in bash.
Registered trademarks are marked with (R) not TM.
Microsoft(R) Windows(R) -- not Microsoft(TM) Windows(TM)
One is still one... However, if they do not understand the concept of finite objects, then there might be a problem.
However, since they still percieve objects as objects, they might choose to count timelines rather than an object-time coordinate. Speaking of which, they still need to have a coordinate system to identify a position in space-time...
If there is a sentience that percieves absolutely everything... Hmm, it is possible that it does not understand the concept of finite numbers.
You have yet to encounter extreme-RISC instruction sets, I take it then.
There is the subtract-and-branch-if-negative variant, that is probably much more painful than the x86.
For systems programming, x86 is horrible. But for applications programming, it is OK. Application programmers don't have to fret over the insane layout of a GDT entry, or the virtual-physical address mode, or the A20 line, or the privillege ring, etc. Then again, with only 8 general purpose registers, it could be considered more difficult.
Still, I believe a load-store (RISC-style) system architecture with a large register file (at least 32, of each type, or 64 if registers are orthogonal in use) is easier to program in.
MIX is used in his published (as of to date) TAOCP book.
MMIX will replace it.
Neither are stack machines. MIX predates 8-bit bytes, to give you an idea; MMIX is a load-store (i.e. RISC-style) machine.
Both don't have a hardware implementation (yet).
Curiously, why this time?
b rot_set&oldid=354283
Google frequently has special logos linked to queries. At least on the order of once a month.
But I suppose this is the first time it was linked to an image search, on fascinating fractals, no less.
For people looking for a laugh: http://en.wikipedia.org/w/wiki.phtml?title=Mandel
And yet, for the same reason there are some things only megacorporations can do, there are some things only governments can do.
Developing a space program is very expensive. Research is expensive. Actually building it is expensive, in material cost and human cost.
Privatising it will encourage the reduction in cost, but the entry barrier is still there. It's a causal loop really, only those which have lots of resources can run a space program, and they don't have a lot of incentive to reduce costs. (Other than a rapidly dwindling budget (or a threat thereof))
"Comic Sans MS"
Comic -- humour.
Sans -- without.
MS -- Microsoft.
"Humour without Microsoft"
Impossible.
I suppose that the Top Secret documents are printed on red paper, to maximise secrecy.
So, you're saying that HTTP authentication is stupid?
l e is just a way of storing the authentication data in the URL. It would be translated to a HTTP request like
/path/file
http://username:password@www.domain.net/path/fi
GET
Authorization: Basic QdXNlcm5hbWU6cGFzc3dvcmQ=
Host: www.domain.net
If you didn't supply the authentication tokens in the URL, you would have been prompted for them, and it would be the same request.
Remember: not every HTTP user agent is an interactive browser with session cookies. HTTP authentication needs neither.
Hmm, what is the origin of this particular piece of flamebait? Seems quite old. Apple System Software and Multifinder died quite some time ago. MacOS cooperative multitasking and no-CLI died with OS 9.
Besides, we all know we can't run EMACS on Amiga... yet.
Better contribution auditing. That is the only thing we can do.
Be cautious of code you recieve. Ask for papers to show copyright ownership. Ask for papers to confirm licensing. etc.
Either that, or don't allow contributed code. Which is about as bad as Shared Source.
When dragging removable (data storage) media to trash, the trash icon magically changes itself to an eject icon.
Not as consistent as it could be.
But most of the time, the eject button on the keyboard can be used to eject a CD. The only time you need to manually eject media is if it is not a CD, or if the keyboard is not working.
Perhaps, because they have been sent to Earth, and keeping them on the Rover would be redundant. Thus, they are the perfect candidates for removal. The problem was apparently inode exhaustion. You can experience this for yourself by running this little shell script: i=0 while true do touch $i i=$(($i + 1)) done
The POWER chips initially required that all instructions took 1 cycle to execute. (They didn't have floating point, as such)
I don't know if this attribute carried down to the early PowerPCs, but I know the newer ones certainly don't.
The Magic SysRq key makes a Request to the Linux System -- Alt + SysRq + (on x86)
/usr/src/linux/Documentation/sysrq.txt
For more details, study
Very useful. Far more convenient than pressing some chord that brings up a GUI screen with buttons to press. The framebuffer may be corrupted, for instance, or the mouse non-responsive, or the GUI server frozen, etc.