Anybody who says the WPS was awesome never tried to write code for it. SOM was like somebody took every bad idea in CORBA, put them in a blender with Stroustrop's evil twin, and then burned any documentation that was accidentally created. It was over a decade ago and I still bear the scars.
Kool-Aid is a cheap flavoring powder added to water to make an almost drinkable concoction. The expression "drink the Kool-Aid" is an reference to the cult leader Jim Jones, who convinced nearly a thousand people to kill themselved by drinking poisoned Flavor-Aid (an even cheaper knockoff of Kool-Aid). It's a warning not to follow a charismatic leader / authority figure blindly, though bizarrely enough management types in the US frequently use it in a boasting sense to indicate they convinced people to follow their plan, as in "that customer totally drank our Kool-Aid, now we can rip them off endlessly".
You'll find that your two examples will compile to the same bytecode in any modern Java compiler.
Do you count the Sun's current JDK release as a "modern compiler"? If so, you may find the following interesting:
(kvigor@mojo tmp)$ cat Component.java public class Component {
private Point location;
public Point getLocation() { return new Point(location); }
public double getDistanceFrom(Component other) {
Point otherLocation = other.getLocation();
int deltaX = otherLocation.getX() - location.getX();
int deltaY = otherLocation.getY() - location.getY();
return Math.sqrt(deltaX*deltaX + deltaY*deltaY);
}
public double getDistanceFromOptimized(Component other) {
Point otherLocation = other.location;
int deltaX = otherLocation.getX() - location.getX();
int deltaY = otherLocation.getY() - location.getY();
return Math.sqrt(deltaX*deltaX + deltaY*deltaY);
} } (kvigor@mojo tmp)$ ~/jdk/jdk1.5.0_05/bin/javac Component.java (kvigor@mojo tmp)$ ~/jdk/jdk1.5.0_05/bin/javap -c Component Compiled from "Component.java" public class Component extends java.lang.Object{ public Component();
Code:
0: aload_0// comment added to pad average line length because slashcode is teh SUXXOR
1: invokespecial #1;//Method java/lang/Object."":()V
4: return// comment added to pad average line length because slashcode is teh SUXXOR
public Point getLocation();
Code:
0: new #2;//class Point
3: dup// comment added to pad average line length because slashcode is teh SUXXOR
4: aload_0// comment added to pad average line length because slashcode is teh SUXXOR
5: getfield #3;//Field location:LPoint;
8: invokespecial #4;//Method Point."":(LPoint;)V
11: areturn
While it was jolly considerate of you to cut-n-paste your brochure for us, Mr. FingerGear, I do wonder what made you think that was appropriate in this context. It seems the FingerGear is in fact a simple USB thumb drive with a Linux distro preinstalled, which is rather a long way from the true standalone system the BlackDog is. For instance, I'd be interested in the CPU the device uses; the web site is strangely silent on the issue.
It certainly can use autorun on Windows, which is very convenient and is a primary reason for using a CD-ROM interface. But the CD image is a.ISO file served off the device's flash, which means it can contain anything. As we ship it, it has an AUTORUN.INF and an X server and some handy networking tools. But it's trivial to change the CD image to hold whatever you may want.
This is an incredibly flexible device. You can change almost anything about it. Heck, if you're up to hacking kernel drivers, you can make it emulate ANY USB device. Thus far, we've had call for emulating CD-ROM and ethernet-over-USB. But there's no reason it can't appear to the host as a mass storage device. Or a keyboard (take *that*, keyloggers!). Or a serial port dongle. Or... anything USB at all.
None of this flexibility, alas, will make any Linux distro in the world dumb enough to autorun software just because somebody presented a CD-ROM, unlike another certain popular OS. So autorunning on a Linux host requires a hotplug script be installed. We provide samples for some major distros and assume that the guy running Gentoo is smart enough to hack his own hotplug (or just run a.SH off the CD-ROM image).
OS-X support is "coming soon", i.e. we plan to support it, have done nothing to prevent it working, but haven't yet spent any test cycles on it. Having got it working with the incredibly balky USB stacks in Windows and Linux, I have no doubt it will be made to work just fine.
(again, *not* an official spokesman for the company, not legally binding, don't sue me please, yadda yadda.)
Ethernet would have made this an instant sale for me. I could develop and demo network appliance products on the road without virtual machines. Without Ethernet, this is just a really portable terminal to me.
Despite the summary claiming the device appears as a CD-ROM, it also appears to the host as a network device (masquerading as a ethernet-over-USB dongle). Sounds to me like it might be downright handy for your application.
The device is seen as *both* a CD-ROM and a network device. As the parent suggested, the CD-ROM is used to launch an X server on the host [1] and the network interface is used to allow that X server to communicate to the applications running on the device.
(disclaimer: I am an employee of Realm Systems but do not speak for the company)
[1] obviously useful only for Windows hosts, since Linux hosts will almost always have an X server running already.
That would also be wrong. strncpy does *not* null-terminate the buffer if it would overflow, so strlen(tempbuf) is not defined in that case. You want something like:
The stack grows downward on the vast majority of architectures, including PowerPC, ARM, SPARC, and MIPS. The only architectures I've come across where stacks grow up are PA-RISC and i960.
So if you cponsider this is a sign of CPU inferiority, well, I hope you're an HPUX fanboy.
Not just Cisco. I formerly worked for a certain very large US telecom equipment manufacturer, and Huawei was selling a unit that was bug-identical to one of ours. The story, as related to me, was that said large company attempted to sue in China (they weren't selling into the US at the time). Said lawsuit came to a screeching halt when it became apparent that the Chinese government was the primary investor and the eventual result of the lawsuit was therefore completely predictable.
I am surprised to see that Cisco settled with them in US court. I expect the company in question, which has phalanxes of lawyers on salary, won't roll over so easily when it comes to defending the domestic market.
Your world of text programming languages is passing, and will one day be a mere footnote.
Yes! Death to text! If you need proof that communication can be more efficiently done in a visual medium, you need look no further than this very web site where we converse exclusively using...
As it happens, Jan Axelson has written this exact book, USB Complete. I am in the process of implementing a USB device and I give it a hearty recommend. The "standard" work is USB Design By Example by John Hyde; it's nice to have, but if you have to pick one, the Axelson is both more readable and more complete.
I've had satellite service for about five years. I get one or two outages a year due to heavy snowfall. These are easily resolved with the high-tech dish-snow-removal tool (the Mk.I broom duct-taped to a pipe). No other weather-related outages ever.
Judging from neighbourhood heresay, this is orders of magnitude more reliable than Comcast's cable service (I'm in Salt Lake City).
And Jacques Anquetil took off three years to fight, then came back to win more Tours. It's widely regarded that had not the World War so rudely interrupted, Anquetil would have been the all-time leader by far with eight Tour wins.
Really? The same Jaques Anquetil who didn't start racing until 1953?
Actually, there are (at least) three. H.323 is the original ITU spec. It is shockingly complex and scales poorly, since it is incredibly chatty. It is, however, feature rich and is probably the most widely implemented standard, though vendor interoperability is spotty at best, due to the complexity of the protocol.
SIP is a reaction to H.323. It has damn near zero features, but is efficient and easy to implement. This is number two with a bullet, and rising fast, despite its serious limitations.
H.248 is the IETF's attempt to fix H.323. I don't know much about it except that it's allegedly much simpler than H.323 while being much more featureful than SIP. It's the cool new buzzword; it remains to be seen if anybody actually implements/uses it.
My understanding was that the signature applied to a particular binary. Thus, a signature would be good only for whatever kernel revision the XBox linux guys submitted for approval. This would rather miss the point of Linux, wouldn't it? All bugs are shallow, but none of them can be fixed without asking Microsoft for approval?
Amazing! The Japanese promo literature is all correctly spelled, while the Slashdot story has at least three egregious errors. Unless "smurf", "shinny", and "Weib" are meant as some kind of sophisticated humor.
Slashdot Mirror
Anybody who says the WPS was awesome never tried to write code for it. SOM was like somebody took every bad idea in CORBA, put them in a blender with Stroustrop's evil twin, and then burned any documentation that was accidentally created. It was over a decade ago and I still bear the scars.
Kool-Aid is a cheap flavoring powder added to water to make an almost drinkable concoction.
The expression "drink the Kool-Aid" is an reference to the cult leader Jim Jones, who convinced nearly a thousand people to kill themselved by drinking poisoned Flavor-Aid (an even cheaper knockoff of Kool-Aid). It's a warning not to follow a charismatic leader / authority figure blindly, though bizarrely enough management types in the US frequently use it in a boasting sense to indicate they convinced people to follow their plan, as in "that customer totally drank our Kool-Aid, now we can rip them off endlessly".
Do you count the Sun's current JDK release as a "modern compiler"? If so, you may find the following interesting:
While it was jolly considerate of you to cut-n-paste your brochure for us, Mr. FingerGear, I do wonder what made you think that was appropriate in this context. It seems the FingerGear is in fact a simple USB thumb drive with a Linux distro preinstalled, which is rather a long way from the true standalone system the BlackDog is. For instance, I'd be interested in the CPU the device uses; the web site is strangely silent on the issue.
It certainly can use autorun on Windows, which is very convenient and is a primary reason for using a CD-ROM interface. But the CD image is a .ISO file served off the device's flash, which means it can contain anything. As we ship it, it has an AUTORUN.INF and an X server and some handy networking tools. But it's trivial to change the CD image to hold whatever you may want.
.SH off the CD-ROM image).
This is an incredibly flexible device. You can change almost anything about it. Heck, if you're up to hacking kernel drivers, you can make it emulate ANY USB device. Thus far, we've had call for emulating CD-ROM and ethernet-over-USB. But there's no reason it can't appear to the host as a mass storage device. Or a keyboard (take *that*, keyloggers!). Or a serial port dongle. Or... anything USB at all.
None of this flexibility, alas, will make any Linux distro in the world dumb enough to autorun software just because somebody presented a CD-ROM, unlike another certain popular OS. So autorunning on a Linux host requires a hotplug script be installed. We provide samples for some major distros and assume that the guy running Gentoo is smart enough to hack his own hotplug (or just run a
OS-X support is "coming soon", i.e. we plan to support it, have done nothing to prevent it working, but haven't yet spent any test cycles on it. Having got it working with the incredibly balky USB stacks in Windows and Linux, I have no doubt it will be made to work just fine.
(again, *not* an official spokesman for the company, not legally binding, don't sue me please, yadda yadda.)
Despite the summary claiming the device appears as a CD-ROM, it also appears to the host as a network device (masquerading as a ethernet-over-USB dongle). Sounds to me like it might be downright handy for your application.
The device currently consumes ~300 milliwatts max and is barely warm to the touch. Naturally it consumes far less when idle.
The device is seen as *both* a CD-ROM and a network device. As the parent suggested, the CD-ROM is used to launch an X server on the host [1] and the network interface is used to allow that X server to communicate to the applications running on the device.
(disclaimer: I am an employee of Realm Systems but do not speak for the company)
[1] obviously useful only for Windows hosts, since Linux hosts will almost always have an X server running already.
The MMUs on various PowerPCs vary; I know that there definitely is an executable bit in the page mapping table for the 40x family.
The stack grows downward on the vast majority of architectures, including PowerPC, ARM, SPARC, and MIPS. The only architectures I've come across where stacks grow up are PA-RISC and i960.
So if you cponsider this is a sign of CPU inferiority, well, I hope you're an HPUX fanboy.
Not just Cisco. I formerly worked for a certain very large US telecom equipment manufacturer, and Huawei was selling a unit that was bug-identical to one of ours. The story, as related to me, was that said large company attempted to sue in China (they weren't selling into the US at the time). Said lawsuit came to a screeching halt when it became apparent that the Chinese government was the primary investor and the eventual result of the lawsuit was therefore completely predictable.
I am surprised to see that Cisco settled with them in US court. I expect the company in question, which has phalanxes of lawyers on salary, won't roll over so easily when it comes to defending the domestic market.
<objection tone="disgusted">
<body>xml is too sodding verbose for any use ever anywhere. Satan himself recoils before its horror.
</body>
</objection>
Yes! Death to text! If you need proof that communication can be more efficiently done in a visual medium, you need look no further than this very web site where we converse exclusively using...
Oh. Never mind. Move along, nothing to see here.
As it happens, Jan Axelson has written this exact book, USB Complete. I am in the process of implementing a USB device and I give it a hearty recommend. The "standard" work is USB Design By Example by John Hyde; it's nice to have, but if you have to pick one, the Axelson is both more readable and more complete.
I've had satellite service for about five years. I get one or two outages a year due to heavy snowfall. These are easily resolved with the high-tech dish-snow-removal tool (the Mk.I broom duct-taped to a pipe). No other weather-related outages ever.
Judging from neighbourhood heresay, this is orders of magnitude more reliable than Comcast's cable service (I'm in Salt Lake City).
You mean like Cateye's GameBike?
Really? The same Jaques Anquetil who didn't start racing until 1953?
The devil you say! Those darn whiny pilots and their "control" and their "not dying in screaming terror because their controls have beem r00t3d"!
Actually, there are (at least) three. H.323 is the original ITU spec. It is shockingly complex and scales poorly, since it is incredibly chatty. It is, however, feature rich and is probably the most widely implemented standard, though vendor interoperability is spotty at best, due to the complexity of the protocol.
SIP is a reaction to H.323. It has damn near zero features, but is efficient and easy to implement. This is number two with a bullet, and rising fast, despite its serious limitations.
H.248 is the IETF's attempt to fix H.323. I don't know much about it except that it's allegedly much simpler than H.323 while being much more featureful than SIP. It's the cool new buzzword; it remains to be seen if anybody actually implements/uses it.
My understanding was that the signature applied to a particular binary. Thus, a signature would be good only for whatever kernel revision the XBox linux guys submitted for approval. This would rather miss the point of Linux, wouldn't it? All bugs are shallow, but none of them can be fixed without asking Microsoft for approval?
Don't be so sure. Ever hear of Alternative Mimimum Tax? It is quite possible to be taxed at considerably more than 100% of your income.
Just because one is a dropout doesn't necessarily make one ignorant: even though you are clearly both, they are in fact independant states.
All programmers, stop selling your code, sell your time.
Most of us do. We call this "working for companies that produce proprietary software".
Of course, as a side effect, this means less software (legally) available to you at reasonable prices. But my family's fed.
Amazing! The Japanese promo literature is all correctly spelled, while the Slashdot story has at least three egregious errors. Unless "smurf", "shinny", and "Weib" are meant as some kind of sophisticated humor.