Symbian does the job that it was designed to do, and does it well. That is, it makes a good phone, with an elementry address book and simple games. This is good for a phone from 1999. The problem is that today, phones do much more and in the near future they will need to do even more so. Among these things are: bluetooth stack, audio and video playback, filesystem support (think sd cards), more advanced applications and games, virtual memory management, advanced process scheduling features, (wireless) USB stack with host/client, mass storage controller stack, input device support (who knows what kind), hotplug capability (expansion cards), tcp/ip networking (perhaps for VOIP). On top of these, you'll want to run advanced web browsers (how about KHTML?), mp3 players and an itunes-compatible DRM client, an address book to sync with Outlook, real games that have a better interface than the lousy phone keypad, Java/brew environment, Vonage client, net stumbler, secure credit card transaction manager, SD card file browser,...
Yes, some of these things are hacked into Symbian now. But think of who Symbian's biggest competitor is: Windows CE. WinCE provides all of the above. Nokia is not an operating system company, and can't afford to be. They can modify Symbian to no end, but the effort required is large. Or they can use a freely available piece of code that does it all already.
As for NetBSD, Nokia is kind of in the anti-Microsoft camp because they fear MS marginalizing them. Like Palm, they've been fighting the invasion of WinCE, and they too realize that having an open system (to which they and their fellows in the anti-MS group) have to contribute benefits everyone.
Principal Skinner: Oh, well, that was wonderful. A good time was had by all; i'm pooped. Superintendent Chalmers: Yes, I should be -- Good Lord! What is happening in there? Skinner: Aurora Borealis. Chalmers: Aurora Borealis? At this time of year, at this time of the day, in this part of the country, localized entirely in your kitchen? Skinner: Yes. Chalmers: May i see it? Skinner: No. Agnes Skinner: Seymour, the house is on fire! Skinner: No, Mother, it's just the Northern Lights
This is different from a TCP/IP offload engine in one critical feature: it is designed so that the network interface is run directly out of user-space, without involving the kernel. It may still do the network stack in software, but it does it in-process. This means that there is no copying from/to userspace and no context switch for every read or write. When you're handling gigabit-speed traffic, this becomes a big issue. Just like video players today use special OS features to open a video port directly to the graphics card without routing it through the windowing system, this does the same with network data. Obviously, you and I won't be buying this thing, unless you happen to work in an environment where you have massive network traffic.
Well, the answer you seem to be getting is "multithreading". That's one aspect. Another is multiple processes. My desktop system (with only me logged in and browsing the web) has 66 processes running. Think of all those producer-consumer processes you create when you invoke a command line that has 'find' pipe to 'grep' pipe to 'awk' pipe to 'sort'... each can be on a separate processor, running in parallel as real pipeline. Some programs have been written to take advantage of process parallelization. make(1) is a good example of this. Given a list of dependencies, it will figure out those that can be run independently and execute those on as many processors as are available. Another is Apache (1.0 series), that forks off server processes. It can scale nearly linearly with added cores/processors (assuming the bottleneck is not disk or network IO).
There are languages and libraries that also include support for explicit paralellism. These are typically found in specialty software written for supercomputers. That's one extreme. The other extreme is compiler-optimized parallelism that the programmer does not have to explicitly code. Today, compilers can optimize your program so that it makes efficient use of execution units in parallel (on a single core). It's logical that this will be extended to the multi-core world, although it will be quite a step up in compiler technology from what we have today. And of course, multithreaded programming (which too is explicit parallelism) is here and will only get more popular with anyone who demands speed.
holy crap, remind me to never hire you. if you know exactly what the string is going to look like, why didn't you just write it? if you expect it to change, why did you hard code the length values into the buffers?
Mayor: What do you mean, "biblical"? Dr Ray Stantz: What he means is Old Testament, Mr. Mayor, real wrath-of-God type stuff. Dr. Peter Venkman: Exactly. Dr Ray Stantz: Fire and brimstone coming down from the skies. Rivers and seas boiling. Dr. Egon Spengler: Forty years of darkness. Earthquakes, volcanoes... Winston Zeddemore: The dead rising from the grave. Dr. Peter Venkman: Human sacrifice, dogs and cats living together - mass hysteria.
This is very true at UCLA (where I went, 10 years ago). The first two years of CS are made to be hell, so that a large part of the class will drop it. This is done partly to enhance the school's ranking, and partly to get rid of ones that aren't really interested in the major. Going from straight A's in highschool to mostly C's in college is a difficult transition for many people. I would say that half of my classmates dropped in the first two years, and that was back when anyone could get big money with a CS degree.
The classes were definitely prone to the Steve Rule, and of the women that were there, there were only one or two per entering year that could both speak English and stuck through the program to the end. It should be noted though that many of these women found they were in such high demand by companies that they didn't really need to go through the CS program. They were more smart than lazy to drop CS for an easier major.
I think this movie proves that a bunch of Star Wars geeks with access to good effects and no acting or writing talent can do just as good of a job as George Lucas. Imagine what would have happened here if they actually had a budget. If anything, it leads me to believe that an all-CG Star Wars film done by fans might turn out far better than the crap that episodes 1 and 2 have been.
In fact, what he is describing is NOT virtual memory, but demand paging. At this point, I stopped reading the article.
Virtual memory is the mapping of physical memory pages to a "virtual" memory address using hardware translation tables. This is done so that every application lives in its own private memory space, and cannot interfere with other applications' memory (or the OS's). Basically, this technique of memory isolation keeps user apps from crashing the system or other applications. Virtual memory support has been added to x86 with the release of the 80386 processors and 32-bit flat memory addressing. Of course, virtual memory has been available for years before this on such OS's as DEC's VMS (the Virtual Memory System), IBM's MVS, UNIX, and a bunch of other systems I'm too young to know about.
The misnaming of demand paging was actually started by Apple (continuing their tradition of calling the box a "CPU") for their swap file management (long before MacOS's support for VM in OSX).
I have, and already built Firefox 1.0.2 while I was out getting dinner. The source was not on any of the mirrors, but emerge pulled it from the Mozilla FTP site.
Strangely, it seemed to have explicitly unset my "-fstack-protector" CFLAG during compile. I don't remember the other Firefox builds doing that, and it seems unfortunate to have had to do it on one of my most exposed networked applications. I hope we get an -r1 with it soon. Seems to be working fine, BTW.
Not quite so clearly. One common strategy often played by big companies when they want to purchase someone (especially when it's done in a hostile fashion), is to sue them beyond their means to defend themselves, and then promise to call off their lawyers if they sell out at the lowball offer price. I've worked for a company that has had this done to them. It should be called extortion, but is perfectly legal. Imagine being the guy getting a lawsuit against himself (the individual), by a multi-billion dollar corporation. What choice do you have?
I'm in. But we're going to have to add support for explicit pointers, with arithmatic and array operations. And gotos. Also, let's throw in operator overloading. And a select(2) call and asynchronous IO implementation. And #defines, and header files. And inline assembly. Also, we'll have to make the garbage collector optional for people who want to handle memory allocation themselves, and implement a malloc(3) and free(3).
We'll call it Java.add(1).
Re:Why would you care for shorter -compile- times?
on
GCC 4.0 Preview
·
· Score: 1
I'm a software engineer, I use gcc every day, and I sometimes compile code 200 times in one day. So, while you can afford to wait 4 hours for your desktop manager to compile, I can't afford to wait two minutes for a build to complete two hundred times per day.
I don't care if it compiles any faster, just as long as it compiles correctly.
It goes without saying that correctness is most important. A compiler that doesn't compile correctly is worthless. Having said that, if I'm working on a big project and a compile takes one minute instead of two, it will save me a couple of workdays of wasted time by my thousandth compilation.
But will it compile C++ any faster? The difference between compile times of C and C++ files is staggering. Compiling Qt/KDE takes forever with gcc 3.x.
Okay, nobody seems to get the point of this change, so let me spell it out for you:
ADVERTISING
They don't care about reading what 12 year olds gossip about, and they don't care about finding criminals, terrorists, or anyone else. They care about *making money* by selling targeted ads to you, and they will figure out what you like by parsing context out of your chat logs. Y'know, like Google does with Gmail and Google Groups. The TOS let them do whatever they want with the data so they can store it, mine it, and sell the results anytime they feel like with no consequences.
Hmm, I recall being stuck in a lab in front of DOS machines in college. The TA would drone on and on and on... Luckily, these machines were networked. So, I brought in a floppy with telnet.exe (no SSH available at the time) for DOS. From there, log into the Engineering dept. servers. From there, run ircii, lynx, micq, talk, and pine all inside their own screen sessions, as well as do my programming assignments using our friends vi and gcc.
Thank $DEITY for those command line distractions, otherwise I would have had to pay attention in class!
Aha, I knew something didn't look right there. Thank you for pointing out my ignorance. What I meant was "nested", not anonymous functions. This is what the OpenBSD people were having trouble with. Anonymous functions don't exist in C, to the best of my knowledge.
This is a gcc extension that is not standard C, and is generally a stupid idea and should not be used by anyone, ever. Now that I've given the disclaimer, they go a little something like this:
#include <stdio.h>
void go(void) { int i;
/* this is a nested function it is only valid in the scope of function go() */ static void nested(int x) { printf("%d\n", x); return; }
My God, who modded you up? That was, perhaps, the least factually correct post I've ever read on Slashdot (even including the troll reports of Stephen King's death).
Basically this is just laziness in the Windows architecture that overlaps the code and data segments. Separate these and the problem is solved with no new hardware, minimal application rework, and the like.
This is true for Linux, FreeBSD, NetBSD, and BeOS. In fact, OpenBSD is the only OS I know of for x86 that does make use of the execution segment division (only recently implemented), and they freely admit that their implementation is only a halfassed solution to the problem. The reality is that this is not a trivial problem. x86 processors only provide a single "line" in the memory which separates instructions and data (unlike AMD64's per-page execute permission bits). With dynamic linking, mmap, and shared memory, this becomes very messy. The OpenBSD implementation broke many applications which execute stack data. Particularly, gcc tends to produce code that executes stack data when it encounters anonymous functions in C (so any program using anonymous functions can potentially have explicit stack execution coded in).
As long as applications are behaved and don't have segment overrides all over the place, this should be just fine.
This is utter nonesense. No 32bit windows program uses the segmented memory model. As another poster pointed out, they all use the flat 4gig memory model.
Then, when you try to jump to an address that's in the stack, the processor will trip a general protection fault (because the stack must be in a segment defined as data, well, stack to be precise).
Jump to an address in the stack, you mean like returning from a function call? Or calling a function pointer? Or in some cases calling a method of a class? Again, nonesense.
Sure, you can use "smart" languages and NX bits and stuff like that, but it's all assembly at some level, and the processor manufactures actually built in sufficient protection decades ago when they came up with segmented memory. (PowerPC architecture can also distinguish between code and non-code).
No, the NX bits on Alpha, Sparc, and now AMD64 processors are at the _hardware_ level, and they do not require segmented memory. There are software-level solutions, but they are beyond the scope of this discussion. The PowerPC code segmentation is as bad as the x86 model, and equally painful. MacOS X does not segment its memory space either!
My friend used to work at Netscape (way back when they were actually coding their own browser). He said that they discovered the reason that Netscape Navigator 3.x took so long to start up was because it performed a big bubble-sort when initializing the Java plugin. Replacing it with quicksort halved the startup time.
You mean self-signed certificates delivered over an insecure connection, right?
No, only delivered over a TRUSTED and SECURE connection. I can establish a secure SSL connection with anyone. How do I know I've established this "secure connection" with the correct host and not the evil man-in-the-middle? Without a third party authentication method, I don't know of one. Yes, there's USB keyrings and key signatures and so on, but they all break down to having a shared secret with the remote host, which does not hold true for every machine you may want to connect to.
The certificates that came with your browser, for the most part, are trusted. You trust the code, and you trust that it was delivered to you in an untampered form (otherwise, you should not be using this browser for secure transactions). While not impossible, it's much less likely that someone is staging an attack on you by tampering with your browser or operating system delivery path (and if they did, you'd have much bigger problems than just ssl snooping).
Otherwise, why is a certificate that I generate and safely install any less secure (transport-wise) than one delivered with my browser? The one that can be verified with the keys that came with your browser is known to be who it claims to be. There is no way to prove this for the cert you generated.
I agree, the sentence is deceptive, but not necessarily wrong. This is true for many witty remarks.
I would love for you to provide a better example though - and although they doubtlessly exist, I've no doubt they can be accomodated for.
No doubt, modifying English to accomodate a computer is easily done. However, that isn't the the trick we're going for. We're trying for a computer program that can parse any English, like the original post in this thread.
Slashdot Mirror
I stand corrected. Thanks.
Yes, but...
...
Symbian does the job that it was designed to do, and does it well. That is, it makes a good phone, with an elementry address book and simple games. This is good for a phone from 1999. The problem is that today, phones do much more and in the near future they will need to do even more so. Among these things are: bluetooth stack, audio and video playback, filesystem support (think sd cards), more advanced applications and games, virtual memory management, advanced process scheduling features, (wireless) USB stack with host/client, mass storage controller stack, input device support (who knows what kind), hotplug capability (expansion cards), tcp/ip networking (perhaps for VOIP).
On top of these, you'll want to run advanced web browsers (how about KHTML?), mp3 players and an itunes-compatible DRM client, an address book to sync with Outlook, real games that have a better interface than the lousy phone keypad, Java/brew environment, Vonage client, net stumbler, secure credit card transaction manager, SD card file browser,
Yes, some of these things are hacked into Symbian now. But think of who Symbian's biggest competitor is: Windows CE. WinCE provides all of the above. Nokia is not an operating system company, and can't afford to be. They can modify Symbian to no end, but the effort required is large. Or they can use a freely available piece of code that does it all already.
As for NetBSD, Nokia is kind of in the anti-Microsoft camp because they fear MS marginalizing them. Like Palm, they've been fighting the invasion of WinCE, and they too realize that having an open system (to which they and their fellows in the anti-MS group) have to contribute benefits everyone.
Principal Skinner: Oh, well, that was wonderful. A good time was had by all; i'm pooped.
Superintendent Chalmers: Yes, I should be -- Good Lord! What is happening in there?
Skinner: Aurora Borealis.
Chalmers: Aurora Borealis? At this time of year, at this time of the day, in this part of the country, localized entirely in your kitchen?
Skinner: Yes.
Chalmers: May i see it?
Skinner: No.
Agnes Skinner: Seymour, the house is on fire!
Skinner: No, Mother, it's just the Northern Lights
This is different from a TCP/IP offload engine in one critical feature: it is designed so that the network interface is run directly out of user-space, without involving the kernel. It may still do the network stack in software, but it does it in-process. This means that there is no copying from/to userspace and no context switch for every read or write. When you're handling gigabit-speed traffic, this becomes a big issue. Just like video players today use special OS features to open a video port directly to the graphics card without routing it through the windowing system, this does the same with network data. Obviously, you and I won't be buying this thing, unless you happen to work in an environment where you have massive network traffic.
Well, the answer you seem to be getting is "multithreading". That's one aspect. Another is multiple processes. My desktop system (with only me logged in and browsing the web) has 66 processes running. Think of all those producer-consumer processes you create when you invoke a command line that has 'find' pipe to 'grep' pipe to 'awk' pipe to 'sort'... each can be on a separate processor, running in parallel as real pipeline. Some programs have been written to take advantage of process parallelization. make(1) is a good example of this. Given a list of dependencies, it will figure out those that can be run independently and execute those on as many processors as are available. Another is Apache (1.0 series), that forks off server processes. It can scale nearly linearly with added cores/processors (assuming the bottleneck is not disk or network IO).
There are languages and libraries that also include support for explicit paralellism. These are typically found in specialty software written for supercomputers. That's one extreme. The other extreme is compiler-optimized parallelism that the programmer does not have to explicitly code. Today, compilers can optimize your program so that it makes efficient use of execution units in parallel (on a single core). It's logical that this will be extended to the multi-core world, although it will be quite a step up in compiler technology from what we have today. And of course, multithreaded programming (which too is explicit parallelism) is here and will only get more popular with anyone who demands speed.
holy crap, remind me to never hire you. if you know exactly what the string is going to look like, why didn't you just write it? if you expect it to change, why did you hard code the length values into the buffers?
#include <stdio.h>
int main(void)
{
puts("#include<stdio.h>\n"
"int main(void)\n"
"{\n"
" puts(\"hello world!\");\n"
" return 0;\n"
"}");
return 0;
}
Mayor: What do you mean, "biblical"?
Dr Ray Stantz: What he means is Old Testament, Mr. Mayor, real wrath-of-God type stuff.
Dr. Peter Venkman: Exactly.
Dr Ray Stantz: Fire and brimstone coming down from the skies. Rivers and seas boiling.
Dr. Egon Spengler: Forty years of darkness. Earthquakes, volcanoes...
Winston Zeddemore: The dead rising from the grave.
Dr. Peter Venkman: Human sacrifice, dogs and cats living together - mass hysteria.
This is very true at UCLA (where I went, 10 years ago). The first two years of CS are made to be hell, so that a large part of the class will drop it. This is done partly to enhance the school's ranking, and partly to get rid of ones that aren't really interested in the major. Going from straight A's in highschool to mostly C's in college is a difficult transition for many people. I would say that half of my classmates dropped in the first two years, and that was back when anyone could get big money with a CS degree.
The classes were definitely prone to the Steve Rule, and of the women that were there, there were only one or two per entering year that could both speak English and stuck through the program to the end. It should be noted though that many of these women found they were in such high demand by companies that they didn't really need to go through the CS program. They were more smart than lazy to drop CS for an easier major.
I think this movie proves that a bunch of Star Wars geeks with access to good effects and no acting or writing talent can do just as good of a job as George Lucas. Imagine what would have happened here if they actually had a budget. If anything, it leads me to believe that an all-CG Star Wars film done by fans might turn out far better than the crap that episodes 1 and 2 have been.
"Let them hate, so long as they fear"
-- Lucius Accius
In fact, what he is describing is NOT virtual memory, but demand paging. At this point, I stopped reading the article.
Virtual memory is the mapping of physical memory pages to a "virtual" memory address using hardware translation tables. This is done so that every application lives in its own private memory space, and cannot interfere with other applications' memory (or the OS's). Basically, this technique of memory isolation keeps user apps from crashing the system or other applications. Virtual memory support has been added to x86 with the release of the 80386 processors and 32-bit flat memory addressing. Of course, virtual memory has been available for years before this on such OS's as DEC's VMS (the Virtual Memory System), IBM's MVS, UNIX, and a bunch of other systems I'm too young to know about.
The misnaming of demand paging was actually started by Apple (continuing their tradition of calling the box a "CPU") for their swap file management (long before MacOS's support for VM in OSX).
I have, and already built Firefox 1.0.2 while I was out getting dinner. The source was not on any of the mirrors, but emerge pulled it from the Mozilla FTP site.
Strangely, it seemed to have explicitly unset my "-fstack-protector" CFLAG during compile. I don't remember the other Firefox builds doing that, and it seems unfortunate to have had to do it on one of my most exposed networked applications. I hope we get an -r1 with it soon. Seems to be working fine, BTW.
Let's not forget the battle-cry of the Chico State Fighting Keggers:
Woooooooooo0000000000OOOOOOOOOOOOOOOooooo!!!!
(ladies, groggily add "I'm so wasted" towards the end).
Hasbro clearly isn't interested in a purchase...
Not quite so clearly. One common strategy often played by big companies when they want to purchase someone (especially when it's done in a hostile fashion), is to sue them beyond their means to defend themselves, and then promise to call off their lawyers if they sell out at the lowball offer price. I've worked for a company that has had this done to them. It should be called extortion, but is perfectly legal. Imagine being the guy getting a lawsuit against himself (the individual), by a multi-billion dollar corporation. What choice do you have?
I'm in. But we're going to have to add support for explicit pointers, with arithmatic and array operations. And gotos. Also, let's throw in operator overloading. And a select(2) call and asynchronous IO implementation. And #defines, and header files. And inline assembly. Also, we'll have to make the garbage collector optional for people who want to handle memory allocation themselves, and implement a malloc(3) and free(3).
We'll call it Java.add(1).
I'm a software engineer, I use gcc every day, and I sometimes compile code 200 times in one day. So, while you can afford to wait 4 hours for your desktop manager to compile, I can't afford to wait two minutes for a build to complete two hundred times per day.
I don't care if it compiles any faster, just as long as it compiles correctly.
It goes without saying that correctness is most important. A compiler that doesn't compile correctly is worthless. Having said that, if I'm working on a big project and a compile takes one minute instead of two, it will save me a couple of workdays of wasted time by my thousandth compilation.
But will it compile C++ any faster? The difference between compile times of C and C++ files is staggering. Compiling Qt/KDE takes forever with gcc 3.x.
Okay, nobody seems to get the point of this change, so let me spell it out for you:
ADVERTISING
They don't care about reading what 12 year olds gossip about, and they don't care about finding criminals, terrorists, or anyone else. They care about *making money* by selling targeted ads to you, and they will figure out what you like by parsing context out of your chat logs. Y'know, like Google does with Gmail and Google Groups. The TOS let them do whatever they want with the data so they can store it, mine it, and sell the results anytime they feel like with no consequences.
Hmm, I recall being stuck in a lab in front of DOS machines in college. The TA would drone on and on and on... Luckily, these machines were networked. So, I brought in a floppy with telnet.exe (no SSH available at the time) for DOS. From there, log into the Engineering dept. servers. From there, run ircii, lynx, micq, talk, and pine all inside their own screen sessions, as well as do my programming assignments using our friends vi and gcc.
Thank $DEITY for those command line distractions, otherwise I would have had to pay attention in class!
Aha, I knew something didn't look right there. Thank you for pointing out my ignorance. What I meant was "nested", not anonymous functions. This is what the OpenBSD people were having trouble with. Anonymous functions don't exist in C, to the best of my knowledge.
/* this is a nested function
This is a gcc extension that is not standard C, and is generally a stupid idea and should not be used by anyone, ever. Now that I've given the disclaimer, they go a little something like this:
#include <stdio.h>
void go(void)
{
int i;
it is only valid in the scope of function go() */
static void nested(int x)
{
printf("%d\n", x);
return;
}
for(i = 0; i < 5; i++) nested(i);
}
int main(void)
{
go();
return 0;
}
My God, who modded you up? That was, perhaps, the least factually correct post I've ever read on Slashdot (even including the troll reports of Stephen King's death).
Basically this is just laziness in the Windows architecture that overlaps the code and data segments. Separate these and the problem is solved with no new hardware, minimal application rework, and the like.
This is true for Linux, FreeBSD, NetBSD, and BeOS. In fact, OpenBSD is the only OS I know of for x86 that does make use of the execution segment division (only recently implemented), and they freely admit that their implementation is only a halfassed solution to the problem. The reality is that this is not a trivial problem. x86 processors only provide a single "line" in the memory which separates instructions and data (unlike AMD64's per-page execute permission bits). With dynamic linking, mmap, and shared memory, this becomes very messy. The OpenBSD implementation broke many applications which execute stack data. Particularly, gcc tends to produce code that executes stack data when it encounters anonymous functions in C (so any program using anonymous functions can potentially have explicit stack execution coded in).
As long as applications are behaved and don't have segment overrides all over the place, this should be just fine.
This is utter nonesense. No 32bit windows program uses the segmented memory model. As another poster pointed out, they all use the flat 4gig memory model.
Then, when you try to jump to an address that's in the stack, the processor will trip a general protection fault (because the stack must be in a segment defined as data, well, stack to be precise).
Jump to an address in the stack, you mean like returning from a function call? Or calling a function pointer? Or in some cases calling a method of a class? Again, nonesense.
Sure, you can use "smart" languages and NX bits and stuff like that, but it's all assembly at some level, and the processor manufactures actually built in sufficient protection decades ago when they came up with segmented memory. (PowerPC architecture can also distinguish between code and non-code).
No, the NX bits on Alpha, Sparc, and now AMD64 processors are at the _hardware_ level, and they do not require segmented memory. There are software-level solutions, but they are beyond the scope of this discussion. The PowerPC code segmentation is as bad as the x86 model, and equally painful. MacOS X does not segment its memory space either!
My friend used to work at Netscape (way back when they were actually coding their own browser). He said that they discovered the reason that Netscape Navigator 3.x took so long to start up was because it performed a big bubble-sort when initializing the Java plugin. Replacing it with quicksort halved the startup time.
You mean self-signed certificates delivered over an insecure connection, right?
No, only delivered over a TRUSTED and SECURE connection. I can establish a secure SSL connection with anyone. How do I know I've established this "secure connection" with the correct host and not the evil man-in-the-middle? Without a third party authentication method, I don't know of one. Yes, there's USB keyrings and key signatures and so on, but they all break down to having a shared secret with the remote host, which does not hold true for every machine you may want to connect to.
The certificates that came with your browser, for the most part, are trusted. You trust the code, and you trust that it was delivered to you in an untampered form (otherwise, you should not be using this browser for secure transactions). While not impossible, it's much less likely that someone is staging an attack on you by tampering with your browser or operating system delivery path (and if they did, you'd have much bigger problems than just ssl snooping).
Otherwise, why is a certificate that I generate and safely install any less secure (transport-wise) than one delivered with my browser?
The one that can be verified with the keys that came with your browser is known to be who it claims to be. There is no way to prove this for the cert you generated.
I agree, the sentence is deceptive, but not necessarily wrong. This is true for many witty remarks.
I would love for you to provide a better example though - and although they doubtlessly exist, I've no doubt they can be accomodated for.
No doubt, modifying English to accomodate a computer is easily done. However, that isn't the the trick we're going for. We're trying for a computer program that can parse any English, like the original post in this thread.