AFAIK, K&R state that int can be 2bytes or 4bytes. long is 4 long long is 8
Try this:
#include <stdio.h>
int main(){
printf("Size of int: %d\n", sizeof(int));
printf("Size of long: %d\n", sizeof(long));
printf("Size of long long: %d\n", sizeof(long long));
return 0; }
On a pure 32-bit machine:
[srdjant@thor ~]$./sizes Size of int: 4 Size of long: 4 Size of long long: 8
So you must be wrong in that regard.
I can't comment for the time being on the C standard saying that sizeof(long) >= 4.
> Also manages to not know about memset and write C code that assumes sizeof(long)==4 and > puts non-inline code in header files. I don't dare look at the C++ code since there's a > large probability it will cause permanent brain damage.
On 32-bit machines, sizeof(long) is 4, though I think on 64-bit machines this is 8 (but I'm not sure).
So I suspect your links is a symlink to lynx. Links links with the math library, while lynx does not. It seems unlikely that math lib is stripped out of links.
They are trying make Yahoo Mail just like using Outlook......
And that's why I don't like Yahoo Mail Beta. Not sure if others don't like it because of that too.
And nonsensical huge flash adverts, that span over 3 or so lines of my inbox, so I can't see the latest emails...
I know that under Linux you can use the nvram driver to send an ioctl() to clear the _user settings_. To do this, you need to issue the ioctl() call with the correct parameter as a privaleged user, such as root.
As to wipeing, replacing or changing the actual BIOS program code, I have no idea if it is possible. I know that you can do low-level read and write operations on memory and IO ports in kernel-space, but I'm unsure if the entirety of the BIOS can be written to by the kernel. If so, only the kernel would have access. The kernel that you boot or a module that you load would have to have code to allow programs write access to the BIOS in order for you to be initially infected. Both of these operations would need root access.
If this was an intentional backdoor, Microsoft would be able to say "oh... would you look at that! We made a slight bug. whoopsie-daisy." since as you say, it's much too elaborate to be intentional.
If they put a backdoor in an obvious place, sooner or later someone snooping will find it and Microsoft will be caught red-handed. Putting it in a less obvious place gives them the power of deniability.
Even the Borg know that... as we found out in Star Trek First Contact.
What's with the Microsoft Excel style graphs? They're not very precise or professional-looking. You would have thought the author would use something better like gnuplot?
The author's opinion "Personally, I still choose XFS for filesystem performance and scalability." is largely irrelevant here and sounds like bias, although the author acknowledges this.
There is no discussion of the results. The text between the graphs only mentions superficially what is obvious to anyone looking at the graphs.
Actually, there's Slackware. Pat has dropped GNOME support from Slackware 10.2, which makes Slackware a KDE-centric distro. And Slackware is great too;-)
[srdjant@tigerclaw ~]$ ps aux USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND [...snip...] srdjant 4897 5.0 21.8 322352 112756 ? S 22:46 0:08/usr/lib/jdk/bin/java -cp/home/srdjant/eclipse/eclipse/./startup.jar org.eclipse.core.launcher.Main -os linux -ws
As can be seen from the 5th column (VSZ), the Java Virtual Machine eats up some 320MB. And this is Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_02-b09).
[srdjant@tigerclaw ~]$ hexdump -C foo.html 00000000 3c 68 74 6d 6c 3e 3c 62 6f 64 79 3e 3c 61 20 68 |<html><body><a h| 00000010 72 65 66 3d 68 74 74 70 73 3a ad ad ad ad ad ad |ref=https: | 00000020 ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad | | 00000030 ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad | | 00000040 ad ad ad ad ad ad 20 3e 62 6c 61 68 3c 2f 61 3e | >blah</a>| 00000050 3c 2f 62 6f 64 79 3e 3c 2f 68 74 6d 6c 3e |</body></html>| 0000005e
Installation and setup are a bit of a trial, however. In fact, with Solaris 10 (although not with Solaris Express), there was too little memory on our system to run the installer. After Googling on the issue, we learned that some BIOSes don't make enough base memory available.
This leads me to believe that the installer is running in Real Mode. If the CPU was in Protected Mode, it would be able to access all of the avaliable system RAM.
Why would the installer run in Real Mode (where it can access at most 1MB), and why does the Solaris kernel not enter protected mode like Linux kernels do?
If Sun can't get even this right, they have no chance of making Open Solaris into a Linux 'killer'.
Or, encrypt your file with PGP or whatever, then rot13. When the police ask for more info and a password, just give them the password - they'll find the archive corrupt.
You might want to chop up the file or whatever as well, to make it look more corrupt. Chop the file into 4 parts, then swap over the parts and put together the file, then rot13 it again.
As long as you know how to bring it back to how it was, and you can tell them its a pgp file (or whatever), you should be able to convince them that it is corrupt.
Or get a program that logically inverts the bits of the i-nodes that belong to the encrypted file - rendering that part of the filesystem useless - corrupt. You'd have to hide the program that does this though...
Slashdot Mirror
We can use this to make machines that have proportions of 1 by 4 by 9.
Thanks for correcting me.
In which case, you are right about sizeof(long)>=4.
However I don't think many compilers define it longer than 4 bytes on
32bit machines.
I should add, that on 64bit machines, this changes.
On 64bit machines, long should be 64bit and thus sizeof(long) on
those machines should be 8.
AFAIK, K&R state that int can be 2bytes or 4bytes.
./sizes
long is 4
long long is 8
Try this:
#include <stdio.h>
int main(){
printf("Size of int: %d\n", sizeof(int));
printf("Size of long: %d\n", sizeof(long));
printf("Size of long long: %d\n", sizeof(long long));
return 0;
}
On a pure 32-bit machine:
[srdjant@thor ~]$
Size of int: 4
Size of long: 4
Size of long long: 8
So you must be wrong in that regard.
I can't comment for the time being on the C standard saying that sizeof(long) >= 4.
> Also manages to not know about memset and write C code that assumes sizeof(long)==4 and
> puts non-inline code in header files. I don't dare look at the C++ code since there's a
> large probability it will cause permanent brain damage.
On 32-bit machines, sizeof(long) is 4, though I think on 64-bit machines this is 8 (but I'm
not sure).
Please don't look at the C++ code..
His main BitFast code is in header files...
Messy and inconsistent indenting (check c/ and c++/ files for differences)
Difference between c/* and c++/* is that c++/ has cout instead of printf()... well done...
This stuff should go to worsethanfailure.com (was TheDailyWTF.com)
From TFA: "...suspected internet tax cheats..."
The Internet is now taxed by the government? Huh?
ldd /usr/bin/lynx /usr/lib/libz.so.1 (0xb7ed8000) /lib/libbz2.so.1.0 (0xb7ec8000) /lib/libncurses.so.5 (0xb7e89000) /usr/lib/libssl.so.0 (0xb7e58000) /usr/lib/libcrypto.so.0 (0xb7d54000) /lib/tls/libc.so.6 (0xb7c38000) /lib/libgpm.so.1 (0xb7c32000) /lib/tls/libdl.so.2 (0xb7c2d000) /lib/ld-linux.so.2 (0xb7f10000)
linux-gate.so.1 => (0xffffe000)
libz.so.1 =>
libbz2.so.1.0 =>
libncurses.so.5 =>
libssl.so.0 =>
libcrypto.so.0 =>
libc.so.6 =>
libgpm.so.1 =>
libdl.so.2 =>
So I suspect your links is a symlink to lynx. Links links with the math library, while lynx does not. It seems unlikely that math lib is stripped out of links.
$ ldd `which links` | grep png
Links links with the png library.
Don't they write/publish Linux Format magazine? That mag is awesome. I hope this won't affect this part of their business.
They are trying make Yahoo Mail just like using Outlook......
And that's why I don't like Yahoo Mail Beta. Not sure if others don't like it because of that too.
And nonsensical huge flash adverts, that span over 3 or so lines of my inbox, so I can't see the latest emails...
I know that under Linux you can use the nvram driver to send an ioctl() to clear the _user settings_.
To do this, you need to issue the ioctl() call with the correct parameter as a privaleged user, such
as root.
As to wipeing, replacing or changing the actual BIOS program code, I have no idea if it is possible.
I know that you can do low-level read and write operations on memory and IO ports in kernel-space,
but I'm unsure if the entirety of the BIOS can be written to by the kernel. If so, only the kernel
would have access. The kernel that you boot or a module that you load would have to have code to
allow programs write access to the BIOS in order for you to be initially infected. Both of these
operations would need root access.
If this was an intentional backdoor, Microsoft would be able to say "oh... would you look at that! We made a slight bug.
whoopsie-daisy." since as you say, it's much too elaborate to be intentional.
If they put a backdoor in an obvious place, sooner or later someone snooping will find it and Microsoft will be caught red-handed. Putting it in a less obvious place gives them the power of deniability.
Even the Borg know that... as we found out in Star Trek First Contact.
What's with the Microsoft Excel style graphs? They're not very precise or professional-looking.
You would have thought the author would use something better like gnuplot?
The author's opinion "Personally, I still choose XFS for filesystem performance and scalability."
is largely irrelevant here and sounds like bias, although the author acknowledges this.
There is no discussion of the results. The text between the graphs only mentions superficially
what is obvious to anyone looking at the graphs.
Seems a far cry from the very nicely done BSD and Linux benchmark at http://bulk.fefe.de/scalability/
Could Hubert Mantel have quit due to Novell making SuSE a GNOME-centred distro instead of keeping it a KDE-centred one?
/ 1620206&tid=223&tid=106
Novell standardise on GNOME: http://linux.slashdot.org/article.pl?sid=05/11/05
Whoops! Gotta preview more often.
s/KDE-centric/KDE-friendly/ perhaps?
Actually, there's Slackware. Pat has dropped GNOME support from Slackware 10.2, which makes Slackware ;-)
b le.txt
a KDE-centric distro. And Slackware is great too
http://slackware.com/changelog/i386/ChangeLog-sta
GNU's Now UNIX
Precisely what I was thinking.
But anyone could break it open and tap off the analogue from the speaker wires. Unless they weld the
whole speaker box up.
I'm sorry to say, Java takes up a lot of RAM.
/usr/lib/jdk/bin/java -cp /home/srdjant/eclipse/eclipse/./startup.jar org.eclipse.core.launcher.Main -os linux -ws
[srdjant@tigerclaw ~]$ ps aux
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
[...snip...]
srdjant 4897 5.0 21.8 322352 112756 ? S 22:46 0:08
As can be seen from the 5th column (VSZ), the Java Virtual Machine eats up some 320MB. And this is
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_02-b09).
Yes it's 2005, and yes Java's kludgy.
I forgot the hexdump:
[srdjant@tigerclaw ~]$ hexdump -C foo.html
00000000 3c 68 74 6d 6c 3e 3c 62 6f 64 79 3e 3c 61 20 68 |<html><body><a h|
00000010 72 65 66 3d 68 74 74 70 73 3a ad ad ad ad ad ad |ref=https: |
00000020 ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad | |
00000030 ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad ad | |
00000040 ad ad ad ad ad ad 20 3e 62 6c 61 68 3c 2f 61 3e | >blah</a>|
00000050 3c 2f 62 6f 64 79 3e 3c 2f 68 74 6d 6c 3e |</body></html>|
0000005e
I entered the html in hex editor as from:
= 13519728
a aaaaaaa/
m files/96083106_1_rvmshv
http://it.slashdot.org/comments.pl?sid=161697&cid
and clicked on the link. The link pointed to:
https://xn--m1aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
and firefox downloaded this:
http://www.srh.noaa.gov/abrfc/archive/1996/aug/rv
Why would the installer run in Real Mode (where it can access at most 1MB), and why does the Solaris kernel not enter protected mode like Linux kernels do?
If Sun can't get even this right, they have no chance of making Open Solaris into a Linux 'killer'.
ARGH! I'm just waiting for the Initio ini9100 driver to be fixed in the 2.6.0 tree, and then i'll be moving to 2.6.0.
Dammit, somebody fix that driver!
Or, encrypt your file with PGP or whatever, then rot13. When the police ask for more info and a password, just give them the password - they'll find the archive corrupt.
You might want to chop up the file or whatever as well, to make it look more corrupt.
Chop the file into 4 parts, then swap over the parts and put together the file, then rot13 it again.
As long as you know how to bring it back to how it was, and you can tell them its a pgp file (or whatever), you should be able to convince them that it is corrupt.
Or get a program that logically inverts the bits of the i-nodes that belong to the encrypted file - rendering that part of the filesystem useless - corrupt. You'd have to hide the program that does this though...