From what I have heard, the problem with P2P is the overall dependence on a single server or a set of servers. Every network to date has been dependent on at least a few common servers, if not a single common server. Even Gnutella has this problem (you have to get host information from database servers, therefore if they fail, the network will inevitably fail). Since BitTorrent is based on trackers, this could also be a problem. Is there any plans to allow the mirroring of trackers to ensure that the network doesnt die out? The other issue is Scalability. How do you plan to work on that?
Anyhow... the problems are this: 1. How will you ensure the availability of trackers? This could be done by mirroring the trackers.
2. How can you scale the system to support more users? Again, this would be able to be done by adding more users and trackers, since each client contributes.
3. Provided BitTorrent becomes popular, how can bad/improper file transfers be prevented? MD5 hashing could be used to find files that are invalid. According to the site, this is already done. MD5 hashes can theoretically be spoofed however.
The bittorrent project seems to be aware of these issues, and are working on them, so they show promise. Which is good.
WRONG. NEITHER AN OpenBSD router or a linksys box are routers. They are both NAT (Network Address Translation). You want a ROUTER. Go buy a CISCO 2605.
Games are not original. They are all variations on the same games. There are hundreds of Breakout clones, thousands of doom clones (and even a TTY version), and... and... Quake... holy crap. Anyway. Games are not original, and havent been since ~1980 or so. Why is it this way? Because it's easy to make a shooter. Oooh. Just hack up Quake and you're good to go.
No talent required. Take a look at CounterStrike. Ha! Quake 1+Halflife. Impressive. NOT. Or maybe Warcraft. THAT was original. Now we have Warcraft 2, Warcraft III, Starcraft, Diablo, Diablo II, Brood War, FreeCraft and thousands of other derivitives. Impressive. Again, nope. It's the same idea, objectives as always. DESTROY THE ENEMY. Who cares? Civ. Civ1. Civ2. Civ3. FreeCiv. And hundreds of others.
DRAKCONNECT SUCKS!!!! REALLY. IT DOES. It's PPP implementation does not WORK. DSL *kinda* works. Drakconnect will not execute PPPd half of the time. See, DrakConnect is a POS, if you didnt know already. HardDrake was decent, but it doesnt provide the damn drivers which doesnt make it n00b friendly. However, I use FreeBSD which is evidentally dead according some people on here. Even Debian's PPP implementation is DECENT. And... urpmi... DONT GET ME STARTED. URPMI SUCKS!!! AAHHH!! IT DOESNT FUCKING DO ANYTHING. YEAH LETS USE URPMI! OH WAIT. IT HAS JACK SHIT FOR PACKAGES. Whoops. Mandrake is a WASTE OF TIME for anyone who actually has CLUE. That is all.
Therefore, operating systems do SUCK, but... as long as you use them within their capacities, they will run fine. It is only until you surpass their capacity that they will fail. If you don't surpass their capacity, and it still crashes, it is a HARDWARE ISSUE.
You can really only find the capacity of a given system on a given box through experience.
...Actually... Linksys doesnt make firewalls. They make shitty NAT appliances. If you want firewalling, then openbsd is an option. If you want it done right, then using ACL on a Cisco 2610 is a better option. And if you're using Cable or DSL and calling that box with linksys or netgear or whatever on it a "router", then you need to have the crap knocked out of you with a cluebat.
Another issue about these NAT appliances made by linksys is... how secure is NAT when dumbasses enable port forwarding? Port forwarding defeats the entire purpose of NAT, IP Masquerade, IP Gatewaying, Transparent Proxying, whatever you wish to call it. NAT gives people an entirely FALSE sense of security, unless properly implemented by the person who configures the NAT appliance.
IF your NAT appliance was set up with your modem, then it was installed by a goon. Cable or DSL goons do not know how to properly config NAT appliances. Almost every cable or DSL installation that I have seen done in this manner has had all ports forwarded to a box, via placing the box in what is known as a "DMZ". Now, given this information, HOW SECURE IS A FIREWALL WHEN IT'S FORWARDING PORTS TO YOUR BOX WITHOUT ASKING QUESTIONS?!
Firewalling is a crummy method of securing internet connections that is usable for home usage, but not normally acceptable for doing anything REAL. If you care about security, then you had better make sure your crappy NAT appliance is configured properly.
Now... on to OpenBSD.
OpenBSD allows ACL, therefore is much nicer for firewalling than a NAT appliance. The problem is that it is dependant on a primarily software solution. Due to the software dependance, it can be exploited. Oops, there goes *BSD, Linux, AIX and anything else you would DARE throw at the problem on the hardware front.
Well... they can't do that... because that would require actual work... and... that isn't plausible for major conglomerates. Since a majority of the users of these cards run windows, which is insecure anyway, they figure their cards dont really need to be secure as well.
Plus... if they do it with hardware... then people will be suspicious about privacy...
There's an issue with that paper.
on
OSI vs SCO
·
· Score: 1
By using GECOS checking, (called something else in that bloatware called UnrealIRCd), people can deny connections from users using certain realname patterns. From what I have seen, Fizzer uses certain gecos information that can be used to identify itself.
Also, By ctcp pinging them, a lot of them will crash, from what I have heard. The website also states this.
Looks like it was written by a high school student, but none the less, what a nice brief. Dont those SCO People work hard?
Plus, their case doesn't hold water because, the SCO that we're talking about is not the same SCO as the SCO that provided Unix, after AT&T.
This SCO is caldera. The old SCO is SCO. Do the math, Caldera != SCO. Therefore, I do not see their grounds at all.
Re:Microsoft's new file format is:
on
Office 2003 and XML
·
· Score: 5, Funny
Oops, i forgot to set the reply to "Code". Please note, your SAX parser probably wont be able to parse this, heh. It is however, theoretically proper XML.
<?xml version="1.0" standalone="yes" encoding="en"> <!DOCTYPE worddoc [ <!ELEMENT document (document_properties, document_section)> <!ELEMENT document_properties (title, author, organization, department, job, generalsummary)> <!ELEMENT title (#PCDATA)> <!ELEMENT author (#PCDATA)> <!ELEMENT organization (#PCDATA)> <!ELEMENT department (#PCDATA)> <!ELEMENT job (#PCDATA)> <!ELEMENT generalsummary (#PCDATA)> <!ELEMENT document_section (sectionsummary, proprietarybinary, unenhancedcrappytext)> <!ELEMENT sectionsummary (#PCDATA)> <!ELEMENT proprietarybinary (#PCDATA)> <!ELEMENT unenhancedcrappytext (#PCDATA)> ]> <document> <document_properties> <title>Crappydoc</title> <author>William H. Gates III</title> <organization>BORG</organization> <department>Unimatrix 0</department> <job>Secondary information processing adjunct</job> <generalsummary>Doc about crappy M$ things.</generalsummary> </document_properties> <document_section> <sectionsummary>Haha, you cant parse this and make it look perty, it's BINARY! You're still screwed!</sectionsummary> <proprietarybinary>firoiorfioeiojvonvonviniooiwnco ncooisoi39f940f9439 0f904390f94390fj904j90j3f09j4fj3490jf30jf040fj03j0 9fj9340fj043j90fj4903fj9043jfj0vjoirejvoojvoerjgoe jgojerogjoejoenmvotnhnoignoengotnhinringuinfi</pro prietarybinary> <unenhancedcrappytext>Hehe, doesnt this text just look ugly? I bet it does, if you arent using M$ WORD!</unenhancedcrappytext> </document_section> </document>
RPM is actually a pretty good packaging system, and RPM-based distros are what will bring linux to the consumer, because the common idiot can run the system.
You _CAN_ use a stock kernel, and they offer many precompiled kernels that you can use. The files are in an acceptable location (i.e. configs are in/etc, so what do you mean there?)
Also, you can DOWNLOAD it, so you arent paying $$$$$. Anyhow, I prefer Debian/Gentoo, so dont call this a redhat plug.
What is the maximum effective strength of encryption? By this I mean, what is the best cipher strength that provides a good tradeoff between speed and security?
Encryption is good, but can be slow, so I'm just wondering.
I'd say more like 60% though. However, i'd also say that 40% of idiots make up statistics to prove their point, and 90% of people know that.
Anyway, I get about 1800 messages a day, total. Messages are ran through procmail and a complex spam filtering perl script that I wrote for myself. about 600-700 messages are blocked per day, therefore being more than 40%.
Since the american car companies have decided to not provide environmentally friendly vehicles, then I think we should boycott them. Why should they make money and not attempt to at least provide new ideas? Ford has backed out, and now GM has as well. I guess those two companies should be boycotted.
It's even worse when a company discontinues a car when people already own them. That's like a slap in the face to owners of the vehicles and also raises questions such as, can we get parts for our car?
Anyway, boycotting can be very effective, and if a boycott of GM and Ford were to occur, they would most likely continue these programs.
Ships run into each other due to the fact that they were running windows and microsoft networking hardware (loss of signal). Once the signal was lost, the ships could not be controlled because the system used to control the ship had a stop error.
any form of inactivity... such as sleeping (for example) will cause a heartattack. I guess standing at the computer and insomnia are good things after all.
.NET is a piece of trash that takes stabs at Java. C# is an annoyance, any of the M$.NET languages suck. Show me a way where visual c++ or C# can handle sockets via arrays, and I might have some respect for.NET. But not until the impossible happens will i have respect for it.
In theory, (if you have a correct socket implementation), this should work:
#include socket.h
int main() {
int socket[512];
struct addrinfo *ai;
struct linger linger;
for each y in socket[y] {
socket = socket(TCP, 0, 0);
bind(socket, "0.0.0.0/0");
listen(socket, 5);
} }
Carewolf writes "Is Windows ready for the desktop? We have heard it year after year, that now is the time for Windows on the desktop. But is it really time? Richard K. Yamauchi at OSNews don't think so and has writen a piece that list a number of issues that needs to be solved before Windows is really ready for the masses and "Joe Longkneck"."
Slashdot Mirror
From what I have heard, the problem with P2P is the overall dependence on a single server or a set of servers. Every network to date has been dependent on at least a few common servers, if not a single common server. Even Gnutella has this problem (you have to get host information from database servers, therefore if they fail, the network will inevitably fail). Since BitTorrent is based on trackers, this could also be a problem. Is there any plans to allow the mirroring of trackers to ensure that the network doesnt die out? The other issue is Scalability. How do you plan to work on that?
Anyhow... the problems are this:
1. How will you ensure the availability of trackers?
This could be done by mirroring the trackers.
2. How can you scale the system to support more users?
Again, this would be able to be done by adding more users and trackers, since each client contributes.
3. Provided BitTorrent becomes popular, how can bad/improper file transfers be prevented?
MD5 hashing could be used to find files that are invalid. According to the site, this is already done. MD5 hashes can theoretically be spoofed however.
The bittorrent project seems to be aware of these issues, and are working on them, so they show promise. Which is good.
WRONG. NEITHER AN OpenBSD router or a linksys box are routers. They are both NAT (Network Address Translation). You want a ROUTER. Go buy a CISCO 2605.
Games are not original. They are all variations on the same games. There are hundreds of Breakout clones, thousands of doom clones (and even a TTY version), and... and... Quake... holy crap. Anyway. Games are not original, and havent been since ~1980 or so. Why is it this way? Because it's easy to make a shooter. Oooh. Just hack up Quake and you're good to go.
No talent required. Take a look at CounterStrike. Ha! Quake 1+Halflife. Impressive. NOT. Or maybe Warcraft. THAT was original. Now we have Warcraft 2, Warcraft III, Starcraft, Diablo, Diablo II, Brood War, FreeCraft and thousands of other derivitives. Impressive. Again, nope. It's the same idea, objectives as always. DESTROY THE ENEMY. Who cares? Civ. Civ1. Civ2. Civ3. FreeCiv. And hundreds of others.
DRAKCONNECT SUCKS!!!! REALLY. IT DOES. It's PPP implementation does not WORK. DSL *kinda* works. Drakconnect will not execute PPPd half of the time. See, DrakConnect is a POS, if you didnt know already. HardDrake was decent, but it doesnt provide the damn drivers which doesnt make it n00b friendly. However, I use FreeBSD which is evidentally dead according some people on here. Even Debian's PPP implementation is DECENT. And... urpmi... DONT GET ME STARTED. URPMI SUCKS!!! AAHHH!! IT DOESNT FUCKING DO ANYTHING. YEAH LETS USE URPMI! OH WAIT. IT HAS JACK SHIT FOR PACKAGES. Whoops. Mandrake is a WASTE OF TIME for anyone who actually has CLUE. That is all.
Not to mention Linux, which has a worse implementation of POSIX.1 threading than Windows (get into thousands of threads and kernel goes boom).
UnixWare, which is a joke that isnt much more stable than Linux when it comes to threads.
Solaris, which can easily crash and choke up on threading applications.
QNX, which has really shitty thread support, etc etc etc.
Anyhow... let's look at some uptimes.
trogdor.nenolod.net (Windows 98SE, I do not use this box) -- 308 days.
matrix.nenolod.net (Linux, slight usage):
12:46:38 up 17 days, 13:31, 2 users, load average: 0.00, 0.01, 0.03
Therefore, operating systems do SUCK, but... as long as you use them within their capacities, they will run fine. It is only until you surpass their capacity that they will fail. If you don't surpass their capacity, and it still crashes, it is a HARDWARE ISSUE.
You can really only find the capacity of a given system on a given box through experience.
...Actually... Linksys doesnt make firewalls. They make shitty NAT appliances. If you want firewalling, then openbsd is an option. If you want it done right, then using ACL on a Cisco 2610 is a better option. And if you're using Cable or DSL and calling that box with linksys or netgear or whatever on it a "router", then you need to have the crap knocked out of you with a cluebat.
Another issue about these NAT appliances made by linksys is... how secure is NAT when dumbasses enable port forwarding? Port forwarding defeats the entire purpose of NAT, IP Masquerade, IP Gatewaying, Transparent Proxying, whatever you wish to call it. NAT gives people an entirely FALSE sense of security, unless properly implemented by the person who configures the NAT appliance.
IF your NAT appliance was set up with your modem, then it was installed by a goon. Cable or DSL goons do not know how to properly config NAT appliances. Almost every cable or DSL installation that I have seen done in this manner has had all ports forwarded to a box, via placing the box in what is known as a "DMZ". Now, given this information, HOW SECURE IS A FIREWALL WHEN IT'S FORWARDING PORTS TO YOUR BOX WITHOUT ASKING QUESTIONS?!
Firewalling is a crummy method of securing internet connections that is usable for home usage, but not normally acceptable for doing anything REAL. If you care about security, then you had better make sure your crappy NAT appliance is configured properly.
Now... on to OpenBSD.
OpenBSD allows ACL, therefore is much nicer for firewalling than a NAT appliance. The problem is that it is dependant on a primarily software solution. Due to the software dependance, it can be exploited. Oops, there goes *BSD, Linux, AIX and anything else you would DARE throw at the problem on the hardware front.
Well... they can't do that... because that would require actual work... and... that isn't plausible for major conglomerates. Since a majority of the users of these cards run windows, which is insecure anyway, they figure their cards dont really need to be secure as well.
Plus... if they do it with hardware... then people will be suspicious about privacy...
Linux was derived from MINIX. That is all.
By using GECOS checking, (called something else in that bloatware called UnrealIRCd), people can deny connections from users using certain realname patterns. From what I have seen, Fizzer uses certain gecos information that can be used to identify itself.
Also, By ctcp pinging them, a lot of them will crash, from what I have heard. The website also states this.
nenolod, OpenIRC Network administrator.
Just more proof that unless you are absolutely like them, even in your thinking that you wont get anywhere. Isn't that just sick?
The opteron has a 32-bit compatibility mode, where the Intel solution does not, so Windows 2003 Server *does* support the Opteron.
and Arthur Anderson (the accounting firm that caused Enron) is teaching a course in corporate responsibility.
Looks like it was written by a high school student, but none the less, what a nice brief. Dont those SCO People work hard?
Plus, their case doesn't hold water because, the SCO that we're talking about is not the same SCO as the SCO that provided Unix, after AT&T.
This SCO is caldera. The old SCO is SCO. Do the math, Caldera != SCO. Therefore, I do not see their grounds at all.
Oops, i forgot to set the reply to "Code". Please note, your SAX parser probably wont be able to parse this, heh. It is however, theoretically proper XML.
o ncooisoi39f940f9439 0f904390f94390fj904j90j3f09j4fj3490jf30jf040fj03j0 9fj9340fj043j90fj4903fj9043jfj0vjoirejvoojvoerjgoe jgojerogjoejoenmvotnhnoignoengotnhinringuinfi</pro prietarybinary>
<?xml version="1.0" standalone="yes" encoding="en">
<!DOCTYPE worddoc [
<!ELEMENT document (document_properties, document_section)>
<!ELEMENT document_properties (title, author, organization, department, job, generalsummary)>
<!ELEMENT title (#PCDATA)>
<!ELEMENT author (#PCDATA)>
<!ELEMENT organization (#PCDATA)>
<!ELEMENT department (#PCDATA)>
<!ELEMENT job (#PCDATA)>
<!ELEMENT generalsummary (#PCDATA)>
<!ELEMENT document_section (sectionsummary, proprietarybinary, unenhancedcrappytext)>
<!ELEMENT sectionsummary (#PCDATA)>
<!ELEMENT proprietarybinary (#PCDATA)>
<!ELEMENT unenhancedcrappytext (#PCDATA)>
]>
<document>
<document_properties>
<title>Crappydoc</title>
<author>William H. Gates III</title>
<organization>BORG</organization>
<department>Unimatrix 0</department>
<job>Secondary information processing adjunct</job>
<generalsummary>Doc about crappy M$ things.</generalsummary>
</document_properties>
<document_section>
<sectionsummary>Haha, you cant parse this and make it look perty, it's BINARY! You're still screwed!</sectionsummary>
<proprietarybinary>firoiorfioeiojvonvonviniooiwnc
<unenhancedcrappytext>Hehe, doesnt this text just look ugly? I bet it does, if you arent using M$ WORD!</unenhancedcrappytext>
</document_section>
</document>
]>
Crappydoc
William H. Gates III
BORG
Unimatrix 0
Secondary information processing adjunct
Doc about crappy M$ things.
Haha, you cant parse this, it's BINARY! You're still screwed!
firoiorfioeiojvonvonviniooiwnconcooisoi39f940f943
And, you go around and troll.
/etc, so what do you mean there?)
RPM is actually a pretty good packaging system, and RPM-based distros are what will bring linux to the consumer, because the common idiot can run the system.
You _CAN_ use a stock kernel, and they offer many precompiled kernels that you can use. The files are in an acceptable location (i.e. configs are in
Also, you can DOWNLOAD it, so you arent paying $$$$$. Anyhow, I prefer Debian/Gentoo, so dont call this a redhat plug.
What is the maximum effective strength of encryption? By this I mean, what is the best cipher strength that provides a good tradeoff between speed and security?
Encryption is good, but can be slow, so I'm just wondering.
I'd say more like 60% though. However, i'd also say that 40% of idiots make up statistics to prove their point, and 90% of people know that.
Anyway, I get about 1800 messages a day, total. Messages are ran through procmail and a complex spam filtering perl script that I wrote for myself. about 600-700 messages are blocked per day, therefore being more than 40%.
I'd also state that most SMB popups are SPAM.
Since the american car companies have decided to not provide environmentally friendly vehicles, then I think we should boycott them. Why should they make money and not attempt to at least provide new ideas? Ford has backed out, and now GM has as well. I guess those two companies should be boycotted.
It's even worse when a company discontinues a car when people already own them. That's like a slap in the face to owners of the vehicles and also raises questions such as, can we get parts for our car?
Anyway, boycotting can be very effective, and if a boycott of GM and Ford were to occur, they would most likely continue these programs.
Ships run into each other due to the fact that they were running windows and microsoft networking hardware (loss of signal). Once the signal was lost, the ships could not be controlled because the system used to control the ship had a stop error.
My mistake, i meant 81 or something.
Second, duh... to handle multiple connections.
Third, never tried that in C? It's a basic nested loop. (not so sure if that's the proper syntax however)
any form of inactivity... such as sleeping (for example) will cause a heartattack. I guess standing at the computer and insomnia are good things after all.
Oh, shoot... it's getting dark in here.
.NET is a piece of trash that takes stabs at Java. C# is an annoyance, any of the M$ .NET languages suck. Show me a way where visual c++ or C# can handle sockets via arrays, and I might have some respect for .NET. But not until the impossible happens will i have respect for it.
In theory, (if you have a correct socket implementation), this should work:
#include socket.h
int main() {
int socket[512];
struct addrinfo *ai;
struct linger linger;
for each y in socket[y] {
socket = socket(TCP, 0, 0);
bind(socket, "0.0.0.0/0");
listen(socket, 5);
}
}
Or something like that.
Carewolf writes "Is Windows ready for the desktop? We have heard it year after year, that now is the time for Windows on the desktop. But is it really time? Richard K. Yamauchi at OSNews don't think so and has writen a piece that list a number of issues that needs to be solved before Windows is really ready for the masses and "Joe Longkneck"."
Replace that with doesn't think so.
How about use NO OS. Write your own system for control. Best way to do it.