So.. your machine has started on the swap, what windowmanager is running? have you tried something lightweight like windowmaker ? KDE/Gnome are very good at dragging the overall performance of X down, and if the windowmanager is sluggish it will make every app running under it seem sluggish too.. X isn`t sluggish atall here, and the machine: top - 15:35:17 up 22 days, 5:14, 28 users, load average: 2.95, 2.89, 2.87 Tasks: 162 total, 4 running, 157 sleeping, 1 stopped, 0 zombie Cpu(s): 75.8% user, 20.7% system, 3.5% nice, 0.0% idle Mem: 773136k total, 700144k used, 72992k free, 160160k buffers Swap: 393584k total, 357856k used, 35728k free, 261656k cached has an old 8mb permedia-2 gfxcard, running in 1024x768 32bpp, even under this load (caused by multiple compiles, courtesy of emerge -u world) it`s not making a noticeable difference to the performance of X, windowmaker is very swift, and the only time it used to slow down was when i had considerably less ram, and parts of X/windowmaker got swapped out. Your system is using 350mb of ram, not counting the 130mb allocated to buffers and the 14mb or so unused, this is a LOT for a machine running only 65 processes, contrast to mine:
total used free shared buffers cached Mem: 773136 705768 67368 0 160736 255328 -/+ buffers/cache: 289704 483432 Swap: 393584 357840 35744 using about 289mb, and 350 swap (which is mostly my ramdisk/dev/shm that i use for/tmp) not only that, but this is a 64bit machine (alpha cpu) so average memory consumption for a given app should be higher... so your ram usage seems WAY too high, i assume your running kde or something? try a more lightweight wm and it should be a lot quicker
Well, X11 doesnt even use the tcp/ip transport when your running local apps, DISPLAY=:0 Ofcourse, if you set DISPLAY to 127.0.0.1:0 then it uses the loopback interface, and is a little sluggish, but why would you do that? When connecting to:0 it uses a shared memory system instead of tcp/ip, the beauty of X is the flexibility.. Whats more, gentoo seems to have disabled the tcp transport when i last updated X, dont know why, but it has.. i`l fix it when i have time to: bert@rocky:/dev/shm$ export DISPLAY=127.0.0.1:0 bert@rocky:/dev/shm$ xterm _X11TransSocketOpen: socket() failed for tcp _X11TransSocketOpenCOTSClient: Unable to open socket for tcp _X11TransOpen: transport open failed for tcp/127.0.0.1:0 xterm Xt error: Can't open display: 127.0.0.1:0
I cant see how yours would be more sluggish than this machine, in theory it should be faster, so i can only assume its either misconfigured or swapping.. what does glxgears -fps benchmark at anyway? and whats $DISPLAY set to?
X is not sluggish atall for me, on a 7yr old 600mhz box with a Permedia-2 8mb displaycard.. infact, its downright speedy. Are you sure X is configured correctly, and your not using the (default) framebuffer drivers, or a generic vga/vesa driver, both of which are very very slow, and used more often than you think, especially in gentoo/slackware which doesnt configure X for you, but rather leaves you with a default config (Framebuffer or vesa, i dont remember which) and makes you configure it properly yourself. Or perhaps your window manager is causing X to seem sluggish,try running something lightweight like windowmaker, and makesure the system isnt swapping as that will greatly increase interface latency. Or perhaps your using a very modern displaycard that has yet to be supported correctly? Either way, on another machine here, an Athlon XP 1900+ with a Radeon 64mb DDR (the first radeon card) X11 is more responsive than XP (i guess that new interface slows it down a bit) and on a par with win2k.
Why not launch DDoS attacks against the spammers themselves. Steal their ddos drones, hack the spam relays, take the true source ip`s of the spams offline, and take offline the websites that the spams promote. Companies will think twice about paying spammers to advertise them if instead of bringing in revenue in the form of customers, it drains revenue in the form of bandwidth costs and lost customers. Spammers also will think twice about their actions if their bandwidth costs go through the roof.
Consider how many machines were vulnerable to the dcom worms... Consider that the exploit was in the wild for several weeks before the worms went out... It`s safe to assume that people seriously interested in launching ddos attacks would have quickly begun compromising hosts as soon as they got hold of the exploit, and most likely patching those hosts against furthur compromise (to prevent the victims from cracking the hosts themselves and deleteing the ddos software) So, lets assume a spammer compromises 20,000 hosts, on a range of connections from dialups to 100mbit university connections and webservers on similarly quick lines... 20,000 is not many considering the millions of unpatched windows 2000/xp hosts connected to the internet, but 20,000 is more than enough to saturate any single datacenter, remember these machines will typically be sending floods at their maximum upstream rate, whereas a site like google usually handles relatively low traffic http transfers. As for tracking the IP`s, an attack would most likely be spoofed, or atleast spoofed as ip`s within the same local address range as the source.
AmigaOS always had ethernet support, just put your hydra.device or ariadne.device into devs: and away you go... As for TCP/IP support, commodore even made a tcp stack for it, called AS2525 or something... hardly anyone used it, opting instead for third party stacks, but it was there...
Because now it mentions a touchpad screen, so it`s applicable to ipaq users etc... Originally the faq only mentioned a keyboard, i remember reading it back when it was first posted up there.
Not so for wordperfect atleast, but not sure about the other apps... Wordperfect had a unix version, i believe for SCO.. available for years, and indeed the first linux version was a port from this.. and it always worked and looked much better than the newer wine-based versions
Well, in some ways.. a secret such as a gpg key or a password is also a form of obscurity, the difference is your trusting yourself to keep the secret and not a third party.. And knowing how the encryption works you know that your secret is the only key, and there is not a set of "master" keys that you dont know about.
Because theres often staff who take their laptops home, get infected, and then bring their infected laptops in to work, where they proceed to populate over the local network. As for software firewalls, i understand many of these cost money, or atleast cost money for commercial use.. So buying 500+ copies to install on every machine of a corporate network would put the accountants off
Actually, they made a port of WordPerfect 9 first, this was done from the SCO version... as an attempt to "test the water" so to speak, it was very much a native-feeling application, and ran pretty well... Later, they ported the entire corel office suite, including wordperfect 10.. but this was done entirely through wine.. the result? a dog slow unstable pile of trash with an inflexible non-native interface that didnt look right or interact properly with other programs, cut+paste never worked properly for instance.
It also shows they realise the poor quality of their own products. If they thought their product could compete on a level playing field with other OS`s, they would let it do so.. And then use this as a marketting tool, in an ideal world, products would have to sell themselves based on quality and overall value for money. Unfortunately, in today`s world, marketting campaigns and strongarm tactics keep people with the established players and raise the barrier of entry too high for a competitor to achieve any great success, remember.. if people dont know about your product, or dont know why its better than someone else`s, then they wont buy it.. and you`l make no money on it, and without money you cant afford the same kind of marketting campaigns as the established players. And end users generally won`t pay for a product they`ve never heard of, if there`s something more familiar available.. "better the devil you know" The only reason linux is gaining a foothold is because it can be downloaded for free, and thus theres nothing to lose by trying it... and secondly, theres no company behind it that needs to make profit in order for linux itself to succeed.
Because HP are dropping the tried and tested PA-RISC architecture in favor of the itanic, which is expensive, not fully compatible, untested, under performing, and seems already to have shown problems due to overly aggressive clocking. Contrast this with Sun, who are continuing to develop the Sparc architecture, the new UltraSparc processors will provide full backwards compatibility with even the first generation of sparc processors, and the new versions of the Solaris OS also provides backwards compatibilty, even with SunOS 4.x, which was pre-solaris and in many ways a totally different system.
Any large business is mostly concerned with profits rather than what`s better.. To them, something is better if it`s more profitable.. Altho, often the benefits to the business as a whole are second to the personal benefits to the decision makers.
$700M saved by the company, and thus accounted for and reinvested into the business, of $1M directly into the back pockets of the people making the decisions? which do you think the self-serving businessman will choose?
And it also opens you up to unstoppable floods of spam from those few companies, because it would be the only source of spam users would pay it more attention, and because it was from the mail provider users would be more likely to trust it.
Also, consider this... spamming is very profitable, so you can pretty much guarantee that spammers would want to pay these companies for the "right" to send spam to their users, ms already does this with hotmail.. tho it`s not publicly acknowleged. Providing a spammer with access to your thousands, even millions, of customers, could prove highly lucrative, virtually any business will jump at this opportunity to increase profits.
Unless you compile it yourself, you cannot be sure the sourcecode your reading is the same code that was used to build the binaries your running.. What, you trust microsoft to tell you that it is the same? maybe you trusted them when they told you that windows has no bugs, and how secure it is, and how easy it is to use and you wont need to hire skilled staff to administer it. This doesnt go together, sure a skilled admin can keep a windows box relatively secure, but this goes against their own marketting, if they want to market a product as being useable by unskilled individuals they have to make damn sure it`s secure by default.
As for the kernel, that was mostly stolen from DEC, the stability of the nt kernel has actually gone down since the days of 3.51, due to all the crap microsoft has shoehorned into it.
Also, what has blaster got to do with sql? blaster attains SYSTEM priveleges regardless of how you configure sql, the two are totally unrelated.. it was slapper that exploited sql, and that worm was months ago. And even running as an unpriveleged user, its possible for a worm to scan and infect other computers, send email, and do all manner of nasty things on the network resulting in angry people targetting your box.
So, despite the increased difficulty of finding vulnerabilities in a closed source program, windows/ie/outlook/iis have had more flaws discovered in them than equivalent opensource applications.. This would suggest that there are a far higher number of vulnerabilities existing in these applications, just waiting to be discovered. Remember what they said a while ago, microsoft code is so fundamentally flawed it would be a risk to national security to disclose it.. What if a cracker breaks into their systems and steals the sourcecode? he will be able to find MANY holes which arent easily findable by poking with the binaries, and ofcourse a blackhat who breaks in and steals sourcecode is unlikely to disclose vulnerabilities he finds, hes far more likely to exploit them and steal other things.
Slashdot Mirror
95 used to take ages on a p233, but 3.11 started almost instantly
So.. your machine has started on the swap, what windowmanager is running? have you tried something lightweight like windowmaker ? KDE/Gnome are very good at dragging the overall performance of X down, and if the windowmanager is sluggish it will make every app running under it seem sluggish too.. /dev/shm that i use for /tmp) not only that, but this is a 64bit machine (alpha cpu) so average memory consumption for a given app should be higher...
X isn`t sluggish atall here, and the machine:
top - 15:35:17 up 22 days, 5:14, 28 users, load average: 2.95, 2.89, 2.87
Tasks: 162 total, 4 running, 157 sleeping, 1 stopped, 0 zombie
Cpu(s): 75.8% user, 20.7% system, 3.5% nice, 0.0% idle
Mem: 773136k total, 700144k used, 72992k free, 160160k buffers
Swap: 393584k total, 357856k used, 35728k free, 261656k cached
has an old 8mb permedia-2 gfxcard, running in 1024x768 32bpp, even under this load (caused by multiple compiles, courtesy of emerge -u world) it`s not making a noticeable difference to the performance of X, windowmaker is very swift, and the only time it used to slow down was when i had considerably less ram, and parts of X/windowmaker got swapped out.
Your system is using 350mb of ram, not counting the 130mb allocated to buffers and the 14mb or so unused, this is a LOT for a machine running only 65 processes, contrast to mine:
total used free shared buffers cached
Mem: 773136 705768 67368 0 160736 255328
-/+ buffers/cache: 289704 483432
Swap: 393584 357840 35744
using about 289mb, and 350 swap (which is mostly my ramdisk
so your ram usage seems WAY too high, i assume your running kde or something? try a more lightweight wm and it should be a lot quicker
Well, X11 doesnt even use the tcp/ip transport when your running local apps, DISPLAY=:0 :0 it uses a shared memory system instead of tcp/ip, the beauty of X is the flexibility..
Ofcourse, if you set DISPLAY to 127.0.0.1:0 then it uses the loopback interface, and is a little sluggish, but why would you do that?
When connecting to
Whats more, gentoo seems to have disabled the tcp transport when i last updated X, dont know why, but it has.. i`l fix it when i have time to:
bert@rocky:/dev/shm$ export DISPLAY=127.0.0.1:0
bert@rocky:/dev/shm$ xterm
_X11TransSocketOpen: socket() failed for tcp
_X11TransSocketOpenCOTSClient: Unable to open socket for tcp
_X11TransOpen: transport open failed for tcp/127.0.0.1:0
xterm Xt error: Can't open display: 127.0.0.1:0
I cant see how yours would be more sluggish than this machine, in theory it should be faster, so i can only assume its either misconfigured or swapping.. what does glxgears -fps benchmark at anyway? and whats $DISPLAY set to?
X is not sluggish atall for me, on a 7yr old 600mhz box with a Permedia-2 8mb displaycard.. infact, its downright speedy.
Are you sure X is configured correctly, and your not using the (default) framebuffer drivers, or a generic vga/vesa driver, both of which are very very slow, and used more often than you think, especially in gentoo/slackware which doesnt configure X for you, but rather leaves you with a default config (Framebuffer or vesa, i dont remember which) and makes you configure it properly yourself.
Or perhaps your window manager is causing X to seem sluggish,try running something lightweight like windowmaker, and makesure the system isnt swapping as that will greatly increase interface latency.
Or perhaps your using a very modern displaycard that has yet to be supported correctly?
Either way, on another machine here, an Athlon XP 1900+ with a Radeon 64mb DDR (the first radeon card) X11 is more responsive than XP (i guess that new interface slows it down a bit) and on a par with win2k.
So, order a windows server, and then install linux onto it yourself, easy.. let ms subsidise your server
Why not launch DDoS attacks against the spammers themselves. Steal their ddos drones, hack the spam relays, take the true source ip`s of the spams offline, and take offline the websites that the spams promote.
Companies will think twice about paying spammers to advertise them if instead of bringing in revenue in the form of customers, it drains revenue in the form of bandwidth costs and lost customers.
Spammers also will think twice about their actions if their bandwidth costs go through the roof.
Consider how many machines were vulnerable to the dcom worms...
Consider that the exploit was in the wild for several weeks before the worms went out...
It`s safe to assume that people seriously interested in launching ddos attacks would have quickly begun compromising hosts as soon as they got hold of the exploit, and most likely patching those hosts against furthur compromise (to prevent the victims from cracking the hosts themselves and deleteing the ddos software)
So, lets assume a spammer compromises 20,000 hosts, on a range of connections from dialups to 100mbit university connections and webservers on similarly quick lines... 20,000 is not many considering the millions of unpatched windows 2000/xp hosts connected to the internet, but 20,000 is more than enough to saturate any single datacenter, remember these machines will typically be sending floods at their maximum upstream rate, whereas a site like google usually handles relatively low traffic http transfers.
As for tracking the IP`s, an attack would most likely be spoofed, or atleast spoofed as ip`s within the same local address range as the source.
AmigaOS always had ethernet support, just put your hydra.device or ariadne.device into devs: and away you go...
As for TCP/IP support, commodore even made a tcp stack for it, called AS2525 or something... hardly anyone used it, opting instead for third party stacks, but it was there...
Because now it mentions a touchpad screen, so it`s applicable to ipaq users etc... Originally the faq only mentioned a keyboard, i remember reading it back when it was first posted up there.
Not so for wordperfect atleast, but not sure about the other apps...
Wordperfect had a unix version, i believe for SCO.. available for years, and indeed the first linux version was a port from this.. and it always worked and looked much better than the newer wine-based versions
The colour screens are also much harder to read on bright sunny days, the old green text on black were fine
OpenBSD is all about hype and arrogance rather than actual coding work.. hence the overflows discovered.
Well, the bible is the largest selling book ever, but the number two spot goes to "Mein Kampf" the book written by Adolf Hitler.
Well, in some ways.. a secret such as a gpg key or a password is also a form of obscurity, the difference is your trusting yourself to keep the secret and not a third party.. And knowing how the encryption works you know that your secret is the only key, and there is not a set of "master" keys that you dont know about.
Because theres often staff who take their laptops home, get infected, and then bring their infected laptops in to work, where they proceed to populate over the local network.
As for software firewalls, i understand many of these cost money, or atleast cost money for commercial use.. So buying 500+ copies to install on every machine of a corporate network would put the accountants off
Actually, they made a port of WordPerfect 9 first, this was done from the SCO version... as an attempt to "test the water" so to speak, it was very much a native-feeling application, and ran pretty well...
Later, they ported the entire corel office suite, including wordperfect 10.. but this was done entirely through wine.. the result? a dog slow unstable pile of trash with an inflexible non-native interface that didnt look right or interact properly with other programs, cut+paste never worked properly for instance.
Or perhaps it caches the content in an attempt to reduce server load, and periodically updates the cache.
It also shows they realise the poor quality of their own products. If they thought their product could compete on a level playing field with other OS`s, they would let it do so.. And then use this as a marketting tool, in an ideal world, products would have to sell themselves based on quality and overall value for money. Unfortunately, in today`s world, marketting campaigns and strongarm tactics keep people with the established players and raise the barrier of entry too high for a competitor to achieve any great success, remember.. if people dont know about your product, or dont know why its better than someone else`s, then they wont buy it.. and you`l make no money on it, and without money you cant afford the same kind of marketting campaigns as the established players. And end users generally won`t pay for a product they`ve never heard of, if there`s something more familiar available.. "better the devil you know"
The only reason linux is gaining a foothold is because it can be downloaded for free, and thus theres nothing to lose by trying it... and secondly, theres no company behind it that needs to make profit in order for linux itself to succeed.
Well, we could always go and mass spam their site around.. and get their hosting provider to close them down
Because HP are dropping the tried and tested PA-RISC architecture in favor of the itanic, which is expensive, not fully compatible, untested, under performing, and seems already to have shown problems due to overly aggressive clocking.
Contrast this with Sun, who are continuing to develop the Sparc architecture, the new UltraSparc processors will provide full backwards compatibility with even the first generation of sparc processors, and the new versions of the Solaris OS also provides backwards compatibilty, even with SunOS 4.x, which was pre-solaris and in many ways a totally different system.
Any large business is mostly concerned with profits rather than what`s better.. To them, something is better if it`s more profitable.. Altho, often the benefits to the business as a whole are second to the personal benefits to the decision makers.
$700M saved by the company, and thus accounted for and reinvested into the business, of $1M directly into the back pockets of the people making the decisions? which do you think the self-serving businessman will choose?
And it also opens you up to unstoppable floods of spam from those few companies, because it would be the only source of spam users would pay it more attention, and because it was from the mail provider users would be more likely to trust it.
Also, consider this... spamming is very profitable, so you can pretty much guarantee that spammers would want to pay these companies for the "right" to send spam to their users, ms already does this with hotmail.. tho it`s not publicly acknowleged. Providing a spammer with access to your thousands, even millions, of customers, could prove highly lucrative, virtually any business will jump at this opportunity to increase profits.
Unless you compile it yourself, you cannot be sure the sourcecode your reading is the same code that was used to build the binaries your running.. What, you trust microsoft to tell you that it is the same? maybe you trusted them when they told you that windows has no bugs, and how secure it is, and how easy it is to use and you wont need to hire skilled staff to administer it. This doesnt go together, sure a skilled admin can keep a windows box relatively secure, but this goes against their own marketting, if they want to market a product as being useable by unskilled individuals they have to make damn sure it`s secure by default.
As for the kernel, that was mostly stolen from DEC, the stability of the nt kernel has actually gone down since the days of 3.51, due to all the crap microsoft has shoehorned into it.
Also, what has blaster got to do with sql? blaster attains SYSTEM priveleges regardless of how you configure sql, the two are totally unrelated.. it was slapper that exploited sql, and that worm was months ago.
And even running as an unpriveleged user, its possible for a worm to scan and infect other computers, send email, and do all manner of nasty things on the network resulting in angry people targetting your box.
So, despite the increased difficulty of finding vulnerabilities in a closed source program, windows/ie/outlook/iis have had more flaws discovered in them than equivalent opensource applications.. This would suggest that there are a far higher number of vulnerabilities existing in these applications, just waiting to be discovered.
Remember what they said a while ago, microsoft code is so fundamentally flawed it would be a risk to national security to disclose it.. What if a cracker breaks into their systems and steals the sourcecode? he will be able to find MANY holes which arent easily findable by poking with the binaries, and ofcourse a blackhat who breaks in and steals sourcecode is unlikely to disclose vulnerabilities he finds, hes far more likely to exploit them and steal other things.