tiny linux kernel, compiled overnight on 1990 hardware, hah, unless it took 48 hours it was not a real program. Try 'make world' on same era hardware and lets compare compile time between your 'hello world' and a mans app.
Well, I dont know about you guys but my 'vi' sed scripts run really fast nowadays also 'make world' runs really really fast compared to when i had a sun3/180 at home.
There is no such law.
He asked for your passport because you wanted to pay with creditcard and he though you look like the kind of guy that might try to present him a stolen card.
I have been to some 30 different countries in business and private travel and the us is the only place that routinely asks for and copies your passport.
I never showed any passport or ID when staying at hotels in Spain, Portugal, Italy, France, Germany, Greece, Slovenja, Belgium, Holland, Denmark, Finland, Norway, Sweden, Ireland, Austria, Poland, Switzerland.
I did show my passport once in England though.
I paid ~1000USD for my first Sony DVD player in ~1998 and that at a time when there were like 5 movies available in region 2.
One of which was Jumanji but what the hell, the kids love jumanji. And I could get a region chip for only ~100USD.
Now I can buy a new no-name region free DVD player in australia for about ~40USD. Times change.
"or add signatures in their virus definitions for"
I am still waiting for when someone studies and figures out how a virus scanner works and adapts the virus to it.
For example, does anyone think a virus scanner will scan the entire executable when scanning a file?
No, that would take ages, what the scanner instead does is reading the first x kb of the exe and the last x kb of the exe to check if someone has been modifying any of the tables for load-linking. And most virii gets added to the end of the exe since that is the easiest place to put the virii.
What a virus would need to do is to understand the filestructure of an exe and insert itself somewhere random in the middle of the exe, where the scanners never check.
This would require the virus to be more intelligent and understand the exe filestructure to modify it and add a block of code to somewhere in the middle of the file.
The only solution to that is to change virus scanners to start scanning the entire file instead of just a few kb at select places. And you think todays virus scanners make the computer slow...
Well,
MPAA has never been in Beijing or most of Asia, that is for sure.
As someone that lives in Asia I can guarantee you that the movies available in the corner store in mainland china (or most Asian countries except Thailand) is a perfect DVD copy including all the DVD extras etc made from master copies for distribution for preescreening.
Trying to flog a camcorder copy with bleeding colors and non-perfect 5.1 sound will get you laughed at in Beijing and Shanghai.
I can not even imagine how crappy the quality would be if filmed by a camcorder. Doesnt it make your eyes bleed?
(the Exception Thailand: there in the touristy areas there seems to be popular to sell incredibly poor quality camcorder crap. Totally unviewable)
"In an IDE (interactive!), adding/* TODO */ comments or suchlike are
automatically, and in real-time, added to a task list. When comments are removed
or the task is clicked off on the GUI (and possibly in combination with revision
control) you can see what stuff has been done and has not been done. In real
time. From an IDE."
Whatever.
Ive been doing this in my IDE for 15 years.
Add/*XXX something */ wherever there is something for later to do.
Cycle through the TODO list using M^ ^S XXX ^S*
Big deal.
overflowing the mac table in the switch is one way to sniff a switch as you say, and a managed switch will send an SNMP trap to that. correct.
A much easier way to sniff a switch, which you will NOT get an SNMP trap for is using ETTERCAP or HUNT.
Or do you mean that you dont use ARP on your network? Isnt it a lot of work keeping all the/etc/ethers file up to date?
no it is a switch inside it, a two port switch connectiong the two hubs.
you can easily see this if you connect 4 computers to the device. 3 of them at 100mbit and the fourth at 10mbit.
ping one 100mbit host from another 100mbit host. Capture the traffic from the third 100mbit host.
you can see the ping, hence you are on the same collission domain as the others i.e. its a 100mbit hub.
at the same time capture from the 10mbit host as well.
notice you will not see the packets? thats because the 10mbit host sits on a different collission domain.
that is not true for a hub. a hub will make all devices run at the slowest speed.
what you are thinking of is a s.c. dual-speed hub / switched hub 10/100 hub .
these devices are not hubs.
they consist of :
one 10mbit/s hub
one 100mbit/s hub
inside the same enclosure connected together internally by a 2 port switch.
a hub is a layer-1 device it creates a broadcast domain and a collission domain that is the set of all physical ports on the device.
a switch is a layer-2 device it creates/emulates the same broadcast domain as a hub would but creates one independent collisison domain (physical link) on each port.
Since only one nic is attached to one port/one physical domain collissions can not happen and thus the collission detection circuitry is disabled hence full duplex.
Dont be silly. Why do you want it to end?
When it does, all this excellent entertainment will dissaepar down memory lane.
Actually, I am disturbed and dissapointed by the recent lack of pressreleases issued by SCO. Can they not afford pressreleases anymore?
Lets set up a fund and collect money for SCO, money earmarked so that SCO can only use it to pay for pressreleases.
I worry a lot, for the day when SCO will no longer ease.
With the exception of AbsolutelyFabolous, this has been the most entertaining and funniest thing I have ever seen in the last few years.
Remember all the times we have collectively rolled around on the floor laughing at something Darl has said. He is indeed a genious of comedy.
I dont want to sit here 1 year from now and say "remember when SCO was still around. Didnt we have a great time then laughing at all those crack hallucinations. I havent had a really good laygh since they died. If only I had contributed money to the keep-sco-alive fund. If only I knew how boring life would be without sco..."
Do you want to risk that? No I thought so. Keep sending in your donations to SCO.
I think we should start a fund to collect money for SCO instead.
Look, they are running out of money. When they do, the entertainment will cease.
As all good entertainment, tv or elsewhere, for a really good quality laugh you have to put your money on the table.
Think about it for a while, these guys have enlightened our world for 1 year plus, and are still going strong. We have cried, we have laughed but most of all we have had a good time. Now we must step up and save them or we will go back to just boring it news once again.
Application response time is a littel bit more than just "data passed in direction X on the session" "other data passed in direction Y" measure the delta.
For single threaded apps thats fine, but explain how that can possibly work with NFS where the normal case is multithreaded commands?
Try Analyze/ServiceResponseTime for ONCRPC/NFS in ethereal and you will see why that other product sucks. It cant even tell me the difference in response times between GETATTRs and WRITEs. Pathetic.
As for SMB and CIFS. Why can that other tool not even dissect SAMR or LSA? Why do NAI themself use ethereal in their own classes ina dvanced active directory and cifs training they give to customers? Can it be because something that cant even dissect basic DCERPC services such as SAMR/LSA/NETLOGON is pretty useless when it comes to post-netbios cifs traffic?
How do I enable SAMR LSA and NETLOGON dissection in that other tool? i would like to know since it would help me find authentication problems in my domain. I would also like to reassemble the full DCERPC PDUs for large ones so i can easily see the response code (last 4 bytes) for large PDUs spanning multiple smb transaction/read/write/readandx/writeandx packets.
i really would like to know how to enable this. it would really help me troubleshoot and diagnose problems!
Bloody h*ll, that pos cant even decode the SMB protocol properly or DCERPC.
Hello world? Calling? SMB and DCERPC are the two MOST common protocol families in the world? Whats the bloody point in a sniffer if it cant even decode windows protocols?
Too bad then that you couldnt figure out how to code sniffer so that it would even be able to detect simple retransmissions on the tcp layer reliably.
Or why not implement something like service response times that actual professionals use every day to diagnose performance problems.
Etherela is a protocol analyzer. Sniffer is a network monitoring tool. If all you want is a pretty picture of who downloads the most porn, well, that wont help you when your NFS clients starts stalling due to lock manager problems.
Slashdot Mirror
Is that you Bill?
Question: "My question is what do they use to manage such a huge complex code base?" Answer: flamefest-fuel
tiny linux kernel, compiled overnight on 1990 hardware, hah, unless it took 48 hours it was not a real program. Try 'make world' on same era hardware and lets compare compile time between your 'hello world' and a mans app.
Well, I dont know about you guys but my 'vi' sed scripts run really fast nowadays also 'make world' runs really really fast compared to when i had a sun3/180 at home.
Oh, thats not entirely correct. Hotels in mainland china ALWAYS take your passport for a day or two.
There is no such law. He asked for your passport because you wanted to pay with creditcard and he though you look like the kind of guy that might try to present him a stolen card. I have been to some 30 different countries in business and private travel and the us is the only place that routinely asks for and copies your passport. I never showed any passport or ID when staying at hotels in Spain, Portugal, Italy, France, Germany, Greece, Slovenja, Belgium, Holland, Denmark, Finland, Norway, Sweden, Ireland, Austria, Poland, Switzerland. I did show my passport once in England though.
You can get a PC today for 50USD excluding monitor.
Problem is, it wont play doom3 very well.
Problem solved.
You just sell 100 milion players instead and give the extra 99milion to the shareholders in dividends.
You think that is bad?
I paid ~1000USD for my first Sony DVD player in ~1998 and that at a time when there were like 5 movies available in region 2.
One of which was Jumanji but what the hell, the kids love jumanji. And I could get a region chip for only ~100USD.
Now I can buy a new no-name region free DVD player in australia for about ~40USD. Times change.
You also have an S3, i love that chip.
Do you have, as I have, it on a VesaLocal bus card, (high end) , or are you using a low-end ISA card?
"or add signatures in their virus definitions for"
...
I am still waiting for when someone studies and figures out how a virus scanner works and adapts the virus to it.
For example, does anyone think a virus scanner will scan the entire executable when scanning a file?
No, that would take ages, what the scanner instead does is reading the first x kb of the exe and the last x kb of the exe to check if someone has been modifying any of the tables for load-linking. And most virii gets added to the end of the exe since that is the easiest place to put the virii.
What a virus would need to do is to understand the filestructure of an exe and insert itself somewhere random in the middle of the exe, where the scanners never check.
This would require the virus to be more intelligent and understand the exe filestructure to modify it and add a block of code to somewhere in the middle of the file.
The only solution to that is to change virus scanners to start scanning the entire file instead of just a few kb at select places.
And you think todays virus scanners make the computer slow
Well,
MPAA has never been in Beijing or most of Asia, that is for sure.
As someone that lives in Asia I can guarantee you that the movies available in the corner store in mainland china (or most Asian countries except Thailand) is a perfect DVD copy including all the DVD extras etc made from master copies for distribution for preescreening.
Trying to flog a camcorder copy with bleeding colors and non-perfect 5.1 sound will get you laughed at in Beijing and Shanghai.
I can not even imagine how crappy the quality would be if filmed by a camcorder. Doesnt it make your eyes bleed?
(the Exception Thailand: there in the touristy areas there seems to be popular to sell incredibly poor quality camcorder crap. Totally unviewable)
So what.
TODO FIXME XXX and QQQ has been in my sourcecode and my editor/ide/os for at least 15+ years.
Use M^lessthan ^S XXX ^S* to cycle through them
"In an IDE (interactive!), adding /* TODO */ comments or suchlike are
automatically, and in real-time, added to a task list. When comments are removed
or the task is clicked off on the GUI (and possibly in combination with revision
control) you can see what stuff has been done and has not been done. In real
time. From an IDE."
Whatever.
Ive been doing this in my IDE for 15 years.
Add /*XXX something */ wherever there is something for later to do.
Cycle through the TODO list using M^ ^S XXX ^S*
Big deal.
overflowing the mac table in the switch is one way to sniff a switch as you say, and a managed switch will send an SNMP trap to that. correct. A much easier way to sniff a switch, which you will NOT get an SNMP trap for is using ETTERCAP or HUNT. Or do you mean that you dont use ARP on your network? Isnt it a lot of work keeping all the /etc/ethers file up to date?
no it is a switch inside it, a two port switch connectiong the two hubs. you can easily see this if you connect 4 computers to the device. 3 of them at 100mbit and the fourth at 10mbit. ping one 100mbit host from another 100mbit host. Capture the traffic from the third 100mbit host. you can see the ping, hence you are on the same collission domain as the others i.e. its a 100mbit hub. at the same time capture from the 10mbit host as well. notice you will not see the packets? thats because the 10mbit host sits on a different collission domain.
that is not true for a hub. a hub will make all devices run at the slowest speed. what you are thinking of is a s.c. dual-speed hub / switched hub 10/100 hub . these devices are not hubs. they consist of : one 10mbit/s hub one 100mbit/s hub inside the same enclosure connected together internally by a 2 port switch. a hub is a layer-1 device it creates a broadcast domain and a collission domain that is the set of all physical ports on the device. a switch is a layer-2 device it creates/emulates the same broadcast domain as a hub would but creates one independent collisison domain (physical link) on each port. Since only one nic is attached to one port/one physical domain collissions can not happen and thus the collission detection circuitry is disabled hence full duplex.
there is absolutely no difference in security between a hub and a switch.
Dont be silly. Why do you want it to end? When it does, all this excellent entertainment will dissaepar down memory lane. Actually, I am disturbed and dissapointed by the recent lack of pressreleases issued by SCO. Can they not afford pressreleases anymore? Lets set up a fund and collect money for SCO, money earmarked so that SCO can only use it to pay for pressreleases.
I worry a lot, for the day when SCO will no longer ease. With the exception of AbsolutelyFabolous, this has been the most entertaining and funniest thing I have ever seen in the last few years. Remember all the times we have collectively rolled around on the floor laughing at something Darl has said. He is indeed a genious of comedy. I dont want to sit here 1 year from now and say "remember when SCO was still around. Didnt we have a great time then laughing at all those crack hallucinations. I havent had a really good laygh since they died. If only I had contributed money to the keep-sco-alive fund. If only I knew how boring life would be without sco..." Do you want to risk that? No I thought so. Keep sending in your donations to SCO.
I think we should start a fund to collect money for SCO instead. Look, they are running out of money. When they do, the entertainment will cease. As all good entertainment, tv or elsewhere, for a really good quality laugh you have to put your money on the table. Think about it for a while, these guys have enlightened our world for 1 year plus, and are still going strong. We have cried, we have laughed but most of all we have had a good time. Now we must step up and save them or we will go back to just boring it news once again.
Application response time is a littel bit more than just "data passed in direction X on the session" "other data passed in direction Y" measure the delta.
For single threaded apps thats fine, but explain how that can possibly work with NFS where the normal case is multithreaded commands?
Try Analyze/ServiceResponseTime for ONCRPC/NFS in ethereal and you will see why that other product sucks.
It cant even tell me the difference in response times between GETATTRs and WRITEs. Pathetic.
As for SMB and CIFS. Why can that other tool not even dissect SAMR or LSA? Why do NAI themself use ethereal in their own classes ina dvanced active directory and cifs training they give to customers? Can it be because something that cant even dissect basic DCERPC services such as SAMR/LSA/NETLOGON is pretty useless when it comes to post-netbios cifs traffic?
How do I enable SAMR LSA and NETLOGON dissection in that other tool? i would like to know since it would help me find authentication problems in my domain. I would also like to reassemble the full DCERPC PDUs for large ones so i can easily see the response code (last 4 bytes) for large PDUs spanning multiple smb transaction/read/write/readandx/writeandx packets.
i really would like to know how to enable this. it would really help me troubleshoot and diagnose problems!
Bloody h*ll, that pos cant even decode the SMB protocol properly or DCERPC. Hello world? Calling? SMB and DCERPC are the two MOST common protocol families in the world? Whats the bloody point in a sniffer if it cant even decode windows protocols?
Too bad then that you couldnt figure out how to code sniffer so that it would even be able to detect simple retransmissions on the tcp layer reliably. Or why not implement something like service response times that actual professionals use every day to diagnose performance problems. Etherela is a protocol analyzer. Sniffer is a network monitoring tool. If all you want is a pretty picture of who downloads the most porn, well, that wont help you when your NFS clients starts stalling due to lock manager problems.
I only use ethereal. In true unix style, if you want a proper tool that works you have to write it yourself.