Hmm. I'd say "operator error". I just read the specs from HP, and indeed "Linux" is not listed as a client OS. However, I suspect that this is an omission in the specs than an indication that they don't really run "lpd" on the machine.
It's pretty clear. They took the cheapest hardware they could find, put Linux on it and are selling it. It would cost EFFORT to make it not work with Linux as a client OS, as that's standard. My bet is they just use lpd (Confirmed!), and it would cost effort to prevent it from working....
Hmm. Disproved here.
Now that I can decide what projects to take and which I won't, I'm no longer taking projects involving microsoft.
However, during your education, getting to know how to write portable code is really a very important thing.
The only way to learn writing portable code is to port your own code a few times, so that you figure out what constructs are counter-productive, as they don't port easily.
So my recommendation: Write the code on Linux/g++, and then "port" the code to whatever the boss wants.
There are two articles in my current Scientific american about SETI. The first was pretty pro-SETI. The second one put things in perspective:
Extra terristrial societies that have found a way to capture ALL the energy from their sun and use all that to broadcast "here we are", can be detected across about half the galaxy. Well, we can assume that this is simply not possible. Read "ringworld" by Niven, and do the math.
Societies like ours, who are happily minding our own business, but leaking some radiation into the surrounding space can be detected from about 100 light-years. There are a handfull of stars in that area....
An ordinary cubic kilometer of ocean water contains 14 tons of gold (this was on the radio again yesterday). However, it's uneconomic to go and get all that gold from the ocean.
Assume that the concentration is similar to the gold in the ocean. So then the size of the moon can be calculated as siomilar to a block of 100x100x10 km. That's way smaller than the real thing, so we have just concluded that the concentration is WAY smaller!
Extracting it is not trivial.
So, where everybody can just drive over and put a plant on the nearest ocean shore to go and extract the gold from the ocean, you're expecting people to build a plant 300000 km away and haul the stuff back here, and make a profit? No way!
Getting it here is not trivial.
By the way, I must tell you that I know of an even larger store of fusion-fuel just a bit further away. The concentrations are much more manageable! Problem this time is that the stuff is fusing already....
On the + side you get oxygen. On the - side, you get hydrogen. Burn the hydrogen, and you get presto chango... Water.
I hate to break this to ya, but if you electrolyse normal salty water, you get hydrogen and Chloride.
Although the potentials would indicate that the oxygen would form before the chloride, somehow there is ONE exception to the "potentials can predict what products will form" and this is it.
But, yes, you can make hydrogen in your kitchen with this. Makes pretty little bangs if you light it. And if you do it professionally you can certainly produce the oxygen too, but I'm not sure how. (No salt-> low conductivity -> higher voltages? Different salt?)
Remember, it's not the volts, nor the amps that kill you. It's the two together.
Not really. It's the amps that kill you. In fact about 20mA of them is the "critical" threshold. Note that you can drive much more if it only goes through an arm or something. The heart is the critical organ here.
However the body has resistance. Normal hand-to-other-hand resistance is on the order of 100k -> About 2000V is deadly.
If your hands are wet that resistance drops quickly, and the deadly voltage can reach "home voltages" (110/200).
Oh. You don't really do much "permanent" damage with say a 50ma current through the heart. So after a few seconds of "relaxation", the heart would work again just fine, but it just needs some external impulse to get started.
My father was working in the garden with an electrical drill. Seems the outside of the drill was under tension, which he didn't notice until he put his knee in a puddle of mud. He managed to put out a yelp, which alerted my mother who teaches CPR. No pulse, no breathing. CPR -> Dad's still alive.
This does not open up any more remote exploitable holes, but rather makes it give you root rather then your "nobody" user when you break a program like sendmail that uses this sort of security
No. Its worse than that. You can specify programs that sendmail should run in CF files, you can specify an alternate CF file. Sendmail tries to drop privs if you do these things. It (dropping privs) doesn't work in 2.2.15 if you toggle the bits correctly....
The company notifies eBay of suspect auctions and asks them to terminate the auctions within 24 hours.
And Microsoft is perfectly allowed to do so.
If all they do is ask politely, that's perfectly legal. I could ask slashdot to remove this story, and that's legally just fine. If for whatever reason, slashdot decides to honor my request, that's their decision.
So, it seems that Ebay is responding to Microsoft's polite requests. The Ebay usage policy probably states that Ebay can pull any auction they want to pull. In that case, there is just about nothing you can do about this. Sorry.
The problem with DCMA kinds of acts is that they make the ISP responsible if they don't "take you down" after getting notified of "copyright violations".
The problem with the internet, whatever method you choose, is that you'll always need an ISP.
Suppose you put your server on a boat in the ocean. You'll need an ISP to hook it up to the internet. THEY have a "home country" and when they get threatened by lawyers talking their legal language, they'll pull your plug really fast, even if you're in international waters. Sure, nobody could sue you in the end, but your ISP IS land-based.
Same holds when you have your own server behind a T1 in your home. Your ISP can still be forced to pull the plug on you. They just firewall off your port 80, and once you move your server, they fireall you off alltogether. Then what?
What law would you say it could possibly violate? Remember we don't have things like UCITA and DCMA here [yet].
If I remember correctly, there is an English court ruling effectively with the same results as the DCMA: If a copyright holder informs an ISP of a copyright violation, the ISP should take appropriate action.
On the subject of "lawyer fees", it would be fun to send a few law-students into a fight over this with the MPAA (or whomever). Costs the MPAA real money, the university has a chance of training the students "for real", with a very, very small chance of losing the case....
However, once a software house is taking out full page adverts claiming that their software will be good for a company, they have an obligation to uphold those claims and make sure the software works.
I think that this is the right "track".
If you pay for software, you can expect it to do roughly what it was supposed to do. You should NOT be able to sue the company for stuff that simply doesn't do what you expected it to do, but if the software seriously malfunctions, liability starts.
In my opinion, the claimant has a good case: They had a history of using the software, and knew it would work.
The software company also showed "gross negligence" in finding a bug and not notifying their customers. Sure the "disclaim all liability" clause is valid, and software is known to have bugs, but if as a software manufacturer you KNOW the software has a serious bug and not act upon it by sending your customer base a new version, then you're taking on resposibility.
Remember those floppy drives a wile back? They caused data corruption in certain cases. That's OK. But continuing to ship them for 10 years is NOT.
If micorsoft ships a version of word which occasionally changes "dear boss" into "flatulent jerk", that's an honest bug. When they have been notified about this and don't take efforts to correct the situation, then they should be liable.
In microsoft's case a few ads in big papers stating that this bug has been fixed in the next version may be enough. In this case the software producer had an address of the customer, and should've contacted them.
One of the important things is that a (G)UI is good for both beginners and experts. This means that an expert can use keyboard accellerators and a beginner should be able to use the menus.
But that's not all: Just binding keyboard acellerators to the most often used menu-options doesn't cut it: The transition from "beginner" to "expert" should be smoothened by having the keyboard accellerators show in the menus.
Not even that is served from the root servers. All the root servers serve is IP addresses of the nameservers for the domain of the host being looked up, its up to the domains nameservers to deal out any actual IP's, including for their own domain.
You look up marthastewart.com, your nameserver asks one of the root nameservers where the nameservers for marthastewart.com is, it then asks them for the IP to marthastewart.com.
Not even THAT is asked from the root server. If I'm not mistaken, your dns server asks for the address of "www.marthastewart.com" and gets the reply "you should be asking ns.marthastewart.com and before you'll ask me, the IP of ns.marthastewart.com is: x.y.z.w".
the price is too high, but that's something more easily changeable.)
If you invest $7000M, you'll need to take about $350M per year in interest payments into account for the calculation of your "operating costs".
If you manage to buy the network of satelites from the bankrupt iridium company for say $7000, you need to factor in about $350 per year in interest payments on that. Suddenly the actual operating costs of the satelites (manning the ground-stations, interconnect to PSTN) overwhelms the costs in operating the thing.
So, the original owners are looking at a minimum of $350M of operational costs per year, and selling services on that network for "cheap" is simply not an option.
I think larger L1 cache will compensate it. BTW - it's strange how people only keep in mind L2 cache when talking about caches. Considering that L1 cache does 90% (or so) of the speeding job.
Not sure where you are getting your numbers on this one, [...]
The rule-of-thumb is that every level of cache catches (hits) about 90% of the accesses thrown at it. This means that for 100 accesses, you spend: 90*[time to access L1] + 9 * [time to access L2] + 1 * [time to access Main RAM] . (And so the average is 0.01 times that. )
So, no L2 speed does not really matter all that much. Sure you can generate a benchmark that fits in L2 cache, but not in L1. Then it will depend highly on the L2 speed. But in general, the on-package L2's are good enough.
A heavy-lift booster in this class could throw a pretty good-sized payload to Mars for a "Mars Direct" type of manned mission.
Ehhhhm. One problem: The "heavy lift" of these guys lifts around 37400 pounds into LEO (low earth orbit). That's almost 17 tons.
Compare that to space shuttle: 23 tons, magnum: 80 tons (The rocket they are developing for mars direct), and Saturn V: > 130 tons (the requirement for a mars-direct type mission).
I would recommend setting up a site where users, professors, kernel developers and students can enter projects.
Then you have to get people to enter projects in the database, which can be done by different types of students: Small assignment for an OS course: small projects. Graduating projects: larger projects.
With a bit of hacking, jitterbug could handle this kind of job nicely. Professors wanting to "sort" requests to directories that they approve or disapprove for a certain assignment, can be given an account on jitterbug.... (The hacking is required to allow different professors to have different ratings for the projects)
Is a 36 inch, 200 ppi display. Of course, we wouldn't have any chance of taking advantage of that resolution at that size.
[...] I think that we should be concentrating on increasing the bandwidth that we can send to the monitor. Why not run a fiber-optic cable from your computer to your monitor? Put that SCSI interface to work providing you with the bandwidth you need, to your monitor, not your CD-ROM!
I just did the math. You need to pump about 5G per second to your 36 inch 200 ppi screen to get a 70Hz refresh rate. You wanted to use your spare 80Mb per second SCSI controller for that? You need 65 of them to achieve enough bandwidth. Ah, you have your CDROM drive connected to your 160Mbps lvd SCSI chain. Ah. that changes things. You need only 32 of them to drive a display like that.
No, SCSI and video hardware don't really compare in the transfer rates.
They have a display made from four displays 1/4 the size, but would it be more or less expensive to combine 9 1/9 size or 16 1/16 size displays?
My guess is that they need one edge at the top-or-bottom and one at the left-or-right to connect the panel. This makes 4 the max.
My guess is that they have something like half a pixel or a whole pixel of empty space between the two screens. The adjacent pixels need to be a bit brighter to compensate. You need to be at a sufficient distance not to notice.
it sucks; almost everything you can get wrong with a spam law is wrong in it.
No. To me it doesn't suck. I have several mailboxes. The one where mail with "ADV:" in the subject goes, I scan several times a week. The one where "the rest" goes, says "beep" and usually gets noticed/answered on 3 minutes notice.
It prevents me from getting distracted from my work by the beep and scanning my mail for every spam that I get.
I still think someone should setup an MX host in colorado, and make a profit from sueing the spammers. hehe.
Business idea for someone in Colorado or one of the other states that require spam to follow certain rules:
Set up a mail server say "spamcatcher" (spamcatcher.com is already taken), allow ISPs to set their the primary MX to "spamcatcher". Make all mailers in the MX chain ignore the MX records for spamcatcher.
Mail will then automatically be routed through the host in colorado, as long as it's up. If it goes down: no worries, the old MX chain will pick up.
Now at spamcatcher in colorado, have every mail tagged with a unique number. Allow people to authenticate as a reciever of a message through the spamcatcher, and mark it as spam.
Spamcatcher requires just a "one machine + internet link" investement and can make a profit off the $10 per spam message that goes through them!
Roger.
P.S. I would appreciate a share of your company if you decide to try it..;-)
Slashdot Mirror
It's pretty clear. They took the cheapest hardware they could find, put Linux on it and are selling it. It would cost EFFORT to make it not work with Linux as a client OS, as that's standard. My bet is they just use lpd (Confirmed!), and it would cost effort to prevent it from working.... Hmm. Disproved here.
Roger.
Hi,
Now that I can decide what projects to take and which I won't, I'm no longer taking projects involving microsoft.
However, during your education, getting to know how to write portable code is really a very important thing.
The only way to learn writing portable code is to port your own code a few times, so that you figure out what constructs are counter-productive, as they don't port easily.
So my recommendation: Write the code on Linux/g++, and then "port" the code to whatever the boss wants.
Roger.
There are two articles in my current Scientific american about SETI. The first was pretty pro-SETI. The second one put things in perspective:
Extra terristrial societies that have found a way to capture ALL the energy from their sun and use all that to broadcast "here we are", can be detected across about half the galaxy. Well, we can assume that this is simply not possible. Read "ringworld" by Niven, and do the math.
Societies like ours, who are happily minding our own business, but leaking some radiation into the surrounding space can be detected from about 100 light-years. There are a handfull of stars in that area....
Roger.
An ordinary cubic kilometer of ocean water contains 14 tons of gold (this was on the radio again yesterday). However, it's uneconomic to go and get all that gold from the ocean.
Assume that the concentration is similar to the gold in the ocean. So then the size of the moon can be calculated as siomilar to a block of 100x100x10 km. That's way smaller than the real thing, so we have just concluded that the concentration is WAY smaller!
Extracting it is not trivial.
So, where everybody can just drive over and put a plant on the nearest ocean shore to go and extract the gold from the ocean, you're expecting people to build a plant 300000 km away and haul the stuff back here, and make a profit? No way!
Getting it here is not trivial.
By the way, I must tell you that I know of an even larger store of fusion-fuel just a bit further away. The concentrations are much more manageable! Problem this time is that the stuff is fusing already....
Roger.
On the + side you get oxygen. On the - side, you get hydrogen. Burn the hydrogen, and you get presto chango... Water.
I hate to break this to ya, but if you electrolyse normal salty water, you get hydrogen and Chloride.
Although the potentials would indicate that the oxygen would form before the chloride, somehow there is ONE exception to the "potentials can predict what products will form" and this is it.
But, yes, you can make hydrogen in your kitchen with this. Makes pretty little bangs if you light it. And if you do it professionally you can certainly produce the oxygen too, but I'm not sure how. (No salt-> low conductivity -> higher voltages? Different salt?)
Roger.
and that I'm getting high ping times and good traceroutes.
I can get you high pingtimes really cheap. Had a period with ping times as high as 3 minutes out here about a week and a half ago.
Good traceroutes? Nice and long? I can statically
route your packets to make nice long traceroutes too!
Roger.
Applix does not convert DOC, it converts RTF
Not true. Applix does convert DOC. It doesn't write doc however.
At least the version that I use dayly.
Roger.
Remember, it's not the volts, nor the amps that kill you. It's the two together.
Not really. It's the amps that kill you. In fact about 20mA of them is the "critical" threshold. Note that you can drive much more if it only goes through an arm or something. The heart is the critical organ here.
However the body has resistance. Normal hand-to-other-hand resistance is on the order of 100k -> About 2000V is deadly.
If your hands are wet that resistance drops quickly, and the deadly voltage can reach "home voltages" (110/200).
Oh. You don't really do much "permanent" damage with say a 50ma current through the heart. So after a few seconds of "relaxation", the heart would work again just fine, but it just needs some external impulse to get started.
My father was working in the garden with an electrical drill. Seems the outside of the drill was under tension, which he didn't notice until he put his knee in a puddle of mud. He managed to put out a yelp, which alerted my mother who teaches CPR. No pulse, no breathing. CPR -> Dad's still alive.
Roger.
This does not open up any more remote exploitable holes, but rather makes it give you root rather then your "nobody" user when you break a program like sendmail that uses this sort of security
No. Its worse than that. You can specify programs that sendmail should run in CF files, you can specify an alternate CF file. Sendmail tries to drop privs if you do these things. It (dropping privs) doesn't work in 2.2.15 if you toggle the bits correctly....
Roger.
The company notifies eBay of suspect auctions and asks them to terminate the auctions within 24 hours.
And Microsoft is perfectly allowed to do so.
If all they do is ask politely, that's perfectly legal. I could ask slashdot to remove this story, and that's legally just fine. If for whatever reason, slashdot decides to honor my request, that's their decision.
So, it seems that Ebay is responding to Microsoft's polite requests. The Ebay usage policy probably states that Ebay can pull any auction they want to pull. In that case, there is just about nothing you can do about this. Sorry.
Roger.
The problem with DCMA kinds of acts is that they make the ISP responsible if they don't "take you down" after getting notified of "copyright violations".
The problem with the internet, whatever method you choose, is that you'll always need an ISP.
Suppose you put your server on a boat in the ocean. You'll need an ISP to hook it up to the internet. THEY have a "home country" and when they get threatened by lawyers talking their legal language, they'll pull your plug really fast, even if you're in international waters. Sure, nobody could sue you in the end, but your ISP IS land-based.
Same holds when you have your own server behind a T1 in your home. Your ISP can still be forced to pull the plug on you. They just firewall off your port 80, and once you move your server, they fireall you off alltogether. Then what?
Roger.
What law would you say it could possibly violate? Remember we don't have things like UCITA and DCMA here [yet].
If I remember correctly, there is an English court ruling effectively with the same results as the DCMA: If a copyright holder informs an ISP of a copyright violation, the ISP should take appropriate action.
On the subject of "lawyer fees", it would be fun to send a few law-students into a fight over this with the MPAA (or whomever). Costs the MPAA real money, the university has a chance of training the students "for real", with a very, very small chance of losing the case....
Roger.
Comments are owned by the poster.
Yes, and what I understand is that the DMCA gives ISPs the right/obligation to remove copyright infringements to prevent further damage.
Roger.
However, once a software house is taking out full page adverts claiming that their software will be good for a company, they have an obligation to uphold those claims and make sure the software works.
I think that this is the right "track".
If you pay for software, you can expect it to do roughly what it was supposed to do. You should NOT be able to sue the company for stuff that simply doesn't do what you expected it to do, but if the software seriously malfunctions, liability starts.
In my opinion, the claimant has a good case: They had a history of using the software, and knew it would work.
The software company also showed "gross negligence" in finding a bug and not notifying their customers. Sure the "disclaim all liability" clause is valid, and software is known to have bugs, but if as a software manufacturer you KNOW the software has a serious bug and not act upon it by sending your customer base a new version, then you're taking on resposibility.
Remember those floppy drives a wile back? They caused data corruption in certain cases. That's OK. But continuing to ship them for 10 years is NOT.
If micorsoft ships a version of word which occasionally changes "dear boss" into "flatulent jerk", that's an honest bug. When they have been notified about this and don't take efforts to correct the situation, then they should be liable.
In microsoft's case a few ads in big papers stating that this bug has been fixed in the next version may be enough. In this case the software producer had an address of the customer, and should've contacted them.
Roger.
One of the important things is that a (G)UI is good for both beginners and experts. This means that an expert can use keyboard accellerators and a beginner should be able to use the menus.
But that's not all: Just binding keyboard acellerators to the most often used menu-options doesn't cut it: The transition from "beginner" to "expert" should be smoothened by having the keyboard accellerators show in the menus.
Netscape and XEmacs both follow that strategy.
Roger.
Not even that is served from the root servers. All the root servers serve is IP addresses of the nameservers for the domain of the host being looked up, its up to the domains nameservers to deal out any actual IP's, including for their own domain.
You look up marthastewart.com, your nameserver asks one of the root nameservers where the nameservers for marthastewart.com is, it then asks them for the IP to marthastewart.com.
Not even THAT is asked from the root server. If I'm not mistaken, your dns server asks for the address of "www.marthastewart.com" and gets the reply "you should be asking ns.marthastewart.com and before you'll ask me, the IP of ns.marthastewart.com is: x.y.z.w".
Roger.
the price is too high, but that's something more easily changeable.)
If you invest $7000M, you'll need to take about $350M per year in interest payments into account for the calculation of your "operating costs".
If you manage to buy the network of satelites from the bankrupt iridium company for say $7000, you need to factor in about $350 per year in interest payments on that. Suddenly the actual operating costs of the satelites (manning the ground-stations, interconnect to PSTN) overwhelms the costs in operating the thing.
So, the original owners are looking at a minimum of $350M of operational costs per year, and selling services on that network for "cheap" is simply not an option.
Roger.
I think larger L1 cache will compensate it. BTW - it's strange how people only keep in mind L2 cache when talking about caches. Considering that L1 cache does 90% (or so) of the speeding job.
Not sure where you are getting your numbers on this one, [...]
The rule-of-thumb is that every level of cache catches (hits) about 90% of the accesses thrown at it. This means that for 100 accesses, you spend: 90*[time to access L1] + 9 * [time to access L2] + 1 * [time to access Main RAM] . (And so the average is 0.01 times that. )
So, no L2 speed does not really matter all that much. Sure you can generate a benchmark that fits in L2 cache, but not in L1. Then it will depend highly on the L2 speed. But in general, the on-package L2's are good enough.
Roger.
A heavy-lift booster in this class could throw a pretty good-sized payload to Mars for a "Mars Direct" type of manned mission.
Ehhhhm. One problem: The "heavy lift" of these guys lifts around 37400 pounds into LEO (low earth orbit). That's almost 17 tons.
Compare that to space shuttle: 23 tons, magnum: 80 tons (The rocket they are developing for mars direct), and Saturn V: > 130 tons (the requirement for a mars-direct type mission).
Roger.
I'll do the "different professors get to classify projects into their own categories later on, if there is interest.
Roger.
I would recommend setting up a site where users, professors, kernel developers and students can enter projects.
Then you have to get people to enter projects in the database, which can be done by different types of students: Small assignment for an OS course: small projects. Graduating projects: larger projects.
With a bit of hacking, jitterbug could handle this kind of job nicely. Professors wanting to "sort" requests to directories that they approve or disapprove for a certain assignment, can be given an account on jitterbug.... (The hacking is required to allow different professors to have different ratings for the projects)
Roger.
Is a 36 inch, 200 ppi display. Of course, we wouldn't have any chance of taking advantage of that resolution at that size.
[...]
I think that we should be concentrating on increasing the bandwidth that we can send to
the monitor. Why not run a fiber-optic cable from your computer to your monitor? Put that
SCSI interface to work providing you with the bandwidth you need, to your monitor, not
your CD-ROM!
I just did the math. You need to pump about 5G per second to your 36 inch 200 ppi screen to get a 70Hz refresh rate. You wanted to use your spare 80Mb per second SCSI controller for that? You need 65 of them to achieve enough bandwidth. Ah, you have your CDROM drive connected to your 160Mbps lvd SCSI chain. Ah. that changes things. You need only 32 of them to drive a display like that.
No, SCSI and video hardware don't really compare in the transfer rates.
They have a display made from four displays 1/4 the size, but would it be more or less expensive to combine 9 1/9 size or 16 1/16 size displays?
My guess is that they need one edge at the top-or-bottom and one at the left-or-right to connect the panel. This makes 4 the max.
My guess is that they have something like half a pixel or a whole pixel of empty space between the two screens. The adjacent pixels need to be a bit brighter to compensate. You need to be at a sufficient distance not to notice.
Roger.
it sucks; almost everything you can get wrong with a spam law is wrong in it.
No. To me it doesn't suck. I have several mailboxes. The one where mail with "ADV:" in the subject goes, I scan several times a week. The one where "the rest" goes, says "beep" and usually gets noticed/answered on 3 minutes notice.
It prevents me from getting distracted from my work by the beep and scanning my mail for every spam that I get.
I still think someone should setup an MX host in colorado, and make a profit from sueing the spammers. hehe.
Roger.
Of course it's going to be difficult to fight it,
;-)
Business idea for someone in Colorado or one of the other states that require spam to follow certain rules:
Set up a mail server say "spamcatcher" (spamcatcher.com is already taken), allow ISPs to set their the primary MX to "spamcatcher". Make all mailers in the MX chain ignore the MX records for spamcatcher.
Mail will then automatically be routed through the host in colorado, as long as it's up. If it goes down: no worries, the old MX chain will pick up.
Now at spamcatcher in colorado, have every mail tagged with a unique number. Allow people to authenticate as a reciever of a message through the spamcatcher, and mark it as spam.
Spamcatcher requires just a "one machine + internet link" investement and can make a profit off the $10 per spam message that goes through them!
Roger.
P.S. I would appreciate a share of your company if you decide to try it..