I agree with you, but I think that FTP also ought to be criticized for breaking layering, which naturally makes it brittle in the face of changes to other layers.
End-to-end bandwidth across the Internet is already often inadequate to saturate links at the edges of the network. When I was working for an ISP, I regularly fielded calls from customers who thought there was something wrong with their T1s because they were only getting 15K/second or whatever from their favorite site.
Ciscos report a per-interface five-minute input/output rate. So I would usually get the customer's permission to ping-flood them, and then I'd do so from the router on our end of the link. Meanwhile, I'd have the customer do a "show int" on the router on their side of the link. After five minutes, the rate would creep up to about 1.5 megabits, and the customer would go away satisfied.
I almost did this for the NYPD once, but they believed me when I explained about net congestion, so I didn't have to.
And perhaps bytes will cost more at peak times than at off-peak times.
Perhaps even we'll be able to request the type of performance we want out of the network, and tell it how much we're willing to pay, and get service based on that -- after all, I might not care how long it takes my FTP session to complete, while I want my stock trade to go through as fast as possible, and my streaming audio to arrive with any reasonable delay so long as the jitter is bounded.
I'm not sure yet if I agree with him, but intuitively his argument makes sense to me, and it's certainly a thought-provoking assertion.
If I had the book handy, I'd quote him verbatim, and at more length. It's a terrific book, and I encourage anyone interested in networks to take a look at it.
Given that the credit card companies eat the cost of fraud, they have every motive to minimize it.
I do think the ball's in their court to develop more secure systems, but remember that they have a vast amount of infrastructure and investment to deal with.
I haven't looked into the details of American Express's new one-time-use CC numbers, but at first glance it looks to me like they are moving in the right direction.
What's this "us vs. them" crap? I am no more part of a "we" that includes skr1pt k1dd13s than I am part of a "we" that includes Western Union. Give it a rest.
Some network client code has buffer overflow or other security problems. Running network client code as root is therefore riskier than running it as an unprivileged user.
Why does Linux have to be gratuitously incompatible with Windows like this?
--
Re:Someone beat this guy with a clue stick
on
Lawsuits Suck
·
· Score: 1
I live in New York State, which is pretty much guaranteed to go Democratic, so I don't fear that I will help throw the election to Bush by supporting Nader.
I thought it's ORBS who are known for behaving that way?
--
Re:Someone beat this guy with a clue stick
on
Lawsuits Suck
·
· Score: 1
If a party knows you will cast your vote for them no matter what, then you can hardly influence them much.
By voting for Nader, I hope to send a message to the Democrats that their practice of creeping rightwards year after year is alienating me.
--
Re:Shooting the messenger
on
Lawsuits Suck
·
· Score: 1
Politicians know that very few people bother to write snail mail. So they figure that every piece of snail mail they receive probably also represents $BIGNUM constituents who didn't bother to write.
Given that they "only care about one thing: being re-elected", writing to them is a pretty effective way to make them sit up and take notice. Especially if you can get a few dozen other people to do it too.
Now I understand what you were saying. My apologies for the misunderstanding.
I'm not sure that Java is a bad choice for Apple. It's been around long enough that I think it might not be a passing fad -- though I'm not saying I'm sure. Anyway, I'm under the impression that pushing obscure languages works better for monopolies, while underdogs want to stress compatibility, interoperability, and openness.
As for their decision to drop ObjC support: I can see why you're offended. That seems like a typical Apple foot-shooting bit.
Oh, one other thing -- about the cross-platform OS. Of course one shouldn't base any plans on the emergence of a cross-platform OS from Apple, but that doesn't mean there's no chance for the future. Maybe they just don't want to piss off Microsoft yet.
I haven't seen anyone point this out yet: Solaris runs on Intel, and that doesn't seem to have hurt Sun's hardware sales.
Of course, Suns and Macs are used for different things, and I'm not suggesting that the same would necessarily hold true for Apple, but it's a point to consider.
> Mac OS X [...] has become increasingly Mac hardware specific.
Do you mean that the code has become more tied to the hardware, or just that Apple's stopped talking about supporting other platforms? There's a difference.
> They're also dropping Objective-C, their most useful foundation technology in my opinion, for future versions of WebObjects.
Objective-C doesn't even compare to WebObjects. One is a programming language, the other is a kind of middleware. Apple is pushing Java bindings to the OPENSTEP^WYellow Box^W^WCocoa APIs, but I believe the ObjC bindings will still be usable. From a marketing perspective, this makes a lot more sense than trying to get everyone to learn ObjC. I know marketing is a dirty word around here, but a platform without developers is not a happy platform.
Yes, you're missing something. Although BSD packages are distributed in files ending with ".tgz", they do have internal structure -- there are certain files in certain formats that must be present in the tarball, or it is not a package.
Slashdot Mirror
I figure it can be abbreviated "MOSX" and hence pronounced "Mosix", to highlight its unix heritage.
Mosix would also be a good name for a jewish version of Jesux, I suppose.
--
Maybe you can sue them under the DMCA. ;-)
--
I agree with you, but I think that FTP also ought to be criticized for breaking layering, which naturally makes it brittle in the face of changes to other layers.
--
Damn, I thought I put a smiley in that post.
Looks like my unintentional troll caught a fish (not you, the other responder).
--
"And he is probably wrong."
Would you care to elaborate on that?
Tell me more about your profit-sharing idea.
--
There are laws about redundancy and availability being built in to emergency services. I wouldn't worry about this, at least not if you're in the USA.
--
Yes.
Check out Bellovin and Cheswick's Firewalls and Internet Security: Repelling the Wily Hacker if you don't believe me.
--
End-to-end bandwidth across the Internet is already often inadequate to saturate links at the edges of the network. When I was working for an ISP, I regularly fielded calls from customers who thought there was something wrong with their T1s because they were only getting 15K/second or whatever from their favorite site.
Ciscos report a per-interface five-minute input/output rate. So I would usually get the customer's permission to ping-flood them, and then I'd do so from the router on our end of the link. Meanwhile, I'd have the customer do a "show int" on the router on their side of the link. After five minutes, the rate would creep up to about 1.5 megabits, and the customer would go away satisfied.
I almost did this for the NYPD once, but they believed me when I explained about net congestion, so I didn't have to.
--
And perhaps bytes will cost more at peak times than at off-peak times.
Perhaps even we'll be able to request the type of performance we want out of the network, and tell it how much we're willing to pay, and get service based on that -- after all, I might not care how long it takes my FTP session to complete, while I want my stock trade to go through as fast as possible, and my streaming audio to arrive with any reasonable delay so long as the jitter is bounded.
--
... in his book An Engineering Approach to Computer Networking that the answer to this problem is ultimately economic: flat-rate pricing has got to go.
I'm not sure yet if I agree with him, but intuitively his argument makes sense to me, and it's certainly a thought-provoking assertion.
If I had the book handy, I'd quote him verbatim, and at more length. It's a terrific book, and I encourage anyone interested in networks to take a look at it.
--
> A file system is a file system, even if it is based on persistant objects in virtual memory space.
Wasn't that one of the lessons of MULTICS?
--
If Western Union can warn all 50K customers, they should much more easily be able to warn a couple of credit card companies.
--
Given that the credit card companies eat the cost of fraud, they have every motive to minimize it.
I do think the ball's in their court to develop more secure systems, but remember that they have a vast amount of infrastructure and investment to deal with.
I haven't looked into the details of American Express's new one-time-use CC numbers, but at first glance it looks to me like they are moving in the right direction.
--
What's this "us vs. them" crap? I am no more part of a "we" that includes skr1pt k1dd13s than I am part of a "we" that includes Western Union. Give it a rest.
--
Some network client code has buffer overflow or other security problems. Running network client code as root is therefore riskier than running it as an unprivileged user.
--
Why does Linux have to be gratuitously incompatible with Windows like this?
--
I live in New York State, which is pretty much guaranteed to go Democratic, so I don't fear that I will help throw the election to Bush by supporting Nader.
--
I thought it's ORBS who are known for behaving that way?
--
If a party knows you will cast your vote for them no matter what, then you can hardly influence them much.
By voting for Nader, I hope to send a message to the Democrats that their practice of creeping rightwards year after year is alienating me.
--
Politicians know that very few people bother to write snail mail. So they figure that every piece of snail mail they receive probably also represents $BIGNUM constituents who didn't bother to write.
Given that they "only care about one thing: being re-elected", writing to them is a pretty effective way to make them sit up and take notice. Especially if you can get a few dozen other people to do it too.
--
Without GNOME, would TrollTech have GPL'd Qt? I think you're being too harsh on RMS.
--
Now I understand what you were saying. My apologies for the misunderstanding.
I'm not sure that Java is a bad choice for Apple. It's been around long enough that I think it might not be a passing fad -- though I'm not saying I'm sure. Anyway, I'm under the impression that pushing obscure languages works better for monopolies, while underdogs want to stress compatibility, interoperability, and openness.
As for their decision to drop ObjC support: I can see why you're offended. That seems like a typical Apple foot-shooting bit.
Oh, one other thing -- about the cross-platform OS. Of course one shouldn't base any plans on the emergence of a cross-platform OS from Apple, but that doesn't mean there's no chance for the future. Maybe they just don't want to piss off Microsoft yet.
--
I haven't seen anyone point this out yet: Solaris runs on Intel, and that doesn't seem to have hurt Sun's hardware sales.
Of course, Suns and Macs are used for different things, and I'm not suggesting that the same would necessarily hold true for Apple, but it's a point to consider.
--
> Mac OS X [...] has become increasingly Mac hardware specific.
Do you mean that the code has become more tied to the hardware, or just that Apple's stopped talking about supporting other platforms? There's a difference.
> They're also dropping Objective-C, their most useful foundation technology in my opinion, for future versions of WebObjects.
Objective-C doesn't even compare to WebObjects. One is a programming language, the other is a kind of middleware. Apple is pushing Java bindings to the OPENSTEP^WYellow Box^W^WCocoa APIs, but I believe the ObjC bindings will still be usable. From a marketing perspective, this makes a lot more sense than trying to get everyone to learn ObjC. I know marketing is a dirty word around here, but a platform without developers is not a happy platform.
--
Yes, you're missing something. Although BSD packages are distributed in files ending with ".tgz", they do have internal structure -- there are certain files in certain formats that must be present in the tarball, or it is not a package.
--