Open source RDBMS's are good solutions for many, perhaps even most, problems. But there are still some situations where I'd want to stick with Oracle's strength and maturity and not take chances.
PostgreSQL isn't mature? It's a direct descendant of Ingres, the original relational database. Ingres was written in 1977 at Berkeley. Bob Miner, Ed Oates, and Bruce Scott saw the commercial potential of RDBMS and founded a company later in 1977 called Software Development Laboratories. Larry Ellison joined up with them several months later.
It wasn't until two years later in 1979 that the first version of Oracle was released (SDL had since changed its name to Relational Software Inc.).
In 1983, Relational Software changed its company name to Oracle.
The funny part is that Berkeley UNIX (i.e. BSD) started out as a modification to AT&T UNIX to provide a better OS to run Ingres on...
I won't use.NET for game development, period. I guess I'm old fashioned, but I like my SDK's as simple as possible, something Microsoft doesn't seem to like making anymore.
But... How would you write efficient code if your functions didn't have helpful names like DDLockBufferAndBlitToSurfaceAndPrayThePointerIsVal id_u ?
If you need to give a site your email addy, leave in a reference to that site. eg slashdot@myname.ath.cx. That way if someone sells your address, an address leaks, or whatver, you know EXACTLY who is responsible, and you can block junk mail without affecting legitimate email.
I've been using this technique too for quite some time (except I use subdomains so I don't even have to accept SMTP connections for dead accounts anymore). The only spam I get is on mailing list accounts and I rotate those every once in a while. It's also great for automatically sorting your email -- think sieve. Combine with SSL+IMAP access to your mailbox folders and life is good.
And there's no way in hell I'm going to pay for a certificate for each one of my 50+ addresses. That guy is an idiot.
Perhaps sendmail et. al. could implement a thoughtful pause (of a configurable value) before it gives the 550. If my mail server slept for 60 seconds before issuing the error, it might slow them down, or cause them to run out of threads.
Postfix does this by default. I think it pauses 10-15 seconds before issuing the error message. And the more errors you have in a single session, the longer the pause before each one gets. You can also configure it to automatically terminate the connection after X amount of errors.
You have a point, what I mean is that for there to be the kind of growth that we want, it will need to be even.
Oh, I agree completely there. Nature tries very hard to be balanced and get to a state of equilibrium, probably unbalanced things generally don't work in the very long term. Any "modifications" we make would have to be balanced as well or they would ultimately fail.
I see your point about everything being a tradeoff. More brainpower requires more energy, and in our current biology, more oxygen (and blood) to the brain. That's why the corotid arteries have so much more capacity than anything else in our body. The more energy/cells/whatever you devote to one area of specialization, the less you have to apply to others.
Heck, we could end up looking like those aliens in X-Files, big eyes, large head, small nose, mouth and body.
That would be amusing. Even if evolution happened naturally, we would probably evolve into a species adapted to living in a city and being dependant on industrial and technological means. It could be disasterous if that technology ever failed.
The only true way to improve ourselves that I can think of is good ol' natural evolution, which is slow, but works!
One question I've found myself pondering is, does evolution really work anymore?
Think about it. Our ideal of society is to protect the weak from harm, and give everyone the right to reproduce. This runs completely contrary to the theory of natural selection promoting genetic evolution, which depends on survival of the fittest and the weak not being able to reproduce as effectively.
So, if you subscribe to the theory of natural selection being the root cause of evolution, then we've already pretty much beaten it. Our genetic makeup, instead of becoming stronger, would spread out in all directions and possibly become stagnant.
There are only a few solutions to this problem. One of them, a particularly unpopular one (for good reason), would be so-called "genetic cleansing", or eradicating those with weak genes. Or in the cases we've seen, perceived to be weak by those in power, which probably isn't necessarily true -- i.e. someone in a wheelchair may have been put there by circumstance and actually have really good genes.
Genetic engineering is another. If natural evolution doesn't apply anymore, maybe the only way for the human race to advance is if we take matters into our own hands. There's all kinds of sticky issues here, like at what point do we cease being "human" (and would that also apply to naturally-evolved "super"-humans as well?)? Or should we restrict it only to small changes with a long-term goal, more closely following the path that natural evolution would.
It's not an easy topic, and I'm not sure if I actually believe any of that. Anyway, just something to think about...
I have to reboot my XP machine every few hours most days.
Dude, what did you do to that poor XP box? Granted, I hate XP's interface and will never use it, but it is based on the NT kernel. My coworker's Win2k box is up to 52 days uptime, and is neck-and-neck with my FreeBSD 5.0 desktop in our uptime battle. I managed to hit 120 days once in Win2k back when I ran it on my desktop -- his record is 97.
I use SpeakEasy DSL via Covad. This service is technically residential, because my servers are sitting in my house. But I have a legitimate domain, and static IPs on my servers. However, reverse DNS lookups return "dslwww-xxx-yyy-zzz.phl.yadayadayada," NOT my registered domain name.
If the reverse DNS does become a problem, I've found that Speakeasy is more than happy to set up your reverse DNS to the name that you want. All you have to do is ask.
Re:Listen, don't worry about the UNIX thing
on
Legacy-Free PCs
·
· Score: 1
I think the Starship Enterprise runs OpenBSD myself.
God, I hope not, what with all the remote root expoits the Enterprise has had to date...
In comparison, I think Voyager must run Windows 2003 Server.
And all of the holodeck systems run Mac OS -- they work fine when it's not imporant and always fail when you need them most.
I sniffed the packets from this duplicate article, and the evil bit was not set! I demand Slashdot immediately implement RFC3514 for all duplicate articles!
They're waiting for Debian to release updated packages...
Basically, Trent, (ie, Verisign, Thawte or others) signs a certificate for Bob indicating his domain. Alice sends Bob a request for the certificate, Bob sends Alice the certificate. Alice verifies that the certificate is properly signed. Alice then uses that certificate to encrypt all communication with Bob.
If you control both the gateway and the client machine (as in a corporate / govt. network), you can MITM SSL fairly easily.
Let's say that Vader is the big bad imperial gateway, it works like this:
Alice sends Bob a request for the certificate, which is intercepted by Vader because he is a transparent proxy. Vader proxies the request and gets the domain name from Bob. Vader creates a new certificate with Bob's domain name and signs it with Vader's key. The new cert is passed on to Alice, who has Vader's public key in her trusted CA list (as per company policy). So Alice encrypts data with Vader's key, who then decrypts it, scans the content, and re-encrypts it using a different key to send on to Bob. Higher latency, but it works.
So combine that with blocking any outbound traffic that can't be scanned (somebody brought a laptop from home, sorry, too bad, against company policy) and you're all set.
(the header of all of my programs) typedef apstring gaypstring; typedef apvect gaypvect;
Ah, yes, gaypstring. I remember it well from my high school CS class. There was a bug in it that caused a certain series of string copies and concatenations to end up with the string having a length different than the actual size of the memory buffer. It was pretty rare, but could cause memory corruption and odd results. I remember banging my head against the wall trying to find the bug in one of my programs before I finally traced into the library itself and located the problem (somebody forgot to update the string length during a copy in the case that it was shorter than it used to be).
There was also gaypvect that did an unnecessary new[] and free[] for almost every single operation. I swear it ran in like O(10*n^2) time or something. I got so fed up with it that I wrote a class called nongayvect that used a pretty simple-minded chuncked allocator and realloc() where possible. On average, even that was 500%-1000% faster, at least on a benchmark that created and copied random arrays. Normal programs showed a perceptible increase in speed.
Of course, the whole class was pretty bad. The "book" was a beta version written by a Pascal programmer, printed out chapter by chapter, and was full of errors. He even slipped a couple times and used pascal syntax (a:= b) in C++ programs!
I guess my point is that even "safe" APIs sometimes have bugs. There have been Java VM implementation problems in the pase that let Java programs break out of the sandbox and execute native code. I'm sure it's only a matter of time before something similar happens with C#. Being careful and using tools to help make things more secure is a Good Thing(TM), but don't assume that the API is infallible. Remember that the more layers of abstraction you add, the more complex things become, and the more likely it is that there is a bug or an invalid assumption somewhere in there.
Or is this just a baseless rumour? I haven't found any concrete proof supporting this claim.
Partition, I doubt it. But if you're using grub and it has a stage 1.5 loader stored right after the MBR, I could see how it might get corrupted by C-Dilla...
Thats why they made that little dohicky that works via RF - you wear a token and you can only read your encrypted files if you're within 5 feet or so of your laptop.
Hey, cool. I can't want to walk around and steal people's private keys with a little device that pretends to be their laptop.
That's just moronic. The 2k and XP machines I have regularly have uptimes longer than the 95 timer rollover crash time. And that wasn't found for years because nobody had a 95 machine that ever stayed up that long. If you want stable without IE than at least get something in the NT line.
We need to figure out a way to rip IE out of 2k (replacement shell, probably). Call it NT 5.
Guess you could always use cmd.exe as your shell:P
Slashdot Mirror
Open source RDBMS's are good solutions for many, perhaps even most, problems. But there are still some situations where I'd want to stick with Oracle's strength and maturity and not take chances.
PostgreSQL isn't mature? It's a direct descendant of Ingres, the original relational database. Ingres was written in 1977 at Berkeley. Bob Miner, Ed Oates, and Bruce Scott saw the commercial potential of RDBMS and founded a company later in 1977 called Software Development Laboratories. Larry Ellison joined up with them several months later.
It wasn't until two years later in 1979 that the first version of Oracle was released (SDL had since changed its name to Relational Software Inc.).
In 1983, Relational Software changed its company name to Oracle.
The funny part is that Berkeley UNIX (i.e. BSD) started out as a modification to AT&T UNIX to provide a better OS to run Ingres on...
I won't use .NET for game development, period. I guess I'm old fashioned, but I like my SDK's as simple as possible, something Microsoft doesn't seem to like making anymore.
l id_u ?
But... How would you write efficient code if your functions didn't have helpful names like DDLockBufferAndBlitToSurfaceAndPrayThePointerIsVa
The irony is that NSI is now sending out messages, warning their customers about companies doing exactly that to get them to switch away from NSI...
It would be quite understandable that a company didn't want the image that staff wearing sidearms creates.
Hey, that would be pretty cool. I think I'll have to talk to my company about getting the dress code changed to be fatigues and sidearms...
crazy-go-nuts RMS style
Is RMS a graduate of Crazy Go Nuts University?
GO CGNU!
If you need to give a site your email addy, leave in a reference to that site. eg slashdot@myname.ath.cx. That way if someone sells your address, an address leaks, or whatver, you know EXACTLY who is responsible, and you can block junk mail without affecting legitimate email.
I've been using this technique too for quite some time (except I use subdomains so I don't even have to accept SMTP connections for dead accounts anymore). The only spam I get is on mailing list accounts and I rotate those every once in a while. It's also great for automatically sorting your email -- think sieve. Combine with SSL+IMAP access to your mailbox folders and life is good.
And there's no way in hell I'm going to pay for a certificate for each one of my 50+ addresses. That guy is an idiot.
Perhaps sendmail et. al. could implement a thoughtful pause (of a configurable value) before it gives the 550. If my mail server slept for 60 seconds before issuing the error, it might slow them down, or cause them to run out of threads.
Postfix does this by default. I think it pauses 10-15 seconds before issuing the error message. And the more errors you have in a single session, the longer the pause before each one gets. You can also configure it to automatically terminate the connection after X amount of errors.
You have a point, what I mean is that for there to be the kind of growth that we want, it will need to be even.
Oh, I agree completely there. Nature tries very hard to be balanced and get to a state of equilibrium, probably unbalanced things generally don't work in the very long term. Any "modifications" we make would have to be balanced as well or they would ultimately fail.
I see your point about everything being a tradeoff. More brainpower requires more energy, and in our current biology, more oxygen (and blood) to the brain. That's why the corotid arteries have so much more capacity than anything else in our body. The more energy/cells/whatever you devote to one area of specialization, the less you have to apply to others.
Heck, we could end up looking like those aliens in X-Files, big eyes, large head, small nose, mouth and body.
That would be amusing. Even if evolution happened naturally, we would probably evolve into a species adapted to living in a city and being dependant on industrial and technological means. It could be disasterous if that technology ever failed.
The only true way to improve ourselves that I can think of is good ol' natural evolution, which is slow, but works!
One question I've found myself pondering is, does evolution really work anymore?
Think about it. Our ideal of society is to protect the weak from harm, and give everyone the right to reproduce. This runs completely contrary to the theory of natural selection promoting genetic evolution, which depends on survival of the fittest and the weak not being able to reproduce as effectively.
So, if you subscribe to the theory of natural selection being the root cause of evolution, then we've already pretty much beaten it. Our genetic makeup, instead of becoming stronger, would spread out in all directions and possibly become stagnant.
There are only a few solutions to this problem. One of them, a particularly unpopular one (for good reason), would be so-called "genetic cleansing", or eradicating those with weak genes. Or in the cases we've seen, perceived to be weak by those in power, which probably isn't necessarily true -- i.e. someone in a wheelchair may have been put there by circumstance and actually have really good genes.
Genetic engineering is another. If natural evolution doesn't apply anymore, maybe the only way for the human race to advance is if we take matters into our own hands. There's all kinds of sticky issues here, like at what point do we cease being "human" (and would that also apply to naturally-evolved "super"-humans as well?)? Or should we restrict it only to small changes with a long-term goal, more closely following the path that natural evolution would.
It's not an easy topic, and I'm not sure if I actually believe any of that. Anyway, just something to think about...
Offtopic, but how did you manage that? I've had zero luck getting 9i to run under FreeBSD -- we had to set up a Gentoo box for our test environment...
I have to reboot my XP machine every few hours most days.
Dude, what did you do to that poor XP box? Granted, I hate XP's interface and will never use it, but it is based on the NT kernel. My coworker's Win2k box is up to 52 days uptime, and is neck-and-neck with my FreeBSD 5.0 desktop in our uptime battle. I managed to hit 120 days once in Win2k back when I ran it on my desktop -- his record is 97.
I can't imagine XP being that much worse.
I use SpeakEasy DSL via Covad. This service is technically residential, because my servers are sitting in my house. But I have a legitimate domain, and static IPs on my servers. However, reverse DNS lookups return "dslwww-xxx-yyy-zzz.phl.yadayadayada," NOT my registered domain name.
If the reverse DNS does become a problem, I've found that Speakeasy is more than happy to set up your reverse DNS to the name that you want. All you have to do is ask.
I think the Starship Enterprise runs OpenBSD myself.
God, I hope not, what with all the remote root expoits the Enterprise has had to date...
In comparison, I think Voyager must run Windows 2003 Server.
And all of the holodeck systems run Mac OS -- they work fine when it's not imporant and always fail when you need them most.
I sniffed the packets from this duplicate article, and the evil bit was not set! I demand Slashdot immediately implement RFC3514 for all duplicate articles!
They're waiting for Debian to release updated packages...
SSL is secured against man in the middle attacks.
Basically, Trent, (ie, Verisign, Thawte or others) signs a certificate for Bob indicating his domain. Alice sends Bob a request for the certificate, Bob sends Alice the certificate. Alice verifies that the certificate is properly signed. Alice then uses that certificate to encrypt all communication with Bob.
If you control both the gateway and the client machine (as in a corporate / govt. network), you can MITM SSL fairly easily.
Let's say that Vader is the big bad imperial gateway, it works like this:
Alice sends Bob a request for the certificate, which is intercepted by Vader because he is a transparent proxy. Vader proxies the request and gets the domain name from Bob. Vader creates a new certificate with Bob's domain name and signs it with Vader's key. The new cert is passed on to Alice, who has Vader's public key in her trusted CA list (as per company policy). So Alice encrypts data with Vader's key, who then decrypts it, scans the content, and re-encrypts it using a different key to send on to Bob. Higher latency, but it works.
So combine that with blocking any outbound traffic that can't be scanned (somebody brought a laptop from home, sorry, too bad, against company policy) and you're all set.
...and all was laid to burnination!!!!
I believe it was called APSTRING.
:= b) in C++ programs!
(the header of all of my programs)
typedef apstring gaypstring;
typedef apvect gaypvect;
Ah, yes, gaypstring. I remember it well from my high school CS class. There was a bug in it that caused a certain series of string copies and concatenations to end up with the string having a length different than the actual size of the memory buffer. It was pretty rare, but could cause memory corruption and odd results. I remember banging my head against the wall trying to find the bug in one of my programs before I finally traced into the library itself and located the problem (somebody forgot to update the string length during a copy in the case that it was shorter than it used to be).
There was also gaypvect that did an unnecessary new[] and free[] for almost every single operation. I swear it ran in like O(10*n^2) time or something. I got so fed up with it that I wrote a class called nongayvect that used a pretty simple-minded chuncked allocator and realloc() where possible. On average, even that was 500%-1000% faster, at least on a benchmark that created and copied random arrays. Normal programs showed a perceptible increase in speed.
Of course, the whole class was pretty bad. The "book" was a beta version written by a Pascal programmer, printed out chapter by chapter, and was full of errors. He even slipped a couple times and used pascal syntax (a
I guess my point is that even "safe" APIs sometimes have bugs. There have been Java VM implementation problems in the pase that let Java programs break out of the sandbox and execute native code. I'm sure it's only a matter of time before something similar happens with C#. Being careful and using tools to help make things more secure is a Good Thing(TM), but don't assume that the API is infallible. Remember that the more layers of abstraction you add, the more complex things become, and the more likely it is that there is a bug or an invalid assumption somewhere in there.
Or is this just a baseless rumour? I haven't found any concrete proof supporting this claim.
Partition, I doubt it. But if you're using grub and it has a stage 1.5 loader stored right after the MBR, I could see how it might get corrupted by C-Dilla...
Yes, pen and paper.
Except that it was the 'f' key, and also the star (*).
It's a safety measure to keep you from accidentally typing rm -rf *
The idea was to use a Win2K disk on a WinXP box and the Win2K thinks it is a "corrupt" install.
After seeing WinXP in action, I would tend to agree with the Win2k disk on its assessment...
Thats why they made that little dohicky that works via RF - you wear a token and you can only read your encrypted files if you're within 5 feet or so of your laptop.
Hey, cool. I can't want to walk around and steal people's private keys with a little device that pretends to be their laptop.
Okay, time to wire up my door frame to emit a focused EM pulse every time someone enters/leaves my house.
Come to think of it, that would be a great way to deal with people who bring in cell phones too...
how did slashdot know what size you wear?
kevin told them back when they interviewed him. Kind of like that one guy's paypal password.
That's just moronic. The 2k and XP machines I have regularly have uptimes longer than the 95 timer rollover crash time. And that wasn't found for years because nobody had a 95 machine that ever stayed up that long. If you want stable without IE than at least get something in the NT line.
:P
We need to figure out a way to rip IE out of 2k (replacement shell, probably). Call it NT 5.
Guess you could always use cmd.exe as your shell