We've seen this over and over again. If it's not archival and offline, it's a mirror... not a backup.
Unfortunately this is such a common misconception that there's just not enough demand for inexpensive high density offline storage to make archival storage technologies (like, say, tape) viable at the low end.
Re:GPL'd code available only by request?
on
Phoenix BIOSOS?
·
· Score: 1
Downloading it now. Big old tarball. Going to take some study to figure what bits are theirs and what bits are everyone else's.
So private individuals are prohibited from profiting from mash-ups involving large numbers of original works?
They're just not enabled directly by this clever scheme. It's not trying to solve every possible problem with copyright, it's simply trying to expand the domain of free culture by an increment large enough to be useful.
The original James Bond, by Ian Fleming, was a parody of the "superspy" literature of the time. Adjusted for cultural inflation, James Bond was Austin Powers for a quieter time. James Bond is ALL ABOUT jumping the shark... and that has carried over to the movies right from the start.
So you contact the human licensing the good and say "I'm making a calendar, and can't afford to pay 12 people 8% of the gross each, will you accept 1.5% (18% of the gross shared among 12 people)?".
And if you're publishing an actual printed encyclopedia you ought to be able to afford actual lawyers.
Given that the protocols aren't interchangeable in their feature sets what you state isn't even a desirable situation.
The point of the UNIX pipe and socket API is that they hide that kind of detail from applications.
A pipe, an AF_UNIX socket, a local disk file, a file on an NFS share, an OpenNET connection to a named pipe on MSNET or DECNET servers, a serial port, raw disk partition, and all kinds of other objects are all presented to a program using the same abstraction. I have had software running under Eunice on VMS talking over OSI CONS to a Xenix named pipe, and without changing a line of code used the same code to talk to a local serial port, an IPv4 network port, and over a multiplexed file to a load tester. I've written the same code talking to a local raw disk partition, a local floppy disk, a remote raw disk partition over OpenNET, a stream file on a VMS server over DECNET, and so on.
So there is a huge interchangeable subset in the feature sets of any network that can maintain the file and socket abstraction.
But knowing the internals of the addressing scheme isn't necessary for a properly written program.
Most properly written programs were written using gethostbyname(), not even gethostbyname2(). And even getaddrinfo() doesn't support AF_UNIX let alone Lan Manager/NetBIOS, nor does it hide the semantics of the endpoint namespace (integer in IPv4 and IPv6, but not specified in networks using file system namespaces like AF_UNIX or OpenNET, and a character string in networks that use the named pipe abstraction).
Anyone who writes an application that needs to know the details of addresses is doing it wrong. Sockets don't require any particular knowledge of the underlying network protocols.
$ man getaddrinfo ...
The hostname and servname arguments are either pointers to NUL-terminated
strings or the null pointer. An acceptable value for hostname is either
a valid host name or a numeric host address string consisting of a dotted
decimal IPv4 address or an IPv6 address. The servname is either a deci-
mal port number or a service name listed in services(5). At least one of
hostname and servname must be non-null. ... $ man gethostbyaddr ...
You can do anything you want with an already opened socket without knowing if the underlying network layer uses IPv4, IPv6, X.400, DECNET, OpenNET, FutureNet, SMB, NetBIOS, Netware, or ID4 nework addresses or end-point identifiers.
You can't open a socket without implicit knowledge about AF_INET, AF_INET6, or AF_UNIX addresses. There's been SOME improvements over the past twenty-odd years, so if your application is relatively new and you're on top of things it's not too much work to handle IPv4 and IPv6, but damn...
The API should never have exposed anything but an anonymous character string as address and end-point identifiers for anything but applications like network scanners that are inherently protocol-sensitive.
They need an "IE Lite". One with no ActiveX support, no "comet cursor" support, and so one... one that only allows helper applications and URL handlers to be explicit plugins that explicitly register with IE as "secure". THe API for running an application would have more in common with the fork/exec in the POSIX subsystem than the Win32 ShellExecute API that requires the application to reverse-engineer quoting. It would call the HTML control (and preferably a new "secure" HTML control) using a new API that calls back to the IE shell for anything beyond rendering content and running hard-sandboxed scripts. Yes, that includes following URLs and downloading embedded images and calling plugins.
This would not have all the features of the "legacy IE", particularly at first as few applications use the secure new APIs. But that's OK, you'd be able to fall back to full IE for pages that really needed it.
With some work they could make it even more inherently secure than Firefox and Safari, with no need for a leaky "reduced privilege" sandbox. Plus, with a new secure HTML control you'd get people writing Firefox extensions to call the Microsoft HTML control without people like me going "YOU CRAZY FOOL, WHY YOU DO A STUPID THING LIKE THAT?", and be able to get more effective browser share even with people who use the Firefox shell.
The socket API... or rather the UNIX file descriptor API... has been extended many times. Sockets are already one such extension, and there's no reason you couldn't do something like mmap() a socket to map the buffers into user space directly. Heck, udp sockets already diverge from the read/write paradigm.
The problem with sockets is at a higher level. They're not mapped into the file system name space. You should be able to open a socket by calling open() on something like "/dev/tcp/address-or-name/port-or-name" and completely hide the details of gethostbyname(), bind(), and so on from the application layer. If they'd done that we'd already be using IPv6 for everything because applications wouldn't have to know about the details of addresses because they'd just be arbitrary strings like file names already are.
Like I said, I don't MIND most web ads, they're what pay for the site and they're far less annoying than TV ads. So I'm not interested in turning everything off. These are an exception, they're like you're watching an episode of "House" and every few seconds another talk-bubble pops up advertising penis pills and obscuring the action.
How about something specific and selective... I don't mind seeing banner ads, but I want something that eliminates rollover ads completely. And don't give them the option of asking to be disabled.
Ideally they could ask to be excluded but the software would deliver electric shocks to the advertiser's groin instead.
It's going to be pretty damn rare that someone's going to break a counterfeit Windows install nearly as badly as box makers already break genuine Windows installs with the bloatware they stuff in there. I suspect that more of these "genuine" systems don't even come with a clean install disk, let alone give as good an experience as the original product, than counterfeits do.
"counterfeit software delivers a poor experience and impacts customer satisfaction with our products, particularly if users do not know that their software is non-genuine."
Since it's byte for byte identical whether it's "counterfeit" or "real", what does that say about Windows 7?
They will have to either replace legacy software (which is difficult for some clients) or buy a high-end computer just to get a decent upgrade for Vista/XP.
I'd rather get a database-in-a-box that works so well you forget it's there until your receptionist calls you because it's detected a bad drive and ordered a replacement for you... rather than one where you have to find the magic dongle headphones before you can turn it on.
There's two completely contradictory definitions of "trojan", one refers to the infection method, the other refers to the existence of a backdoor. Apparently different researchers were thinking of different parts of the Trojan Horse myth: tricking the Trojans to bring the horse into the city, or opening a gate to let the Greek army in.
At least we should be glad they haven't decided to use the condom metaphor as well.
Even if public transport was "only" 25% cheaper than a car, I'd still prefer it. Why? Because of the value of my time. Until we have fully automated self-driving cars, I can't read, nap, work, or simply daydream in a car. I have to actively drive it, and I'm pretty damn sure at least 50% of the other drivers are trying to read, nap, work, or daydream behind the wheel.
Slashdot Mirror
We've seen this over and over again. If it's not archival and offline, it's a mirror... not a backup.
Unfortunately this is such a common misconception that there's just not enough demand for inexpensive high density offline storage to make archival storage technologies (like, say, tape) viable at the low end.
Downloading it now. Big old tarball. Going to take some study to figure what bits are theirs and what bits are everyone else's.
What Would Cory Do?
More practically, have you talked to Google, this seems like a clear opportunity for them to microtune their ranking algorithm.
(yes, I'm serious)
So private individuals are prohibited from profiting from mash-ups involving large numbers of original works?
They're just not enabled directly by this clever scheme. It's not trying to solve every possible problem with copyright, it's simply trying to expand the domain of free culture by an increment large enough to be useful.
When the competition is Windows, you don't need to be Marcus Ranum or Bruce Schnier to stroll over the hurdles... with crutches.
* Powered by a 533Mhz 3D accelerator plus flash engine
What's this in geek?
Why are they pointing to some web forum instead of the manufacturer, anyway?
* The new console boasts a 533MHz ARM9 CPU with 3D acceleration.
That's better, I think.
One could argue that James Bond jumped the shark
The original James Bond, by Ian Fleming, was a parody of the "superspy" literature of the time. Adjusted for cultural inflation, James Bond was Austin Powers for a quieter time. James Bond is ALL ABOUT jumping the shark... and that has carried over to the movies right from the start.
So you contact the human licensing the good and say "I'm making a calendar, and can't afford to pay 12 people 8% of the gross each, will you accept 1.5% (18% of the gross shared among 12 people)?".
And if you're publishing an actual printed encyclopedia you ought to be able to afford actual lawyers.
Yes, except for being twenty years too late.
Given that the protocols aren't interchangeable in their feature sets what you state isn't even a desirable situation.
The point of the UNIX pipe and socket API is that they hide that kind of detail from applications.
A pipe, an AF_UNIX socket, a local disk file, a file on an NFS share, an OpenNET connection to a named pipe on MSNET or DECNET servers, a serial port, raw disk partition, and all kinds of other objects are all presented to a program using the same abstraction. I have had software running under Eunice on VMS talking over OSI CONS to a Xenix named pipe, and without changing a line of code used the same code to talk to a local serial port, an IPv4 network port, and over a multiplexed file to a load tester. I've written the same code talking to a local raw disk partition, a local floppy disk, a remote raw disk partition over OpenNET, a stream file on a VMS server over DECNET, and so on.
So there is a huge interchangeable subset in the feature sets of any network that can maintain the file and socket abstraction.
But knowing the internals of the addressing scheme isn't necessary for a properly written program.
Most properly written programs were written using gethostbyname(), not even gethostbyname2(). And even getaddrinfo() doesn't support AF_UNIX let alone Lan Manager/NetBIOS, nor does it hide the semantics of the endpoint namespace (integer in IPv4 and IPv6, but not specified in networks using file system namespaces like AF_UNIX or OpenNET, and a character string in networks that use the named pipe abstraction).
I'm still bitter about Plan 9. Damn, that bitch had one hot set of APIs.
I'm bitter about Alpha, too. And Amiga. And SGML. And Palm OS. And Power PC.
*sigh*
Anyone who writes an application that needs to know the details of addresses is doing it wrong. Sockets don't require any particular knowledge of the underlying network protocols.
You can do anything you want with an already opened socket without knowing if the underlying network layer uses IPv4, IPv6, X.400, DECNET, OpenNET, FutureNet, SMB, NetBIOS, Netware, or ID4 nework addresses or end-point identifiers.
You can't open a socket without implicit knowledge about AF_INET, AF_INET6, or AF_UNIX addresses. There's been SOME improvements over the past twenty-odd years, so if your application is relatively new and you're on top of things it's not too much work to handle IPv4 and IPv6, but damn...
The API should never have exposed anything but an anonymous character string as address and end-point identifiers for anything but applications like network scanners that are inherently protocol-sensitive.
They need an "IE Lite". One with no ActiveX support, no "comet cursor" support, and so one... one that only allows helper applications and URL handlers to be explicit plugins that explicitly register with IE as "secure". THe API for running an application would have more in common with the fork/exec in the POSIX subsystem than the Win32 ShellExecute API that requires the application to reverse-engineer quoting. It would call the HTML control (and preferably a new "secure" HTML control) using a new API that calls back to the IE shell for anything beyond rendering content and running hard-sandboxed scripts. Yes, that includes following URLs and downloading embedded images and calling plugins.
This would not have all the features of the "legacy IE", particularly at first as few applications use the secure new APIs. But that's OK, you'd be able to fall back to full IE for pages that really needed it.
With some work they could make it even more inherently secure than Firefox and Safari, with no need for a leaky "reduced privilege" sandbox. Plus, with a new secure HTML control you'd get people writing Firefox extensions to call the Microsoft HTML control without people like me going "YOU CRAZY FOOL, WHY YOU DO A STUPID THING LIKE THAT?", and be able to get more effective browser share even with people who use the Firefox shell.
But... this is a pipe dream, isn't it?
Reminds me of something, I'll get it in a moment...
Do you get villages of minions to torture, or do you have to be satisfied with PKs?
Right, we don't want to have excessive chances for success.
The socket API... or rather the UNIX file descriptor API... has been extended many times. Sockets are already one such extension, and there's no reason you couldn't do something like mmap() a socket to map the buffers into user space directly. Heck, udp sockets already diverge from the read/write paradigm.
The problem with sockets is at a higher level. They're not mapped into the file system name space. You should be able to open a socket by calling open() on something like "/dev/tcp/address-or-name/port-or-name" and completely hide the details of gethostbyname(), bind(), and so on from the application layer. If they'd done that we'd already be using IPv6 for everything because applications wouldn't have to know about the details of addresses because they'd just be arbitrary strings like file names already are.
Like I said, I don't MIND most web ads, they're what pay for the site and they're far less annoying than TV ads. So I'm not interested in turning everything off. These are an exception, they're like you're watching an episode of "House" and every few seconds another talk-bubble pops up advertising penis pills and obscuring the action.
I'm not sure what you mean by roll-over ads
Those companies like Vibrant that put pseudo-pop-up ads in double-underlined rollovers in the text of subscribing sites.
How about something specific and selective... I don't mind seeing banner ads, but I want something that eliminates rollover ads completely. And don't give them the option of asking to be disabled.
Ideally they could ask to be excluded but the software would deliver electric shocks to the advertiser's groin instead.
It's going to be pretty damn rare that someone's going to break a counterfeit Windows install nearly as badly as box makers already break genuine Windows installs with the bloatware they stuff in there. I suspect that more of these "genuine" systems don't even come with a clean install disk, let alone give as good an experience as the original product, than counterfeits do.
"counterfeit software delivers a poor experience and impacts customer satisfaction with our products, particularly if users do not know that their software is non-genuine."
Since it's byte for byte identical whether it's "counterfeit" or "real", what does that say about Windows 7?
They will have to either replace legacy software (which is difficult for some clients) or buy a high-end computer just to get a decent upgrade for Vista/XP.
Or just stick with XP.
I'd rather get a database-in-a-box that works so well you forget it's there until your receptionist calls you because it's detected a bad drive and ordered a replacement for you... rather than one where you have to find the magic dongle headphones before you can turn it on.
There's two completely contradictory definitions of "trojan", one refers to the infection method, the other refers to the existence of a backdoor. Apparently different researchers were thinking of different parts of the Trojan Horse myth: tricking the Trojans to bring the horse into the city, or opening a gate to let the Greek army in.
At least we should be glad they haven't decided to use the condom metaphor as well.
Even if public transport was "only" 25% cheaper than a car, I'd still prefer it. Why? Because of the value of my time. Until we have fully automated self-driving cars, I can't read, nap, work, or simply daydream in a car. I have to actively drive it, and I'm pretty damn sure at least 50% of the other drivers are trying to read, nap, work, or daydream behind the wheel.