Undemocratic? Are you aware that if you took a vote of people in the UK, most would say that the license fee is basically a good thing? That's what "democratic" means. It does not mean that you're free to opt out of paying, though of course you are; you're free to not have a TV at all (a good move anyway; leaves more time for coding and slashdot.)
I'd be startled if that was covered. Invoking external viewer apps was a common technique used in the first versions of Mosaic, and which definitely predate the file-date by a long way. It was probably common even before the graphical web at all, but I didn't work at a place with IP-based networking back then...
Suffice to say, what you describe had masses of prior art (there were many similar examples in other pieces of software such as mail and usenet readers dating waaay back too). Which is why the Eolas patent doesn't cover it.
The right way to deal with this, and which will drive the FSF up the wall, is to fork any code that switches to the v3 license at the point just before the switch. Any code that is created from scratch with the v3 license can just be cloned wholesale with a different license.
Of course, this ends up looking very much like the Unix Vendor wars, but that's what you get when you allow some third-party organization to dictate the terms of your licensing. Think for yourselves!
Computational resources, by themselves, aren't particularly valuable or hard to obtain; even bandwidth resources are beginning to become expendable if you're smart about how you use them.
FWIW, it really depends on how much you need. If you need a lot of computing power, it tends to be very expensive, and the same is true of bandwidth. (Sometimes it is possible to do tricks like those done by SETI@home, but many problems just aren't decomposable that way.) But if you're willing to put up with just using resources in "single machine-loads" then yes, computing is dirt cheap.
Your other points are good ones; like you, I'm certainly not willing to let blackhats anywhere near any of my systems...
It's laziness in general that causes people to forego robustness in coding.
Not just laziness, but also schedules that are too short and (closely related, but not quite the same) workloads that are too high. Also add in the fact that new hires can take quite a long time (we reckon six months for our mainline systems) to come up to full speed, which makes taking on someone new very difficult. If there's too much to do in too little time and not enough (knowledgable) people to do it, something has got to give and it is usually code quality, whether intentionally or not.
If so, Ben Goldacre has been like this (going by the tone of the regular column he's had in The Guardian) for years.
What was more interesting though was the fact that there was a piece (the Bad Science article was moved from its usual location because it was a lot longer; in its usual place was the piece I'm referring to now) from the editor of the paper about how highly regarded by the rest of the staff for the quality of his critical thinking and his habit of not letting the other journalists get away with sloppy work. Good for him!
So it makes a deep hole. Think about it. The comet is a mile in diameter, and even if it is the consistency of a snowdrift, there is a load of mass in a one-mile-long-but-otherwise-fridge-sized cross-section of snowdrift. The impactor wasn't going nearly fast enough to stand a chance of punching all the way through, and its vaporization is really no surprise at all.
if there were a statue in New York harbor it would be of Queen Victoria
Speaking as an Englishman living in a country where there are lots of statues of QV (all of which make her look short, dumpy and thoroughly Not Amused) I have to admit that an alternate history with that substitution would be particularly bizarre.
You young whipper-snappers should learn from decades of experience with SIMD processors in the scientific supercomputing community: if you want high-speed parallel code, you use Fortran.
Curiously, as I understand it, hydrogen is considerably less explosive than the hydrocarbons that are the primary components of conventional gasoline. Indeed, as hydrogen gas doesn't pond in depressions (being much lighter than air) the explosion risk should be much reduced. But this just indicates that gasoline is dangerous stuff, and not that hydrogen is risk free; my impression is that no material with a high and reasonably available energy density is really safe.
Many years ago I was on a business trip to London to help support a technical demo, and shortly after I got there I became really ill, the sickest I've ever been.
You cannot block port 445 (which zotob uses) since that is what is used in part for file and print sharing.
Whyever not? Or are you claiming that file and printer sharing (as opposed to using one of the stronger client-server protocols for these things) is a good idea?
For a commodity OS, developed by many, used by many and owned by none, the GPL is the best license out there. Under the BSD license, we could see a "splintering" of the OS exactly as we did with the old *nixes. Under the GPL, that is impossible.
So the number of different Linux distros with subtle differences between them are a figment of my imagination then?
Most companies will simply walk away from code that has licensing issues. Who's contributing to the GPL codebase then?
Re:The biggest annoyance with DevStudio
on
Visual Studio Hacks
·
· Score: 2, Informative
Part of the problem for VE is that Java doesn't have a simple XY layout model
Actually it does. Set the layout manager to null and you can put components at any location you want and set the size to anything you want. Of course, if you've got the ability to change font sizes (e.g. to support people with visual difficulties, which is a legal requirement in some places) then absolute layouts suck horribly...
Thanks for that. I know now not to get excited over it.:-D
It occurred to me that the usual technique used by us C hackers (tags + nm + grep + a sane coding style) might not work too well in languages with argument-type polymorphism (e.g. C++ or Java). Of course, I also suspect that the complexity of C++ means that such tools are required too (templates, operator overloading and very loose binding between filenames and class/function names would make this a tricky task for simpler approaches, yes?)
Assume I've been living under a rock for a while: what is this "integrated browse info" of which you speak? How is integrated? What info are you browsing? Why is using the compiler to generate it a good idea? How would this differ from having a small script to post-process the output of 'nm' run over the generated object files?
Slashdot Mirror
Undemocratic? Are you aware that if you took a vote of people in the UK, most would say that the license fee is basically a good thing? That's what "democratic" means. It does not mean that you're free to opt out of paying, though of course you are; you're free to not have a TV at all (a good move anyway; leaves more time for coding and slashdot.)
I must be in the other half then, as I immediately thought about a number of chickens between six and eight...
I'd be startled if that was covered. Invoking external viewer apps was a common technique used in the first versions of Mosaic, and which definitely predate the file-date by a long way. It was probably common even before the graphical web at all, but I didn't work at a place with IP-based networking back then...
Suffice to say, what you describe had masses of prior art (there were many similar examples in other pieces of software such as mail and usenet readers dating waaay back too). Which is why the Eolas patent doesn't cover it.
The right way to deal with this, and which will drive the FSF up the wall, is to fork any code that switches to the v3 license at the point just before the switch. Any code that is created from scratch with the v3 license can just be cloned wholesale with a different license.
Of course, this ends up looking very much like the Unix Vendor wars, but that's what you get when you allow some third-party organization to dictate the terms of your licensing. Think for yourselves!
An anti-aircraft missile?! Don't be ridiculous, AJAX is a war rocket!
Computational resources, by themselves, aren't particularly valuable or hard to obtain; even bandwidth resources are beginning to become expendable if you're smart about how you use them.
FWIW, it really depends on how much you need. If you need a lot of computing power, it tends to be very expensive, and the same is true of bandwidth. (Sometimes it is possible to do tricks like those done by SETI@home, but many problems just aren't decomposable that way.) But if you're willing to put up with just using resources in "single machine-loads" then yes, computing is dirt cheap.
Your other points are good ones; like you, I'm certainly not willing to let blackhats anywhere near any of my systems...
It's laziness in general that causes people to forego robustness in coding.
Not just laziness, but also schedules that are too short and (closely related, but not quite the same) workloads that are too high. Also add in the fact that new hires can take quite a long time (we reckon six months for our mainline systems) to come up to full speed, which makes taking on someone new very difficult. If there's too much to do in too little time and not enough (knowledgable) people to do it, something has got to give and it is usually code quality, whether intentionally or not.
Of course this sucks. That's life.
Reminds me of my local bookstore, where the the New Age section is between the general Fiction section and the Comedy section.
If so, Ben Goldacre has been like this (going by the tone of the regular column he's had in The Guardian) for years.
What was more interesting though was the fact that there was a piece (the Bad Science article was moved from its usual location because it was a lot longer; in its usual place was the piece I'm referring to now) from the editor of the paper about how highly regarded by the rest of the staff for the quality of his critical thinking and his habit of not letting the other journalists get away with sloppy work. Good for him!
So it makes a deep hole. Think about it. The comet is a mile in diameter, and even if it is the consistency of a snowdrift, there is a load of mass in a one-mile-long-but-otherwise-fridge-sized cross-section of snowdrift. The impactor wasn't going nearly fast enough to stand a chance of punching all the way through, and its vaporization is really no surprise at all.
if there were a statue in New York harbor it would be of Queen Victoria
Speaking as an Englishman living in a country where there are lots of statues of QV (all of which make her look short, dumpy and thoroughly Not Amused) I have to admit that an alternate history with that substitution would be particularly bizarre.
Ugh. I mean ScuttleMonkey.
Need more coffee before posting...
... to metamoderate ScuttleButt out of existance as an editor? Please?
You young whipper-snappers should learn from decades of experience with SIMD processors in the scientific supercomputing community: if you want high-speed parallel code, you use Fortran.
3. Hydrogen conversion: Risk: [...] pretty explosive
Curiously, as I understand it, hydrogen is considerably less explosive than the hydrocarbons that are the primary components of conventional gasoline. Indeed, as hydrogen gas doesn't pond in depressions (being much lighter than air) the explosion risk should be much reduced. But this just indicates that gasoline is dangerous stuff, and not that hydrogen is risk free; my impression is that no material with a high and reasonably available energy density is really safe.
Must be "level playing field" as in "crushed flat underneath the jack-booted heel of Microsoft".
Well, going by the current crop of disposable razor ads, we next have Ultimate Extreme Turbo 4D Mach Plus. With Supersized Fries.
Many years ago I was on a business trip to London to help support a technical demo, and shortly after I got there I became really ill, the sickest I've ever been.
Here's a quick tip: don't eat donner kebabs.
You cannot block port 445 (which zotob uses) since that is what is used in part for file and print sharing.
Whyever not? Or are you claiming that file and printer sharing (as opposed to using one of the stronger client-server protocols for these things) is a good idea?
Thanks for not reading your own site, CmdrTaco
For a commodity OS, developed by many, used by many and owned by none, the GPL is the best license out there. Under the BSD license, we could see a "splintering" of the OS exactly as we did with the old *nixes. Under the GPL, that is impossible.
So the number of different Linux distros with subtle differences between them are a figment of my imagination then?
You've never worked with companies I see.
Most companies will simply walk away from code that has licensing issues. Who's contributing to the GPL codebase then?
Part of the problem for VE is that Java doesn't have a simple XY layout model
Actually it does. Set the layout manager to null and you can put components at any location you want and set the size to anything you want. Of course, if you've got the ability to change font sizes (e.g. to support people with visual difficulties, which is a legal requirement in some places) then absolute layouts suck horribly...
Thanks for that. I know now not to get excited over it. :-D
It occurred to me that the usual technique used by us C hackers (tags + nm + grep + a sane coding style) might not work too well in languages with argument-type polymorphism (e.g. C++ or Java). Of course, I also suspect that the complexity of C++ means that such tools are required too (templates, operator overloading and very loose binding between filenames and class/function names would make this a tricky task for simpler approaches, yes?)
Assume I've been living under a rock for a while: what is this "integrated browse info" of which you speak? How is integrated? What info are you browsing? Why is using the compiler to generate it a good idea? How would this differ from having a small script to post-process the output of 'nm' run over the generated object files?