I've read your original comment, and your attacks against ewhac, and wish to say the following:
Without purporting to speak for Be, Inc., yet in an attempt to obviate the need for you to undertake the strenous research that would make the following statement painfully obvious for even the most discerning waterborne marsupial, I will vouchsafe for "Bols Ewhac", that he is who he says he is, and that he is, in fact, an employee of Be. If you want more proof, see The Be Team, and connect the dots. (Of course, I won't tell you who I am though. Life isn't fair sometimes.)
Further, I hope that you'll agree that his prior contributions to Slashdot are sufficient to prove his integrity.
Finally, I disagree with your comment that what ewhac said is "childish and unprofessional." Many professional experts and Slashdot gurus say far worse of Microsoft or other benchmark-stuffers than what ewhac has said, and with far less justification or wit. And they weren't speaking on their own time with their own voice, defending the integrity of their work, as ewhac is.
To answer your original comment about what is worth moderating up, healthy skepticism of any benchmark is fully warranted. Now, you may feel that completely unsupported accusations and groundless rumormongering constitute good entertainment value -- and it seems, from my experience, that the majority of Americans, at least, would agree with you:-). They certainly do not, however, constitute healthy skepticism, and are the last thing I want to see encouraged on Slashdot. That is the reason the post in question should be moderated into oblivion, not because The Man, who happens to be wearing a Be cap over his long hair, tells you to. I can't speak for you, but I would vastly prefer to see hard evidence or reason backing up a piece of juicy criticism before said criticism gets spewed into the net, not after. Perhaps I differ from some of my Slashdot colleagues, however, in this admittedly farfetched belief.
As far as bottlenecks go, Jason Sams and the other 3D god{esse}?s in these parts are the ultimate authority, but I could think of several places besides the hardware where the performance of OpenGL would be affected. Anything to shorten or optimize the rendering pipeline in software (using SIMD, dynamic generation of the pipeline for critical sections, native OpenGL vs. Glide support in the driver, etc.) is going to make a big difference for most interesting OpenGL work, and no, not every implementation of OpenGL has all of these software issues sorted out. So perhaps these numbers are not as thin-air as you think.
On a different tack, if you did do the equivalent of dropping "triangles" in the audio world, and could do it without anybody noticing, it would be heralded as "next-generation compression based on psychoacoustic perception heuristics," and the computing community would clap you on the back. Perhaps Be does have something up their sleeve after all! ( Now now, that's of course a groundless rumor, and I shouldn't have said it. Please don't take it seriously.:-P )
I agree with you. Of course, I don't want everyone to know my location -- just as I don't want everyone to know my cell phone number either! But most of the time I use a cell phone is to communicate my location to people anyway ("no, I'm in a meeting right now," "yes, I'm on my way over," etc.). And if I don't want people to find me, chances are excellent that I don't want them calling me either. So for me, the two issues are practically the same thing...
My point was, knowing location is an important ability for mobile communication. So, better ability to communicate location is a Good Thing -- provided that steps are taken to protect your location from simply being handed out to destinations without your knowledge and consent.
Constantly broadcasting your location is a serious invasion of your privacy.
But, is broadcasting the location of your cell phone an invasion of privacy? This may seem like splitting hairs, but if you can turn your cell phone off or leave it behind when you don't want to be found, it seems like a lot less of an issue. Besides, isn't half of the point of having a cell phone making sure that you can be found?
This whole privacy issue is also easily surmountable if you must grant permission for your location to be disclosed as a security setting on the phone (disregarding the FCC/EMS requirements, since there's no evidence that they're about to start up an advertising racket).
High-precision GPS will be an absolute godsend when it comes to things like navigation. I wouldn't want to exclude the possibility of having this technology in mobile phones and devices because of privacy worries, as long as we, the users, have control of who gets that information.
And as a final note, I think advertisers are barking up the wrong tree with this. They're far more likely to catch (and irritate) my eye with a huge billboard in my face than they are from a little chirp and cheesy low-res display on a device that's sitting comfortably in my backpack.:-P
From my experience, using so-called "beginner" languages like LOGO and BASIC was a waste of time. I found 300+ line BASIC listings impossible to read, and quickly lost patience with making the turtle run around in circles. It wasn't until I got to college and got to play with C and Lisp that I felt that I could really grok the code and do something useful with it. Although I think VB and OO-LOGO are probably much easier to deal with, I still have to get over that initial revulsion before I get anywhere near them.
Although it's a snap to get started in LISP, and I love the language personally, LISP becomes subtle and arcane quite quickly, at least in my experience with Commmon LISP (Scheme is, I'm sure, much easier). For example, try explaining how defmacro works to a kid, and see how far you get! C++ has the same problem when you start talking about copy constructors vs. assignment operators, templates, and multiple inheritance. It's not that a child couldn't get good mileage out of these languages, but I think they pose very challenging problems to the instructor, to make sure that the student don't get bogged down with tricky features in the beginning.
I think one of the most important things to consider is not the language, but the libraries and code that you can build upon. I would get pretty bored if I had to go through a year of implementing data structures, graphics libraries, and regexp matching algorithms before I could do anything with the language. And the learning scale on some existing libraries is absolutely daunting. So, for example, I wouldn't sic a kid on the Win32 C API, or start them off with the standard C/C++ library either.
Find out something your child would like to learn how to create (this could be a game, a mathematical exercise, or whatever), and tailor your choice of language and environment to help them accomplish that goal. It will be much more relevant to them, and will hopefully teach them lots of stuff that relates to the field they're playing with in the process (digital a/v, graphics, server issues, etc.)
Consider BeOS as a beginning programming environment. It's quite simple to learn how to program, there's relatively little legacy in the form of mysterious incantations, there's lots of sample code and libraries, you're using real pro-powered tools, and it's a blast to use.
Although I've also observed the trend of increasing diversity in musical listener's tastes, I wouldn't necessarily attribute it to Napster or MP3s. I could argue that this 'trend' has been evolving over the last 20 years or more. Of course, maybe I'm just extrapolating this trend from my own personal development as a music listener, and the fact that acquaintances around me have tended to age as well. Perhaps we need another "Entertainment Marketing Solution" to measure this trend as well, eh?
I disagree, however, with your premise that "variety is the enemy of profits." Note that:
The recording industry already has a mechanism for controlling the costs of a diversifying industry: spend less (or in some cases, next to nothing) on the bands that you don't think will sell hot. I've seen several examples of bands that you'd think would have it made because they managed to sign with a major record label. Sadly, it just ain't so.
Nothing is ever a sure deal in the recording industry (or, indeed, any industry). Even established bands can have a crappy run of it in the studio, and end up with a DOA album. So, it doesn't make sense for a record label to stack all its chips on one horse, even if it would superficially look like economies of scale were on their side.
Increasing variety gives a record label an opportunity to differentiate itself and artists from competing record labels and artists. If record labels can find and exploit these emerging markets, then the record labels win.
A wider, more diverse base of artists also gives a record label opportunities to do bundled promotions, antholgies, collections, mixes, and other ways to get you to buy more of their stuff.
A wider, more diverse base of artists means that the listener never gets bored. As a record label, you definitely don't want to be on the receiving end of a bored consumer.
So, my take is that variety is good -- even for the big guys. Apparently, they have the same take too, as is evidenced by the large number of sublabels and spin-off subsidiaries that focus on a wide variety of stuff.
Getting back to Napster, it is definitely the case that online music exchange makes obtaining rarities and obscure music a heckuvalot easier. This can only help folks in our position, who crave exposure to music that we might never otherwise be able to hear -- and by extension, artists whose music might never otherwise be heard. I'm talking about studio droppings and out-of-print stuff from bands that never get to the box-set stage of their careers. However, these are the parts of the industry that the RIAA already conveniently overlooks (see the "big labels doesn't necessarily mean big money" spin above). So, I'd say that Goldmine and $30 bootleggers stand to lose more from Napster than the RIAA ever will.
Slashdot Mirror
To answer your original comment about what is worth moderating up, healthy skepticism of any benchmark is fully warranted. Now, you may feel that completely unsupported accusations and groundless rumormongering constitute good entertainment value -- and it seems, from my experience, that the majority of Americans, at least, would agree with you :-). They certainly do not, however, constitute healthy skepticism, and are the last thing I want to see encouraged on Slashdot. That is the reason the post in question should be moderated into oblivion, not because The Man, who happens to be wearing a Be cap over his long hair, tells you to. I can't speak for you, but I would vastly prefer to see hard evidence or reason backing up a piece of juicy criticism before said criticism gets spewed into the net, not after. Perhaps I differ from some of my Slashdot colleagues, however, in this admittedly farfetched belief.
As far as bottlenecks go, Jason Sams and the other 3D god{esse}?s in these parts are the ultimate authority, but I could think of several places besides the hardware where the performance of OpenGL would be affected. Anything to shorten or optimize the rendering pipeline in software (using SIMD, dynamic generation of the pipeline for critical sections, native OpenGL vs. Glide support in the driver, etc.) is going to make a big difference for most interesting OpenGL work, and no, not every implementation of OpenGL has all of these software issues sorted out. So perhaps these numbers are not as thin-air as you think.
On a different tack, if you did do the equivalent of dropping "triangles" in the audio world, and could do it without anybody noticing, it would be heralded as "next-generation compression based on psychoacoustic perception heuristics," and the computing community would clap you on the back. Perhaps Be does have something up their sleeve after all! ( Now now, that's of course a groundless rumor, and I shouldn't have said it. Please don't take it seriously. :-P )
My point was, knowing location is an important ability for mobile communication. So, better ability to communicate location is a Good Thing -- provided that steps are taken to protect your location from simply being handed out to destinations without your knowledge and consent.
This whole privacy issue is also easily surmountable if you must grant permission for your location to be disclosed as a security setting on the phone (disregarding the FCC/EMS requirements, since there's no evidence that they're about to start up an advertising racket).
High-precision GPS will be an absolute godsend when it comes to things like navigation. I wouldn't want to exclude the possibility of having this technology in mobile phones and devices because of privacy worries, as long as we, the users, have control of who gets that information.
And as a final note, I think advertisers are barking up the wrong tree with this. They're far more likely to catch (and irritate) my eye with a huge billboard in my face than they are from a little chirp and cheesy low-res display on a device that's sitting comfortably in my backpack. :-P
- From my experience, using so-called "beginner" languages like LOGO and BASIC was a waste of time. I found 300+ line BASIC listings impossible to read, and quickly lost patience with making the turtle run around in circles. It wasn't until I got to college and got to play with C and Lisp that I felt that I could really grok the code and do something useful with it. Although I think VB and OO-LOGO are probably much easier to deal with, I still have to get over that initial revulsion before I get anywhere near them.
- Although it's a snap to get started in LISP, and I love the language personally, LISP becomes subtle and arcane quite quickly, at least in my experience with Commmon LISP (Scheme is, I'm sure, much easier). For example, try explaining how defmacro works to a kid, and see how far you get! C++ has the same problem when you start talking about copy constructors vs. assignment operators, templates, and multiple inheritance. It's not that a child couldn't get good mileage out of these languages, but I think they pose very challenging problems to the instructor, to make sure that the student don't get bogged down with tricky features in the beginning.
- I think one of the most important things to consider is not the language, but the libraries and code that you can build upon. I would get pretty bored if I had to go through a year of implementing data structures, graphics libraries, and regexp matching algorithms before I could do anything with the language. And the learning scale on some existing libraries is absolutely daunting. So, for example, I wouldn't sic a kid on the Win32 C API, or start them off with the standard C/C++ library either.
- Find out something your child would like to learn how to create (this could be a game, a mathematical exercise, or whatever), and tailor your choice of language and environment to help them accomplish that goal. It will be much more relevant to them, and will hopefully teach them lots of stuff that relates to the field they're playing with in the process (digital a/v, graphics, server issues, etc.)
- Consider BeOS as a beginning programming environment. It's quite simple to learn how to program, there's relatively little legacy in the form of mysterious incantations, there's lots of sample code and libraries, you're using real pro-powered tools, and it's a blast to use.
Hope this helps!I disagree, however, with your premise that "variety is the enemy of profits." Note that:
- The recording industry already has a mechanism for controlling the costs of a diversifying industry: spend less (or in some cases, next to nothing) on the bands that you don't think will sell hot. I've seen several examples of bands that you'd think would have it made because they managed to sign with a major record label. Sadly, it just ain't so.
- Nothing is ever a sure deal in the recording industry (or, indeed, any industry). Even established bands can have a crappy run of it in the studio, and end up with a DOA album. So, it doesn't make sense for a record label to stack all its chips on one horse, even if it would superficially look like economies of scale were on their side.
- Increasing variety gives a record label an opportunity to differentiate itself and artists from competing record labels and artists. If record labels can find and exploit these emerging markets, then the record labels win.
- A wider, more diverse base of artists also gives a record label opportunities to do bundled promotions, antholgies, collections, mixes, and other ways to get you to buy more of their stuff.
- A wider, more diverse base of artists means that the listener never gets bored. As a record label, you definitely don't want to be on the receiving end of a bored consumer.
So, my take is that variety is good -- even for the big guys. Apparently, they have the same take too, as is evidenced by the large number of sublabels and spin-off subsidiaries that focus on a wide variety of stuff.Getting back to Napster, it is definitely the case that online music exchange makes obtaining rarities and obscure music a heckuvalot easier. This can only help folks in our position, who crave exposure to music that we might never otherwise be able to hear -- and by extension, artists whose music might never otherwise be heard. I'm talking about studio droppings and out-of-print stuff from bands that never get to the box-set stage of their careers. However, these are the parts of the industry that the RIAA already conveniently overlooks (see the "big labels doesn't necessarily mean big money" spin above). So, I'd say that Goldmine and $30 bootleggers stand to lose more from Napster than the RIAA ever will.