Springer-Verlag actually requires you to sign over the copyright. The copyright! You're not licensing the work to them to publish, but actually giving it away.
It depends a lot on the publisher, but the more enlightened ones only require a non-exclusive grant of the copyright. They have to have the copyright in some form so they can publish it, after all. Of course most of the enlightened publishers are really professional societies in the IT sector, like the IEEE and ACM. Support your professional society! They at least have other ways to make money than holding your papers hostage.
3D textures and texture-dependent lookups are completely reasonable for rendering rectilinear grids. It's only silly if you have unstructured data. But people are doing that on the GPU now as well. There were 2 papers on it at the IEEE Visualization conference this year. Given that this is still an active research topic, perhaps a formal standardized API would be unwise at the moment.
Image packing and unpacking are pro features. Think parallel image compositing. The problem with making additional "Pro"
features part of the mandatory spec is that it pretty much
guarantees game cards won't have a compliant OpenGL driver. And if you can't have a compliant driver, why have one at all? The beauty of OpenGL has always been its extendability. It's just been too long since all the extensions that pretty much everyone implements have been pushed back into the required part of the spec.
I know that programmable pixel shaders etc. are useful, but why does OpenGL not specify things like raytraced and radiosity lighting models, along with voxel primitives, and features for window and page oriented output of arbitary geometry (including support for true curves/surfaces etc.) ala Postscript.
So, from last to first: SGI actually had the beginnings of a PostScript output extension (called GLS) working and there was a note on their page that said something to the effect of "if people are interested, we could make it a formal GL extension." Apparently, no one was interested. It's kind of moot now anyway, since all the programmability features make a PostScript engine very difficult and near-useless for many applications.
Voxel primitives are there: 3D textures.
Radiosity and raytracing are difficult to put in hardware because they would require you to have a full description of the scene in memory. This flies agains a fundamental assumption of OpenGL that geometry may be streamed to the video card -- keep the state small. A scene description language that performs radiosity calculations should be a layer on top of OpenGL, not subsumed into it. OpenGL's design intent has always been "Provide a wafer-thin layer of sanity over the graphics hardware so that people can make things go really fast." It has never been "Make programming graphics easy." Making graphics programming easy is the job of other layers such as Crystal Space (games) or VTK (visualization).
It's not necessarily a bad thing to have an exam that is so hard. There's nothing carved in stone saying that exams must be planned so that at least one student will correctly answer 90 percent of the questions.
I know some professors who like to put unanswered questions about their current research on exams; it's not meant to flunk the students, but to see how students go about answering the question. You don't have to be correct to show that you know how to organize your thoughts -- which is often the most valuable skill you will receive from an education, if you are lucky enough to get it at all.
It used to be that a product was feature complete when you could use it to check your e-mail. Since cell phones had that first, it seems they're upping the ante:
A product is not feature-complete unless you can play Doom with it
"On two occasions I have been asked [by members of Parliament!], `Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question."
- Charles Babbage
I don't think this would be so beneficial for more modern cars, as they are pretty advanced in this particular field. The new Mercedes R230 (SL) has a really low cw value of 0.29 for an open roadster.
It's not that they can't reduce the drag coefficient any more... that's relatively easy.
The tradeoff is that you lose downforce (and thus
stability). If you don't have downforce on the
front tires, you can't steer. If you don't have
it in the rear, you can't accelerate.
Remember, skin friction (between the air and
the car's skin) slows the car down, but the
change in the momentum of the air as it
slows down increases the pressure on the car, pushing it on to the ground (or away from the ground if you're not a very good automotive engineer:-).
That's why F1 cars have drag coefficients ~1 and
huge fins in front & back... they want to push
the tires down and have power to spare.
Since IBM already sells an operating system for
enterprise use (AIX) and desktop use (Windows),
where is it that you intend to push Linux as a
solution? Or is the goal of supporting Linux
more to accomodate customers who want it than to
introduce it as a preferred solution?
Re:Vertical markets with nice profit margins
on
Compaq Shifts Focus
·
· Score: 1
Yeah, there's a lot of markup on the price. But a
lot of it goes to cover insurance costs on many
of the applications you mentioned. Especially
medical devices. When a computer controls
something that could hurt people (or through
inaction allow harm), insurance companies start
charging very high rates.
> Hmmmm.... Since it's all about information wanting to be free,
>maybe RMS should offer his courses free of charge. I digress, however.
MIT just announced free course materials over
the web (http://web.mit.edu/ocw/). You could
say that attending class was like answering
support questions on the material.
Despite the story intro, I've seen many posts
asking for an Exchange-alike. Let me say:
Please don't make some huge monolithic copy of
a huge monolithic microsoft product;
if you want to do two tasks, make two tools with
well-defined interfaces. Don't just
integrate calendaring and e-mail so that anyone
who wants to use your calendar service has to
use your e-mail too. The power of Unix is
small tools that talk to each other. If you want
your database to interact with Exchange clients,
write a separate backend rather than constraining
the whole project to deal with microsoft's
structure.
Please don't reinvent the wheel. Use standards
(like LDAP for a contact database, for instance).
Everyone seems to be intent on MPEG encoding the audio from the headphone jack if that's what it takes to copy encrypted music. I wouldn't be surprised if watermarks could be used to make re-compressed music sound awful. Ever xeroxed a check or a college transcript? Not very usable, is it? If people are going to make digital copies of watermarked music, I suspect it will have to be losslessly compressed. Not that that will be a big deal as even current digital media (DVD-RAM) holds much more than a CD's worth anyway.
Although the DOJ asked for an OS company and an apps company, they were not clear at all about where the line should be drawn. They have a definition of Operating System, but it's rather simple. There are a lot of marginal areas where at least parts of services could arguably be included in the OS and a lot of areas where Windows contains more than the definition of OS (like the GUI), but would be very difficult for code to be excised. If they aren't careful, Microsoft will move tons of stuff into the OS group to avoid having to publish APIs:
File sharing: SMB stuff really shouldn't be considered part of the OS... it should be a userland daemon with some kernel support, like NFS. Of course, it's already plastered into Windows like bubble gum to the bottom of a school desk.
Encryption services. Although authentication services need to be in any multi-user OS (though 95/98 are debatably NOT), encryption services really don't have anything to do with the OS as the DOJ has defined it. This is a particularly tricky issue if file sharing (SMB) gets to stay in the OS.
User interface: Why should a widget set and session manager be part of the operating system?
Audio and video codecs should not be in the OS except the drivers for hardware implementations.
Gaming libraries. Yes, direct hardware access like DirectX is an OS item, but things like joystick configuration management aren't.
Scripting languages. No, Visual BASIC doesn't have anything to do with allocating access to devices or CPU time. I wouldn't be surprised if some type of scripting language got plopped in the OS group. Rather than document the fast OS calls, they could issue an API for a slow wrapper.
Servers: databases and web servers should not be in the kernel (although khttpd does make it a little arguable for static content).
The most popular games today seem to be first person shooters. This wasn't always so... until real-time 3D software (and now hardware) developed, the shoot-em-up games shared popularity much more evenly with other types. What does interactive fiction need to give it a boost like the shoot-em-up genre has seen? Is it
Technology ("Natural language software" is still an oxymoron).
A story that captures a large group of people's imagination.
A group of authors large enough (or one with enough caffeine) to branch a story so that it's tall enough to be interesting and wide enough to not seem guided or linear.
Recently, there's been a lot of noise over the WTO, IMF, and world bank. What do you think of the "transparency" buzzword being floated as a solution to the problems these organizations are having with their public relations? If transparency is the solution, was freedom of information the problem? If not, is the freedom to participate the problem? I ask because these problems got me started thinking about Free Software. I know that you stress free speech as an analogy to free software, but doesn't free software really consist of two freedoms:
information: how the software accomplishes what it does, and
participation: the ability to change the software to suit your needs?
The GPL provides both, the LGPL can be used to restrict the first (by closing the source to an application using the library), and Open Source software can restrict the second (by requiring royalties, for example).
I don't know if these are just less well known or whether people just don't like them as much but I like:
Contact by Carl Sagan. This one was way down the page, but I want to point out that the movie left out the most wonderful ending. I felt completely ripped off by an otherwise awesome movie. I guess they thought it was too technical for people to understand. (I disagree of course.)
Double Star by Heinlen. It's a pretty light read, but very fun, especially when you keep in mind when it was written.
The Puppet Masters by Heinlen.
Beggars in Spain by Nancy Kress. This book has some excellent insight into what might happen when people start fiddling with genetics on a large scale. The rest of the books in the series didn't appeal to me as much.
Interface by Stephen Bury (which is a pseudonym for Neal Stephenson and his uncle). With the presidential election coming up, this is a really entertaining read.
The Final Reflection by John M. Ford is an excellent Star Trek book and I do not usually enjoy series-style books. It only obliquely references the Star Trek characters from the TV show.
The Princess Bride translated/edited by William Goldman. I know this was mentioned below but I couldn't resist. It's perfect.
Her Majesty's Wizard by Christopher Stasheff is an excellent fantasy.
Surely You're Joking Mr. Feynman isn't science fiction, but it reads like it. It communicates Feynman's fascination with the way the universe works and his rather quirky approach to living in it.
-- I will make you shorter by the head. -- Elizabeth I
get leather smell when you add leather bucket seats option
For estimating value of trade-ins: you enter odometer reading and use "odor-o-meter" slider to identify the condition of your trade-in. Ranges from exhaust+burnt oil+unburnt gas to new car smell.
Slashdot Mirror
3D textures and texture-dependent lookups are completely reasonable for rendering rectilinear grids. It's only silly if you have unstructured data. But people are doing that on the GPU now as well. There were 2 papers on it at the IEEE Visualization conference this year. Given that this is still an active research topic, perhaps a formal standardized API would be unwise at the moment.
Voxel primitives are there: 3D textures.
Radiosity and raytracing are difficult to put in hardware because they would require you to have a full description of the scene in memory. This flies agains a fundamental assumption of OpenGL that geometry may be streamed to the video card -- keep the state small. A scene description language that performs radiosity calculations should be a layer on top of OpenGL, not subsumed into it. OpenGL's design intent has always been "Provide a wafer-thin layer of sanity over the graphics hardware so that people can make things go really fast." It has never been "Make programming graphics easy." Making graphics programming easy is the job of other layers such as Crystal Space (games) or VTK (visualization).
It's not necessarily a bad thing to have an exam that is so hard. There's nothing carved in stone saying that exams must be planned so that at least one student will correctly answer 90 percent of the questions.
I know some professors who like to put unanswered questions about their current research on exams; it's not meant to flunk the students, but to see how students go about answering the question. You don't have to be correct to show that you know how to organize your thoughts -- which is often the most valuable skill you will receive from an education, if you are lucky enough to get it at all.
... have been around a long time:
"On two occasions I have been asked [by members of Parliament!], `Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question."
- Charles Babbage
Sewer, Gas, and Electric and Fool On the Hill.
Remember, skin friction (between the air and the car's skin) slows the car down, but the change in the momentum of the air as it slows down increases the pressure on the car, pushing it on to the ground (or away from the ground if you're not a very good automotive engineer :-).
That's why F1 cars have drag coefficients ~1 and huge fins in front & back... they want to push the tires down and have power to spare.
Since IBM already sells an operating system for enterprise use (AIX) and desktop use (Windows), where is it that you intend to push Linux as a solution? Or is the goal of supporting Linux more to accomodate customers who want it than to introduce it as a preferred solution?
Yeah, there's a lot of markup on the price. But a lot of it goes to cover insurance costs on many of the applications you mentioned. Especially medical devices. When a computer controls something that could hurt people (or through inaction allow harm), insurance companies start charging very high rates.
>maybe RMS should offer his courses free of charge. I digress, however.
MIT just announced free course materials over the web (http://web.mit.edu/ocw/). You could say that attending class was like answering support questions on the material.
See the DVD FAQ. The shiny stuff on CD-R's isn't so reflective at the frequency that DVD lasers use.
If you're having trouble with your GeForce2 MX, try this FAQ.
Please don't make some huge monolithic copy of a huge monolithic microsoft product; if you want to do two tasks, make two tools with well-defined interfaces. Don't just integrate calendaring and e-mail so that anyone who wants to use your calendar service has to use your e-mail too. The power of Unix is small tools that talk to each other. If you want your database to interact with Exchange clients, write a separate backend rather than constraining the whole project to deal with microsoft's structure.
Please don't reinvent the wheel. Use standards (like LDAP for a contact database, for instance).
Everyone seems to be intent on MPEG encoding the audio from the headphone jack if that's what it takes to copy encrypted music. I wouldn't be surprised if watermarks could be used to make re-compressed music sound awful. Ever xeroxed a check or a college transcript? Not very usable, is it? If people are going to make digital copies of watermarked music, I suspect it will have to be losslessly compressed. Not that that will be a big deal as even current digital media (DVD-RAM) holds much more than a CD's worth anyway.
- information: how the software accomplishes what it does, and
- participation: the ability to change the software to suit your needs?
The GPL provides both, the LGPL can be used to restrict the first (by closing the source to an application using the library), and Open Source software can restrict the second (by requiring royalties, for example).- Contact by Carl Sagan. This one was way down the page, but I want to point out that the movie left out the most wonderful ending. I felt completely ripped off by an otherwise awesome movie. I guess they thought it was too technical for people to understand. (I disagree of course.)
- Double Star by Heinlen. It's a pretty light read, but very fun, especially when you keep in mind when it was written.
- The Puppet Masters by Heinlen.
- Beggars in Spain by Nancy Kress. This book has some excellent insight into what might happen when people start fiddling with genetics on a large scale. The rest of the books in the series didn't appeal to me as much.
- Interface by Stephen Bury (which is a pseudonym for Neal Stephenson and his uncle). With the presidential election coming up, this is a really entertaining read.
- The Final Reflection by John M. Ford is an excellent Star Trek book and I do not usually enjoy series-style books. It only obliquely references the Star Trek characters from the TV show.
- The Princess Bride translated/edited by William Goldman. I know this was mentioned below but I couldn't resist. It's perfect.
- Her Majesty's Wizard by Christopher Stasheff is an excellent fantasy.
- Surely You're Joking Mr. Feynman isn't science fiction, but it reads like it. It communicates Feynman's fascination with the way the universe works and his rather quirky approach to living in it.
--I will make you shorter by the head. -- Elizabeth I