Well it's great to have some development libraries for the SGI video hardware without having to pay for them, now that SGI no longer makes video hardware. Sort of like getting to look at the source code to Mozilla after Netscape folded up, getting to look at the source code to Hexen after Quake III came out. They're consistantly products that the developers gave up on because the users didn't care about them anymore. And so I don't really care about IRIS performer even though it's still nice to not have to pay for something.
Never knew there was a race to decode gene sequences using computers. There is a race for low paid women to load the sequencers but the "decoding" of the sequence is not the limiting factor. You've got to be damn good to get into those labs. Harvard PhD quality.
Well you understand that a significant portion of Slashdot's operating budjet comes from Andover.net and they most certainly determine the narrowness of content. Slashdot of 2 years ago would never get financed simply because the stories were too technical and unusual. Probably they could have stuck with the shared T1 in Michigan and we would all be posting to segfault or some other portal, but unlike those portals the Slashdot people dropped the obscure and unusual and started quoting CNN and MSNBC in exchange for more money. Nothing comes without a cost, whether the programmer borrows $100,000 to pay for it or the user pays for it. I remember when a virtual server costed $25 a month with unlimited bandwidth. Nowadays you can't get that for $200 a month. You can't depend on corporations to stay up and keep providing the same scope of the content.
Well what you're in the market for was called a "capture card" and designed specifically to record video, and not usually designed to drive your X server. You should spend no more than $80 on a capture board. In 2002 all American broadcasters are required to broadcast in HDTV and in 2006 all American broadcasters are required to offer only HDTV. Unfortunately no capture boards support HDTV today and probably never will. No capture boards are fully supported in Linux but you can get close. The Hauppage WinTV was about as well supported as you could get. The problem with what you're doing is that less and less people are doing it so your options are decreasing. SGI is an e-commerce company not a multimedia company anymore. IOmega and a lot of companies stopped making capture boards years ago. Even Avid is moving away from video and into e-commerce. So don't get too attached to it.
DVD-RAM's as they're called have been around for about 2 years. All you needed to record TV shows was a recording program and CD recording software. On a 5 gig DVD-RAM you can store 2 hours of 640x480 15fps MJPA movies, or a virtually unlimited size MPEG-I movie. The hardware only costs $500 today. A blank DVD-RAM costs $25, and is fixed to match the cost of official movie releases.
You know we keep hearing about how such and such a side effect of open sourcing something makes it better: well you can put some energy into answering phones instead of coding all day and increase its value. You can conceptually increase its value by expanding its user base. You can sell t-shirts with your product's name on it and increase its value. It seems as people get more experienced with software they find engineering to amount to less and less of a product's value. When do we finally tell engineers to shove it and focus on marketing instead?
5 years ago the standard prodecure was to recompile your program for the new Pentium architecture and instantly get a 10x speed boost. Today the standard prodecure is assembly language so it depends on where we are in the punctuated evolution. After the next recession they'll work their assess off to speed up processors and we'll once again benefit from recompiling more than optimizing. If you're doing database and web applications don't bother optimizing but write a game instead.
Most of the Linux users out there would rather die than use a GUI to get a job done. They'd much rather perform hand calculations and do everything with a small set of utilities. Whether or not the level of graphical automation has anything to do with the quality of the output is up to speculation. Despite the existance of the Linux command line for the last 9 years industries which only existed in the last 2 years have defaulted to consistantly producing TV shows on graphical interfaces running Win NT.
If all you did was buy $1000 worth of VALinux stock at their IPO price you would have about $7000 when it peaked on its first day. Think of the software you could write with that money. Take three months off and code Linux software. Compare to getting pages and pages of certification and credentials together, trying to find a job with only a 1% chance of success, and writing Visual Basic scripts for 3 months instead. Of course, I didn't get the letter.
Well there's http://cg.cs.tu-berlin.de/~ki/engines.html a comprehensive list of many 3D graphics engines.
If you want the best stuff you have to break a few patents and read a few research papers. The latest motion vectoring techniques and modelling techniques are usually only in IEEE journals or something really dry in your college library. They're also usually patented, so you can either spend lots of money on the latest patented algorithms or spend lots of money coding obsolete, unpatented algorithms.
This will no doubt result in more layoffs from SGI's engineering team. We remember how they laid off 3000 engineers in April. Now is a bad time to be in anything but e-commerce. SGI's e-commerce strategy is definitely to get rid of everything else. That's the method of operation in Silicon Valley: if it doesn't work don't fix it. Get rid of it.
Well computer magazines have changed over the years so whether or not they're dead depends on how you define them. When I was a kid COMPUTE! magazine was 300 pages long and published complete listings of programs and the only way to get them was to type in the listings. The writing was very technical back then and well above what a modern magazine could get away with. Then they stopped publishing software entirely and started writing philosophical articles on a very technical level. Now they're either reviewing software or writing about biotech or the meaning of life. So the days of COMPUTE! magazine are definitely over but computer magazines are just adapting to match today's less technically oriented audience.
This is just the reaction you'd expect from someone not involved in UNIX for very long. A lack of vendor support is nothing new. The turnover rate is so high for UNIX users that every year you get new Windows converts overreacting to vendor apathy they have never seen before and as old users drop out every year, the overall feedback from UNIX users stays the same year after year.
The people you run into on trade show floors are usually just out of college and putting their CS degree to good use in sales. They have no clue what's going on and they're certainly not going to know what UNIX is. If they ever start questioning the Microsoft way of life, it'll be long after they've been promoted out of sales. Next year you'll have completely new people on the floor and get the same reaction from whoever just converted over from Windows.
Well the benchmarks still give only a 33% improvement over the trusty 550Mhz Celeron. A A dual Celeron 550 would still give you more MIPS. For encoding massive amounts of mp3s and rendering video in parallel the dual Celeron 550 has held its mark longer than any configuration since its time.
Rather dissapointing that we're at a point where every week a new graphics card comes out with a proprietary chipset yet we can only run games on one chipset introduced 5 years ago that costs twice what most chipsets cost nowadays. 3DImage 975: $30 cheap. Voodoo 3: $73 obsolete.
Well, I believe there have been several 100 gigabit fibers running the entire width of the country for quite some time. Maybe if Microsoft does it it's supposed to be an innovation but Microsoft is really only one company. There are many other companies building networks much faster than Microsoft's.
Well I'm still a staunch opponent of using goto. Maybe goto makes you feel superior because you're using Microsoft's crown jewel of BASIC programming but unless you've got employers to impress Microsoft isn't everything.
How to program the rapidly advancing chipsets
on
New ATi 3D Chip
·
· Score: 3
We're entering a time when the only way to program 3D cards is manipulating the registers directly. There's no way you can get these cards supported by a library in time for the next generation of cards. We're leaving the age of abstraction and it's becoming more important for software to access the graphics hardware directly.
Having worked with the gcc compiler for many years I can say that math fluctuations are increadibly easy to introduce through optimization. Floating point operations, though rarely used in CS education, are rampant in the Seti code. The problem isn't in producing bad data but in differences in the output that most floating point optimizations produce. Floating point operations are much more prone to fluctuations due to optimization. The optimized seti clients are guaranteed to produce different results than the original clients.
Since they haven't been able to collect enough data by orders of magnitude to feed their clients, there's little use to optimized Seti clients outside of learning how the Seti client works.
Funny how global media got an award for tweeking RealServer for Linux and providing a fancier package for RealPlayer. Global Media first and foremost sells CD's, competing directly with the free mp3 portals, and their repackaged version of RealPlayer isn't even part of their total business strategy.
When Linus graduated, he was asked why he didn't take a job at RedHat devoting his full time to writing the kernel. His response was that he didn't want to be swayed by his employer's interests into decisions that would harm the kernel's functionality. He joined Transmeta to keep his employment completely seperate from his Linux hobby.
Well what he's doing today is very close to what he wanted to avoid. Transmeta developes an embedded processor. Every time he speeks he chants about getting Linux into embedded systems and today he actually said Linux is more practical for embedded systems than a desktop. You have to wonder how much pressure he experiences working for an embedded systems company to divert his efforts from other kernel improvements to fitting the kernel into Transmeta's agenda.
When the Gimp 1.0 came out in early 1998 that spurred a frenzy of students to produce open source applications to equal The Gimp both in size and complexity, AudioTechQue, AbiWord, GnoMoney, Gnuotes, etc. One of these mega apps was Mnemonic, intended to be the world's first completely open source web browser. Mnemonic was mentioned on Slashdot every day and it was even predicted to kill Internet Explorer 4. Well a lot changes in 2 years. Hardly anyone using Linux today has ever heard of Mnemonic and most of those Gimp spinoffs have either died off or slowed to geological rates of progression.
I know someone who used to work for them back when they were one of the first virtual hosting providers. Hard to believe they stuck with BSDi as long as they did, when the amount of competition in that business has gone from 0 to infinity. Hopefully the Utah weather hasn't gotten any worse than it already was.
"We want to keep as many of those people as we c"
on
Red Hat Gets New CEO
·
· Score: 3
"We want to keep as many of those people as we can find productive and useful and challenging [jobs] for,"
In other words, don't be surprised to see the first big layoffs in the Linux world. Hundreds of CS students who worked like mad for 4 years to get this far and it's all over in one single swipe of the pen. They'll join last week's 300 layoffs at Avid and last month's 3000 layoffs at SGI. Thank god I don't work for Cygnus.
The best way to get newbies to the next level is to keep them using it after college. All the evidence points to very transient usage. They pick it up in college and give it up when they start working so we have a user base which is primarily interested in starting the basic operating system and trying out different distributions but rarely getting to the point of actually doing anything with it.
Slashdot Mirror
Well it's great to have some development libraries for the SGI video hardware without having to pay for them, now that SGI no longer makes video hardware. Sort of like getting to look at the source code to Mozilla after Netscape folded up, getting to look at the source code to Hexen after Quake III came out. They're consistantly products that the developers gave up on because the users didn't care about them anymore. And so I don't really care about IRIS performer even though it's still nice to not have to pay for something.
Never knew there was a race to decode gene sequences using computers. There is a race for low paid women to load the sequencers but the "decoding" of the sequence is not the limiting factor. You've got to be damn good to get into those labs. Harvard PhD quality.
Well you understand that a significant portion of Slashdot's operating budjet comes from Andover.net and they most certainly determine the narrowness of content. Slashdot of 2 years ago would never get financed simply because the stories were too technical and unusual. Probably they could have stuck with the shared T1 in Michigan and we would all be posting to segfault or some other portal, but unlike those portals the Slashdot people dropped the obscure and unusual and started quoting CNN and MSNBC in exchange for more money. Nothing comes without a cost, whether the programmer borrows $100,000 to pay for it or the user pays for it. I remember when a virtual server costed $25 a month with unlimited bandwidth. Nowadays you can't get that for $200 a month. You can't depend on corporations to stay up and keep providing the same scope of the content.
Well what you're in the market for was called a "capture card" and designed specifically to record video, and not usually designed to drive your X server. You should spend no more than $80 on a capture board. In 2002 all American broadcasters are required to broadcast in HDTV and in 2006 all American broadcasters are required to offer only HDTV. Unfortunately no capture boards support HDTV today and probably never will. No capture boards are fully supported in Linux but you can get close. The Hauppage WinTV was about as well supported as you could get. The problem with what you're doing is that less and less people are doing it so your options are decreasing. SGI is an e-commerce company not a multimedia company anymore. IOmega and a lot of companies stopped making capture boards years ago. Even Avid is moving away from video and into e-commerce. So don't get too attached to it.
DVD-RAM's as they're called have been around for about 2 years. All you needed to record TV shows was a recording program and CD recording software. On a 5 gig DVD-RAM you can store 2 hours of 640x480 15fps MJPA movies, or a virtually unlimited size MPEG-I movie. The hardware only costs $500 today. A blank DVD-RAM costs $25, and is fixed to match the cost of official movie releases.
You know we keep hearing about how such and such a side effect of open sourcing something makes it better: well you can put some energy into answering phones instead of coding all day and increase its value. You can conceptually increase its value by expanding its user base. You can sell t-shirts with your product's name on it and increase its value. It seems as people get more experienced with software they find engineering to amount to less and less of a product's value. When do we finally tell engineers to shove it and focus on marketing instead?
5 years ago the standard prodecure was to recompile your program for the new Pentium architecture and instantly get a 10x speed boost. Today the standard prodecure is assembly language so it depends on where we are in the punctuated evolution. After the next recession they'll work their assess off to speed up processors and we'll once again benefit from recompiling more than optimizing. If you're doing database and web applications don't bother optimizing but write a game instead.
Most of the Linux users out there would rather die than use a GUI to get a job done. They'd much rather perform hand calculations and do everything with a small set of utilities. Whether or not the level of graphical automation has anything to do with the quality of the output is up to speculation. Despite the existance of the Linux command line for the last 9 years industries which only existed in the last 2 years have defaulted to consistantly producing TV shows on graphical interfaces running Win NT.
If all you did was buy $1000 worth of VALinux stock at their IPO price you would have about $7000 when it peaked on its first day. Think of the software you could write with that money. Take three months off and code Linux software. Compare to getting pages and pages of certification and credentials together, trying to find a job with only a 1% chance of success, and writing Visual Basic scripts for 3 months instead. Of course, I didn't get the letter.
Well there's http://cg.cs.tu-berlin.de/~ki/engines.html
a comprehensive list of many 3D graphics engines.
If you want the best stuff you have to break a few patents and read a few research papers. The latest motion vectoring techniques and modelling techniques are usually only in IEEE journals or something really dry in your college library. They're also usually patented, so you can either spend lots of money on the latest patented algorithms or spend lots of money coding obsolete, unpatented algorithms.
This will no doubt result in more layoffs from SGI's engineering team. We remember how they laid off 3000 engineers in April. Now is a bad time to be in anything but e-commerce. SGI's e-commerce strategy is definitely to get rid of everything else. That's the method of operation in Silicon Valley: if it doesn't work don't fix it. Get rid of it.
Well computer magazines have changed over the years so whether or not they're dead depends on how you define them. When I was a kid COMPUTE! magazine was 300 pages long and published complete listings of programs and the only way to get them was to type in the listings. The writing was very technical back then and well above what a modern magazine could get away with. Then they stopped publishing software entirely and started writing philosophical articles on a very technical level. Now they're either reviewing software or writing about biotech or the meaning of life. So the days of COMPUTE! magazine are definitely over but computer magazines are just adapting to match today's less technically oriented audience.
This is just the reaction you'd expect from someone not involved in UNIX for very long. A lack of vendor support is nothing new. The turnover rate is so high for UNIX users that every year you get new Windows converts overreacting to vendor apathy they have never seen before and as old users drop out every year, the overall feedback from UNIX users stays the same year after year.
The people you run into on trade show floors are usually just out of college and putting their CS degree to good use in sales. They have no clue what's going on and they're certainly not going to know what UNIX is. If they ever start questioning the Microsoft way of life, it'll be long after they've been promoted out of sales. Next year you'll have completely new people on the floor and get the same reaction from whoever just converted over from Windows.
Well the benchmarks still give only a 33% improvement over the trusty 550Mhz Celeron. A A dual Celeron 550 would still give you more MIPS. For encoding massive amounts of mp3s and rendering video in parallel the dual Celeron 550 has held its mark longer than any configuration since its time.
Rather dissapointing that we're at a point where every week a new graphics card comes out with a proprietary chipset yet we can only run games on one chipset introduced 5 years ago that costs twice what most chipsets cost nowadays. 3DImage 975: $30 cheap. Voodoo 3: $73 obsolete.
Well, I believe there have been several 100 gigabit fibers running the entire width of the country for quite some time. Maybe if Microsoft does it it's supposed to be an innovation but Microsoft is really only one company. There are many other companies building networks much faster than Microsoft's.
Well I'm still a staunch opponent of using goto. Maybe goto makes you feel superior because you're using Microsoft's crown jewel of BASIC programming but unless you've got employers to impress Microsoft isn't everything.
We're entering a time when the only way to program 3D cards is manipulating the registers directly. There's no way you can get these cards supported by a library in time for the next generation of cards. We're leaving the age of abstraction and it's becoming more important for software to access the graphics hardware directly.
Having worked with the gcc compiler for many years I can say that math fluctuations are increadibly easy to introduce through optimization. Floating point operations, though rarely used in CS education, are rampant in the Seti code. The problem isn't in producing bad data but in differences in the output that most floating point optimizations produce. Floating point operations are much more prone to fluctuations due to optimization. The optimized seti clients are guaranteed to produce different results than the original clients.
Since they haven't been able to collect enough data by orders of magnitude to feed their clients, there's little use to optimized Seti clients outside of learning how the Seti client works.
Funny how global media got an award for tweeking RealServer for Linux and providing a fancier package for RealPlayer. Global Media first and foremost sells CD's, competing directly with the free mp3 portals, and their repackaged version of RealPlayer isn't even part of their total business strategy.
When Linus graduated, he was asked why he didn't take a job at RedHat devoting his full time to writing the kernel. His response was that he didn't want to be swayed by his employer's interests into decisions that would harm the kernel's functionality. He joined Transmeta to keep his employment completely seperate from his Linux hobby.
Well what he's doing today is very close to what he wanted to avoid. Transmeta developes an embedded processor. Every time he speeks he chants about getting Linux into embedded systems and today he actually said Linux is more practical for embedded systems than a desktop. You have to wonder how much pressure he experiences working for an embedded systems company to divert his efforts from other kernel improvements to fitting the kernel into Transmeta's agenda.
When the Gimp 1.0 came out in early 1998 that spurred a frenzy of students to produce open source applications to equal The Gimp both in size and complexity, AudioTechQue, AbiWord, GnoMoney, Gnuotes, etc. One of these mega apps was Mnemonic, intended to be the world's first completely open source web browser. Mnemonic was mentioned on Slashdot every day and it was even predicted to kill Internet Explorer 4. Well a lot changes in 2 years. Hardly anyone using Linux today has ever heard of Mnemonic and most of those Gimp spinoffs have either died off or slowed to geological rates of progression.
I know someone who used to work for them back when they were one of the first virtual hosting providers. Hard to believe they stuck with BSDi as long as they did, when the amount of competition in that business has gone from 0 to infinity. Hopefully the Utah weather hasn't gotten any worse than it already was.
"We want to keep as many of those people as we can find productive and useful and challenging [jobs] for,"
In other words, don't be surprised to see the first big layoffs in the Linux world. Hundreds of CS students who worked like mad for 4 years to get this far and it's all over in one single swipe of the pen. They'll join last week's 300 layoffs at Avid and last month's 3000 layoffs at SGI. Thank god I don't work for Cygnus.
The best way to get newbies to the next level is to keep them using it after college. All the evidence points to very transient usage. They pick it up in college and give it up when they start working so we have a user base which is primarily interested in starting the basic operating system and trying out different distributions but rarely getting to the point of actually doing anything with it.