This is a big deal, a very big deal. "Standard: something considered by an authority or by general consent as a basis of comparison; an approved model." How does OOXML fit into that? It doesn't, it's half baked, no one can implement it including MS from what I read. Yet I bet they keep calling their files OOXML, but of course no one else can read it. If in 20 years there are loads important MS OOXML files and all we have to work out how to read this is the ISO standard, then the data is lost. This is what standards are to avoid. ISO is a quality stamp, OOXML being ISO stamped calls into question the quality of the ISO stamp. And worse this all happened through skullduggery and corruption.
If the software is done on the cheap, it will crash even if the system doesn't underneath it. Expensive hardware for cheap development costs, but as I said, bet the cost of keeping it limping on makes it all work out more expensive. In the long run, I would be very surprised if it doesn't work out cheaper to do it with cheap kit, cut down Linux and decent developers.
The point is that the system can be stripped to only what is required. You can not strip windows in the same way. Less stuff running is less things to compromise, less hardware requirements, less things to maintain, etc etc. Bet it's mainly the cheap crappy companies that use windows as the basis of their hardware/software solution, and also bet in the long run the nightmare mess they create costs to maintain makes it expensive. False economy.
I don't want any child of mine's head filled with every silly idea that apposes a valid scientific hypothesis. That has no place in a science class room. They going to start insisting the math classes state the value for PI could be 3 because that's what the bible states?
I'm sure most programming nerds program something at some point that mutates with random then selects from it? I wrote a simple ant java thing where each ant started with random rules of how to get to the food and back to the hive. If they don't get to the food, and then the hive, after a set time, they get fresh random rules. After 30 min of so, each ant has the same rules. Real life clearly has mutations, and clearly has selection forces, no brainer. It's so easy to see logically it's true, that's why creationists fear it so much.
Why on earth would I want a touch screen keyboard? You can't feel the keys! I touch type solely on the feel of the keys. Why would I want to have to go back to looking down? If I using my hands to input, having to look at them as they do so is wasting my time. Yer they look good on a set of star trek, but in reality that ship would have been destroyed long ago but villains with keyboards they don't have to look down at to press fire. Until the touch screen raises where buttons are, you are using one sense less while working, and if you aren't using that touch screen to look at, what is the point?
* get lots of data
* use computers to search for pattens in that data
* use those patterns
instead of thinking up models and testing them.
But don't you have to have some model to define the parameters of the data you are collecting? I'm not sure there is anything new here..... Isn't it just data mining?
This is bait, but hey, it's tasty even with the hook.
The reason open source will win out is because you really can fix it. It's not work related, but a good example. I was processing some video and the java app doing it crashed out telling what line the problem was it was a simple -1 index problem. I could just stick a index check in and carry on (the bad index was due to bad data I was running the app to fix!). A good proportion of crashes are something silly like that, or a null pointer being used/freed, any programmer can see it and fix it with out having to get to know the whole code base. I get crashes in closed software at work some times, and I can't do a thing about it! I can submit it as a bug to the vendor, and if I'm very very lucky I will get a patch, but most of the time your told it will be fixed in the next version and maybe a work round. I did programmer support for a middleware API for years, I use to do exactly that to customers myself. If there is a work round, ok, unless it's not really workable, but if not, what then? I'm willing to fix it but unable to because I don't have the source. You then get into "Are you willing to pay mega bucks to get it fixed?", again, I use to have to do this to customers myself. With open source you can fix it yourself if need be, or pay to get it done just like you would in closed source. So basically with open source you have all of the safety net of closed source (if you pay for support that is) with the extra advantage that you can help yourself.
Writing apps for huge number of processors is a bit of a nightmare, but ray tracing is a problem that fits very neatly in this world. A ray tracer can be very small, so once loaded on a processor, it can just munch through data (the scene) until it's finished. A collection of stream processors with read access to main memory (or a large pool of memory where the scene lives) would be perfect. If you have a pixel shader with read access to main memory, draw a single quad over the view, and have each ray tracer in pixel shader form and you get what I mean.
Which is why ARM is going to be laughing. ARM and Linux is going to be huge on the UMPC market if the battery life is going to be as good as promised. If you running Linux why do you care about chip architecture? Users want a cheap machine with a battery that lasts for ages with a powerful fully functional operating system. That must scare Intel as well as MS. MS will do what? Port XP to ARM, upgrade CE to full OS standards? What advantage has Windows over Linux without being able to run standard Windows software? Intel, well er, really they should have a line of processor where power use is the number one issue, not x86 compatibility (though x86 chips that need less power are also needed, just not in the UMPC market). Could be very interesting times.
I'm not sure it will be history repeating because size and power matter more in these devices then in the desktop. It will be hard for x86 machines to win out on price, size and power over ARM. They are paying a heavy price for x86, all for compatibility I really don't think we need.
No-one wants an OS that requires a beefy machine just to run. We want a tiny OS that leaves us most of the machine to do with as we will. The desktop is morphing into a form of client via the browser. Who would want a fat OS to run their browser when a thin one will also run their browser, but faster on the same client machine. This is what the sub-laptops are all about. Thin OS, means you can have a thin/cheap client. Using a old OS isn't really a solution, you want one that scales with the hardware. I.e Linux. If MS wants to compete, they need MinWin. Not that I want them to, I want source code for when the docs fail me (though the Wine source is quite good for Win32 questions if msdn isn't answering you;-) ).
"Linux excels in various code structure metrics, but lags in code style. This could be attributed to the work of brilliant motivated programmers who aren't however efficiently managed to pay attention to the details of style. In contrast, the high marks of WRK in code style and low marks in code structure could be attributed to the opposite effect: programmers who are efficiently micro-managed to care about the details of style, but are not given sufficient creative freedom to structure their code in an appropriate manner.
"
How ever I was left wondering how it was possible to compare fairly? He already stated:
"Excluded from the kernel code are the device drivers, and the plug-and-play, power management, and virtual DOS subsystems. The missing parts explain the large size difference between the WRK and the other three kernels."
and reading I see even more of the drivers aren't there:
"The NT Hardware Abstraction Layer, file systems, network stacks, and device drivers are implemented separately from NTOS and loaded into kernel mode as dynamic libraries. Sources for these dynamic components are not included in the WRK. "
So it's not like for like. Maybe you would draw different conclusions if it was, maybe the Linux style issue is because of all the drivers the WRK lacks. So even though I think his conclusion sounds probable, I don't feel I can state it as so with any confidence.
I've heard people here say nice things about 3DSMax, and I must inform them that 3DSMax is a very bad peice of software. Sure as a interface it seams ok for knocking things together, but you'll hit a wall at some point, and it will really hurt. I am amazed this CAD package with a mountain of hacks on it is respected by anyone.
It's tech sucks very very very very badly. Get hold of the debug source (not complete source code, but quite a bit, much of it dated 1994). It sucks terribly. In the sdk, you are given the memory of objects directly, up casting and down casting every where (we just talking of (*) casting pointers, and people make mistakes.......). You seen the NULL function call error? Because of a bad upcast somewhere (in biped I think). Dealing with a editable mesh is different then dealling with a editable poly, dealing with a biped is completely different then dealing with a transform hierarchy, dealling with a xref object you must deference it first, ensuring it has a value (crash I've seen in Max). All the interface stuff is hard coded Win32. Defining parameter blocks is like a text book way of how it shouldn't be done. The scene graph is awful, no plug system for connections. From what source I've seen I can only feel sorry for a developer that has to work with it. I feel sorry for me when I have to use the sdk (just headers of the source they let you see, not a proper sdk).
3DSMax is terrible for animation houses because:
* it's complete lack of referencing of heirarchies. (Referencing is a must for large work, or rig update requires changing of n files).
* biped is a biped, try and do something more then that and your heading for trouble. (People do horrific things like bult together two bipeds trying to do what they should just use better software for).
* toolies have a nightmare supporting you.
We have finally dropped it for main stream animation because of years of nightmares supporting it while the maya teams breaze along.
Not looked at Blender, don't do computer art anymore and we don't use it here, but it can't be worse. As it's open source, if/when it reaches critical mass, it will leave Maya and XSI behind (can't have not left 3DSMax behind) as armies of toolies fix bugs and add features.
but it's part of the way programs used to be written. What's this use to be written?
We had a script kiddy here who made a horrid tangled mess, hard pathing everything, including to a directory called "Not_Used_Anymore". He didn't know better.
The horrors of hard pathing are alive and well.
Hard pathing is just coding in a path into your source like any other constant. Sometimes it makes sense to do it (small personal batch script for instance) and some will always do it where they shouldn't (in something compiled or large).
buy a ps2 and lego starwars. Even if you don't like starwars. My partner and I played through both games together and it was great fun. We are waiting for lego indiana jones! Most fun co-op series I've ever played.
If you want to play games, buy a games consol. Don't give the fat lazy software houses more resource to do the same thing, or the apps will continue to get fatter and slower.
Open source doesn't suffer as bad from this laziness because trimming the fat is always one of the number one things coders (not sales reps and managers) care about.
I really won't be surprised if we end up with pc (as in crap for games), consols (specifically for games) and workstations (great for games, but really for media work). We are pretty much there now, your crazy (or rich) if you splash out for a workstation to play games.
I've got a bee in my bonnet to day for modern big fat slow apps doing the same thing, only slower, then I was doing on my Acorn 15 years ago with 4Mhz and 1MB! It was all BBC BASIC with some ARM asm, on such slow spec, how could it be faster! HOW! I blame pc gamers and fat lazy software houses! GRRR!!!
I would say, depends... Different languages have different uses. What need is your current languages not fulfilling?
Might be worth looking into asm just so you think about what your code really boils down to.
On the other hand, if you just want a language to throw around and knock stuff together quickly, python is good. Really good. I could rave all day about python and it's interactive shell and flexibility.:-) (though seriously, it's worth having a script language under your belt for complex batch tasks at least)
If you only have done C++, might be worth doing pure C, working object oriented with structs, again it makes you think about what your C++ code boils down to.
However, what I would say is that some times rather then learning a new language, might be worth picking/debugging through an open source project that you have no idea how it works. (May I recommend a emulator with JIT). I think I've learnt far more from other people's source then I have from changing languages, and some times what I've learnt is the language, even if I thought I knew it!
but but but Microsoft hate open source, right from the start have fought against it.
Often on ground of protecting their intellectual property.
But here they are using some else's intellectual property (lego) claiming to be for open source.
WTF!
Is this related to publicly showing the horrors that are the old Office formats?
Policy change or new mask? While they refuse to play nice with standards and stop vender lock in, I won't believe it.
Open the source to current software, like Windows and Office and maybe I would beleive it.
It's all cosmetics.
Slashdot Mirror
Exactly! It is hopelessly naive to think this doesn't matter.
This is a big deal, a very big deal. "Standard: something considered by an authority or by general consent as a basis of comparison; an approved model." How does OOXML fit into that? It doesn't, it's half baked, no one can implement it including MS from what I read. Yet I bet they keep calling their files OOXML, but of course no one else can read it. If in 20 years there are loads important MS OOXML files and all we have to work out how to read this is the ISO standard, then the data is lost. This is what standards are to avoid. ISO is a quality stamp, OOXML being ISO stamped calls into question the quality of the ISO stamp. And worse this all happened through skullduggery and corruption.
If the software is done on the cheap, it will crash even if the system doesn't underneath it. Expensive hardware for cheap development costs, but as I said, bet the cost of keeping it limping on makes it all work out more expensive. In the long run, I would be very surprised if it doesn't work out cheaper to do it with cheap kit, cut down Linux and decent developers.
The point is that the system can be stripped to only what is required. You can not strip windows in the same way. Less stuff running is less things to compromise, less hardware requirements, less things to maintain, etc etc. Bet it's mainly the cheap crappy companies that use windows as the basis of their hardware/software solution, and also bet in the long run the nightmare mess they create costs to maintain makes it expensive. False economy.
http://en.wikipedia.org/wiki/Evolution_of_the_eye http://en.wikipedia.org/wiki/Cambrian_explosion
I don't want any child of mine's head filled with every silly idea that apposes a valid scientific hypothesis. That has no place in a science class room. They going to start insisting the math classes state the value for PI could be 3 because that's what the bible states?
I'm sure most programming nerds program something at some point that mutates with random then selects from it? I wrote a simple ant java thing where each ant started with random rules of how to get to the food and back to the hive. If they don't get to the food, and then the hive, after a set time, they get fresh random rules. After 30 min of so, each ant has the same rules. Real life clearly has mutations, and clearly has selection forces, no brainer. It's so easy to see logically it's true, that's why creationists fear it so much.
Why on earth would I want a touch screen keyboard? You can't feel the keys! I touch type solely on the feel of the keys. Why would I want to have to go back to looking down? If I using my hands to input, having to look at them as they do so is wasting my time. Yer they look good on a set of star trek, but in reality that ship would have been destroyed long ago but villains with keyboards they don't have to look down at to press fire. Until the touch screen raises where buttons are, you are using one sense less while working, and if you aren't using that touch screen to look at, what is the point?
It seams to basically be advocating:
* get lots of data
* use computers to search for pattens in that data
* use those patterns
instead of thinking up models and testing them.
But don't you have to have some model to define the parameters of the data you are collecting? I'm not sure there is anything new here..... Isn't it just data mining?
Exactly the points I would have made, but made better. ;-)
You can get Adobe Flash on the ARM (you think Adobe are going to miss the mobile market that's all ARM?) Plus there is Gnash that's cross platform....
I know (I've written code for Cell), but I was thinking more of a sea of stream processors. It's a great way of using them all. :-)
This is bait, but hey, it's tasty even with the hook. The reason open source will win out is because you really can fix it. It's not work related, but a good example. I was processing some video and the java app doing it crashed out telling what line the problem was it was a simple -1 index problem. I could just stick a index check in and carry on (the bad index was due to bad data I was running the app to fix!). A good proportion of crashes are something silly like that, or a null pointer being used/freed, any programmer can see it and fix it with out having to get to know the whole code base. I get crashes in closed software at work some times, and I can't do a thing about it! I can submit it as a bug to the vendor, and if I'm very very lucky I will get a patch, but most of the time your told it will be fixed in the next version and maybe a work round. I did programmer support for a middleware API for years, I use to do exactly that to customers myself. If there is a work round, ok, unless it's not really workable, but if not, what then? I'm willing to fix it but unable to because I don't have the source. You then get into "Are you willing to pay mega bucks to get it fixed?", again, I use to have to do this to customers myself. With open source you can fix it yourself if need be, or pay to get it done just like you would in closed source. So basically with open source you have all of the safety net of closed source (if you pay for support that is) with the extra advantage that you can help yourself.
Writing apps for huge number of processors is a bit of a nightmare, but ray tracing is a problem that fits very neatly in this world. A ray tracer can be very small, so once loaded on a processor, it can just munch through data (the scene) until it's finished. A collection of stream processors with read access to main memory (or a large pool of memory where the scene lives) would be perfect. If you have a pixel shader with read access to main memory, draw a single quad over the view, and have each ray tracer in pixel shader form and you get what I mean.
Which is why ARM is going to be laughing. ARM and Linux is going to be huge on the UMPC market if the battery life is going to be as good as promised. If you running Linux why do you care about chip architecture? Users want a cheap machine with a battery that lasts for ages with a powerful fully functional operating system. That must scare Intel as well as MS. MS will do what? Port XP to ARM, upgrade CE to full OS standards? What advantage has Windows over Linux without being able to run standard Windows software? Intel, well er, really they should have a line of processor where power use is the number one issue, not x86 compatibility (though x86 chips that need less power are also needed, just not in the UMPC market). Could be very interesting times.
I'm not sure it will be history repeating because size and power matter more in these devices then in the desktop. It will be hard for x86 machines to win out on price, size and power over ARM. They are paying a heavy price for x86, all for compatibility I really don't think we need.
I've been waiting for ARM laptop thing. Real battery life! Why do I need x86 compatibility? Give me battery life every time.
No-one wants an OS that requires a beefy machine just to run. We want a tiny OS that leaves us most of the machine to do with as we will. The desktop is morphing into a form of client via the browser. Who would want a fat OS to run their browser when a thin one will also run their browser, but faster on the same client machine. This is what the sub-laptops are all about. Thin OS, means you can have a thin/cheap client. Using a old OS isn't really a solution, you want one that scales with the hardware. I.e Linux. If MS wants to compete, they need MinWin. Not that I want them to, I want source code for when the docs fail me (though the Wine source is quite good for Win32 questions if msdn isn't answering you ;-) ).
"Linux excels in various code structure metrics, but lags in code style. This could be attributed to the work of brilliant motivated programmers who aren't however efficiently managed to pay attention to the details of style. In contrast, the high marks of WRK in code style and low marks in code structure could be attributed to the opposite effect: programmers who are efficiently micro-managed to care about the details of style, but are not given sufficient creative freedom to structure their code in an appropriate manner. "
How ever I was left wondering how it was possible to compare fairly? He already stated:
"Excluded from the kernel code are the device drivers, and the plug-and-play, power management, and virtual DOS subsystems. The missing parts explain the large size difference between the WRK and the other three kernels."
and reading I see even more of the drivers aren't there:
"The NT Hardware Abstraction Layer, file systems, network stacks, and device drivers are implemented separately from NTOS and loaded into kernel mode as dynamic libraries. Sources for these dynamic components are not included in the WRK. "
http://www.microsoft.com/resources/sharedsource/licensing/researchkernel.mspx
So it's not like for like. Maybe you would draw different conclusions if it was, maybe the Linux style issue is because of all the drivers the WRK lacks. So even though I think his conclusion sounds probable, I don't feel I can state it as so with any confidence.
I've heard people here say nice things about 3DSMax, and I must inform them that 3DSMax is a very bad peice of software. Sure as a interface it seams ok for knocking things together, but you'll hit a wall at some point, and it will really hurt. I am amazed this CAD package with a mountain of hacks on it is respected by anyone.
It's tech sucks very very very very badly. Get hold of the debug source (not complete source code, but quite a bit, much of it dated 1994). It sucks terribly. In the sdk, you are given the memory of objects directly, up casting and down casting every where (we just talking of (*) casting pointers, and people make mistakes.......). You seen the NULL function call error? Because of a bad upcast somewhere (in biped I think). Dealing with a editable mesh is different then dealling with a editable poly, dealing with a biped is completely different then dealing with a transform hierarchy, dealling with a xref object you must deference it first, ensuring it has a value (crash I've seen in Max). All the interface stuff is hard coded Win32. Defining parameter blocks is like a text book way of how it shouldn't be done. The scene graph is awful, no plug system for connections. From what source I've seen I can only feel sorry for a developer that has to work with it. I feel sorry for me when I have to use the sdk (just headers of the source they let you see, not a proper sdk).
3DSMax is terrible for animation houses because: * it's complete lack of referencing of heirarchies. (Referencing is a must for large work, or rig update requires changing of n files). * biped is a biped, try and do something more then that and your heading for trouble. (People do horrific things like bult together two bipeds trying to do what they should just use better software for). * toolies have a nightmare supporting you.
We have finally dropped it for main stream animation because of years of nightmares supporting it while the maya teams breaze along.
Not looked at Blender, don't do computer art anymore and we don't use it here, but it can't be worse. As it's open source, if/when it reaches critical mass, it will leave Maya and XSI behind (can't have not left 3DSMax behind) as armies of toolies fix bugs and add features.
Rant over, I feel much better now.
We had a script kiddy here who made a horrid tangled mess, hard pathing everything, including to a directory called "Not_Used_Anymore". He didn't know better.
The horrors of hard pathing are alive and well.
Hard pathing is just coding in a path into your source like any other constant. Sometimes it makes sense to do it (small personal batch script for instance) and some will always do it where they shouldn't (in something compiled or large).
buy a ps2 and lego starwars. Even if you don't like starwars. My partner and I played through both games together and it was great fun. We are waiting for lego indiana jones! Most fun co-op series I've ever played.
If you want to play games, buy a games consol. Don't give the fat lazy software houses more resource to do the same thing, or the apps will continue to get fatter and slower.
Open source doesn't suffer as bad from this laziness because trimming the fat is always one of the number one things coders (not sales reps and managers) care about.
I really won't be surprised if we end up with pc (as in crap for games), consols (specifically for games) and workstations (great for games, but really for media work). We are pretty much there now, your crazy (or rich) if you splash out for a workstation to play games.
I've got a bee in my bonnet to day for modern big fat slow apps doing the same thing, only slower, then I was doing on my Acorn 15 years ago with 4Mhz and 1MB! It was all BBC BASIC with some ARM asm, on such slow spec, how could it be faster! HOW! I blame pc gamers and fat lazy software houses! GRRR!!!
I would say, depends... Different languages have different uses. What need is your current languages not fulfilling? :-) (though seriously, it's worth having a script language under your belt for complex batch tasks at least)
Might be worth looking into asm just so you think about what your code really boils down to.
On the other hand, if you just want a language to throw around and knock stuff together quickly, python is good. Really good. I could rave all day about python and it's interactive shell and flexibility.
If you only have done C++, might be worth doing pure C, working object oriented with structs, again it makes you think about what your C++ code boils down to.
However, what I would say is that some times rather then learning a new language, might be worth picking/debugging through an open source project that you have no idea how it works. (May I recommend a emulator with JIT). I think I've learnt far more from other people's source then I have from changing languages, and some times what I've learnt is the language, even if I thought I knew it!
but but but Microsoft hate open source, right from the start have fought against it. Often on ground of protecting their intellectual property. But here they are using some else's intellectual property (lego) claiming to be for open source. WTF! Is this related to publicly showing the horrors that are the old Office formats? Policy change or new mask? While they refuse to play nice with standards and stop vender lock in, I won't believe it. Open the source to current software, like Windows and Office and maybe I would beleive it. It's all cosmetics.