I'm not sure if you are just playing devil's advocate here, but pay attention to what is going on in California. Citizens don't always roll over and play dead when they get screwed.
Are you referring to the recall campaign? Yes, people are getting screwed alright: by the Republicans, who think little of wasting enormous amounts of state money because they think it is advantageous for them to have the election take place on a different day. Let's hope citizens will remember that and hand the recall campaign a stinging defeat. Whether or not Davis is a good governor, the kind of reckless behavior exhibited by the recall campaign should not go politically unpunished.
Microsoft is attempting to spin the situation even in admitting defeat, by referring to Linux as "non-commercial" software.
Open source software is, of course, "commercial" software: it's at least as good as closed source software, it's used by many commercial enterprises, and it's sold commercially.
Do you remember the savings and loan bailout? Do you remember who was responsible for it? Do you even remember how much money we paid for it and still are paying for it?
Issues like these get drowned in useless debate and fingerpointing. By the time the election comes around, everybody has forgotten about it, and those responsible will be able to weasel out with phrases like "oh, that old thing, we already showed that what we did was right", etc.
I thought the entier point of a hobby was to spend time with it.
Robotics research isn't a hobby, and the point of robotics research is not to "spend time with it", it's to achieve results.
this is nothing more than a sales pitch.
No, it's a web page from a robotics research group about a tool they use for their research. Then some Slashdot editor picked it up as a story about a nifty commercial gadget, like many other stories about nifty commercial gadgets posted on Slashdot. If you can't deal with it, don't read Slashdot because that's what a lot of content here is about.
Holy shit. This person was trying to keep the discussion limited to, God fucking forbid, the actual question put forth by the author,
When someone asks "does X scale", comparisons with other systems, including open source software, are completely legitimate, as I explained in my response.
Absolutely no qualifying statments were made with regards to Microsoft or open-source (such as "Microsoft R0x0rz! or "Linux 1z t3h $ux0r5"), and yet you are able to glean the fact that the grandparent poster is irrationally, willfully enslaved by Microsoft.
If "shut up and don't talk about open source software" isn't a bias against open source software, I don't know what is. The profanity and insults used by the post also should give you a clue.
Unlike commercial software, which is supported by billions of dollars in marketing and PR, open source software lives by word of mouth; trying to exclude it from these kinds of discussions, where it is clearly relevant, is a direct attack on OSS.
and yet you are able to glean the fact that the grandparent poster is irrationally, willfully enslaved by Microsoft.
I have no idea whether the poster is "willfully enslaved by Microsoft". I suspect he actually believes what he is saying; most of the opinions like his are probably based on stupidity, not "willful enslavement".
The amount of money they'd save using OSS would be enough to buy at least one more whole box (SQL server ain't cheap)
But that isn't even the main reason for using OSS. Often, the real reason for using OSS is that it is cheaper to maintain and performs better. Saving licensing fees is icing on the cake.
Re:Why are they running Windows then?
on
Can .NET Really Scale?
·
· Score: 2, Insightful
A) This consultant, it sounds like, is largely or exclusively MS. He's not going to suggest Open Source software to his client because that will mean a loss in business.
That's an idiotic argument. For consultants, OSS is often at least as much of a money maker as Microsoft software. Furthermore, there are mature non-OSS alternatives (e.g., Java) available.
B) Oftentimes a commercial solution to some problems exists where a free one does not. The cost of development and maintanance means that the balance is not strictly in terms of free and non-free; after all, your developers' time costs quite a bit as well and home-grown or open source solutions may need more time taken in administration.
Yeah, and "oftentimes" the commercial solution actually performs less well, is less reliable, requires more hardware, and requires more administration. A lot of Microsoft products fall into that category. Products like MS SQL Server and MS Exchange are prime examples of what a money pit commercial software can be.
Face it, people use OSS not because they save on licensing costs, but because it works better and is easier to maintain.
If you don't have any experience with the scalability of.NET, I advise you to keep your mouth shut.
The question "does.NET scale" means "how much room for improvement is there in how.NET uses a given chunk of hardware"; that's a question that is most easily answered by comparing.NET performance with the performance of other systems.
For example, since.NET has a similar architecture to Java and Java scales fairly well, that tells us that there is nothing in principle that keeps.NET from scaling. (Of course, that also naturally raises the question why Cliff isn't using Java, but that's for him to work out.)
What's your major malfunction?
Well, we know what your major malfunction is: you are one of those Microsoft zealots. Your support for that company is so irrational that you just don't want people looking at alternatives.
The signal/noise ratio is bad enough already.
Yes, and you just contributed to the noise portion of it.
There is nothing in the.NET architecture that keeps it from scaling in principle. I mean, it's basically Java with a few bugs fixed in the language definition.
But don't hold your breath for Microsoft to actually deliver a complete and efficient solution. It took Sun and IBM years to even get the flaky, bloated, and mediocre system they have now (Java 2 and J2EE). Microsoft is further behind and less experienced with these things than Sun and IBM.
Engineering is a game of tradeoffs. If you want an all Microsoft and all.NET solution, you have to pay the price for it. If you have a constrained budget, you need to make other tradeoffs to handle the same load.
Java might be a slightly better tradeoff for you. Like.NET, it's proprietary and bloated, but it's quite a bit more mature, and you could still stay on Windows.
If you really want your system to perform well on limited hardware, ditch the C#/Java "frameworks", stick with plain Apache as much as possible, and write any dynamic content as close to the metal as you can (and, yes, you may have to implement your own caching and connection management--it isn't rocket science). You may also want to ditch the relational database--they are a huge bottleneck and many dynamic web applications just don't need the features.
You gotta read Atlas Shrugged, or at least The Fountainhead to get into Rand's philosophy known as Objectivism,
Or you could avoid putting yourself through the pain of reading a couple of cheap, poorly written romance novels by a writer with no intellectual depth and instead just look up the entry on Objectvissm in the Wikipedia.
Many people would disagree with the tenets of Objectivism because they have different moral and ethical beliefs and values. But whether one agrees or disagrees with the stated tenets of Objectivism, a more fundamental problem with Objectivism is that it is naive about the fact that many of those goals are in conflict with one another. Most of the work in philosophy (and law, for that matter) over the last few thousand years has been trying to address how to resolve those conflicts, while Rand's philosophy just ignores it. Therefore, Objectivism just fails to be a coherent system of philosophical or moral thought.
Objectivism really suffers from the same problem as Communism: it sounds good to its ardent supporters, it might theoretically work if everybody was behaving properly, but it fails miserably in the real world. And just like Communism, Objectivism can be used to justify many kinds of anti-social behavior and social ills.
True, balance in a robot is a challenge in itself, but I just wouldn't start with a system whose design centers around maintaining balance with a human rider (at least if money was a factor), since you have to throw away so much of their technology.
What do you have to throw away? A two-wheeled self-balancing vehicle requires the same kind of sensors, whether a human is standing on top of it or not.
And the reason why you want a two-wheeled self-balancing robot is because it can stand up; a three- or four-wheeled passively balanced design requires a much larger footprint and still won't be as stable.
Why the fuck would someone want to tear apart an umpteen thousand dollar toy and, in the process, make it even more useless?
Have you looked at the prices of ready-made robot platforms? $5000 or whatever isn't a lot of money for a robot platform that's self-balancing and can carry around 300 pounds or more.
it's more like a Segway sales pitch targeted at overbudgeted academics with too much time on their hands.
Building your own takes less time? I don't think so. Building mechanical devices is time-consuming and expensive.
What/are/ you doing, or what platform are you compiling on? I've only seen errors remotely related to what you are talking about when I was playing with ARM processors.
I'm just compiling for regular Athlon and P4 machines. I've been compiling Linux kernels since around 1995 (and UNIX kernels since the early 1980's).
Where are the problems in 2.4? There have been compilation problems in the IEEE drivers, inability to use fb consoles after upgrading, bugs introduced into the USB2 drivers, numerous non-ANSI C constructs that GNU C starts rejecting in later versions (gcc-3.3 will not compile 2.4.19), and on and on. I'm sure many of those are documented in the release notes, and people do trip over them.
I can just repeat: my impression is that kernel development is collapsing under its own weight; I think the Linux kernel and kernel source need a major architectural overhaul.
And again, if 64% of the budget is social programs then saying that the past debt was incurred by military spending is less than honest.
Most of the social programs are like a retirement account: people pay their money into them and they have a right to expect to get the money out. That money should really be accounted for completely separately from the rest of the budget. The fact that politicians keep dipping into it to make up for spending elsewhere is itself an abuse of the system by governments. If the US government didn't provide those social programs, taxes would have to be cut drastically so that people could try to get the equivalent retirement benefits on the free market.
If you remove those programs and just look at discretionary spending, the money for which you and I pay income tax and that gets divided up among different programs by Congress every year, somewhere between a quarter and one half of that money goes to the military. US military spending is completely out of control.
In fact, the net amount of money annually flowing into the US from other nations (the increase in the US foreign debt) corresponds rather nicely to total US military spending. In effect, Europe, China, and Japan are financing the US military. Without that influx of money, the US could not afford a strong military.
In reality, you can't say what caused the debt other than "total spending."
Do you apply that same reasoning to your personal budget? "Let's see, I spend $1000 more per month than I get. I can't really say which is responsible, the mortgage, the 401k contributions, groceries, car payments, or the fact that I am a gun collector and keep buying guns every month. Really, it could be anything. Maybe I should stop paying the mortgage."
I'd rather have C# with templates. C# has eliminated most of the cruft in C++ without adding significant overhead. The only important feature that's in C++ but not in C# is templates. C# still retains all the abilities to access every bit of memory and perform unsafe operations. However, unlike C++, safe and unsafe operations are operations that can be distinguished lexically.
C++ templates have turned out to be surprisingly powerful and useful. C++ templates are not at all like generics in other programming languages because C++ templates respect overloading. That means that template instantiations may do completely different things for different classes. On the one hand, that requires extra care to avoid bloat during expansion, on the other hand it enables expressiveness that is unavailable in just about any other language. I hope Microsoft won't make the mistake of adding a watered-down generics facility to C#, instead of full template support.
If people want to beat C++ into shape, I think here are some things that would have to be done:
Add a notion of "safe" compilation units, compilation units in which all unsafe operations (pointer dereferencing, pointer subscripting, some uses of the reference operator) are prohibited. (How do you get anything done? You use data structures that are internally using unsafe operations but are externally safe.)
Require the presence of a garbage collector in every C++ implementation.
Improve reflection features.
Make error and bounds checking mandatory and the default in all standard libraries, in particular the STL, but provide options or operations that allow the user to turn off error checking selectively.
The biggest reason for not using C# for everything for me is that stupid virtual machine; Microsoft's CLR and Mono seem ever more eager to copy the sluggish startup times of the Java JVM. And, frankly, I don't give a damn about being able to move byte codes across architectures, so why should I pay the overhead?
If we had a native compiler like gcj for C#, I'd probably be using C# for everything.
All stable kernel series take a while to sort themselves out. Stable series doesn't mean bug-free, it means working toward such
I dunno--I think 2.4 has actually deteriorated throughout its lifetime. For example, for my hardware, 2.4.20 seems to have serious bugs in USB2, while 2.4.19 is working fine. Also, I have encountered numerous compilation problems in different 2.4 kernels with different configuration settings.
It runs 24/7 and has crashed/frozen exactly zero times.
Yes, in my experience, there hasn't been much flakiness in those kernels--if you manage to find one that configures correctly, compiles without complaints, and boots up, chances are that it will work well. But if my experience is any guide, that's a big "if".
Altogether, I think Linux kernel development is really having some serious problems.
Scientists have been shipping around disks (well, first disk packs, then disks) with experimental data for years, and doing so is more popular than ever. It doesn't take a genius to figure that out: let's see, we have to download 100G of data, at the current rate that's going to take whatever, so let's just ask the good folks to send us a tape/disk/whatever.
Of course, in many cases, it is still actually more convenient, cheaper, and less time consuming to just rsync something than to send a disk; the bandwidth may be lower, but it takes fewer man hours. I suppose that may not matter to Gray, who probably can command a bunch of assistants to do his bidding, but for normal people, it does matter. I rather type a couple of commands and forget about them for a week than go out, buy a disk, hook it up, transfer stuff, pack it, stamp it, and take it to the post office/FedEx.
This leads me to wonder: has Gray done anything scientifically or technically interesting lately? Just buying lots of disks and installing MS SQL Server on them for various Microsoft marketing gimmicks doesn't count.
while more traditional scientists - including some funded by NASA - view them as nothing more than contraptions harnessing ionic winds.
Yes, and airplanes are nothing more than contraptions harnessing aerodynamic lift, and the people who designed them originally also didn't fully understand the physics involved. If "ionic wind engines" can be made practical and acceptably efficient, they might give rise to a new class of airborn vehicles.
And perhaps there are other uses as well. For example, electric fields and magnetic currents might be useful for shaping and redirecting the hot air that occurs during reentry from space. Or, the same technology might find uses not for pushing around large amounts of air for propulsion, but instead for changing the properties of the thin layer of air right above the surface of a traditional plane or vehicle--this could perhaps be used to reduce turbulences and improve performance.
"The RJ45 connector is a universal outlet," he said. "It's the only one identical in Asia, Europe and the US."
The same is true for FireWire and USB, and those are far more widely available. In fact, USB is increasingly becoming the standard for recharging portable devices, with USB power cables for most PDAs and cell phones available. FireWire might be better for this than USB because it can deliver more power, but maybe the USB standard will catch up.
The RJ-45 connectors used with Ethernet really don't do it for me; I think both USB and FireWire connectors are far better designed. I think POE (Power over Ethernet) will remain a niche market.
... the majority of people (ie. non Slashdot crowd) who will buy a PDA don't give a flying toss about what the underlying operating system is.
Most people drive SUV's, eat at MacDonalds, and don't exercise. Yet, we still have manufacturers of electric cars and sports cars, fine restaurants, and health clubs. The fact is that for a business, it doesn't matter what "the majority of people" want. What matters is that a business finds a niche market that works for it.
Linux handhelds can be enormously useful in vertical applications: science, engineering, medicine, etc. For that, the quality of their PDA functions doesn't matter. What does matter is making it really easy to develop software for them because almost all the software will be custom-written. Linux-based PDAs help with that. And PDAs that run both Linux and X11 are even better at that.
CompactFlash port, in particular is very bulky. As for plugins like a camera, USB would be a better interface.
With a CF card, the device disappears almost entirely into the handheld. That's a good thing. With SDIO or MS expansion devices, the device mostly sits outside the handheld and is prone to breaking off.
Doesn't it already have a USB port for the desktop?
USB connections are not symmetric, so having a USB hotsync port tells you nothing about whether it can control a camera. But I think the Yopy, very unusally for a PDA, actually can act as a USB host controller. That alone would give it a big leg up for many mobile Linux applications.
Of course, you can plug a USB controller into the CF port and use the same USB devices you would use with your desktop machine, since the Linux kernel already has all the drivers...
IRDA is also not really good for anything. People have to wiggle their devices for minutes just to send a business card.
IrDA works and it works well. Yes, you have to point the two devices roughly at each other, but that's all. IrDA also costs very little to add to a handheld, every handheld has it, and it is expected. Any PDA manufacturer that were to leave off IrDA just couldn't be taken seriously.
I have a Zaurus SL-5500. It's nice for a few special applications: Wireless/Ethernet sniffing, computer remote control, a small music-over-IP receiver. And the keyboard on the thing is great.
But as a PDA, I think the Zaurus line is not all that useful. I think the user interface on the Zaurus frankly sucks. It looks like they copied all the misfeatures of Windows CE: task bars, menu bars, title bars, 3D shaded widgets, wide scroll bars, etc; with a 320x240 screen, you can't afford to waste a lot of pixels on useless eye candy. The interaction is also not very well thought out: dialog buttons are placed in title bars (but only sometimes) and the difference between switching between applications and closing them is unobvious. Much of the user interface is not "direct manipulation"; for example, where in the Palm Calendar application, you just click on a line your your appointment book and write in an appointment, on the Zaurus, you have to pop up the "New Appointment" dialog. And Qt eats up lots of memory and CPU.
Furthermore, those problems are hard to fix: the Zaurus is built on Qt/Embedded, which means that you are stuck with the toolkit and that almost nobody is developing non-Qt applications for it. You may think that there is "plenty of software" for it, but a large fraction of that software is console-based, and the amount of software that is available for it pales in comparison to the amount of software that would be available for it if it ran the same window system as the Linux desktop. Just in terms of development alone, with X11 on a handheld, you can choose among many convenient scripting languages with UI capabilities, on the Zaurus, you get--PyQt and maybe a few even more oddball things. You can install the handhelds.org distribution on it (with X11), but that's a major operation and few people do it.
I have yet to see a really good Linux PDA--a PDA that does the core PDA functions as well as Palm. I think Linux handhelds really have a different market for now: ports of desktop applications, prototyping, diskless controllers, etc. And in that market, the Zaurus fails pretty miserably because it is so non-standard. The Yopy may or may not make a better PDA than the Zaurus (it probably won't be any worse), but it sounds like it will make a much better Linux handheld because it actually runs more standard Linux software.
Just about every PDA around today has the specs of a high-end UNIX workstation of a few years ago. Your average PDA today has a 200MHz RISC chip, 64Mbytes of RAM, hundreds of megabytes of Compact Flash storage, etc. Many UNIX workstation had 1/10th the compute power, memory, and disk storage.
Furthermore, Linux and X11 aren't "high-footprint" at all by modern standards. Sure, on your desktop machine, they use lots of memory; that's because they can: people configure every feature into them and then they go on using lots of cache. On a PDA, you can squeeze a Linux kernel (I don't know about 2.4, but certainly older kernels) into a few hundred kbytes, and the X11 server and toolkit into less than a Mbyte. That's less than Windows CE or Qt/Embedded. It probably is even less than PalmOS 5. In fact, if you really want a small footprint OS, ucLinux is another option; it can even run on old Palm hardware (no MMU).
Slashdot Mirror
I'm not sure if you are just playing devil's advocate here, but pay attention to what is going on in California. Citizens don't always roll over and play dead when they get screwed.
Are you referring to the recall campaign? Yes, people are getting screwed alright: by the Republicans, who think little of wasting enormous amounts of state money because they think it is advantageous for them to have the election take place on a different day. Let's hope citizens will remember that and hand the recall campaign a stinging defeat. Whether or not Davis is a good governor, the kind of reckless behavior exhibited by the recall campaign should not go politically unpunished.
Microsoft is attempting to spin the situation even in admitting defeat, by referring to Linux as "non-commercial" software.
Open source software is, of course, "commercial" software: it's at least as good as closed source software, it's used by many commercial enterprises, and it's sold commercially.
Here is a link on some of the history of the S&L bailout.
Do you remember the savings and loan bailout? Do you remember who was responsible for it? Do you even remember how much money we paid for it and still are paying for it?
Issues like these get drowned in useless debate and fingerpointing. By the time the election comes around, everybody has forgotten about it, and those responsible will be able to weasel out with phrases like "oh, that old thing, we already showed that what we did was right", etc.
to be a Pennsylvanian state legislator when the tax payers find out about all this
Do you really believe voters will remember? I think more often than not they don't.
I thought the entier point of a hobby was to spend time with it.
Robotics research isn't a hobby, and the point of robotics research is not to "spend time with it", it's to achieve results.
this is nothing more than a sales pitch.
No, it's a web page from a robotics research group about a tool they use for their research. Then some Slashdot editor picked it up as a story about a nifty commercial gadget, like many other stories about nifty commercial gadgets posted on Slashdot. If you can't deal with it, don't read Slashdot because that's what a lot of content here is about.
Holy shit. This person was trying to keep the discussion limited to, God fucking forbid, the actual question put forth by the author,
When someone asks "does X scale", comparisons with other systems, including open source software, are completely legitimate, as I explained in my response.
Absolutely no qualifying statments were made with regards to Microsoft or open-source (such as "Microsoft R0x0rz! or "Linux 1z t3h $ux0r5"), and yet you are able to glean the fact that the grandparent poster is irrationally, willfully enslaved by Microsoft.
If "shut up and don't talk about open source software" isn't a bias against open source software, I don't know what is. The profanity and insults used by the post also should give you a clue.
Unlike commercial software, which is supported by billions of dollars in marketing and PR, open source software lives by word of mouth; trying to exclude it from these kinds of discussions, where it is clearly relevant, is a direct attack on OSS.
and yet you are able to glean the fact that the grandparent poster is irrationally, willfully enslaved by Microsoft.
I have no idea whether the poster is "willfully enslaved by Microsoft". I suspect he actually believes what he is saying; most of the opinions like his are probably based on stupidity, not "willful enslavement".
The amount of money they'd save using OSS would be enough to buy at least one more whole box (SQL server ain't cheap)
But that isn't even the main reason for using OSS. Often, the real reason for using OSS is that it is cheaper to maintain and performs better. Saving licensing fees is icing on the cake.
A) This consultant, it sounds like, is largely or exclusively MS. He's not going to suggest Open Source software to his client because that will mean a loss in business.
That's an idiotic argument. For consultants, OSS is often at least as much of a money maker as Microsoft software. Furthermore, there are mature non-OSS alternatives (e.g., Java) available.
B) Oftentimes a commercial solution to some problems exists where a free one does not. The cost of development and maintanance means that the balance is not strictly in terms of free and non-free; after all, your developers' time costs quite a bit as well and home-grown or open source solutions may need more time taken in administration.
Yeah, and "oftentimes" the commercial solution actually performs less well, is less reliable, requires more hardware, and requires more administration. A lot of Microsoft products fall into that category. Products like MS SQL Server and MS Exchange are prime examples of what a money pit commercial software can be.
Face it, people use OSS not because they save on licensing costs, but because it works better and is easier to maintain.
If you don't have any experience with the scalability of .NET, I advise you to keep your mouth shut.
.NET scale" means "how much room for improvement is there in how .NET uses a given chunk of hardware"; that's a question that is most easily answered by comparing .NET performance with the performance of other systems.
.NET has a similar architecture to Java and Java scales fairly well, that tells us that there is nothing in principle that keeps .NET from scaling. (Of course, that also naturally raises the question why Cliff isn't using Java, but that's for him to work out.)
The question "does
For example, since
What's your major malfunction?
Well, we know what your major malfunction is: you are one of those Microsoft zealots. Your support for that company is so irrational that you just don't want people looking at alternatives.
The signal/noise ratio is bad enough already.
Yes, and you just contributed to the noise portion of it.
There is nothing in the .NET architecture that keeps it from scaling in principle. I mean, it's basically Java with a few bugs fixed in the language definition.
.NET solution, you have to pay the price for it. If you have a constrained budget, you need to make other tradeoffs to handle the same load.
.NET, it's proprietary and bloated, but it's quite a bit more mature, and you could still stay on Windows.
But don't hold your breath for Microsoft to actually deliver a complete and efficient solution. It took Sun and IBM years to even get the flaky, bloated, and mediocre system they have now (Java 2 and J2EE). Microsoft is further behind and less experienced with these things than Sun and IBM.
Engineering is a game of tradeoffs. If you want an all Microsoft and all
Java might be a slightly better tradeoff for you. Like
If you really want your system to perform well on limited hardware, ditch the C#/Java "frameworks", stick with plain Apache as much as possible, and write any dynamic content as close to the metal as you can (and, yes, you may have to implement your own caching and connection management--it isn't rocket science). You may also want to ditch the relational database--they are a huge bottleneck and many dynamic web applications just don't need the features.
You gotta read Atlas Shrugged, or at least The Fountainhead to get into Rand's philosophy known as Objectivism,
Or you could avoid putting yourself through the pain of reading a couple of cheap, poorly written romance novels by a writer with no intellectual depth and instead just look up the entry on Objectvissm in the Wikipedia.
Many people would disagree with the tenets of Objectivism because they have different moral and ethical beliefs and values. But whether one agrees or disagrees with the stated tenets of Objectivism, a more fundamental problem with Objectivism is that it is naive about the fact that many of those goals are in conflict with one another. Most of the work in philosophy (and law, for that matter) over the last few thousand years has been trying to address how to resolve those conflicts, while Rand's philosophy just ignores it. Therefore, Objectivism just fails to be a coherent system of philosophical or moral thought.
Objectivism really suffers from the same problem as Communism: it sounds good to its ardent supporters, it might theoretically work if everybody was behaving properly, but it fails miserably in the real world. And just like Communism, Objectivism can be used to justify many kinds of anti-social behavior and social ills.
True, balance in a robot is a challenge in itself, but I just wouldn't start with a system whose design centers around maintaining balance with a human rider (at least if money was a factor), since you have to throw away so much of their technology.
What do you have to throw away? A two-wheeled self-balancing vehicle requires the same kind of sensors, whether a human is standing on top of it or not.
And the reason why you want a two-wheeled self-balancing robot is because it can stand up; a three- or four-wheeled passively balanced design requires a much larger footprint and still won't be as stable.
Why the fuck would someone want to tear apart an umpteen thousand dollar toy and, in the process, make it even more useless?
Have you looked at the prices of ready-made robot platforms? $5000 or whatever isn't a lot of money for a robot platform that's self-balancing and can carry around 300 pounds or more.
it's more like a Segway sales pitch targeted at overbudgeted academics with too much time on their hands.
Building your own takes less time? I don't think so. Building mechanical devices is time-consuming and expensive.
What /are/ you doing, or what platform are you compiling on? I've only seen errors remotely related to what you are talking about when I was playing with ARM processors.
I'm just compiling for regular Athlon and P4 machines. I've been compiling Linux kernels since around 1995 (and UNIX kernels since the early 1980's).
Where are the problems in 2.4? There have been compilation problems in the IEEE drivers, inability to use fb consoles after upgrading, bugs introduced into the USB2 drivers, numerous non-ANSI C constructs that GNU C starts rejecting in later versions (gcc-3.3 will not compile 2.4.19), and on and on. I'm sure many of those are documented in the release notes, and people do trip over them.
I can just repeat: my impression is that kernel development is collapsing under its own weight; I think the Linux kernel and kernel source need a major architectural overhaul.
And again, if 64% of the budget is social programs then saying that the past debt was incurred by military spending is less than honest.
Most of the social programs are like a retirement account: people pay their money into them and they have a right to expect to get the money out. That money should really be accounted for completely separately from the rest of the budget. The fact that politicians keep dipping into it to make up for spending elsewhere is itself an abuse of the system by governments. If the US government didn't provide those social programs, taxes would have to be cut drastically so that people could try to get the equivalent retirement benefits on the free market.
If you remove those programs and just look at discretionary spending, the money for which you and I pay income tax and that gets divided up among different programs by Congress every year, somewhere between a quarter and one half of that money goes to the military. US military spending is completely out of control.
In fact, the net amount of money annually flowing into the US from other nations (the increase in the US foreign debt) corresponds rather nicely to total US military spending. In effect, Europe, China, and Japan are financing the US military. Without that influx of money, the US could not afford a strong military.
In reality, you can't say what caused the debt other than "total spending."
Do you apply that same reasoning to your personal budget? "Let's see, I spend $1000 more per month than I get. I can't really say which is responsible, the mortgage, the 401k contributions, groceries, car payments, or the fact that I am a gun collector and keep buying guns every month. Really, it could be anything. Maybe I should stop paying the mortgage."
C++ templates have turned out to be surprisingly powerful and useful. C++ templates are not at all like generics in other programming languages because C++ templates respect overloading. That means that template instantiations may do completely different things for different classes. On the one hand, that requires extra care to avoid bloat during expansion, on the other hand it enables expressiveness that is unavailable in just about any other language. I hope Microsoft won't make the mistake of adding a watered-down generics facility to C#, instead of full template support.
If people want to beat C++ into shape, I think here are some things that would have to be done:
- Add a notion of "safe" compilation units, compilation units in which all unsafe operations (pointer dereferencing, pointer subscripting, some uses of the reference operator) are prohibited. (How do you get anything done? You use data structures that are internally using unsafe operations but are externally safe.)
- Require the presence of a garbage collector in every C++ implementation.
- Improve reflection features.
- Make error and bounds checking mandatory and the default in all standard libraries, in particular the STL, but provide options or operations that allow the user to turn off error checking selectively.
The biggest reason for not using C# for everything for me is that stupid virtual machine; Microsoft's CLR and Mono seem ever more eager to copy the sluggish startup times of the Java JVM. And, frankly, I don't give a damn about being able to move byte codes across architectures, so why should I pay the overhead?If we had a native compiler like gcj for C#, I'd probably be using C# for everything.
All stable kernel series take a while to sort themselves out. Stable series doesn't mean bug-free, it means working toward such
I dunno--I think 2.4 has actually deteriorated throughout its lifetime. For example, for my hardware, 2.4.20 seems to have serious bugs in USB2, while 2.4.19 is working fine. Also, I have encountered numerous compilation problems in different 2.4 kernels with different configuration settings.
It runs 24/7 and has crashed/frozen exactly zero times.
Yes, in my experience, there hasn't been much flakiness in those kernels--if you manage to find one that configures correctly, compiles without complaints, and boots up, chances are that it will work well. But if my experience is any guide, that's a big "if".
Altogether, I think Linux kernel development is really having some serious problems.
Scientists have been shipping around disks (well, first disk packs, then disks) with experimental data for years, and doing so is more popular than ever. It doesn't take a genius to figure that out: let's see, we have to download 100G of data, at the current rate that's going to take whatever, so let's just ask the good folks to send us a tape/disk/whatever.
Of course, in many cases, it is still actually more convenient, cheaper, and less time consuming to just rsync something than to send a disk; the bandwidth may be lower, but it takes fewer man hours. I suppose that may not matter to Gray, who probably can command a bunch of assistants to do his bidding, but for normal people, it does matter. I rather type a couple of commands and forget about them for a week than go out, buy a disk, hook it up, transfer stuff, pack it, stamp it, and take it to the post office/FedEx.
This leads me to wonder: has Gray done anything scientifically or technically interesting lately? Just buying lots of disks and installing MS SQL Server on them for various Microsoft marketing gimmicks doesn't count.
while more traditional scientists - including some funded by NASA - view them as nothing more than contraptions harnessing ionic winds.
Yes, and airplanes are nothing more than contraptions harnessing aerodynamic lift, and the people who designed them originally also didn't fully understand the physics involved. If "ionic wind engines" can be made practical and acceptably efficient, they might give rise to a new class of airborn vehicles.
And perhaps there are other uses as well. For example, electric fields and magnetic currents might be useful for shaping and redirecting the hot air that occurs during reentry from space. Or, the same technology might find uses not for pushing around large amounts of air for propulsion, but instead for changing the properties of the thin layer of air right above the surface of a traditional plane or vehicle--this could perhaps be used to reduce turbulences and improve performance.
"The RJ45 connector is a universal outlet," he said. "It's the only one identical in Asia, Europe and the US."
The same is true for FireWire and USB, and those are far more widely available. In fact, USB is increasingly becoming the standard for recharging portable devices, with USB power cables for most PDAs and cell phones available. FireWire might be better for this than USB because it can deliver more power, but maybe the USB standard will catch up.
You can even get a USB toothbrush.
The RJ-45 connectors used with Ethernet really don't do it for me; I think both USB and FireWire connectors are far better designed. I think POE (Power over Ethernet) will remain a niche market.
... the majority of people (ie. non Slashdot crowd) who will buy a PDA don't give a flying toss about what the underlying operating system is.
Most people drive SUV's, eat at MacDonalds, and don't exercise. Yet, we still have manufacturers of electric cars and sports cars, fine restaurants, and health clubs. The fact is that for a business, it doesn't matter what "the majority of people" want. What matters is that a business finds a niche market that works for it.
Linux handhelds can be enormously useful in vertical applications: science, engineering, medicine, etc. For that, the quality of their PDA functions doesn't matter. What does matter is making it really easy to develop software for them because almost all the software will be custom-written. Linux-based PDAs help with that. And PDAs that run both Linux and X11 are even better at that.
CompactFlash port, in particular is very bulky. As for plugins like a camera, USB would be a better interface.
With a CF card, the device disappears almost entirely into the handheld. That's a good thing. With SDIO or MS expansion devices, the device mostly sits outside the handheld and is prone to breaking off.
Doesn't it already have a USB port for the desktop?
USB connections are not symmetric, so having a USB hotsync port tells you nothing about whether it can control a camera. But I think the Yopy, very unusally for a PDA, actually can act as a USB host controller. That alone would give it a big leg up for many mobile Linux applications.
Of course, you can plug a USB controller into the CF port and use the same USB devices you would use with your desktop machine, since the Linux kernel already has all the drivers...
IRDA is also not really good for anything. People have to wiggle their devices for minutes just to send a business card.
IrDA works and it works well. Yes, you have to point the two devices roughly at each other, but that's all. IrDA also costs very little to add to a handheld, every handheld has it, and it is expected. Any PDA manufacturer that were to leave off IrDA just couldn't be taken seriously.
I have a Zaurus SL-5500. It's nice for a few special applications: Wireless/Ethernet sniffing, computer remote control, a small music-over-IP receiver. And the keyboard on the thing is great.
But as a PDA, I think the Zaurus line is not all that useful. I think the user interface on the Zaurus frankly sucks. It looks like they copied all the misfeatures of Windows CE: task bars, menu bars, title bars, 3D shaded widgets, wide scroll bars, etc; with a 320x240 screen, you can't afford to waste a lot of pixels on useless eye candy. The interaction is also not very well thought out: dialog buttons are placed in title bars (but only sometimes) and the difference between switching between applications and closing them is unobvious. Much of the user interface is not "direct manipulation"; for example, where in the Palm Calendar application, you just click on a line your your appointment book and write in an appointment, on the Zaurus, you have to pop up the "New Appointment" dialog. And Qt eats up lots of memory and CPU.
Furthermore, those problems are hard to fix: the Zaurus is built on Qt/Embedded, which means that you are stuck with the toolkit and that almost nobody is developing non-Qt applications for it. You may think that there is "plenty of software" for it, but a large fraction of that software is console-based, and the amount of software that is available for it pales in comparison to the amount of software that would be available for it if it ran the same window system as the Linux desktop. Just in terms of development alone, with X11 on a handheld, you can choose among many convenient scripting languages with UI capabilities, on the Zaurus, you get--PyQt and maybe a few even more oddball things. You can install the handhelds.org distribution on it (with X11), but that's a major operation and few people do it.
I have yet to see a really good Linux PDA--a PDA that does the core PDA functions as well as Palm. I think Linux handhelds really have a different market for now: ports of desktop applications, prototyping, diskless controllers, etc. And in that market, the Zaurus fails pretty miserably because it is so non-standard. The Yopy may or may not make a better PDA than the Zaurus (it probably won't be any worse), but it sounds like it will make a much better Linux handheld because it actually runs more standard Linux software.
Just about every PDA around today has the specs of a high-end UNIX workstation of a few years ago. Your average PDA today has a 200MHz RISC chip, 64Mbytes of RAM, hundreds of megabytes of Compact Flash storage, etc. Many UNIX workstation had 1/10th the compute power, memory, and disk storage.
Furthermore, Linux and X11 aren't "high-footprint" at all by modern standards. Sure, on your desktop machine, they use lots of memory; that's because they can: people configure every feature into them and then they go on using lots of cache. On a PDA, you can squeeze a Linux kernel (I don't know about 2.4, but certainly older kernels) into a few hundred kbytes, and the X11 server and toolkit into less than a Mbyte. That's less than Windows CE or Qt/Embedded. It probably is even less than PalmOS 5. In fact, if you really want a small footprint OS, ucLinux is another option; it can even run on old Palm hardware (no MMU).