If you're doing it entirely for love, you might as well GPL it.
If you're doing at least partially for the money, does it really make sense? First you have a much smaller percentage of desktop linux users. Then out of that group you have to eliminate those who don't want use a non-GPL'd program and those who don't want to buy software. That doesn't leave a lot of paying customers.
Of course, IBM was investigated for a decade for anti-trust violations. Unlike MS, however, they were up-to-date on their protection payments, so they never got caught.
If by "stinks" you mean buggy I disagree. I believe there's a tradition of higher standards in embedded software than other types due to the fact that historically fixing bugs had a much higher price than for traditional software. If you were lucky the software was in proms that had sockets so you could send them a new version in the mail instead of requiring the customer to send the whole unit back for repair.
Today of course, many embedded systems support on-board reprogramming with the update binary provided online, but it's still harder than it would be to patch a desktop or server app.
I agree. There's a lot of boiler-plate JavaScript code that merely makes up for the fact that the web wasn't designed to support applications. Common web app functions should be available as first class elements, not hacks.
Looking at Jakob Nielsen's bio, I don't see why his advice should be considered particularly qualified.
Are you sure they were talking about being "secure"? I remember lots of talk about stability, not so much about security. Of course Windows NT was more stable than Windows 98 but that wasn't a consumer OS.
I expect that if the government needed and wanted the XP source code, they could get it. I don't think the government would want everybody else to have access to it though.
The Atari 2600 didn't have a frame buffer, only video registers.
For the background there were two 8-bit registers and one 4-bit register. Those registers established the data for the left half of the screen and were reused in the right half of the screen in a way that depended on the mode you established. The right half could either duplicate the left side or mirror it.
Needless to say, the background "pixels" were very wide. The height could be as small as 1 scan line if you updated the registers for each line.
was that they illustrated the performance using a jpeg file. Yes, the compression artifacts are different.
It's rather ironic that there are more bits of data in the poorly rendered compressed version than there were pixels on the screen when you played a game on the real hardware.
The difference is that in a real-time system time windows aren't vague but known precisely and the software is designed and tested accordingly.
In general, if you make an existing real-time system run faster, it will probably fail. It's a bit like batting in baseball - swinging early is no better than swinging late.
just because you're trying to optimize the performance of your system doesn't make it a real-time one. If it were a real-time system, missing a timing window would result in your data being incorrect, not just tardy.
If you can produce code that fulfills your project's requirements, you're qualified to program it.
In any particular case, it may require specific knowledge, but there is no universal list of necessary and sufficient skills to program.
In my first ten years of programming in embedded systems none of the projects I worked on required sorting, searching, or sophisticated data structures.
That's not to say that knowledge of such things is a waste of time, but it isn't always required.
I wouldn't write the software in the first place if I assumed that customization and service were the only ways to make money. It's a lot easier to find multiple customers willing to pay a modest sum for a software license than it is to find even a single customer willing to pay a significant sum for customization.
Slashdot Mirror
You can download and run apps in Linux without manually invoking chmod.
And how exactly does this prevent a careless user from downloading and running malware?
than I can say that the cheapest version of Linux is about $100.
Is this opposed to the Turbo Netbooks with 30 inch monitor and water-cooled processor?
Translation: I tried Windows 7 on the least powerful machine it could theoretically run on.
If you're doing it entirely for love, you might as well GPL it.
If you're doing at least partially for the money, does it really make sense? First you have a much smaller percentage of desktop linux users. Then out of that group you have to eliminate those who don't want use a non-GPL'd program and those who don't want to buy software. That doesn't leave a lot of paying customers.
If one is really into "do it yourself" and operating systems, one would write one's own.
God, I sound like Robin Williams in Bicentennial Man.
Of course, IBM was investigated for a decade for anti-trust violations. Unlike MS, however, they were up-to-date on their protection payments, so they never got caught.
Now that Richard Teversham is cloaked in the RDF he no longer suffers the taint of Microsoft that many Slashdotters would otherwise sniff out.
Thanks. Yes, that's what I meant - most people under 40 won't get it.
I'm afraid you're joking to the lawn with this one.
If by "stinks" you mean buggy I disagree. I believe there's a tradition of higher standards in embedded software than other types due to the fact that historically fixing bugs had a much higher price than for traditional software. If you were lucky the software was in proms that had sockets so you could send them a new version in the mail instead of requiring the customer to send the whole unit back for repair.
Today of course, many embedded systems support on-board reprogramming with the update binary provided online, but it's still harder than it would be to patch a desktop or server app.
Didn't the EU essentially put a gun to MS's head? That's not usually a precursor to an "embrace".
I agree. There's a lot of boiler-plate JavaScript code that merely makes up for the fact that the web wasn't designed to support applications. Common web app functions should be available as first class elements, not hacks.
Looking at Jakob Nielsen's bio, I don't see why his advice should be considered particularly qualified.
Are you sure they were talking about being "secure"? I remember lots of talk about stability, not so much about security. Of course Windows NT was more stable than Windows 98 but that wasn't a consumer OS.
I expect that if the government needed and wanted the XP source code, they could get it. I don't think the government would want everybody else to have access to it though.
I just made the mistake I've been complaining about with others - it wasn't a prosecution because it wasn't a criminal case.
Indirectly the government did give them billions through prosecuting MS.
The Atari 2600 didn't have a frame buffer, only video registers.
For the background there were two 8-bit registers and one 4-bit register. Those registers established the data for the left half of the screen and were reused in the right half of the screen in a way that depended on the mode you established. The right half could either duplicate the left side or mirror it.
Needless to say, the background "pixels" were very wide. The height could be as small as 1 scan line if you updated the registers for each line.
was that they illustrated the performance using a jpeg file. Yes, the compression artifacts are different.
It's rather ironic that there are more bits of data in the poorly rendered compressed version than there were pixels on the screen when you played a game on the real hardware.
The difference is that in a real-time system time windows aren't vague but known precisely and the software is designed and tested accordingly.
In general, if you make an existing real-time system run faster, it will probably fail. It's a bit like batting in baseball - swinging early is no better than swinging late.
just because you're trying to optimize the performance of your system doesn't make it a real-time one. If it were a real-time system, missing a timing window would result in your data being incorrect, not just tardy.
If you can produce code that fulfills your project's requirements, you're qualified to program it.
In any particular case, it may require specific knowledge, but there is no universal list of necessary and sufficient skills to program.
In my first ten years of programming in embedded systems none of the projects I worked on required sorting, searching, or sophisticated data structures.
That's not to say that knowledge of such things is a waste of time, but it isn't always required.
In that case they aren't using the GPL license.
I wouldn't write the software in the first place if I assumed that customization and service were the only ways to make money. It's a lot easier to find multiple customers willing to pay a modest sum for a software license than it is to find even a single customer willing to pay a significant sum for customization.
YKUTW "own". IDNTIMWYTIM.