That's a nice theoretical argument, but what matters is how each comes out-of-the-box, to users who are not expected to do further work to secure their PDA. If Palm comes with 30 services written in C listening on the network by default and WinCE has 0, WinCE will be more secure, in the real word (or vice-versa).
But that's not the case. They both come out-of-the-box with no further securing necessary.
What I'm saying is that the platform has virtually nothing to do with the state of worms/viruses/trojans. Is it possible to run a secure PDA with PocketPC or Palm? Sure. Is is possible to run an insecure PDA with either? Sure.
Hello, transcriber? Palm has had Graffiti since inception. What do you call graffit but an instant transcriber.
I thought graffiti was a custom input method which had to be learned. Yes, it is very similar to standard manuscript, but it's not the same. Microsoft Transcriber allows you to use standard manuscript and even handwriting. Granted, it isn't quite as accurate as graffiti, at least not in my experience, but it is an additional feature to standard graffiti.
Netscape was a truly innovative, world-changing company, and if they weren't destroyed by MS the world would be a much different place today.
Let's see, would we be better off with an open source browser run by the non-profit charity Mozilla Foundation, or with a shareware browser which can't even be used legally for commercial purposes or for more than a 30 day trial without paying a fee?
Killing Netscape, the for-profit company, was a good thing.
Considering most people use Windows it makes sense to initially develop a program for Windows. It's a responsible use of tax money.
No, what would be a responsible use of tax money would be to create an API, and let whoever wants to create software for that API to use it, on any platform. In other words, what the US government did.
How about the Linux community builds their own open source version? I'm sure it would easily be ported to OS X.
Is there even an open API available for the Linux community to build such a product? If there is, then yes, the Linux community might want to do this, but that's not an excuse for the government to provide people with incentive to buy Windows.
I mean, it would be more disturbing if they only had a version for Be OS, right?
For people who have Windows and don't have Linux, I'd think it would be equally disturbing.
It apears that Tampa is a very different place in that case.
Probably so. And New Jersey, where I lived most of my life, is probably even more different.
If people really do speed up when crowded form behind around there I can only be thankfull most people around here aren't that stupid.
I'm not sure where I said they do that. Some people do it, I suppose. If I'm in the left lane and I see someone coming really fast behind me I'll usually speed up to pass the car on my right and get over.
Most of the time the loss occures when the lane jocky finaly traps himself behind someone going a bit slower than traffic average because he jumped lanes at the first sign of slowdown only to guessed wrong.
Now you're talking about something completely different. I'm just saying that oftentimes speeding gets you from point to point signficantly faster.
If you drive emergency services you get a very distorted view of what 'normal' people see as traffic. I hope that's not confusing things here.
No, not at all. I was a volunteer for a short period of time in my life, over a year ago, in New Jersey.
My experience is predominately in the St. Louis area, with a little in california while visisting relatives.
I've used interstate highways complete with right side on-ramps and off-ramps for most of my life as a commuter. I drove significantly faster than the rest of the traffic the vast majority of the time, never got in an accident or even close, and only got one speeding ticket (which was in Connecticut on a road I wasn't very familiar with).
I've actually slowed down quite a bit lately, in the past six months I've driven a lot in the right lane and probably been passed more often than I've passed (whereas a couple years ago I probably passed 95 cards for every 5 that passed me). My life is not as rushed any more, I have a much shorter distance commute, and more flexible hours. But I don't have any illusions that I'm not taking more time getting from point A to point B as a result.
I believe these are futures contracts, which means they expire on a certain date. So you can hold on to them until expiration, and not actually use them, thus losing your entire investment, or you can sell them to someone else before expiration, possibly make a profit, but cause no effect whatsoever to the environment.
Ok, right. But if it were to take more energy to produce than they generate, they would be sold for less than their cost to manufacture.
Almost certainly not. Non-rechargable batteries take more energy to produce than they generate, but batteries are still sold at a profit.
So you'd have to pay $225 to save $150; and of course you wouldn't.
I wouldn't, but there's a sucker born every minute. Moreover, solar panels are generally not used to power houses, but to bring power to places where running wires to the grid isn't feasible.
Well, there is the issue of portability. Three laptops (or did you mean three backpacks) are a lot heavier (or more awkward) than one, and there's likely no room for more solar panels on a single backpack.
And if your laptop and your wifi card are anything like mine, they won't both fit in your laptop at the same time. So that means using USB probably, and means one more thing hanging off your laptop.
I'm not sure how practical this idea is in the first place. I guess for someone who lives in the city and walks everywhere, but I have thought about rigging up one of these things for my car. I'm probably better off just taking out the EVDO card every day and putting it in my iPAQ. Of course this wouldn't provide wireless access for anyone else (unless I managed to find a thin wifi PCMCIA card and bought the double PCMCIA adapter to replace my single one), but if I needed to I could use the serial port adapter to hook the iPAQ up to another computer like a wireless modem.
Section 2 of the GPL grants the right to prepare derivative works. Clause 2b limits that by requiring "any work that you distribute or publish" to be licensed under the GPL.
Wow. You're right. That's a pretty gaping wide hole there. Sorry about that.
By "use", I mean as a component of the software being developed for distribution. If a developer, for whatever reason, needs or wants to distribute under non-GPL terms, then that developer cannot include GPL components, which means that the user will not have the option of using what would have resulted if the developer were able to include those components.
So can't the developer just distribute the software without the GPL components, and let the end-user link in the GPLed components herself?
It's legal to create non-GPL derived works from GPL sources; you just can't distribute them.
I don't see that anywhere in the GPL. It may be how things work in practice, but copyright law covers the preparation of derivative works regardless of whether or not you distribute them, and like I said, I don't see where the GPL gives you permission to prepare a derivative work without causing the derivative to be released under the GPL.
Furthermore, if the GPL allows you to create non-GPLed derivative works, then there would be no point to having an LGPL at all.
The only real difference for a consumer is a trickle-down effect of the restrictions on developers; if the developer couldn't or wouldn't use a GPLed component for whatever reason, the user can't use whatever derived work would have resulted.
Why can't the developer use whatever component she wants? The GPL doesn't restrict use.
If I buy a music CD, then sell it to you, are you now able to press your own copies and sell those? Of course not, the RIAA would make sure of it.
But if you make 100 music CDs, then sell them to me, I can then resell them.
If I buy GPL-licensed software and then sell it to you, why are you supposed to be treated by the author(s) of the GPL'd work differently than I was?
Because you don't have to agree to the GPL in order to resell a GPLed product. So you don't have to distribute the source code.
If you mean that somehow the GPL license becomes void after First Sale, then that means you don't have the right to distribute the GPL software at all unless you negotiate a new license with the author(s).
No, here's what I'm saying. Company A takes a GPLed product and modifies it. They then sell 100 copies to Company B, and include source code on a separate CD. Company B then destroys the source code and resells the 100 copies. Since they have not agreed to the GPL, and are merely exercising their distribution rights under First Sale, they are not required to distribute the source code. Meanwhile, Company A did distribute the source code, so they haven't broken the law either.
The GPL doesn't limit anything, copyright law does. And in addition to distribution, copyright law limits copying and the preparation of derivative works. According to the FSF, dynamic linking is the preparation of a derivative work, and most likely they are correct.
You want to be able to do so while not contributing back and improving the software for the rest of the users- you want all the benefits of open source with none of the responsibilities. Thats not cool.
I think his point was that he can't contribute back the source code of the entire derivative work, because parts of it were created by others and not licensed under the GPL. It'd be nice if the GPL could exempt people under these particular circumstances, but it probably wouldn't be possible to create a license which did that without providing a huge loophole for someone who wanted to create a proprietary derivative (for instance by creating separate corporations to do the distribution and development).
Of course, I'd argue that the GPL already contains this loophole when combined with the doctrine of First Sale.
My biggest problem with the GPL is the FSF's position that even dynamically linking against a library under GPL is enough to make the resulting code a derivative work (and thus also subject to the GPL).
Be careful here. It doesn't make the program subject to the GPL, it simply doesn't exempt you from copyright infringement if you link a GPLed library without releasing the derivative under the GPL.
Note also that the FSF's interpretation may not be binding, but it hasn't been tested in court (that I'm aware of, and I recently attended a symposium on this very topic). So, in my mind, it creates an unacceptable exposure for anyone who wants to develop software but not adopt the GPL.
Was this symposium sponsored by SCO? Seriously, this is only a problem if you're making software which is specifically intended to be linked with a GPLed library. Then you might get hit with contributory copyright infringement (though as you've said the GPL hasn't been tested in court even for direct infringement, let alone contributory infringement). But if you're creating software which is intended to be legally linked against a license-compatible library, then you don't even have to worry about that. Sure, the end-user who links the software against a GPLed library is breaking the law, but no one is going to prosecute that.
How many companies, as opposed to not-for-profit organisations have actually released software as BSD?
It's probably in proportion to the number of companies that have released software under the GPL, if you consider the number of GPLed vs. BSDed software as a whole.
With the GPL the company is assured of getting any improvements back.
That's not at all true. The GPL just assures them that it's not legal to distribute binaries without distributing source. That's much different from requiring that improvements be given to the company.
For a company *releasing* software under an open-source license BSD has no real advantages and many disadvantages.
The same could be said of the GPL, though, and it'd be about as accurate. Other than for public relations purposes, there aren't any real advantages (from a profit perspective) to release new software under any free license. For old software there's often an advantage to give your modifications back to the original author, so that they can be incorporated into future versions, but that's about it (again, from a purely profit perspective, which is what most large companies have to go by).
Since neither license is different to an end-user, explain how the BSD license is better for the consumer?
They are different. If someone creates a plugin for that GPL software, and the plugin isn't released under the GPL, then it's illegal for you to install that plugin and create a derivative work.
The most popular licenses "in between" GPL and BSD are probably ShareAlike and Attribution-ShareAlike. Both are strong copylefts, in the derivative works must be released under the original license, but neither require releasing the source code. So if you want to modify and use the code in an embedded device, there's no need to release your source. But if you get popular enough for people to care, the code is going to get reverse engineered eventually anyway.
There's usually enough incentive to contribute back code that anyone not contributing modifications are unlikely to use GPLed software anyway. By contributing the modifications back you ensure that they are contained in all the future revisions of the code. Forking an active project is very difficult due to the constant merges that have to take place.
I like the GPL because it's a copyleft. If you take my work and modify it, that's fine, but you can't turn around and tell me I can't copy or distribute your modified version.
On the other hand, I think the BSD license is advantageous because it doesn't have the requirement to release source code. In this sense, I feel the BSD license is more free.
I've created my own license, the QingPL, which I consider the perfect license. Basically, you can do anything you want with the code without any restrictions, as long as you don't sue anyone for copyright infringement over it or a derivative of it. About a year after I created this license Creative Commons released ShareAlike 1.0, which is essentially the same thing which a bunch of extra legalese thrown in.
Unfortunately Creative Commons decided to drop ShareAlike from its 2.0 series of licenses. The next best alternative is Attribution-ShareAlike 2.0. That one adds the requirement to attribute the author(s) of the work, which is fine if you're talking about one or two people, but can get to be a pain in the ass if you're talking about 5000 people.
Sometimes I use the QingPL for the works I create (for instance my Slashdot posts are licensed under the QingPL). But lately I've just been using ShareAlike 1.0, as that has a bit more credibility. For works that are already created I'm willing to use just about any Free license, at least until a more Free license can be found.
Slashdot Mirror
That's a nice theoretical argument, but what matters is how each comes out-of-the-box, to users who are not expected to do further work to secure their PDA. If Palm comes with 30 services written in C listening on the network by default and WinCE has 0, WinCE will be more secure, in the real word (or vice-versa).
But that's not the case. They both come out-of-the-box with no further securing necessary.
What I'm saying is that the platform has virtually nothing to do with the state of worms/viruses/trojans. Is it possible to run a secure PDA with PocketPC or Palm? Sure. Is is possible to run an insecure PDA with either? Sure.
Hello, transcriber? Palm has had Graffiti since inception. What do you call graffit but an instant transcriber.
I thought graffiti was a custom input method which had to be learned. Yes, it is very similar to standard manuscript, but it's not the same. Microsoft Transcriber allows you to use standard manuscript and even handwriting. Granted, it isn't quite as accurate as graffiti, at least not in my experience, but it is an additional feature to standard graffiti.
So there are no network-attack vectors for either platform?
Not in the OS. Of course, if you install insecure server software, you're going to wind up with an insecure device.
That sounds hard to believe.
Well, I guess you know the answer to your own question then. Or maybe it wasn't a question at all, but just flamebait.
It's completely equivalent. If you install bad software, your PDA will do bad things. If you don't, it won't.
I'm not sure if this is possible with the Palm devices, but I've got Linux running on my iPaq.
Netscape was a truly innovative, world-changing company, and if they weren't destroyed by MS the world would be a much different place today.
Let's see, would we be better off with an open source browser run by the non-profit charity Mozilla Foundation, or with a shareware browser which can't even be used legally for commercial purposes or for more than a 30 day trial without paying a fee?
Killing Netscape, the for-profit company, was a good thing.
Considering most people use Windows it makes sense to initially develop a program for Windows. It's a responsible use of tax money.
No, what would be a responsible use of tax money would be to create an API, and let whoever wants to create software for that API to use it, on any platform. In other words, what the US government did.
How about the Linux community builds their own open source version? I'm sure it would easily be ported to OS X.
Is there even an open API available for the Linux community to build such a product? If there is, then yes, the Linux community might want to do this, but that's not an excuse for the government to provide people with incentive to buy Windows.
I mean, it would be more disturbing if they only had a version for Be OS, right?
For people who have Windows and don't have Linux, I'd think it would be equally disturbing.
If it's truly web-based, then yes, it should work with Lynx.
It apears that Tampa is a very different place in that case.
Probably so. And New Jersey, where I lived most of my life, is probably even more different.
If people really do speed up when crowded form behind around there I can only be thankfull most people around here aren't that stupid.
I'm not sure where I said they do that. Some people do it, I suppose. If I'm in the left lane and I see someone coming really fast behind me I'll usually speed up to pass the car on my right and get over.
Most of the time the loss occures when the lane jocky finaly traps himself behind someone going a bit slower than traffic average because he jumped lanes at the first sign of slowdown only to guessed wrong.
Now you're talking about something completely different. I'm just saying that oftentimes speeding gets you from point to point signficantly faster.
If you drive emergency services you get a very distorted view of what 'normal' people see as traffic. I hope that's not confusing things here.
No, not at all. I was a volunteer for a short period of time in my life, over a year ago, in New Jersey.
My experience is predominately in the St. Louis area, with a little in california while visisting relatives.
I've used interstate highways complete with right side on-ramps and off-ramps for most of my life as a commuter. I drove significantly faster than the rest of the traffic the vast majority of the time, never got in an accident or even close, and only got one speeding ticket (which was in Connecticut on a road I wasn't very familiar with).
I've actually slowed down quite a bit lately, in the past six months I've driven a lot in the right lane and probably been passed more often than I've passed (whereas a couple years ago I probably passed 95 cards for every 5 that passed me). My life is not as rushed any more, I have a much shorter distance commute, and more flexible hours. But I don't have any illusions that I'm not taking more time getting from point A to point B as a result.
I believe these are futures contracts, which means they expire on a certain date. So you can hold on to them until expiration, and not actually use them, thus losing your entire investment, or you can sell them to someone else before expiration, possibly make a profit, but cause no effect whatsoever to the environment.
Ok, right. But if it were to take more energy to produce than they generate, they would be sold for less than their cost to manufacture.
Almost certainly not. Non-rechargable batteries take more energy to produce than they generate, but batteries are still sold at a profit.
So you'd have to pay $225 to save $150; and of course you wouldn't.
I wouldn't, but there's a sucker born every minute. Moreover, solar panels are generally not used to power houses, but to bring power to places where running wires to the grid isn't feasible.
It's bad, but it's not quite that bad.
Pinging www.slashdot.org [66.35.250.151] with 32 bytes of data:
Reply from 66.35.250.151: bytes=32 time=280ms TTL=44
Reply from 66.35.250.151: bytes=32 time=370ms TTL=44
Reply from 66.35.250.151: bytes=32 time=300ms TTL=44
Reply from 66.35.250.151: bytes=32 time=303ms TTL=44
Verizon's BroadbandAccess in Tampa, Florida.
Well, there is the issue of portability. Three laptops (or did you mean three backpacks) are a lot heavier (or more awkward) than one, and there's likely no room for more solar panels on a single backpack.
And if your laptop and your wifi card are anything like mine, they won't both fit in your laptop at the same time. So that means using USB probably, and means one more thing hanging off your laptop.
I'm not sure how practical this idea is in the first place. I guess for someone who lives in the city and walks everywhere, but I have thought about rigging up one of these things for my car. I'm probably better off just taking out the EVDO card every day and putting it in my iPAQ. Of course this wouldn't provide wireless access for anyone else (unless I managed to find a thin wifi PCMCIA card and bought the double PCMCIA adapter to replace my single one), but if I needed to I could use the serial port adapter to hook the iPAQ up to another computer like a wireless modem.
Section 2 of the GPL grants the right to prepare derivative works. Clause 2b limits that by requiring "any work that you distribute or publish" to be licensed under the GPL.
Wow. You're right. That's a pretty gaping wide hole there. Sorry about that.
By "use", I mean as a component of the software being developed for distribution. If a developer, for whatever reason, needs or wants to distribute under non-GPL terms, then that developer cannot include GPL components, which means that the user will not have the option of using what would have resulted if the developer were able to include those components.
So can't the developer just distribute the software without the GPL components, and let the end-user link in the GPLed components herself?
It's legal to create non-GPL derived works from GPL sources; you just can't distribute them.
I don't see that anywhere in the GPL. It may be how things work in practice, but copyright law covers the preparation of derivative works regardless of whether or not you distribute them, and like I said, I don't see where the GPL gives you permission to prepare a derivative work without causing the derivative to be released under the GPL.
Furthermore, if the GPL allows you to create non-GPLed derivative works, then there would be no point to having an LGPL at all.
The only real difference for a consumer is a trickle-down effect of the restrictions on developers; if the developer couldn't or wouldn't use a GPLed component for whatever reason, the user can't use whatever derived work would have resulted.
Why can't the developer use whatever component she wants? The GPL doesn't restrict use.
If I buy a music CD, then sell it to you, are you now able to press your own copies and sell those? Of course not, the RIAA would make sure of it.
But if you make 100 music CDs, then sell them to me, I can then resell them.
If I buy GPL-licensed software and then sell it to you, why are you supposed to be treated by the author(s) of the GPL'd work differently than I was?
Because you don't have to agree to the GPL in order to resell a GPLed product. So you don't have to distribute the source code.
If you mean that somehow the GPL license becomes void after First Sale, then that means you don't have the right to distribute the GPL software at all unless you negotiate a new license with the author(s).
No, here's what I'm saying. Company A takes a GPLed product and modifies it. They then sell 100 copies to Company B, and include source code on a separate CD. Company B then destroys the source code and resells the 100 copies. Since they have not agreed to the GPL, and are merely exercising their distribution rights under First Sale, they are not required to distribute the source code. Meanwhile, Company A did distribute the source code, so they haven't broken the law either.
The GPL doesn't limit anything, copyright law does. And in addition to distribution, copyright law limits copying and the preparation of derivative works. According to the FSF, dynamic linking is the preparation of a derivative work, and most likely they are correct.
You want to be able to do so while not contributing back and improving the software for the rest of the users- you want all the benefits of open source with none of the responsibilities. Thats not cool.
I think his point was that he can't contribute back the source code of the entire derivative work, because parts of it were created by others and not licensed under the GPL. It'd be nice if the GPL could exempt people under these particular circumstances, but it probably wouldn't be possible to create a license which did that without providing a huge loophole for someone who wanted to create a proprietary derivative (for instance by creating separate corporations to do the distribution and development).
Of course, I'd argue that the GPL already contains this loophole when combined with the doctrine of First Sale.
My biggest problem with the GPL is the FSF's position that even dynamically linking against a library under GPL is enough to make the resulting code a derivative work (and thus also subject to the GPL).
Be careful here. It doesn't make the program subject to the GPL, it simply doesn't exempt you from copyright infringement if you link a GPLed library without releasing the derivative under the GPL.
Note also that the FSF's interpretation may not be binding, but it hasn't been tested in court (that I'm aware of, and I recently attended a symposium on this very topic). So, in my mind, it creates an unacceptable exposure for anyone who wants to develop software but not adopt the GPL.
Was this symposium sponsored by SCO? Seriously, this is only a problem if you're making software which is specifically intended to be linked with a GPLed library. Then you might get hit with contributory copyright infringement (though as you've said the GPL hasn't been tested in court even for direct infringement, let alone contributory infringement). But if you're creating software which is intended to be legally linked against a license-compatible library, then you don't even have to worry about that. Sure, the end-user who links the software against a GPLed library is breaking the law, but no one is going to prosecute that.
How many companies, as opposed to not-for-profit organisations have actually released software as BSD?
It's probably in proportion to the number of companies that have released software under the GPL, if you consider the number of GPLed vs. BSDed software as a whole.
With the GPL the company is assured of getting any improvements back.
That's not at all true. The GPL just assures them that it's not legal to distribute binaries without distributing source. That's much different from requiring that improvements be given to the company.
For a company *releasing* software under an open-source license BSD has no real advantages and many disadvantages.
The same could be said of the GPL, though, and it'd be about as accurate. Other than for public relations purposes, there aren't any real advantages (from a profit perspective) to release new software under any free license. For old software there's often an advantage to give your modifications back to the original author, so that they can be incorporated into future versions, but that's about it (again, from a purely profit perspective, which is what most large companies have to go by).
Since neither license is different to an end-user, explain how the BSD license is better for the consumer?
They are different. If someone creates a plugin for that GPL software, and the plugin isn't released under the GPL, then it's illegal for you to install that plugin and create a derivative work.
Why people keep forgetting the LGPL in these flamewars I don't know
Because the LGPL is for libraries, and we weren't really talking about libraries.
The most popular licenses "in between" GPL and BSD are probably ShareAlike and Attribution-ShareAlike. Both are strong copylefts, in the derivative works must be released under the original license, but neither require releasing the source code. So if you want to modify and use the code in an embedded device, there's no need to release your source. But if you get popular enough for people to care, the code is going to get reverse engineered eventually anyway.
There's usually enough incentive to contribute back code that anyone not contributing modifications are unlikely to use GPLed software anyway. By contributing the modifications back you ensure that they are contained in all the future revisions of the code. Forking an active project is very difficult due to the constant merges that have to take place.
I like the GPL because it's a copyleft. If you take my work and modify it, that's fine, but you can't turn around and tell me I can't copy or distribute your modified version.
On the other hand, I think the BSD license is advantageous because it doesn't have the requirement to release source code. In this sense, I feel the BSD license is more free.
I've created my own license, the QingPL, which I consider the perfect license. Basically, you can do anything you want with the code without any restrictions, as long as you don't sue anyone for copyright infringement over it or a derivative of it. About a year after I created this license Creative Commons released ShareAlike 1.0, which is essentially the same thing which a bunch of extra legalese thrown in.
Unfortunately Creative Commons decided to drop ShareAlike from its 2.0 series of licenses. The next best alternative is Attribution-ShareAlike 2.0. That one adds the requirement to attribute the author(s) of the work, which is fine if you're talking about one or two people, but can get to be a pain in the ass if you're talking about 5000 people.
Sometimes I use the QingPL for the works I create (for instance my Slashdot posts are licensed under the QingPL). But lately I've just been using ShareAlike 1.0, as that has a bit more credibility. For works that are already created I'm willing to use just about any Free license, at least until a more Free license can be found.