Why do they turn off the incremental updates, rather than keeping them working forever? It seems that if it worked at all they could keep the "update to version x.yy" button on the updates panel, and you could click that through all the intermediate versions to update to the current one.
I had to reinstall because I had not updated my Ubuntu machine for too long and I finally got some software (google earth) that required the new system. This was rather annoying, in particular because it insisted on partitioning my existing Ubuntu partition in half to preserve the old system (I instead wiped the Windows install because I had not ever used it and it did seem to be happy with having the entire disk be Ubuntu).
Linux is encumbered by it's own (non-Free, btw) licensing requirements.
Okay now I know who I am talking to.
Too bad, I was really hoping there would be clear information, it seemed perfectly possible that there was actual usable api that OSS programs were ignoring for whatever reason. But it turns out you are another astroturfer.
Maybe "ESP" for "Exchange Server Protocols" (the title of the page) is close. Other abbreviations I see are "MS-OXDOCO".
A bit more searching found that "EAS" stands for "Exchange Active Sync". Sorry I did not catch that, as I saw plenty of "Active Sync" documentation but not with "Exchange" in front of it. That's closed anyway.
Also this paragraph was damn easy to find, apparently "not all is under the open specification promise":
Patents. Microsoft has patents that may cover your implementations of the technologies described in the Open Specifications. Neither this notice nor Microsoft's delivery of the documentation grants any licenses under those or any other Microsoft patents. However, a given Open Specification may be covered by Microsoft's Open Specification Promise (available here: http://www.microsoft.com/interop/osp) or the Community Promise (available here: http://www.microsoft.com/interop/cp/default.mspx). If you would prefer a written license, or if the technologies described in the Open Specifications are not covered by the Open Specifications Promise or Community Promise, as applicable, patent licenses are available by contacting iplg@microsoft.com.
the "bug" is likely the way the underlying libraries handle the situation, not the application itself. That doesn't make it any better, but MS is great about keeping old code around. Nobody has probably looked at that code in literally years. No freshmen CS is going to be able to outcode management's blind decisions.
Actually this should be possible. If you can move ONE message, then the freshman should be able to fix the program so that it divides the selection up into individual messages and tells the database to do each move independently. Unless a single message is too big, this will work.
I think this is what the original poster was getting at, and I agree it should be possible to fix this bug no matter what the back end is doing.
In other words, standard FAT file path conventions. The most used file system in the world. As used by about 90% of people's desktop computers.
Um, "FAT file path conventions" means 8.3 filenames and no slashes at all, and restricted to a very small set of ASCII characters. Not even Windows has used this in a long time. I think you are maybe talking about "MSDOS conventions".
Windows accepts both forward and backward slashes in filenames and treats them equivalently. Backslash is used as an escape character in many programming languages and for this reason alone they have to support an alternative. I'm guessing Symbian does as well but I don't know.
Drive letters *should* have been eliminated by a similar way backslashes were. My suggestion is that "/A/" should have meant the same as "A:/". Then readdir("/") could have returned all the disks, and we would still have a single interface to list directories even when they added networking, something Unix has had for decades. But Microsoft was too stupid to fix this in MSDOS 2 and later seemed to have turned completely hostile to Unix compatibility even when it was braindead obvious it would improve their platform.
Although OSX might be better than Linux, I have not seen any examples of what you are talking about.
Download & install of an application on OSX seems to always consist of:
1. download the ".dmg" file 2. Double-click it and be a bit mystified until you find the "box" icon. Do not double-click anything that looks like the app or it might end up screwed up! 3. Double-click that and hit ok a bunch and install it (this step always works nicely) 4. For some software you will fuck up and it will not run unless you first double-click the "dmg" to open it. Not sure what that is, but an MP3 encoder did this and so did a version of Maya (but the new one is fixed). 5. Throw away the.dmg file so it does not clutter your desktop, after confirming that the software works.
No system does a "user friendly" thing. Windows users think it is "easy" to first "download" and then "install" is a sure sign of brainwashing. This is what I would consider user-friendly (though there are obvious security problems, but this is the most complicated I would consider to actually be user-friendly):
1. User clicks on the "run this software" button on the website 2. SOFTWARE RUNS NOW! 3. And it is installed, too, or asks user if they want it installed.
The article pretty clearly states that the global warming is due to the energy needed to grow the food needed to feed the dog, which actually can exceed the energy needed to both build and fuel a car. Do you really completely support your dog with food that you would otherwise have thrown away if you did not have a dog?
I think it was interesting. However the right-wing nutjobs have already started claiming this is an actual attempt to ban pets, rather than the more interesting observation that reducing global warming is much more complex and difficult than anybody thinks. That is combined with the total clueless idea that it is exhaled CO2 that is causing global warming, espoused in some of the letters to that article and amazingly enough not refuted by a single one!
What about closed shops, where corporate executives decide to sign an agreement with union executives whereby all workers must join a specific union and pay dues to that union in order to work for that company?
What? You are saying that such free agreements between parties should be illegal? Unions should be "regulated" then?
No wonder libertarians get no respect. Statements like this just show that you have absolutely no consistency and your supposed "reasons" are thrown away at the slightest whim when they disagree with your underlying far-right philosophy.
No, the average person listening to Fox or CNN or NPR has never heard of a "libertariain".
All my knowledge of libertarinism is from a little rag called "Ergo" at MIT, which espoused Ayn-Rand "objectivist" policy and was so nutso that some felt it was in fact a communist plot to brainwash the engineers into believing capitalism was unstoppably evil.
The libertarians have themselves to blame for their own perception.
Libertarians generally want government force behind contracts, in order to make it practical to do business. I certainly can't speak for them, but I would think at least some libertarians like government enforced copyright laws because they are a simple and inexpensive way to get the types of contracts that differentiate their ideas from anarchy. Instant copyright like we have now but with a seriously limited time period (like a year) would probably be the the most in keeping with pure non-anarchy libertarian ideas.
Also the BSD crowd make a pretty persuasave argument that there would be plenty of free software without GPL.
You may be right, I actually was not going to get the nvidia driver to work at all, because I had too old of a video card. However I could not figure out how to restore Ubuntu to a state where it would offer to install the driver without doing a reinstall.
What happened is that it ran the 'nv' driver. You could pick the "unsupported hardware" icon (or whatever it was called) and it listed two grayed-out Nvidia drivers. I picked one, then the other, and in both cases X refused to work (I also had to restore it by editing xorg.conf, which certainly would have stopped a lot of users). They obviously knew they would not work and perhaps could have done something stronger than graying them out, but I'm sure I would have just presevered and screwed up my installation no matter what...
Anyway I then did a search and found that my old graphics card required the older driver from Nvidia, so I got that and attemted to install it. It did the "download the kernel headers and recompile" step but the driver refused to load, saying the kernel was wrong. This has *always* happened to me when I try to use the nvidia downloaded driver. Had to edit xorg.conf to put 'nv' back.
I finally realized that I had a newer card (note both cards are about 4 years old and I actually put the old one in the machine because the old Mandrake install did not like the "new" one 4 years ago) and put that in and determined it was the type that the modern driver recognized. However rather than try the nvidia driver, I attempted to get Ubuntu to recognize and install the modern nvidia driver from the repository. No luck fooling it into doing so, so I finally reinstalled Ubuntu. It then recognized the card immediately, downloaded and installed the driver, and it works great. Still a frustrating experience.
The percentage of Linux desktops using Nvidia is probably higher than the percentage of Windows machines, therefore any accurate count of usage of Nvidia drivers would over-estimate Linux desktop usage. The reason is that Linux 3D is enormously better on Nvidia verses other cards on Linux, while the difference is not so much on Windows. People wanting working graphics greatly outnumber purists who reject the closed drivers (though it is likely that an open Nvidia driver would make the percentage for them far greater).
I agree that the percentage downloaded from the Nvidia site is meaningless, because very very few Linux desktops with Nvidia cards do so. You can blame it on the fact that the downloaded driver never installs correctly (always needing to be recompiled because the kernel is wrong) and then fails, while the one provided by the distribution always works.
WIndows update runs a program that downloads from Nvidia.
However I would agree about the OEM thing, unless the OEM install actually downloads the driver on first boot, which seems unlikely. And an awful lot of Nvidia cards are sold as OEM installs.
I completely screwed my Ubuntu install by trying to install the downloaded Nvidia driver (mistakenly thinking that I needed to do so to get Nvidia graphics to work). As there was nothing on the machine I reinstalled and let Ubuntu do it's automatic thing, and it worked perfectly, I strongly recommend doing that from now on, and I'm sure the vast majority of Ubuntu installs with nvidia cards do exactly that.
I am working at Rhythm & Hues, which is about 95% Linux on the desktop (all SUSE). Windows is used to run scanners, Macs are used for painting.
I was also at Digital Domain, which is closer to 60% Linux on the desktop. There Windows is used to run Maya, and also for painting. Macs hardly used at all.
I also worked at The Foundry in London selling the Nuke compositing system. Sales are approximatey 1/3 each for WIndows, Linux, and OS/X, this is for the interactive version. Since that costs about 50x as much as a render-only version used by a renderfarm you can be pretty damn certain that shows the approximate usage on desktops and not your bogus "renderfarm" claim. Digital Domain gets theirs for free so they don't count, and R&H does not use Nuke so that is a huge collection of Linux that is not being counted. ILM and Weta are also huge users of Linux and they have site licenses so that may also not be counted.
Indeed this bug in OO was used as an excuse by Microsoft to make their ODF output be incompatible. Their argument was basically "since these two different ODF programs treat this sample document differently, we are allowed to write a completely different third implementation". This bogus argument was wrapped in a disgusting and somewhat horrifying amount of obfuscated technobabble by obviously intelligent but amoral individuals at Microsoft (you can find several repeated links to this posted by astroturfers right here on Slashdot).
Truly I used to think that Microsoft was just somewhat incompetent and rushed so that their programmers tended to reinvent things without finding out there was a standard already. But this deliberate outright lying, with enough wording to make a pointed-headed boss think it is some complex technical argument, convinced me that there really are evil people there. Scary indeed.
I think the worry is that the sunspot or solar storm is going to produce a velocity of the gas that is greater than "walking speed". So it would shift the lines by that amount.
I think they must be averaging over a long enough period and for the entire star that this is not a worry, however it does seem like this noise would swamp any such observations.
Then you tell the contributor you cannot accept their code unless they fix the license they are returning it under. Same thing would happen no matter what license you use for your code, somebody can try to give you code with some restrictions you cannot use, and that just means that you cannot use it. But you can ask the contributor to fix it.
You seem to be under the impression that the GPL somehow "infects" the code and cannot be removed, even by the author. This is just more FUD.
As for SGI, it certainly is true that if the contributor thinks they are giving up too much by contributing their code, then they won't do it. That could be true even if you only GPL'd your code, there are programmers that hate the GPL and will refuse to contriute code that will be distributed GPL (see CDDL and MSPL).
This is kind of wrong. The GPL2/3 both have quite a lot of language excluding stuff that people might consider covered by copyright laws. The most obvious is the "mere aggregation" clause, but also stuff about "code commonly included with the OS" and lots of other stuff.
Therefore it is pretty clear that the GPL2/3 is NOT "intended to cover as much as possible". If it was then no such language would be included.
No, you have violated the copyright of the original author. The end user has no rights to demand anything from you (though the fact that they may complain to the copyright holder may make it look like they do).
The GPL is a license that says "I grant you an exception to my copyrights ONLY if you do these extra steps". If you don't do the steps you are not granted the exception by the license, so you are violating the copyrights held by the original author.
This is absolutely false and a huge piece of FUD from Microsoft, and the basis of the "viral" claim for the GPL.
A judge will find you have violated copyright, and place the legal penalty for copyright violations on you, which is that monetary compensation be paid to the copyright owner and that you cease distribution.
It is true that you can often avoid going before a judge by agreeing to some demand of the copyright holder, which might be "release your source code". But the demand could be anything.
That's nonsense. You can dual-license something all you want. You can make up a license that has nothing to do with GPL and contradicts it everywhere, but as long as you dual-license your code you are compatbile.
To answer your idea directly, where you seem to want to allow GPL3 but not GPL2, you dual-license using your "foo license" and the GPL3.
The GPL3 does mean that a derivative can drop the "foo license" requirement from the distribution. The "foo license", since you write it yourself, could say anything, such as even requiring that redistriubtions not be GPL3 and certainly not requiring it.
Slashdot Mirror
Since the next message was them trying a second time, the first three letters sent were "LOL"
Why do they turn off the incremental updates, rather than keeping them working forever? It seems that if it worked at all they could keep the "update to version x.yy" button on the updates panel, and you could click that through all the intermediate versions to update to the current one.
I had to reinstall because I had not updated my Ubuntu machine for too long and I finally got some software (google earth) that required the new system. This was rather annoying, in particular because it insisted on partitioning my existing Ubuntu partition in half to preserve the old system (I instead wiped the Windows install because I had not ever used it and it did seem to be happy with having the entire disk be Ubuntu).
Linux is encumbered by it's own (non-Free, btw) licensing requirements.
Okay now I know who I am talking to.
Too bad, I was really hoping there would be clear information, it seemed perfectly possible that there was actual usable api that OSS programs were ignoring for whatever reason. But it turns out you are another astroturfer.
I do not see the letters "EAS" or "EWS" in there.
Maybe "ESP" for "Exchange Server Protocols" (the title of the page) is close. Other abbreviations I see are "MS-OXDOCO".
A bit more searching found that "EAS" stands for "Exchange Active Sync". Sorry I did not catch that, as I saw plenty of "Active Sync" documentation but not with "Exchange" in front of it. That's closed anyway.
Also this paragraph was damn easy to find, apparently "not all is under the open specification promise":
Patents. Microsoft has patents that may cover your implementations of the technologies described in the Open Specifications. Neither this notice nor Microsoft's delivery of the documentation grants any licenses under those or any other Microsoft patents. However, a given Open Specification may be covered by Microsoft's Open Specification Promise (available here: http://www.microsoft.com/interop/osp) or the Community Promise (available here: http://www.microsoft.com/interop/cp/default.mspx). If you would prefer a written license, or if the technologies described in the Open Specifications are not covered by the Open Specifications Promise or Community Promise, as applicable, patent licenses are available by contacting iplg@microsoft.com.
Care to point to something describing this?
All my searches seem to reveal that Exchange communicates with something called "MAPI/RPC" and can also emulate POP3 and IMAP protocols.
There is also OWA, which is apparently the only reliable way to communicate with it and is what the email readers are actually using.
the "bug" is likely the way the underlying libraries handle the situation, not the application itself. That doesn't make it any better, but MS is great about keeping old code around. Nobody has probably looked at that code in literally years. No freshmen CS is going to be able to outcode management's blind decisions.
Actually this should be possible. If you can move ONE message, then the freshman should be able to fix the program so that it divides the selection up into individual messages and tells the database to do each move independently. Unless a single message is too big, this will work.
I think this is what the original poster was getting at, and I agree it should be possible to fix this bug no matter what the back end is doing.
In other words, standard FAT file path conventions. The most used file system in the world. As used by about 90% of people's desktop computers.
Um, "FAT file path conventions" means 8.3 filenames and no slashes at all, and restricted to a very small set of ASCII characters. Not even Windows has used this in a long time. I think you are maybe talking about "MSDOS conventions".
Windows accepts both forward and backward slashes in filenames and treats them equivalently. Backslash is used as an escape character in many programming languages and for this reason alone they have to support an alternative. I'm guessing Symbian does as well but I don't know.
Drive letters *should* have been eliminated by a similar way backslashes were. My suggestion is that "/A/" should have meant the same as "A:/". Then readdir("/") could have returned all the disks, and we would still have a single interface to list directories even when they added networking, something Unix has had for decades. But Microsoft was too stupid to fix this in MSDOS 2 and later seemed to have turned completely hostile to Unix compatibility even when it was braindead obvious it would improve their platform.
Although OSX might be better than Linux, I have not seen any examples of what you are talking about.
Download & install of an application on OSX seems to always consist of:
1. download the ".dmg" file .dmg file so it does not clutter your desktop, after confirming that the software works.
2. Double-click it and be a bit mystified until you find the "box" icon. Do not double-click anything that looks like the app or it might end up screwed up!
3. Double-click that and hit ok a bunch and install it (this step always works nicely)
4. For some software you will fuck up and it will not run unless you first double-click the "dmg" to open it. Not sure what that is, but an MP3 encoder did this and so did a version of Maya (but the new one is fixed).
5. Throw away the
No system does a "user friendly" thing. Windows users think it is "easy" to first "download" and then "install" is a sure sign of brainwashing. This is what I would consider user-friendly (though there are obvious security problems, but this is the most complicated I would consider to actually be user-friendly):
1. User clicks on the "run this software" button on the website
2. SOFTWARE RUNS NOW!
3. And it is installed, too, or asks user if they want it installed.
Free Range
Free Wheel
Free play (as applied to mechanical connections)
Free Choice
Free Enterprise
I think there are a lot counter examples to your claim. Of course I guess it is a Free Country and you can apply your own Free Interpretation.
The article pretty clearly states that the global warming is due to the energy needed to grow the food needed to feed the dog, which actually can exceed the energy needed to both build and fuel a car. Do you really completely support your dog with food that you would otherwise have thrown away if you did not have a dog?
I think it was interesting. However the right-wing nutjobs have already started claiming this is an actual attempt to ban pets, rather than the more interesting observation that reducing global warming is much more complex and difficult than anybody thinks. That is combined with the total clueless idea that it is exhaled CO2 that is causing global warming, espoused in some of the letters to that article and amazingly enough not refuted by a single one!
What about closed shops, where corporate executives decide to sign an agreement with union executives whereby all workers must join a specific union and pay dues to that union in order to work for that company?
What? You are saying that such free agreements between parties should be illegal? Unions should be "regulated" then?
No wonder libertarians get no respect. Statements like this just show that you have absolutely no consistency and your supposed "reasons" are thrown away at the slightest whim when they disagree with your underlying far-right philosophy.
No, the average person listening to Fox or CNN or NPR has never heard of a "libertariain".
All my knowledge of libertarinism is from a little rag called "Ergo" at MIT, which espoused Ayn-Rand "objectivist" policy and was so nutso that some felt it was in fact a communist plot to brainwash the engineers into believing capitalism was unstoppably evil.
The libertarians have themselves to blame for their own perception.
Libertarians generally want government force behind contracts, in order to make it practical to do business. I certainly can't speak for them, but I would think at least some libertarians like government enforced copyright laws because they are a simple and inexpensive way to get the types of contracts that differentiate their ideas from anarchy. Instant copyright like we have now but with a seriously limited time period (like a year) would probably be the the most in keeping with pure non-anarchy libertarian ideas.
Also the BSD crowd make a pretty persuasave argument that there would be plenty of free software without GPL.
You may be right, I actually was not going to get the nvidia driver to work at all, because I had too old of a video card. However I could not figure out how to restore Ubuntu to a state where it would offer to install the driver without doing a reinstall.
What happened is that it ran the 'nv' driver. You could pick the "unsupported hardware" icon (or whatever it was called) and it listed two grayed-out Nvidia drivers. I picked one, then the other, and in both cases X refused to work (I also had to restore it by editing xorg.conf, which certainly would have stopped a lot of users). They obviously knew they would not work and perhaps could have done something stronger than graying them out, but I'm sure I would have just presevered and screwed up my installation no matter what...
Anyway I then did a search and found that my old graphics card required the older driver from Nvidia, so I got that and attemted to install it. It did the "download the kernel headers and recompile" step but the driver refused to load, saying the kernel was wrong. This has *always* happened to me when I try to use the nvidia downloaded driver. Had to edit xorg.conf to put 'nv' back.
I finally realized that I had a newer card (note both cards are about 4 years old and I actually put the old one in the machine because the old Mandrake install did not like the "new" one 4 years ago) and put that in and determined it was the type that the modern driver recognized. However rather than try the nvidia driver, I attempted to get Ubuntu to recognize and install the modern nvidia driver from the repository. No luck fooling it into doing so, so I finally reinstalled Ubuntu. It then recognized the card immediately, downloaded and installed the driver, and it works great. Still a frustrating experience.
I think that is false.
The percentage of Linux desktops using Nvidia is probably higher than the percentage of Windows machines, therefore any accurate count of usage of Nvidia drivers would over-estimate Linux desktop usage. The reason is that Linux 3D is enormously better on Nvidia verses other cards on Linux, while the difference is not so much on Windows. People wanting working graphics greatly outnumber purists who reject the closed drivers (though it is likely that an open Nvidia driver would make the percentage for them far greater).
I agree that the percentage downloaded from the Nvidia site is meaningless, because very very few Linux desktops with Nvidia cards do so. You can blame it on the fact that the downloaded driver never installs correctly (always needing to be recompiled because the kernel is wrong) and then fails, while the one provided by the distribution always works.
WIndows update runs a program that downloads from Nvidia.
However I would agree about the OEM thing, unless the OEM install actually downloads the driver on first boot, which seems unlikely. And an awful lot of Nvidia cards are sold as OEM installs.
I completely screwed my Ubuntu install by trying to install the downloaded Nvidia driver (mistakenly thinking that I needed to do so to get Nvidia graphics to work). As there was nothing on the machine I reinstalled and let Ubuntu do it's automatic thing, and it worked perfectly, I strongly recommend doing that from now on, and I'm sure the vast majority of Ubuntu installs with nvidia cards do exactly that.
I am working at Rhythm & Hues, which is about 95% Linux on the desktop (all SUSE). Windows is used to run scanners, Macs are used for painting.
I was also at Digital Domain, which is closer to 60% Linux on the desktop. There Windows is used to run Maya, and also for painting. Macs hardly used at all.
I also worked at The Foundry in London selling the Nuke compositing system. Sales are approximatey 1/3 each for WIndows, Linux, and OS/X, this is for the interactive version. Since that costs about 50x as much as a render-only version used by a renderfarm you can be pretty damn certain that shows the approximate usage on desktops and not your bogus "renderfarm" claim. Digital Domain gets theirs for free so they don't count, and R&H does not use Nuke so that is a huge collection of Linux that is not being counted. ILM and Weta are also huge users of Linux and they have site licenses so that may also not be counted.
Indeed this bug in OO was used as an excuse by Microsoft to make their ODF output be incompatible. Their argument was basically "since these two different ODF programs treat this sample document differently, we are allowed to write a completely different third implementation". This bogus argument was wrapped in a disgusting and somewhat horrifying amount of obfuscated technobabble by obviously intelligent but amoral individuals at Microsoft (you can find several repeated links to this posted by astroturfers right here on Slashdot).
Truly I used to think that Microsoft was just somewhat incompetent and rushed so that their programmers tended to reinvent things without finding out there was a standard already. But this deliberate outright lying, with enough wording to make a pointed-headed boss think it is some complex technical argument, convinced me that there really are evil people there. Scary indeed.
I think the worry is that the sunspot or solar storm is going to produce a velocity of the gas that is greater than "walking speed". So it would shift the lines by that amount.
I think they must be averaging over a long enough period and for the entire star that this is not a worry, however it does seem like this noise would swamp any such observations.
Then you tell the contributor you cannot accept their code unless they fix the license they are returning it under. Same thing would happen no matter what license you use for your code, somebody can try to give you code with some restrictions you cannot use, and that just means that you cannot use it. But you can ask the contributor to fix it.
You seem to be under the impression that the GPL somehow "infects" the code and cannot be removed, even by the author. This is just more FUD.
As for SGI, it certainly is true that if the contributor thinks they are giving up too much by contributing their code, then they won't do it. That could be true even if you only GPL'd your code, there are programmers that hate the GPL and will refuse to contriute code that will be distributed GPL (see CDDL and MSPL).
This is kind of wrong. The GPL2/3 both have quite a lot of language excluding stuff that people might consider covered by copyright laws. The most obvious is the "mere aggregation" clause, but also stuff about "code commonly included with the OS" and lots of other stuff.
Therefore it is pretty clear that the GPL2/3 is NOT "intended to cover as much as possible". If it was then no such language would be included.
No, you have violated the copyright of the original author. The end user has no rights to demand anything from you (though the fact that they may complain to the copyright holder may make it look like they do).
The GPL is a license that says "I grant you an exception to my copyrights ONLY if you do these extra steps". If you don't do the steps you are not granted the exception by the license, so you are violating the copyrights held by the original author.
a judge will make you release the source
This is absolutely false and a huge piece of FUD from Microsoft, and the basis of the "viral" claim for the GPL.
A judge will find you have violated copyright, and place the legal penalty for copyright violations on you, which is that monetary compensation be paid to the copyright owner and that you cease distribution.
It is true that you can often avoid going before a judge by agreeing to some demand of the copyright holder, which might be "release your source code". But the demand could be anything.
That's nonsense. You can dual-license something all you want. You can make up a license that has nothing to do with GPL and contradicts it everywhere, but as long as you dual-license your code you are compatbile.
To answer your idea directly, where you seem to want to allow GPL3 but not GPL2, you dual-license using your "foo license" and the GPL3.
The GPL3 does mean that a derivative can drop the "foo license" requirement from the distribution. The "foo license", since you write it yourself, could say anything, such as even requiring that redistriubtions not be GPL3 and certainly not requiring it.