My second thought was 'Wait, there are people who still use Fedora Core?
Not everybody loves Ubuntu, you know. I use Fedora Core, and it works fine. Never had any significant problems with it. In contrast, the Ubuntu cd locks up on boot on all machines I tried it on.
Actually, the Fedora Project have been counting the number of unique ip addresses that have requested updates, and reached more than one million, so there certainly are quite a few systems running Fedora Core.
Actually, a codec buddy is planned for Fedora 7. This would launch when the users tried to play certain codecs, and allow the user to install them after informing the user why they cannot be included in the main distribution.
If a distro claims to be a fully-featured desktop OS, and it can't do stuff that an average user would consider part of a fully-featured OS (that is, play MP3s and videos, or have fully-featured drivers even if they're binary), then the failure to do that is a flaw that is legitimate to flame against. Fedora may or may not portray itself as a fully-featured desktop OS, the fact is that if it can't play MP3s it is in the eyes of many users _not_ a fully-featured OS. And it certainly isn't an option for "switchers" who have lots of MP3s they'd want to use. Making those users feel like criminals for wanting to do so is also not a way to win friends and influence people.
Note that this isn't actually the fault of Red Hat/Fedora. The reason why they can't supply certain common media codecs is called software patents. The reasons why they cannot supply proprietary graphics drivers are 1) Their policy of only shipping free software and 2) EULAs of those drivers denying users the right of redistribution.
Vendors of non-free systems like Microsoft and Apple can purchase patent licenses for each customer, since each customers actually pay them money. It is hard to purchase a patent license for an unlimited and unknown number of users getting the product for free.
Right now it appears Ubuntu is becoming more popular than Fedora (or at least there's that PERCEPTION)
A good reason for staying with Fedora if you like it. With all the people singing "Ubuntu, Ubuntu, Ubuntu...", I feel like they are out to become the Linux monopoly, and I can't support that. I like Fedora, it works very well for me, so I'm sticking with it for the time being. But if I would try something else, it would surely not be Ubuntu, just out of sheer spite. Besides, Ubuntu locks up during kernel boot on all systems (granted, not many) I have tried it on.
ESR after all of this time finally realized that he as an end user might actually want to use a distribution that is NOT BETA or is for END USERS.
Yeah, because Ubuntu never hangs on boot and requires users to boot with noacpi, noapic, nolapic, etc</sarcasm>. Ubuntu have frozen on boot on all computers I have tried it on, while Fedora and CentOS works flawlessly. I refuse to even try booting Ubuntu with a modified kernel command line, since most or all Ubuntu users tout the "ease of use" of Ubuntu, in contrast to such "hard-to-use" distros like Fedora.
Ubuntu won't be installed on any computer near me unless it can be booted without a whole load of kernel parameters. Ubuntu quite frankly sucks.
I am wondering if it would be possible for someone (or company) in the US to preemptively approach the courts to force Microsoft to disclose the patents which are supposedly infringed by Linux.
It might be possible. IANAL, but I know that there is something called a declaratory judgement. From the Wikipedia article:
A declaratory judgment is typically requested when a party is threatened with lawsuit and the threatened lawsuit is not yet filed...
The idea that any program can write its configuration into a centralized system (the registry) could be better than having 100's of configuration files around in different places
How? Most (or all) system-wide configuration files are in the/etc directory tree, which makes them quite easy to find. In addition, they are almost always made up of text, which makes it much easier to understand what they say. In Windows, programs and components usually write into a registry subtree accessed through the component GUID (the CLSID83127-432423-32432-3423423 identifier). How's that for transparency?
I think eventually, we'll reach a point that software per se isn't patentable, the functional ideas that the software expresses will be the patentable aspect.
Are you claiming that this would be better in some way? How would it even differ from the current system?
A common question on various computer forums is how to turn off "that stupid dialog that asks me to confirm certain things, such as installing software" in Vista. A common complaint is "since it is my computer, it should obey me without asking me to prove that I am its owner".
Computer security is fundamentally incompatible with computer illiterates.
didn't that turn out to be against local law (TPB was working within the law?).
That hasn't been determined yet. The police have imaged their servers, but refuse to give them back. They probably intend to keep them forever, maybe sell them to some criminals to make a tidy profit, as has been recently reported in local newspapers.
The case main prosecutor actually wrote in a PM six months before the raid that TPB was probably not doing anything illegal, or at least not doing something that they could prosecute. You can't prosecute someone for contributory copyright infringement without also prosecuting someone for the primary infringement. Finding the primary infringers would be a problem, and that's why TPB was off the hook, at least until the MPAA called the administration and demanded that something had to be done, legal or not.:)
What if you could make a standard for the batteries themselves and fuel stations offered quick change (not charge) capabilities where you pull in and replace your battery.
What advantage would such a system have over a system where you plug your car into an electrical outlet for five minutes?
40% of the heat from fuel turned into movement power?
I'd think that it would be interesting to also consider the energy spent in moving the fuels from their sources to the consumers when computing efficiency. In many cases, oil and gasoline are transported long distances to reach the consumers, and that could certainly be computed into the efficiency metric.
Those bits have compatibility problems because they are designed to run as admin.
I thought that a major part of UAC was to run older programs in a sandbox with a virtualized harddrive and a virtualized registry avoid most or all of those "written to run as admin" incompatibilities that would otherwise appear.
It is pretty much their style to promise the moon when a competitor comes out with a good product so that people hold off on buying it, even though they never deliver said moon.
Sounds like their Cairo project. It was used to dissuade people from buying into the NextSTEP platform, as Cairo would be so much better and would be released only a year later. It was never finished. Most technologies developed for Cairo did make it into other Microsoft products later on though, but the object based file system is still absent.
The reason why I'm going to take issue with this is the view point of the user, the PHB, the non-developer.
End-users shouldn't use development versions unless they understand that they are unfinished and may not do what they want. If they download development versions without understanding this, they themselves are only to blame. Part of this could in theory be solved with user education, so that users understand that a release in the OSS world isn't really the same as a release in the proprietary world. But in practice, I think user education is more or less impossible, just as it is almost impossible to teach users about computer security, especially where it will degrade their convenience.
Maybe the solution here is to create a new category of release that is descriptive enough. Call it, "Code Release for Development".
It already exists, in the form of versions below 1.0.
A CRD would have the major features of the app wired up to work as a proto-type, with enough documentation so other coders can work on it. It would also have a clear and understandable directive that binaries should not be released to users until X number of features are fully implemented, and key features for minimal release configuration would be A, B, C, & D, and have clear metrics for what determines "properly working".
Sounds like way to much work. With such a requirement, the risk that developers would refrain from releasing their programs to the community would be substantial. But that might be what you really want.
If everyone released horrid, non-finished apps into the wild, OSS will soon be viewed as nothing more than a poor sub-culture.
Better to release non-finished apps than not releasing anyting at all.
The the parent to your post seems to think that if I write something in my spare time that isn't polished enough for use by end-users, it shouldn't be released at all. Even for development versions he suggest a whole bunch of minimal requirements that would need to be filled, making it rather probable that developers would refrain from releasing it at all.
A good example is Linux. I Actually never looked at the 0.01 release, but what I heard it didn't conform to any of his criteria for release. But if it hadn't been released, nobody knows if Linus would have brought it to the point of fulfilling these criteria by himself.
That is no different than propietary source -... a competitor/customer can always come up with their own propietary solution to run circles around predatory pricing.
Except that they would have to start from scratch. Customers using F/OSS projects can take the last free version and continue from there, which is a serious head start compared with starting from scratch.
The company that makes the IDE for a living however, is seriously screwed, even if they make a better product.
By the same reasoning, OpenOffice.org should refrain from releasing their office suite, since the use of OO.o cuts into Microsoft's profits.
Maybe you like SCO. They suggested to lawmakers that developers should be required to ask for money for their products, so that competitors that don't want to release their software for free would have an easier time to compete.
Slashdot Mirror
Not everybody loves Ubuntu, you know. I use Fedora Core, and it works fine. Never had any significant problems with it. In contrast, the Ubuntu cd locks up on boot on all machines I tried it on.
Actually, the Fedora Project have been counting the number of unique ip addresses that have requested updates, and reached more than one million, so there certainly are quite a few systems running Fedora Core.
It doesn't follow Debian, duh! ;)
Actually, a codec buddy is planned for Fedora 7. This would launch when the users tried to play certain codecs, and allow the user to install them after informing the user why they cannot be included in the main distribution.
Note that this isn't actually the fault of Red Hat/Fedora. The reason why they can't supply certain common media codecs is called software patents. The reasons why they cannot supply proprietary graphics drivers are 1) Their policy of only shipping free software and 2) EULAs of those drivers denying users the right of redistribution.
Vendors of non-free systems like Microsoft and Apple can purchase patent licenses for each customer, since each customers actually pay them money. It is hard to purchase a patent license for an unlimited and unknown number of users getting the product for free.
A good reason for staying with Fedora if you like it. With all the people singing "Ubuntu, Ubuntu, Ubuntu...", I feel like they are out to become the Linux monopoly, and I can't support that. I like Fedora, it works very well for me, so I'm sticking with it for the time being. But if I would try something else, it would surely not be Ubuntu, just out of sheer spite. Besides, Ubuntu locks up during kernel boot on all systems (granted, not many) I have tried it on.
Yeah, because Ubuntu never hangs on boot and requires users to boot with noacpi, noapic, nolapic, etc</sarcasm>. Ubuntu have frozen on boot on all computers I have tried it on, while Fedora and CentOS works flawlessly. I refuse to even try booting Ubuntu with a modified kernel command line, since most or all Ubuntu users tout the "ease of use" of Ubuntu, in contrast to such "hard-to-use" distros like Fedora.
Ubuntu won't be installed on any computer near me unless it can be booted without a whole load of kernel parameters. Ubuntu quite frankly sucks.
Good. They could start with Steve Ballmer.
It might be possible. IANAL, but I know that there is something called a declaratory judgement. From the Wikipedia article:
A declaratory judgment is typically requested when a party is threatened with lawsuit and the threatened lawsuit is not yet filed...
Sounds like a perfect fit.
So why does Vista run all setup programs as Administrator then?
How? Most (or all) system-wide configuration files are in the /etc directory tree, which makes them quite easy to find. In addition, they are almost always made up of text, which makes it much easier to understand what they say. In Windows, programs and components usually write into a registry subtree accessed through the component GUID (the CLSID83127-432423-32432-3423423 identifier). How's that for transparency?
And so do most trolls. :)
Are you claiming that this would be better in some way? How would it even differ from the current system?
A common question on various computer forums is how to turn off "that stupid dialog that asks me to confirm certain things, such as installing software" in Vista. A common complaint is "since it is my computer, it should obey me without asking me to prove that I am its owner".
Computer security is fundamentally incompatible with computer illiterates.
Latency - yes, transfer rate - no.
In my university we used so much free software that it would be strange to have a course teaching students that "it's illegal to copy software". :)
That hasn't been determined yet. The police have imaged their servers, but refuse to give them back. They probably intend to keep them forever, maybe sell them to some criminals to make a tidy profit, as has been recently reported in local newspapers.
The case main prosecutor actually wrote in a PM six months before the raid that TPB was probably not doing anything illegal, or at least not doing something that they could prosecute. You can't prosecute someone for contributory copyright infringement without also prosecuting someone for the primary infringement. Finding the primary infringers would be a problem, and that's why TPB was off the hook, at least until the MPAA called the administration and demanded that something had to be done, legal or not. :)
There is a difference between not caring for Linux and being hostile towards Linux.
What advantage would such a system have over a system where you plug your car into an electrical outlet for five minutes?
I'd think that it would be interesting to also consider the energy spent in moving the fuels from their sources to the consumers when computing efficiency. In many cases, oil and gasoline are transported long distances to reach the consumers, and that could certainly be computed into the efficiency metric.
I thought that a major part of UAC was to run older programs in a sandbox with a virtualized harddrive and a virtualized registry avoid most or all of those "written to run as admin" incompatibilities that would otherwise appear.
Sounds like their Cairo project. It was used to dissuade people from buying into the NextSTEP platform, as Cairo would be so much better and would be released only a year later. It was never finished. Most technologies developed for Cairo did make it into other Microsoft products later on though, but the object based file system is still absent.
End-users shouldn't use development versions unless they understand that they are unfinished and may not do what they want. If they download development versions without understanding this, they themselves are only to blame. Part of this could in theory be solved with user education, so that users understand that a release in the OSS world isn't really the same as a release in the proprietary world. But in practice, I think user education is more or less impossible, just as it is almost impossible to teach users about computer security, especially where it will degrade their convenience.
Maybe the solution here is to create a new category of release that is descriptive enough. Call it, "Code Release for Development".It already exists, in the form of versions below 1.0.
A CRD would have the major features of the app wired up to work as a proto-type, with enough documentation so other coders can work on it. It would also have a clear and understandable directive that binaries should not be released to users until X number of features are fully implemented, and key features for minimal release configuration would be A, B, C, & D, and have clear metrics for what determines "properly working".Sounds like way to much work. With such a requirement, the risk that developers would refrain from releasing their programs to the community would be substantial. But that might be what you really want.
If everyone released horrid, non-finished apps into the wild, OSS will soon be viewed as nothing more than a poor sub-culture.Better to release non-finished apps than not releasing anyting at all.
The the parent to your post seems to think that if I write something in my spare time that isn't polished enough for use by end-users, it shouldn't be released at all. Even for development versions he suggest a whole bunch of minimal requirements that would need to be filled, making it rather probable that developers would refrain from releasing it at all.
A good example is Linux. I Actually never looked at the 0.01 release, but what I heard it didn't conform to any of his criteria for release. But if it hadn't been released, nobody knows if Linus would have brought it to the point of fulfilling these criteria by himself.
Except that they would have to start from scratch. Customers using F/OSS projects can take the last free version and continue from there, which is a serious head start compared with starting from scratch.
By the same reasoning, OpenOffice.org should refrain from releasing their office suite, since the use of OO.o cuts into Microsoft's profits.
Maybe you like SCO. They suggested to lawmakers that developers should be required to ask for money for their products, so that competitors that don't want to release their software for free would have an easier time to compete.