MySQL is not closing off its source. It's just choosing not to distribute the source code for one version of its product in one way. It doesn't violate the GPL in any way and if you still really want the source you can get it from their repository.
Thank you for that accurate summary of the situation.
Thing is, many people don't understand the GPL. The GPL never said 'you must distribute your source code to everyone'... you can, for example, make private changes and never give them out. In fact, this is explicitly given as an example of an important freedom by Stallman, Moglen, etc. Similarly, you have the freedom to make changes and give them to only a few people; this is exactly what MySQL are doing. Now, the people that do receive the code are free to further distribute it, according to the GPL, and I am sure we will see the code in some manner (compare to CentOS). But MySQL are well within their legal (and moral) rights to have only part of their GPLed code available on their servers in tarball format for anonymous download.
it seems they are using interference to get individual photons of light to traverse every pathway simultaneously. Even if I am only partially correct there, the photons in the experiment are only detected and are never being used as an instrument for computation.
I guess it depends what you mean by 'computation'. As the article itself says,
So in total we will need [about] (5/64)*N^N photons.
Unsurprisingly, they need an exponential amount of photons (and even N^N, whereas we know the problem is solvable using 2^N or so). In effect, each photon is used to perform one 'calculation', in some sense. Or at least that is what I understand from TFA, IANA Physicist.
I think it was Philip K. Dick who said that, if Jesus had lived in our time instead of 2000 years ago, his followers would later be wearing little pendants of electric chairs, hang up large symbols of electric chairs in their churches, etc.
I believe libtorrent is actually part of the rtorrent project, but I could be wrong.
Yeah, this is a common mistake... there are two projects called 'libtorrent'. Sometimes they are referred to as 'Rasterbar libtorrent' (or 'Sourceforge libtorrent') versus 'Rakshasa libtorrent' (or 'rtorrent libtorrent').
Even more confusingly, there was a time at which they had virtually the same release versions (0.9, I think):) . This might still be true, but I don't follow it so I'm not sure.
Well, consider that utorrent takes around 10 MB (20 MB under wine on Linux), a small fraction of the 60 MB you report for Azureus (and I have seen Azureus use a lot more... it climbs over time).
Yeah, rtorrent is very good, especially recent releases with ecryption etc. When I started working on Deluge, we considered using rtorrent as a backend (but ended up using libtorrent instead).
Additionally, only the main BitTorrent.com tracker would have access to tracker-side protocol updates. So, this then means that the only benefit of using the mainline client is when downloading from the BitTorrent.com tracker!
See my comment above. They can keep the tracker protocols (which really don't matter much) and change the p2p ones, which is where all the important stuff happens.
I would like that to be so, but they now control the utorrent client, and utorrent is the only client that (1) has very reasonable resource usage and (2) has excellent performance. Sadly none of the competitors come close.
For a supposedly 'open' protocol, bittorrent has proven remarkably hard to implement well (basically only utorrent and Azureus do so). See this article.
One difference. They don't operate any of the servers people actually use. Unless they can convice the server operators (most of whom they can't legally even admit exists, which will make negotiations somewhat awkward) to adopt their closed protocol, who will notice any optional dead protocols their 'official' but little used client supports?
No, that isn't quite true.
First, Bittorrent is a peer-to-peer protocol. Only a minor part of it is communication with the server (aka tracker). They might keep the tracker protocol exactly the same, and alter the important p2p part.
Second, this has already been done, and successfully. For example, utorrent came out with a 'PEX' (Peer Exchange) protocol that wasn't in the spec. So it was only used between peers that were both using the utorrent client. This provided a nicer bittorrent experience for utorrent users, especially as utorrent's marketshare rose. Later on, because of utorrent's dominant position, other clients started to implement utorrent PEX (KTorrent, libtorrent-based clients), with varying degrees of success.
A similar issue is Azureus's DHT protocol, which is not in the standard. Although at least Azureus is open source, so you can read the actual code to help in understanding what nonstandard protocols they have invented (but then they also have a very nice wiki).
The point is, it is easy to 'embrace and extend' the bittorrent protocol, even if you don't have control of the servers. Is 'extinguish' next? Probably not, but I for one won't be using the official Bittorrent client.
Is KDE the default on any of these configurations?
No. GNOME is the default on all distros that are likely to be pre-installed these days, Ubuntu, SUSE and Red Hat/Fedora.
You can of course very easily install KDE on these systems, a simple apt-get on Ubuntu for example.
The reason, I suspect, is mostly the licensing of Qt and KDE, which is the GPL (and not LGPL, which GTK+ is). This makes it less corporate-friendly, in a way. It probably explains the big shift to GNOME in the major distros, as well as the focus on GNOME in nearly all the new mobile Linux initiatives.
I see your point, and I mostly agree with you actually. But there are 2 separate issues: (1) Do Alcatel-Lucent own MP3 patents, and (2) How much MP3 patents should cost to license.
If Alcatel-Lucent do own MP3 patents, then they can start to assert them whenever they want, and at whatever price they want... that is the (stupid) system.
Do they actually own the patents? I don't know. Certainly it does seem like they might not. But if they do, the price they charge for them is up to them.
Microsoft made good faith effort to license the technology in question from Fraunhofer for $16 million. But Alcatel-Lucent claimed that Microsoft licensed it from the wrong party. Even if that were the case, then Microsoft should only have been ordered to pay ~$16 million to Alcatel-Lucent, not 1.5 billion.
So if A rents out B's house without permission for 1 dollar per month to C, then C should be able to pay B 1 dollar per month for the place, after the mistake (that it is B's house, not A's) is discovered?
Fraunhofer could have given away free-of-charge licenses for MP3 patents. That doesn't mean that if those patents are actually owned by someone else, that other party has to also give them away for free.
The other example mentioned was Microsoft being ordered to pay $1.5bn over two MP3 related patents. We all know that MP3 is covered by a few hundred patents, and Microsoft paid a few million for a license for all those patents, so one or two patents they missed could never be worth $1.5bn.
What? Even if something is covered by many patents, the fact that you somehow get a great deal on licensing part of them (perhaps Fraunhofer gets free Windows licenses? who knows) doesn't mean the others should be licensed for the same (or even similar) price.
For example, Fraunhofer could decide tomorrow to grant a free license to all of their MP3 patents. Does that mean all other MP3-related patents should also be free of charge?
Amazingly, not only is TFA very scarce on information, it misses the most critical issue. Here is the actual reason for the delay:
Red Hat confirmed on Aug. 3 that it would be delaying the release of the newest member of its desktop Linux family, Red Hat Global Desktop, because the company is seeking to provide certain multimedia codecs. Sources close to Red Hat said obtaining some of these codecs was dependent on Red Hat coming to an agreement with Microsoft.
In other words, Red Hat wants to legally include WMA codecs, and are in the midst of negotiations with Microsoft. Microsoft wants a general patent pact (like with Novell, Xandros) in order to license WMA; Red Hat just wants to pay Microsoft for WMA and nothing more. Yet not to license WMA may be tricky, anti-trust wise (or is Red Hat not agreeing to a general patent pact enough of an excuse? No idea).
If Red Hat gets a license for WMA, then it will be shipping a product very different than Ubuntu (for better or for worse, we can argue). But that is the point of the delay.
"A single voice isn't going to tell girls that they shouldn't want to be pretty. One well-spoken voice might convince a few that they can be pretty and smart."
And this is why women are so totally screwed by modern (western) society. Once they had to be good mothers, nowadays they have to be both good mothers and have good careers. Once they had to be pretty, nowadays they have to be (according to what you just said) both pretty and smart (of course it isn't just you, it is society as a whole).
Women have traditional expectations of them - be mothers, be pretty - and additional expectations are being added on. This creates a terrible burden, in a very disproportionate way in comparison to men.
So 1) you can't use code based on the specification in a GPL V2 or GPL V3 program, because you can't satisfy the patent clause
Regarding code implementing the standard, you are indeed passing on any patent license that you have (a totally worthless one, if it even counts as one). The license (again, for whatever it is worth) is universal and nondiscriminatory. So no GPL problem there.
Regarding code implementing functionality not in the standard but sort of implied by it ("SpacingLikeWord97", etc.), this may indeed be a problem - you might be sued. But you can still write GPL code to implement it, there is no problem with the GPL's patent clause - the problem is with (possible) Microsoft patents. In fact, such code would be just as 'dangeous' as running Linux, given Microsoft's recent statements (or, to be more precise, just as dangerous as running Mono).
Elton at least admits he is a Luddite. He's entitled to his opinions, I guess. Anyhow, not all artists are like him; for example, Therapy? bandmembers live in different countries, and much of their collaboration is done by utilizing the internet: sending each other MP3s of song ideas. Then they meet physically for a few weeks and record the stuff (see interview here).
Considering that "One Cure Fits All" (2006) was among their better albums ever IMHO (and I have been listening to them since they got started around 1990), apparently this 'interweb' thing isn't necessarily as detrimental as Sir Elton believes.
the installed base of Office users is going to make it awfully difficult to supplant it as a file format of interchange.
The size argument is just an excuse. If there were a plugin for MS Office that made it save ODF as default, and worked perfectly, then it might be easy to convince CTOs to switch. And it would be possible for people running OpenOffice to send files to Office users, if the plugin were a single-click installation (much like Flash). That would enable people the choice to use whatever office software they want.
There are such plugins in the works. When they mature, things may get interesting.
Last year, Linux and Max OS X had only meager appeal to the CIOs, CSOs, IT and network administrators surveyed: 2% said they planned to deploy the open-source Linux, while none owned up to Mac OS X plans. July's survey, however, noted a six-fold increase in the total willing to do without Windows on at least some systems: 8% of those polled acknowledged Linux plans and 4% said they would deploy Mac OS X.
Hmm, assuming the data indeed reflects reality, this is a significant development, isn't it?
Nexenta is GNU+OpenSolaris. That is, GNU tools + OpenSolaris kernel, or more specifically, Ubuntu using the OpenSolaris kernel instead of Linux.
"Project Indiana" will use OpenSolaris as the kernel (obviously), but it isn't yet clear what else. Sun has their own set of tools for package management and so forth, so they don't need GNU (and not Debian or Ubuntu which are specialized systems built on GNU). So, it may turn out that Indiana and Nexenta only have the OpenSolaris kernel in common.
However, is this the right thing to do? I don't think so. Nexenta are very wise to build a system on Ubuntu, which was wise to use Debian, which was wise to use GNU. Sun should accept that GNU and systems based on GNU are the best foundation for an operating system these days. Sun's best move would be to just support or even make official the Nexenta project, if they want to compete with Linux.
However, one reason why they might not is control. Despite rumors of GPL-ing OpenSolaris, it hasn't yet happened. Perhaps Sun doesn't want an operating system based on GPL2/3 code, and just for that reason will use their own userspace tools, to whom they own all copyrights, and can pick their licenses as they please.
Let's look at the big picture here of Microsoft monopolizing the Chinese desktop market. The US trade deficit with China is $233 billion. If, in several years, there are (say) 1 billion computers in China, and each pays $100 for Microsoft products (Windows, Office, OneCare, who knows what else by then), then Microsoft will be responsible for $100 billion going in the opposite direction than the $233 billion. That is, Microsoft's income from China will be about the same order as that of the entire trade deficit.
(Of course there are many assumptions and guesses here - I don't think this is a serious economic prediction. But it does show the general idea.)
Two conclusions:
There is massive motivation for the US government to bolster Microsoft in any way possible. Don't expect any antitrust lawsuits in the US any time soon.
China's adoption of Microsoft products may be temporary. Other nations have done it in the past - adopt Western ways, modernize their economies using them, and then replace those technologies with their own (e.g., Japan and the auto market). China sees Microsoft as the quickest way to modernize their computer industry. But, especially as a central authoritative government, they can change strategy later on, when the 'Microsoft Tax' becomes a burden.
Slashdot Mirror
Thing is, many people don't understand the GPL. The GPL never said 'you must distribute your source code to everyone'... you can, for example, make private changes and never give them out. In fact, this is explicitly given as an example of an important freedom by Stallman, Moglen, etc. Similarly, you have the freedom to make changes and give them to only a few people; this is exactly what MySQL are doing. Now, the people that do receive the code are free to further distribute it, according to the GPL, and I am sure we will see the code in some manner (compare to CentOS). But MySQL are well within their legal (and moral) rights to have only part of their GPLed code available on their servers in tarball format for anonymous download.
To attack MySQL about this is very unfair.
Cool, thanks for the correction.
I think it was Philip K. Dick who said that, if Jesus had lived in our time instead of 2000 years ago, his followers would later be wearing little pendants of electric chairs, hang up large symbols of electric chairs in their churches, etc.
Even more confusingly, there was a time at which they had virtually the same release versions (0.9, I think)
Well, consider that utorrent takes around 10 MB (20 MB under wine on Linux), a small fraction of the 60 MB you report for Azureus (and I have seen Azureus use a lot more... it climbs over time).
Yeah, rtorrent is very good, especially recent releases with ecryption etc. When I started working on Deluge, we considered using rtorrent as a backend (but ended up using libtorrent instead).
For a supposedly 'open' protocol, bittorrent has proven remarkably hard to implement well (basically only utorrent and Azureus do so). See this article.
First, Bittorrent is a peer-to-peer protocol. Only a minor part of it is communication with the server (aka tracker). They might keep the tracker protocol exactly the same, and alter the important p2p part.
Second, this has already been done, and successfully. For example, utorrent came out with a 'PEX' (Peer Exchange) protocol that wasn't in the spec. So it was only used between peers that were both using the utorrent client. This provided a nicer bittorrent experience for utorrent users, especially as utorrent's marketshare rose. Later on, because of utorrent's dominant position, other clients started to implement utorrent PEX (KTorrent, libtorrent-based clients), with varying degrees of success.
A similar issue is Azureus's DHT protocol, which is not in the standard. Although at least Azureus is open source, so you can read the actual code to help in understanding what nonstandard protocols they have invented (but then they also have a very nice wiki).
The point is, it is easy to 'embrace and extend' the bittorrent protocol, even if you don't have control of the servers. Is 'extinguish' next? Probably not, but I for one won't be using the official Bittorrent client.
The Java demo worked perfectly over here (Ubuntu 7.04, Firefox 2.0.0.6, Sun Java).
You can of course very easily install KDE on these systems, a simple apt-get on Ubuntu for example.
The reason, I suspect, is mostly the licensing of Qt and KDE, which is the GPL (and not LGPL, which GTK+ is). This makes it less corporate-friendly, in a way. It probably explains the big shift to GNOME in the major distros, as well as the focus on GNOME in nearly all the new mobile Linux initiatives.
I see your point, and I mostly agree with you actually. But there are 2 separate issues: (1) Do Alcatel-Lucent own MP3 patents, and (2) How much MP3 patents should cost to license.
If Alcatel-Lucent do own MP3 patents, then they can start to assert them whenever they want, and at whatever price they want... that is the (stupid) system.
Do they actually own the patents? I don't know. Certainly it does seem like they might not. But if they do, the price they charge for them is up to them.
Fraunhofer could have given away free-of-charge licenses for MP3 patents. That doesn't mean that if those patents are actually owned by someone else, that other party has to also give them away for free.
For example, Fraunhofer could decide tomorrow to grant a free license to all of their MP3 patents. Does that mean all other MP3-related patents should also be free of charge?
If Red Hat gets a license for WMA, then it will be shipping a product very different than Ubuntu (for better or for worse, we can argue). But that is the point of the delay.
"A single voice isn't going to tell girls that they shouldn't want to be pretty. One well-spoken voice might convince a few that they can be pretty and smart."
And this is why women are so totally screwed by modern (western) society. Once they had to be good mothers, nowadays they have to be both good mothers and have good careers. Once they had to be pretty, nowadays they have to be (according to what you just said) both pretty and smart (of course it isn't just you, it is society as a whole).
Women have traditional expectations of them - be mothers, be pretty - and additional expectations are being added on. This creates a terrible burden, in a very disproportionate way in comparison to men.
(I am a guy, btw.)
Regarding code implementing functionality not in the standard but sort of implied by it ("SpacingLikeWord97", etc.), this may indeed be a problem - you might be sued. But you can still write GPL code to implement it, there is no problem with the GPL's patent clause - the problem is with (possible) Microsoft patents. In fact, such code would be just as 'dangeous' as running Linux, given Microsoft's recent statements (or, to be more precise, just as dangerous as running Mono).
Elton at least admits he is a Luddite. He's entitled to his opinions, I guess. Anyhow, not all artists are like him; for example, Therapy? bandmembers live in different countries, and much of their collaboration is done by utilizing the internet: sending each other MP3s of song ideas. Then they meet physically for a few weeks and record the stuff (see interview here).
Considering that "One Cure Fits All" (2006) was among their better albums ever IMHO (and I have been listening to them since they got started around 1990), apparently this 'interweb' thing isn't necessarily as detrimental as Sir Elton believes.
There are such plugins in the works. When they mature, things may get interesting.
"Project Indiana" will use OpenSolaris as the kernel (obviously), but it isn't yet clear what else. Sun has their own set of tools for package management and so forth, so they don't need GNU (and not Debian or Ubuntu which are specialized systems built on GNU). So, it may turn out that Indiana and Nexenta only have the OpenSolaris kernel in common.
However, is this the right thing to do? I don't think so. Nexenta are very wise to build a system on Ubuntu, which was wise to use Debian, which was wise to use GNU. Sun should accept that GNU and systems based on GNU are the best foundation for an operating system these days. Sun's best move would be to just support or even make official the Nexenta project, if they want to compete with Linux.
However, one reason why they might not is control. Despite rumors of GPL-ing OpenSolaris, it hasn't yet happened. Perhaps Sun doesn't want an operating system based on GPL2/3 code, and just for that reason will use their own userspace tools, to whom they own all copyrights, and can pick their licenses as they please.
(Of course there are many assumptions and guesses here - I don't think this is a serious economic prediction. But it does show the general idea.)
Two conclusions: