Let's face it. If linux is really to be widely adopted, the big players will push their own features. Someone like Red Hat or SuSE wants a unified look and feel, with no interoperability problems. Think about having the printer config tool from one open-source project, and the print deamon from another. If they want them to work together, they have to exert a lot of control over the individual components.
This is something the SuSE does as well. And so will IBM - just wait until a patch they write for mainframes isn't accepted by linus for some reason.
It's clear there will be an arms race between potential abusers and google.
Regardless of the "winner", it's clear that having google spend effort on this will drive up the cost of offering the service. One of the earlier news pieces quoted $2/user IIRC. There's no way this figure is going to stay at that if you also have to include the man hours needed to thwart all the possible abuses.
This doesn't mean google will stop offering the service for free. Maybe they'll have a "premium account" deal like other big providers. Maybe they'll increase the advertising dose. Maybe they'll just absorb it and hope to win it back some other way (hey, by the time yahoo and hotmail are drained of all of their paying customers, google can do whatever they please!).
In a way, this is just like how spammers increase email costs for everyone by overloading the servers and pipes. It is a liability which google is knowingly opening themselves up to.
So let's imagine a world where these hurdles are all removed.
Lack of support: you can buy phone support service for linux, on a per-instance or
periodic rate. But it's only guaranteed to work on a particular version of a
kernel and provided you did not install any third-party modules. Any support
company that gives you better guarantees will so go bankrupt as it'll be
paying its employees to do google searches for you on why module X does not
work in configuration Y. Actually this part has largely already happened.
Lack of Roadmap: The features that go into "enterprise"
versions and their timing are determined by big companies
which are the only ones who can develop them in the first place. This applies
to mainframe kernels, big databases, major software packages, etc. This is
the only way to keep the big customers happy.
Licensing caveats. Sooner or later everybody that has a legal department
gets cold feet and orders their IT department never to use any linux
distribution that doesn't offer lawsuit protection.
What do we end up with? A flavor of linux which the enterprise world is
willing to accept - level-headed, release-engineered, supported.
And what happens to the grassroot linux? The lonely hacker coding for fun
into the night. The reckless sysadmin replacing a windows group server with an
old box runing samba. The enthusiastic team making up yet another
distro. Who will take care of them? Will linus keep accepting their lowly
patches? And even if he does, will IBM and Red Hat pay much attention to his
kernel anymore?
I think that having Linux the kernel well-accepted and established is the
worst thing that can happen to Linux the social movement.
Here's what will happen. Some company will buy the specification from Microsoft and will make a linux implementation for CIFS. RedHat and SuSe will buy licenses to it and bundle it in their enterprise editions.
SAMBA will continue to work as they do now, and lots of people will use them for small and cheap workgroup servers. At most, they'll need to move their servers out of the EU but I doubt even that will be necessary. If MS wanted and could take them to court they would have done that already.
Why hadn't they done it previously? I'm not seeing how the EU decision changes anything in this respect. If MS had the will and means to stop SAMBA before, they'd have tried it already.
So let me get this straight - the Fedora people turn off broadcast. This breaks auto-discovery, which the CUPS people assumed will work.
How is this not a "bazaar" issue? The only way to prevent this kind of thing is to have one responsible party which designs the protocol, writes the implementation, builds the UI and does the testing. Cathedral anyone?
There is some more information in this writeup. The few extra numbers should help clarify the "share of attacked servers" vs. "share of successfully attacked servers" issue.
But really, inadequate training on newly-commissioned linux systems seems like the true cause.
There's a good side to this - spammers pay for addresses, meaning their costs go up.
I guess you can get a fairly good list of infected machines, for free, just by tracking nanas.
Just to show you how commercialized the internet has become:)
Firebird, or whatever they choose to name it, is a great browser. Everyone I introduced it to liked it, including MSIE users. And that's not a big surprise, seeing how MSIE is all but EOL-ed Explorer. It's obvious how Microsoft sees it as a waste of money to keep maintaining their flagship browser - it's not sold for money and it doesn't create new markets for them.
So there is a big opportunity here for open source. It seems like an appropriate marketing campaign should really widen the userbase. How can this be done? A paypal account to pay for banner ads (popups too:) )? Free CDs in public libraries? Post it as explorer.crack.exe to file sharing networks?
"really lay down"? The rumors I heard (some right here on slashdot) talked about a 1B Euro fine *at most*. That's hardly Bill's lunch money.
Sure, it's more than the current going rate for politicians and judges in the US, but still small potatos to Microsoft.
Slashdot Mirror
A SqueezeBox can do that, no need for a WinXP box. The latest server software even has a module to let you browse ShoutCast with your remote.
This is something the SuSE does as well. And so will IBM - just wait until a patch they write for mainframes isn't accepted by linus for some reason.
Regardless of the "winner", it's clear that having google spend effort on this will drive up the cost of offering the service. One of the earlier news pieces quoted $2/user IIRC. There's no way this figure is going to stay at that if you also have to include the man hours needed to thwart all the possible abuses.
This doesn't mean google will stop offering the service for free. Maybe they'll have a "premium account" deal like other big providers. Maybe they'll increase the advertising dose. Maybe they'll just absorb it and hope to win it back some other way (hey, by the time yahoo and hotmail are drained of all of their paying customers, google can do whatever they please!).
In a way, this is just like how spammers increase email costs for everyone by overloading the servers and pipes. It is a liability which google is knowingly opening themselves up to.
What do we end up with? A flavor of linux which the enterprise world is willing to accept - level-headed, release-engineered, supported.
And what happens to the grassroot linux? The lonely hacker coding for fun into the night. The reckless sysadmin replacing a windows group server with an old box runing samba. The enthusiastic team making up yet another distro. Who will take care of them? Will linus keep accepting their lowly patches? And even if he does, will IBM and Red Hat pay much attention to his kernel anymore?
I think that having Linux the kernel well-accepted and established is the worst thing that can happen to Linux the social movement.
SAMBA will continue to work as they do now, and lots of people will use them for small and cheap workgroup servers. At most, they'll need to move their servers out of the EU but I doubt even that will be necessary. If MS wanted and could take them to court they would have done that already.
Why hadn't they done it previously? I'm not seeing how the EU decision changes anything in this respect. If MS had the will and means to stop SAMBA before, they'd have tried it already.
So let me get this straight - the Fedora people turn off broadcast. This breaks auto-discovery, which the CUPS people assumed will work. How is this not a "bazaar" issue? The only way to prevent this kind of thing is to have one responsible party which designs the protocol, writes the implementation, builds the UI and does the testing. Cathedral anyone?
But really, inadequate training on newly-commissioned linux systems seems like the true cause.
There's a good side to this - spammers pay for addresses, meaning their costs go up. I guess you can get a fairly good list of infected machines, for free, just by tracking nanas. Just to show you how commercialized the internet has become :)
Looks like you re-discovered Norvig's law.
Firebird, or whatever they choose to name it, is a great browser. Everyone I introduced it to liked it, including MSIE users. And that's not a big surprise, seeing how MSIE is all but EOL-ed Explorer. It's obvious how Microsoft sees it as a waste of money to keep maintaining their flagship browser - it's not sold for money and it doesn't create new markets for them. So there is a big opportunity here for open source. It seems like an appropriate marketing campaign should really widen the userbase. How can this be done? A paypal account to pay for banner ads (popups too :) )? Free CDs in public libraries? Post it as explorer.crack.exe to file sharing networks?
"really lay down"? The rumors I heard (some right here on slashdot) talked about a 1B Euro fine *at most*. That's hardly Bill's lunch money. Sure, it's more than the current going rate for politicians and judges in the US, but still small potatos to Microsoft.