He doesn't claim to work for, he work for. However, yes, that's a proposal to be discussed on Fedora list, and as said in the on going thread and the article, RH tried to change things ( as well as Canonical ), yet they did not managed to have a satisfying solution for everybody. I think the whole "let's make the signing be done by a third party" to be better, but as Matthew say, the whole setup is expensive ( think cost of CA, like the type of cost that prevented cacert to be properly added to firefox due to audit cost )
Not sure why they do not do it. Fedora provide it since a few years, so does Opensuse, so while I know that's a feature geared for power users, so not really the market aimed by Canonical, even a distribution for newbie like Ubuntu could benefit from it, and that do not seems to be very complex to do.
Canonical use s3 ( or at least, they did when i looked, maybe this changed, or maybe I got the detail wrong ), and so they pay for each download. Nothing serious, and I think Mark can afford the money for now, but since they try to be profitable since years, maybe trying to push people on BT would be in their interest, and so in the interest of their community of users.
There is more than webservers. Let's take a bank, you need storage, database or assimilated server ( ldap, like 389ds, red hat project ), you may want to run a messaging for all your business software ( amqp, again, there is something ), you may use a application server ( like jboss, part of red hat ).
You may want to have a certified os ( fips, selinux ), you may want certifications for the hardware. That's problem you can solve with hiring some expert, and that can be solved by taking something that already do it. Given that everybody I know basically say that finding skilled people is hard in IT, that's easy to see where it goes.
For exemple, Debian is a nice os, stable, with lots of feature, and a diverse ecosystem of company to support it. I know lots of people using it, and having been a debian sysadmin, it is not hard to manage, but that's a tradeoff,, you cannot go to the vendor in case of big issues ( not that it happen often, but for some case even 1 problem a year is too much if you lose lots of money when it happen ) . Some company are too big to be addressed by smaller ones, some company prefer to have another company to talk ( for appliance for exemple ), etc.
There is not 1 solution that can solve every case of "computer issues", contrary to what most people seems to believe, there is lots of different solutions, adapted to the different case, and both RHEL, Debian, Centos, etc do solve the same big problem on the server side, in a way that give flexibility to people, and enough room for a healthy competition ( if we except of course OEL ).
This is not Canonical ( with the alleged 10 millions revenue due to goodies ), the details are on the SEC filling ( http://investors.redhat.com/sec.cfm )
Given that for a mac book, you need to pay for a extended warranty, you do not have any docking station ( so can be painful for a company where people want to plug mouse, keyboard and a bigger screen ), my company have seen that a macbook was not cheaper than what we can get from our resseler and lenovo.
Now, maybe you managed to get a better deal from Apple, or differents needs ( ie, people without any desk at all or they do not care not having mouses, keyboards, etc ), or dell do really screw you up.
That's true than for different situation, there is different outcomes ( how shocking, I know ). We could also speak of the Apple case, of the fact that we have cyanogenmod also because some handset makers were forced to release source code.
The real question is "do you trust enough people to contribute back", and that's some negociation.
For example, one group could indeed see that a less stringent requirement would help some contributors ( usually, a company, but not only ) to contribute and so would be beneficial, and so using a BSD like license would help. And using the same exact source code, maybe the contributors would not care about sending upstream the changes, due to management policy, to not share with competitors, etc.
So the choice of license is not tied to the source code, but more to what you want to achieve, and the ecosystem around your software. Some people leverage their copyright and the GPL to sell exception ( Canonical does it for Ubuntu-TV, if you read the small letters ), or to offers "entreprise version". Some do use BSD for the same outcome.
Some companies would not use GPL in their products, because they would be forced to contribute back and do not want ( be it for logistic, commercial or legal reasons ). On the other hand, some companies would not contribute back if that mean giving complete copyright to the owner if that's a competiting commercial entity, but have no problem to contribute to GPL or BSD besides that. Some companies even use GPL to be sure that their code is not taken by competitors.
It all boil down on what you think people will do in the future, IE, do you have a goal, a reason, a business plan around your software ? Can you trust people to not screw you, or if that happens, would it bother you ( depending how it happens ) ?
For example, OpenBSD people do not care about not having people contributing back, they prefer to have the software used everywhere so the internet is more secure. FSF people do prefer to make sure the source code can be seen and the knowledge is alive.
2 different group, 2 different goal, 2 different choice. So choosing a license is really up to you, to what you think will happen, based on what happened to you and others, and what you want to see happen. ( yes, that's slightly much more complex than "foo is good, bar is evil", sorry about the complexity of the world, I am sure we will try to fix that for the next version )
If licenses term are violated, I think you lose the right to use the software under any licenses. That's the basis of most license : "we grant you XXX under XXX conditions", meaning that if you do not fullfill the condition, you lose the right granted to you.
People should maybe read what they agree to do, beit for a Microsoft CLUF, or a Free Software license.
Well, could the problem really be the unease of people on the 'obesity' side ? IE, the fact that more than a third of American would feel their own views to be incompatible with what is perceived as the common view would make them feel cognitive dissonnance ?
There is no achievement, you can already run Debian side by side of Android on most phone. And if you look at the pictures, you just see... Unity on a regular screen. Where is the innovation there ?
Even the idea of having a phone as pc is not their own, seek google for people already doing it with bt keyboard and galaxy s II.
If you start to manage lots of data ( and I think some people enter into this category for their email ), it make sense to use a DB. It may not bring much in term of feature, but if this prevent people from redoing optimisation and so on that are already done by others group, that will permit to free some time to worok on others features. So in short, you will not see much directly; Performances are forgot after a week and start to become the norm, and unless you look at internal and code, you may not see the improvement in term of maintainance.
Does it prevent them from getting data from employers and say "match/no match" ? Because this would not be giving the data, just using what was given to them to provides a service. That's not against their privacy policy, and nothing private leave google servers.
( as said elsehwere, I am working for Red Hat, and I can say that's a great place to work, and we are hiring a lot for cloud related stuff, or to fill position of people who have decided to be moved internally to cloud related thing , see the url ).
I would also take a look at the various others companies listed around Openstack, etc, as I know several of them are hiring ( like Puppetlabs, Opscode, etc )
Working as consultant or support for a linux consulting shop, be it Red Hat, Canonical, Suse or the various companies around free software. Even my previous employer was like this, with us being sent at random client for stuff like Nagios, AFP, Cacti, Apache, etc. ITwas a small consulting company, and we were doing good ( until $BOSS decided to change their mind on the business plan... )
And fortunate that people do not have to select between several mobile phone plan who are all a variation of "we are gonna screw you and be expensive", or fortunate there is one only brand of cars, food and stuff like that.
On the other hand, parent is right, most people do not care about choice outside of their area of expertise ( well, until it is too late usually, ie when they are screwed without knowing why ), and since most people grow thinking "computer is hard", they fear computer more due to choice ( "ie, I need to make a decision, but people told me this is hard, omg, OMG OMG !!!!" ). But that miss the point completely about free software. They are not here to offer choice, nor to save people from commercial company. They are here because people think colleboration work better, because commercialisation of knowledge are bad for humanity in general. Ie, they are here to create a pool of common knowledge under the form of source code.
The rest is a side effect, or a consequence of the aformentioned knowledge. No one say "I am gonna do free software because I will be able to create a great product for people who do not care about it". They more say "Using this, I can have a good product for less money" ( like android ). To this date, making customer pay is the best option to get ressources ( having customer being end users, or being people who pay to show adverisement ). And that's hard to make some of them pay when you do not force them.
Developping for someone requires ressources. Someone has to do the work, and if that's not end users and customer, they have to give something to let you do the job.
For commercial software, that's usually money. Simple, understood by everybody.
For free software, that's contribution, ie people do the work, share libraries, report bugs, etc. That's a different kind of economy. Everybody could have the money needed ( aka time, and knowledge ), but most people do not want. Too bad for them.
Except that using Apple stuff is seen as cool, so everybody think that they must be dumb to not get it. On the other hand, without the heavy marketing around Apple product, Linux/BSD/etc based system are all seen as unsuitable and unusable, even when they offer different advantages.
As you say, people do have hardware problem on windows too, but no one talk about it, because "that's normal". People do have issues because the system is too complex, or because they were not trained or any reasons, but no one complains. On free software, every small issue is seen as the end of the world, not because that's a issue, but because that's a non familiar issue, and because there is no social pressure from everybody to say "this is better" ( ie, not like Apple's product ). I have seen this as work, the setup of VPN was a nightmare on iphone ( 3 differents passwords needed + problem due to wifi etc ). And yet, people were happy. If we had put just half of the problem on another phone or system, people would have loudly complained.
That's not free, that's paid by someone else. Some people have given their time for that, some have been paid for that.
People can't complain that coders do not listen to them ( like for gnome-shell, unity, even if I think they are both good and sane system, even if there is some bug to fix ) when the user do not make any differences. IE, by neither contributing or paying, the user is just a cost. A cost because bandwidth is not free, because you have to pay engineers for handling problem, etc.
And Mint just take Canonical packages that take Debian one. This is rather not sustainable. If Canonical disappear, someone will have to do their job, and that's not mint with enough donation for 1 single engineers that will be able to cope with the work of 150/200 peoples. ( and those people mainly do integration or coding for their own software, the kernel, x11, etc work is done by Red hat, Suse, Intel, and lots of others ).
Note that now xen did the work to be pushed upstream, this is supported fine by Fedora. Also, if Red hat didn't want to support xen, it would not appear as supported by libvirt and the rest of the stack maintained by Red Hat.
There was also technical issue of supporting it, but that's much easier to say "that's a conspiracy" than look at the facts.
Slashdot Mirror
He doesn't claim to work for, he work for. However, yes, that's a proposal to be discussed on Fedora list, and as said in the on going thread and the article, RH tried to change things ( as well as Canonical ), yet they did not managed to have a satisfying solution for everybody. I think the whole "let's make the signing be done by a third party" to be better, but as Matthew say, the whole setup is expensive ( think cost of CA, like the type of cost that prevented cacert to be properly added to firefox due to audit cost )
Already 3 people tried, and likely more will try to package everything. But Unity is evolving too fast, and no one want to take care of compiz.
I stand corrected, seems Ubuntu did finally catch up with Fedora :
https://help.ubuntu.com/community/Metalink
Seems to have been already proposed, but not implemented :
https://wiki.ubuntu.com/metalink
Not sure why they do not do it. Fedora provide it since a few years, so does Opensuse, so while I know that's a feature geared for power users, so not really the market aimed by Canonical, even a distribution for newbie like Ubuntu could benefit from it, and that do not seems to be very complex to do.
Canonical use s3 ( or at least, they did when i looked, maybe this changed, or maybe I got the detail wrong ), and so they pay for each download. Nothing serious, and I think Mark can afford the money for now, but since they try to be profitable since years, maybe trying to push people on BT would be in their interest, and so in the interest of their community of users.
There is more than webservers. Let's take a bank, you need storage, database or assimilated server ( ldap, like 389ds, red hat project ), you may want to run a messaging for all your business software ( amqp, again, there is something ), you may use a application server ( like jboss, part of red hat ).
You may want to have a certified os ( fips, selinux ), you may want certifications for the hardware. That's problem you can solve with hiring some expert, and that can be solved by taking something that already do it. Given that everybody I know basically say that finding skilled people is hard in IT, that's easy to see where it goes.
For exemple, Debian is a nice os, stable, with lots of feature, and a diverse ecosystem of company to support it. I know lots of people using it, and having been a debian sysadmin, it is not hard to manage, but that's a tradeoff,, you cannot go to the vendor in case of big issues ( not that it happen often, but for some case even 1 problem a year is too much if you lose lots of money when it happen ) . Some company are too big to be addressed by smaller ones, some company prefer to have another company to talk ( for appliance for exemple ), etc.
There is not 1 solution that can solve every case of "computer issues", contrary to what most people seems to believe, there is lots of different solutions, adapted to the different case, and both RHEL, Debian, Centos, etc do solve the same big problem on the server side, in a way that give flexibility to people, and enough room for a healthy competition ( if we except of course OEL ).
This is not Canonical ( with the alleged 10 millions revenue due to goodies ), the details are on the SEC filling ( http://investors.redhat.com/sec.cfm )
Given that for a mac book, you need to pay for a extended warranty, you do not have any docking station ( so can be painful for a company where people want to plug mouse, keyboard and a bigger screen ), my company have seen that a macbook was not cheaper than what we can get from our resseler and lenovo.
Now, maybe you managed to get a better deal from Apple, or differents needs ( ie, people without any desk at all or they do not care not having mouses, keyboards, etc ), or dell do really screw you up.
Sorry but asshole is well defined. It cannot be placed anywhere on the body. And if nothing go out of it, this is not a asshole.
For example, ear hole are not ass hole.
That's true than for different situation, there is different outcomes ( how shocking, I know ). We could also speak of the Apple case, of the fact that we have cyanogenmod also because some handset makers were forced to release source code.
The real question is "do you trust enough people to contribute back", and that's some negociation.
For example, one group could indeed see that a less stringent requirement would help some contributors ( usually, a company, but not only ) to contribute and so would be beneficial, and so using a BSD like license would help. And using the same exact source code, maybe the contributors would not care about sending upstream the changes, due to management policy, to not share with competitors, etc.
So the choice of license is not tied to the source code, but more to what you want to achieve, and the ecosystem around your software.
Some people leverage their copyright and the GPL to sell exception ( Canonical does it for Ubuntu-TV, if you read the small letters ), or to offers "entreprise version". Some do use BSD for the same outcome.
Some companies would not use GPL in their products, because they would be forced to contribute back and do not want ( be it for logistic, commercial or legal reasons ). On the other hand, some companies would not contribute back if that mean giving complete copyright to the owner if that's a competiting commercial entity, but have no problem to contribute to GPL or BSD besides that. Some companies even use GPL to be sure that their code is not taken by competitors.
It all boil down on what you think people will do in the future, IE, do you have a goal, a reason, a business plan around your software ? Can you trust people to not screw you, or if that happens, would it bother you ( depending how it happens ) ?
For example, OpenBSD people do not care about not having people contributing back, they prefer to have the software used everywhere so the internet is more secure. FSF people do prefer to make sure the source code can be seen and the knowledge is alive.
2 different group, 2 different goal, 2 different choice. So choosing a license is really up to you, to what you think will happen, based on what happened to you and others, and what you want to see happen. ( yes, that's slightly much more complex than "foo is good, bar is evil", sorry about the complexity of the world, I am sure we will try to fix that for the next version )
If licenses term are violated, I think you lose the right to use the software under any licenses. That's the basis of most license :
"we grant you XXX under XXX conditions", meaning that if you do not fullfill the condition, you lose the right granted to you.
People should maybe read what they agree to do, beit for a Microsoft CLUF, or a Free Software license.
Well, could the problem really be the unease of people on the 'obesity' side ? IE, the fact that more than a third of American would feel their own views to be incompatible with what is perceived as the common view would make them feel cognitive dissonnance ?
There is no achievement, you can already run Debian side by side of Android on most phone. And if you look at the pictures, you just see ... Unity on a regular screen. Where is the innovation there ?
Even the idea of having a phone as pc is not their own, seek google for people already doing it with bt keyboard and galaxy s II.
If you start to manage lots of data ( and I think some people enter into this category for their email ), it make sense to use a DB. It may not bring much in term of feature, but if this prevent people from redoing optimisation and so on that are already done by others group, that will permit to free some time to worok on others features. So in short, you will not see much directly; Performances are forgot after a week and start to become the norm, and unless you look at internal and code, you may not see the improvement in term of maintainance.
Does it prevent them from getting data from employers and say "match/no match" ? Because this would not be giving the data, just using what was given to them to provides a service. That's not against their privacy policy, and nothing private leave google servers.
For Canonical (primary sponsor of Ubuntu) : http://www.canonical.com/about-canonical/careers
For Red Hat : http://www.redhat.com/about/work/
( as said elsehwere, I am working for Red Hat, and I can say that's a great place to work, and we are hiring a lot for cloud related stuff, or to fill position of people who have decided to be moved internally to cloud related thing , see the url ).
I would also take a look at the various others companies listed around Openstack, etc, as I know several of them are hiring ( like Puppetlabs, Opscode, etc )
Working as consultant or support for a linux consulting shop, be it Red Hat, Canonical, Suse or the various companies around free software. Even my previous employer was like this, with us being sent at random client for stuff like Nagios, AFP, Cacti, Apache, etc. ITwas a small consulting company, and we were doing good ( until $BOSS decided to change their mind on the business plan... )
( also, Red Hat is hiring in NC : http://www.youtube.com/watch?v=tWL8miDGghY ( disclosure, I work for them ))
And fortunate that people do not have to select between several mobile phone plan who are all a variation of "we are gonna screw you and be expensive", or fortunate there is one only brand of cars, food and stuff like that.
On the other hand, parent is right, most people do not care about choice outside of their area of expertise ( well, until it is too late usually, ie when they are screwed without knowing why ), and since most people grow thinking "computer is hard", they fear computer more due to choice ( "ie, I need to make a decision, but people told me this is hard, omg, OMG OMG !!!!" ).
But that miss the point completely about free software. They are not here to offer choice, nor to save people from commercial company. They are here because people think colleboration work better, because commercialisation of knowledge are bad for humanity in general. Ie, they are here to create a pool of common knowledge under the form of source code.
The rest is a side effect, or a consequence of the aformentioned knowledge. No one say "I am gonna do free software because I will be able to create a great product for people who do not care about it". They more say "Using this, I can have a good product for less money" ( like android ). To this date, making customer pay is the best option to get ressources ( having customer being end users, or being people who pay to show adverisement ). And that's hard to make some of them pay when you do not force them.
Developping for someone requires ressources. Someone has to do the work, and if that's not end users and customer, they have to give something to let you do the job.
For commercial software, that's usually money. Simple, understood by everybody.
For free software, that's contribution, ie people do the work, share libraries, report bugs, etc. That's a different kind of economy. Everybody could have the money needed ( aka time, and knowledge ), but most people do not want. Too bad for them.
Except that using Apple stuff is seen as cool, so everybody think that they must be dumb to not get it. On the other hand, without the heavy marketing around Apple product, Linux/BSD/etc based system are all seen as unsuitable and unusable, even when they offer different advantages.
As you say, people do have hardware problem on windows too, but no one talk about it, because "that's normal". People do have issues because the system is too complex, or because they were not trained or any reasons, but no one complains. On free software, every small issue is seen as the end of the world, not because that's a issue, but because that's a non familiar issue, and because there is no social pressure from everybody to say "this is better" ( ie, not like Apple's product ).
I have seen this as work, the setup of VPN was a nightmare on iphone ( 3 differents passwords needed + problem due to wifi etc ). And yet, people were happy. If we had put just half of the problem on another phone or system, people would have loudly complained.
I am unfortunately doubtful about most applications being suitable for touchscreen usage.
That's not free, that's paid by someone else. Some people have given their time for that, some have been paid for that.
People can't complain that coders do not listen to them ( like for gnome-shell, unity, even if I think they are both good and sane system, even if there is some bug to fix ) when the user do not make any differences. IE, by neither contributing or paying, the user is just a cost. A cost because bandwidth is not free, because you have to pay engineers for handling problem, etc.
And Mint just take Canonical packages that take Debian one. This is rather not sustainable. If Canonical disappear, someone will have to do their job, and that's not mint with enough donation for 1 single engineers that will be able to cope with the work of 150/200 peoples. ( and those people mainly do integration or coding for their own software, the kernel, x11, etc work is done by Red hat, Suse, Intel, and lots of others ).
Debian kfreebsd ?
Note that now xen did the work to be pushed upstream, this is supported fine by Fedora. Also, if Red hat didn't want to support xen, it would not appear as supported by libvirt and the rest of the stack maintained by Red Hat.
There was also technical issue of supporting it, but that's much easier to say "that's a conspiracy" than look at the facts.
And I doubt they were fixed by anybody who complained.
There is some openstack rpm in Fedora ( see http://fedoraproject.org/wiki/OpenStack ), so they started to port.