It is pretty easy to date the why. In 2006 there was a thing called the Userbox wars. There isn't a good page on wikipedia about this. Prior to 2006 Wikipedia user pages were sort of like myspace pages for wikipedia editors. They had lots of personal information and people chatted. Jimmy Wales wanted userspace to be about the encyclopedia. At the same time he didn't want mass deletions. There were mass deletions and the this wasn't easily reversed. The tone changed. This was one of the big steps towards the deletionists winning control of Wikipedia entirely. But if you want to know when the gender's changed this was a crucial moment.
Of course the deletionists winning even more battles probably didn't help
And that's the real problem - how do you properly draw the line between apps that are legitimate but happen to be similar because one inspired the other, and apps that are pure scamware and trying to undermine the original developer?
I'm an Apple user. I can accuse Microsoft of a lot but yes they are substantially more open than Apple:
a) Their hardware base system is extremely open. Apple provides very limited hardware choice b) Their driver selection is 2nd to none. Incredible. Apple is far worse than Linux and might even be worse than other BSDs. c) Azure (their cloud offering) is probably the most open cloud out there. Certainly among the big players. Apple's cloud is completely tied to their platform and they don't allow other clouds. d) Their enterprise apps tend to play well with others and allow you to mix and match.
Microsoft sells computers without the crapware: www.microsoftstore.com/signature On the one hand they hate what the crapware does to the entire experience. OTOH $75-90 in subsidies per machine per OEM translates into about $150 to the end customer in savings. At an ASP of $550 an increase to $700 would be a 27% increase in price which would definitely harm sales. The value trap is a disaster for Microsoft. One of the points of the new interface is to drive up the price of PCs by making better interface hardware worthwhile and thus cut that number down to a level where they might be able to get rid of crapware.
I completely fail to comprehend why most Slashdotters seem to push everyone towards DRM'ed iPads and Chromebooks that put Palladium to shame instead of more open Windows PCs.
They don't. This generation of/.ers pretty much hates everything except for grey market hardware distributors. They hate Linux for failing on the desktop. They hate Apple for being vertically integrated and expensive. , They hate Microsoft for not being innovative while ignoring Azure and enterprise apps where Microsoft has been innovative. Then when Microsoft brings out their biggest innovation for home / small business they trash that because it runs badly on Windows 7 machine.
They've gotten bitter. Of course it was easier to be positive when/. started and we were in a tech explosion with salaries rapidly rising and jobs plentiful.
Obviously if Intel were to substantial cut prices that changes things. But at least the Power8 prices I saw were competitive. Their entire pitch is that Power8 is moderately better especially for virtualization. They have to know that moderately better doesn't cut it if they are way out of range on price.
Red Hat sells operating systems not development tools. The big initiative for RedHat is designing a cloud based operating system which is open and at the same time supports containers -- OpenStack and Docker. They are a major leader in the DevOps approach. But even in development JBoss is a huge suite of development tools.
In terms of the complaints regarding OSes. RedHat is fine with Developers using Ubuntu for their workstations. They are getting to need something to deploy in production on and that's not going to be Ubuntu most of the time. As far as MySQL, MongoDB, or PostgreSQL they've never been a database company but they support all 3 databases. And in terms of Mongo / Cassandra / Hadoop there is no question they are far far ahead of Ubuntu in terms of deployment technology.
The summary is ridiculous. The article linked is more balanced and mainly advice for RedHat doing partnerships / distribution deals.
Welcome to the "Web 2.0" world, which is where the volume is these days, and consequently most of the money.
IBM makes 85% of their money from fortune 100. From there it falls off fast. The money and especially the margin is at the top.
If one cannot order it cheaply and easily on the web ala Amazon shopping experience, who is going to bother to go through a reseller? That was the model 40 years ago! Kids today do not bother
What kid gets to pick the hardware infrastructure for his company of any size?
Why would I pay the vendor or the reseller higher prices when I can automate hundreds of thousands of servers on x86, in a lights out management datacenters across the globe, to the point of throwaway systems?
The prices aren't higher and the system outperforms thus lowering total cost. This is the whole "why quality saves money" issue that comes up in every industry.
IBM doesn't do that. They should be more transparent but they aren't. They want you ordering through a partner or for larger customers through the sales channel. There is some pricing on the website but the real prices are 20-30% lower.
DELL has become expensive as well. For the price of one DELL server one can easily put together two or three blackbox servers, from motherboard to chassis, made 100% by intel.
Not really relevant. The question was Power vs. x86 not generic vs. name brand.
The only company I know of who has announced they would be offering it as cloud is Ubuntu cloud. IBM's hosting solution has it but so far nothing in the cloud space.
I've seen pricing on Power8 systems they are in line with someone like Dell for rack mounted servers. No they aren't priced out of the market. And BTW the Linux on Power is where they are mentioning the advantages of their virtualization.
Phones take good advantage of systemd. For one thing they get plugged into hardware all the time. They have lots of processes, their situations as far as network is constantly changing. That's the hardware use case where systemd is a no brainer.
Now, that being said, systemd seems to bring on top of the previous the ability to dynamically reconfigure the system upon changing hardware. That is indeed a feature that some people may have, yet not necessarily all. Forcing the baggage of that upon the whole linux population, is the major point of contention,
Systemd is a complex system. Rather than greatest common denominator it goes for least common multiple. Absolutely it is not lightweight. It is moving towards being the daemon equivalent of/bin.
Service startup should be a like a tree, not like a chain.
It is worse than that. It is a long complex cycle for some daemons and one that is changing. So something more like an event handler that just starts from a clean initial state. This is kind of the issue. Again this would have been easier to build on top of OpenRC and had that happened the dam wouldn't have broken.
And servers - do they still care about the servers?
Depends on which kinds. Macminis get used as servers. They are suddenly doing well with the MacPro. They have an excellent small business server product. Also remember they just recommitted to enterprise. Apple is very hard to read on this, but I think in general they would want to be able to support a server quickly not have a long lag.
As for ignoring Gnome I agree. But I also suspect heavy server processes will use it. Docker for example already has built in support for process managers. I suspect OpenStack will use it. While FreeBSD can walk away from Gnome can they long term walk away from containers?
I understand your point about portability. I'm not sure I agree with all the trash / junk... type comments. Gnome applications are going to and should have dependencies on glib. I think interprocess communication is valuable so I'm not going to agree that kdbus is trash.
As far as RedHat driving changes to Linux. I kinda like that, I hope you are right. With the death of mini computing platforms we need systems to replace VMS and OS/400. We don't really have good enterprise Unixes. I kinda like the idea of Linux/systemd being a full featured mini computer OS while Linux/SysV (or better Linux/OpenRC is more like a traditional lightweight Unix). That would be terrific. Someone has to lead and RedHat has over two decades proven itself to be good leader.
As far as the port of systemd, we aren't disagreeing. See what I wrote above.
I get that. Which is what I said that a port of systemd has to be done by those alternative Unixes because they are going to need to make choices about their own daemons, what to emulate and what to implement... The BSDs are used to having to follow in Linux's wake. They'll go through this process of deciding. I'm not sure whether FreeBSD or Darwin will be first but I doubt Apple is going to want huge chunks of Linux code to simply never be able to be ported over to Macports easily. I don't think FreeBSD is going to want huge numbers of applications one simply can't run.
So I don't think long term this has much impact on applications. For the next 5 years those applications that have complex startup will probably have slightly reduced functionality via. init scripts.
Systemd is open source, it can be ported. Right now the version that exists depends on features of the Linux kernel. The other operating systems are going to have to make choices about whether they want to support systemd or not. Basically this isn't something systemd developers are going to do it is something IBM/AIX, FreeBSD, OpenBSD, QNX... are going to do.
Are you really arguing that the huge number of people all over the planet who have complained about structural problems with the Unix init system for the last 50 years are in error because you don't mind 15 minute boots but do mind patch problems?
OpenRC is a really good replacement for SysV init. A serious enhancement that keeps with the spirit, an upgrade rather than a rip out and replace. OpenRC doen't keep daemons alive but that would be easy to add. Ultimately then there is only one hard problem what to do about hardware that changes. OpenRC doesn't architecturally have any good way for handing that while systemd does.
Certainly if the argument were OpenRC vs. systemd instead of SysV init vs. systemd I suspect the advantages of systemd wouldn't have outweighed the huge shift. I hope distributions like Slackware which don't have systemd move to OpenRC so that it gets tested in environments other than Gentoo.
Yes people wanted the Windows registry. The people who were designing Windows applications and had to write complex pif and ini file interpreters to handle bugs.
No wrong. Wayland has remote desktop functionality. You'll be passing far less data around the network using Wayland's remote functionality than you do with X11.
Slashdot Mirror
It is pretty easy to date the why. In 2006 there was a thing called the Userbox wars. There isn't a good page on wikipedia about this. Prior to 2006 Wikipedia user pages were sort of like myspace pages for wikipedia editors. They had lots of personal information and people chatted. Jimmy Wales wanted userspace to be about the encyclopedia. At the same time he didn't want mass deletions. There were mass deletions and the this wasn't easily reversed. The tone changed. This was one of the big steps towards the deletionists winning control of Wikipedia entirely. But if you want to know when the gender's changed this was a crucial moment.
Of course the deletionists winning even more battles probably didn't help
Links:
A few statements on Userboxes but not enough to understand what happened: https://en.wikipedia.org/wiki/...
What "deletionists" are and what Wikipedia was like before them: https://en.wikipedia.org/wiki/...
That's what trademarks are for.
I'm an Apple user. I can accuse Microsoft of a lot but yes they are substantially more open than Apple:
a) Their hardware base system is extremely open. Apple provides very limited hardware choice
b) Their driver selection is 2nd to none. Incredible. Apple is far worse than Linux and might even be worse than other BSDs.
c) Azure (their cloud offering) is probably the most open cloud out there. Certainly among the big players. Apple's cloud is completely tied to their platform and they don't allow other clouds.
d) Their enterprise apps tend to play well with others and allow you to mix and match.
etc...
Microsoft sells computers without the crapware: www.microsoftstore.com/signature
On the one hand they hate what the crapware does to the entire experience. OTOH $75-90 in subsidies per machine per OEM translates into about $150 to the end customer in savings. At an ASP of $550 an increase to $700 would be a 27% increase in price which would definitely harm sales. The value trap is a disaster for Microsoft. One of the points of the new interface is to drive up the price of PCs by making better interface hardware worthwhile and thus cut that number down to a level where they might be able to get rid of crapware.
They don't. This generation of /.ers pretty much hates everything except for grey market hardware distributors. They hate Linux for failing on the desktop. They hate Apple for being vertically integrated and expensive. , They hate Microsoft for not being innovative while ignoring Azure and enterprise apps where Microsoft has been innovative. Then when Microsoft brings out their biggest innovation for home / small business they trash that because it runs badly on Windows 7 machine.
They've gotten bitter. Of course it was easier to be positive when /. started and we were in a tech explosion with salaries rapidly rising and jobs plentiful.
Obviously if Intel were to substantial cut prices that changes things. But at least the Power8 prices I saw were competitive. Their entire pitch is that Power8 is moderately better especially for virtualization. They have to know that moderately better doesn't cut it if they are way out of range on price.
Red Hat sells operating systems not development tools. The big initiative for RedHat is designing a cloud based operating system which is open and at the same time supports containers -- OpenStack and Docker. They are a major leader in the DevOps approach. But even in development JBoss is a huge suite of development tools.
In terms of the complaints regarding OSes. RedHat is fine with Developers using Ubuntu for their workstations. They are getting to need something to deploy in production on and that's not going to be Ubuntu most of the time. As far as MySQL, MongoDB, or PostgreSQL they've never been a database company but they support all 3 databases. And in terms of Mongo / Cassandra / Hadoop there is no question they are far far ahead of Ubuntu in terms of deployment technology.
The summary is ridiculous. The article linked is more balanced and mainly advice for RedHat doing partnerships / distribution deals.
IBM makes 85% of their money from fortune 100. From there it falls off fast. The money and especially the margin is at the top.
What kid gets to pick the hardware infrastructure for his company of any size?
The prices aren't higher and the system outperforms thus lowering total cost. This is the whole "why quality saves money" issue that comes up in every industry.
IBM doesn't do that. They should be more transparent but they aren't. They want you ordering through a partner or for larger customers through the sales channel. There is some pricing on the website but the real prices are 20-30% lower.
Not really relevant. The question was Power vs. x86 not generic vs. name brand.
The only company I know of who has announced they would be offering it as cloud is Ubuntu cloud. IBM's hosting solution has it but so far nothing in the cloud space.
I've seen pricing on Power8 systems they are in line with someone like Dell for rack mounted servers. No they aren't priced out of the market. And BTW the Linux on Power is where they are mentioning the advantages of their virtualization.
Phones take good advantage of systemd. For one thing they get plugged into hardware all the time. They have lots of processes, their situations as far as network is constantly changing. That's the hardware use case where systemd is a no brainer.
Well if you are right about Gentoo switching then OpenRC is dead. Without a distribution it is unlikely to catch on.
Systemd is a complex system. Rather than greatest common denominator it goes for least common multiple. Absolutely it is not lightweight. It is moving towards being the daemon equivalent of /bin.
It is worse than that. It is a long complex cycle for some daemons and one that is changing. So something more like an event handler that just starts from a clean initial state. This is kind of the issue. Again this would have been easier to build on top of OpenRC and had that happened the dam wouldn't have broken.
Depends on which kinds. Macminis get used as servers. They are suddenly doing well with the MacPro. They have an excellent small business server product. Also remember they just recommitted to enterprise. Apple is very hard to read on this, but I think in general they would want to be able to support a server quickly not have a long lag.
As for ignoring Gnome I agree. But I also suspect heavy server processes will use it. Docker for example already has built in support for process managers. I suspect OpenStack will use it. While FreeBSD can walk away from Gnome can they long term walk away from containers?
I understand your point about portability. I'm not sure I agree with all the trash / junk... type comments. Gnome applications are going to and should have dependencies on glib. I think interprocess communication is valuable so I'm not going to agree that kdbus is trash.
As far as RedHat driving changes to Linux. I kinda like that, I hope you are right. With the death of mini computing platforms we need systems to replace VMS and OS/400. We don't really have good enterprise Unixes. I kinda like the idea of Linux/systemd being a full featured mini computer OS while Linux/SysV (or better Linux/OpenRC is more like a traditional lightweight Unix). That would be terrific. Someone has to lead and RedHat has over two decades proven itself to be good leader.
As far as the port of systemd, we aren't disagreeing. See what I wrote above.
I get that. Which is what I said that a port of systemd has to be done by those alternative Unixes because they are going to need to make choices about their own daemons, what to emulate and what to implement... The BSDs are used to having to follow in Linux's wake. They'll go through this process of deciding. I'm not sure whether FreeBSD or Darwin will be first but I doubt Apple is going to want huge chunks of Linux code to simply never be able to be ported over to Macports easily. I don't think FreeBSD is going to want huge numbers of applications one simply can't run.
So I don't think long term this has much impact on applications. For the next 5 years those applications that have complex startup will probably have slightly reduced functionality via. init scripts.
Systemd is open source, it can be ported. Right now the version that exists depends on features of the Linux kernel. The other operating systems are going to have to make choices about whether they want to support systemd or not. Basically this isn't something systemd developers are going to do it is something IBM/AIX, FreeBSD, OpenBSD, QNX... are going to do.
GP was saying he wanted the choice to be made by a large committee. You are just saying you disagree with the large committee that made the choice.
Are you really arguing that the huge number of people all over the planet who have complained about structural problems with the Unix init system for the last 50 years are in error because you don't mind 15 minute boots but do mind patch problems?
OpenRC is a really good replacement for SysV init. A serious enhancement that keeps with the spirit, an upgrade rather than a rip out and replace. OpenRC doen't keep daemons alive but that would be easy to add. Ultimately then there is only one hard problem what to do about hardware that changes. OpenRC doesn't architecturally have any good way for handing that while systemd does.
Certainly if the argument were OpenRC vs. systemd instead of SysV init vs. systemd I suspect the advantages of systemd wouldn't have outweighed the huge shift. I hope distributions like Slackware which don't have systemd move to OpenRC so that it gets tested in environments other than Gentoo.
That's how Debian works. Debian switched to systemd. Given the choice between:
a) Massive reengineer Gnome to support sysV
b) Drop Gnome
c) Switch to systemd
They picked (c) by a pretty clear vote. So the result you are asking for happened.
Yes people wanted the Windows registry. The people who were designing Windows applications and had to write complex pif and ini file interpreters to handle bugs.
X11 fundamentally separates application buffers and video buffers. Copying is an expensive operation. How do you fix that?
No wrong. Wayland has remote desktop functionality. You'll be passing far less data around the network using Wayland's remote functionality than you do with X11.
The systems that are killing Windows make greater use of sharing between application and video buffers not less.