RH is re-inventing the wheel by dropping Btrfs in favor of their own fs, called Stratis.
Nope. Stratis composes existing software so that it adds up to having features people wanted from Btrfs/ZFS.
It uses XFS and the linux devicemapper subsystem to export a D-Bus API. This way, there is new implementation code, but there aren't any new core technologies, and the base filesystem code is already stable.
Btrfs is not going to be a significant part of the future, because it has failed to show itself as worthy. You can't lose data, and you can't change features all the time, if you want to be a filesystem. You have to know what your features are supposed to be, and also how to implement them. People who want to futz with the features and API to get it perfect even when it doesn't yet work that well, they should write browsers or something where that type of practice is accepted. That's the simple fact; considering how long it took Btrfs to get to its current state of development, few people would be willing to use it even if they were told it was finally stable.
I don't doubt there are more than one of you who choose it, I just doubt that if RedHat or IBM goes out collecting technical opinions from professionals that Btrfs will be found to be popular, or to have a use case that it was ever actually stable enough to be assigned to. Stratis already has a more-stable, more-highly-regarded filesystem under the hood.
Nonsense, they already support a lot of different stuff.
This is high end support, they have to hire actual smart people who understand *nix to do that work. Because, the employee actually calling from the client company is often a sysadmin or engineer.
The support they do is mostly professional services, not helpdesk. Though they certainly do offer desktop support too. Actual professionals have no problem at all supporting two things, because they're thinking about real problems that are being presented, they're not reading a script or something, or memorizing a checklist.
Most of the speculation as to the reason centers on the cloud technologies RedHat has, and the associated programmers, and IBMs desire to improve the hybrid public/private cloud ecosystem. So that's exactly the opposite direction as wanting to consolidate offerings.
Mostly IBM has been driving consolidation by discontinuing their famously expensive proprietary offerings, and moving customers onto open source instead! They believe in their ability to add value through high quality professional services, they aren't really doing the lock-in thing these days.
I say they would, because the only way that that scenario comes about is if it is IBM buying Microsoft, and you know they would do it.
Once upon a time, most email was delivered by sendmail. Sysadmins joked that you only need a BS to be a sysadmin, but you need a PhD to configure sendmail! Yes, the config was that bad. So then MS put out Exchange, and lots of people started switching. Horrified, IBM wrote Postfix and open sourced it so that there was a non-MS alternative to sendmail. Because obviously everybody using sendmail wanted to switch, as long as there was something with the needed features.
If IBM bought MS, they would totally open source Explorer. Just for giggles. (You can tell they're giggling when their ties oscillate slightly)
That's perhaps the most absurd reason for changing OSes I've ever heard; "somebody random blamed a software bug in a free OS on having low class hardware."
Lets just say, I do believe you that your computer was locking up, but I don't believe you that you know what caused it, or that it was the "IO scheduler."
I miss the clean thin top and bottom panels with some nice effects from Compiz.
Gnome 2 was awesome, but I don't miss it because Xfce has everything it had, and doesn't bother trying to add new stuff to trip me up.
At first I went to Mate, but they were using language that implied that they were not against the forced changing of paradigms, they just didn't like specific details of the changes. So I don't trust them to stay with the Gnome 2 awesome at all.
Probably he doesn't want to look like an old fart with a dozen xterms.
$ ps ax | grep xterm | wc -l 14
I tried to try the tabbed ones long enough to decide if I liked the paradigm, but all of them that I tried crashed at least one time over a few months of use. xterm has never crashed on me in decades! Also, no weird bugs, because no new features.
You're a bit confused about a particular cultural detail.
See, it is the hobby distros that are principled and ideological, and interested in a "pure" environment.
Business culture, OTOH, is pragmatic and practical.
Nobody knows how many people involved in the "creativity business" use CentOS because the distro doesn't have a culture of virtue-signaling; nobody is impressed by the choice, even other people who made the same choice don't care if you use it or not. Generally though people in the "creativity business" are surrounded by very snobby non-linux users, so they probably avoid talking about it at all, and stick to talking about application-level stuff.
And why would a distro need to target a user group like that? The whole point of RHEL is that they try to make all the stuff work, instead of focusing on doing a bunch of shit differently than everybody else, or appealing to specific demographics. Whatever business you're in, RHEL is about getting work done, not about being pure.
The best speculation I've seen as to why IBM bought RedHat (other than the developer base) is to improve their integration of public and private crowds. That's all about making all the different weird shit work together! Purity is what the existing offerings have. And silos and lock-in suck. IBM gets it.
Apple laptops are expensive devices with high profit margins, so Apple could afford to spend the money on such a connector.
You don't seem to have really internalized what having a high profit margin means.
You're saying, "Gosh, they could just have a low profit margin on their device instead." While true, it isn't a useful point.
As a consumer I certainly don't want to buy a device with a high profit margin! I certainly wouldn't go around expecting such devices to have quality parts. That would be insane.
It isn't as easy as you think; unplug a microphone without turning the gain down. Hear anything?! If you didn't hear a bunch of noise, it means whatever you were unplugging it from didn't use a dumb switch, it either used some silicon or some external passive components to prevent the pop. Anything like a laptop where space is at a premium and it is being manufactured in large quantities, then throwing silicon at the problem is cheaper than the passives. Better results, too.
Once you figure out what "entire computer" means, you'll realize that you're not being sarcastic or insightful, and that it is most certainly "an entire separate computer" but also there are already probably 50 "entire separate computers" on their motherboard.
Even what appears to be an analog op-amp is actually an "entire (digital) computer."
It is a real thing in general contract law, but that isn't going to touch something with specific legislation like DMCA.
More likely, the argument is simply wrong on its claims due to being overly-broad when the statute is actually much narrower. Just because a code is needed doesn't actually make it an access control; many of the ICs in a circuit have to be sent a code during startup, so that the chip can make sure you were trying to start it up before it starts up. And to give you time to get everything ready before telling it to start. Being required at boot doesn't automatically make those things into access controls.
Is the T2 chip really needed to implement a simple hardware disconnect?
You can't just disconnect the wire, you have to gently reduce the signal or you'll generate a bunch of crackle and pop.
Plus, you'd at least need a transistor. So, an IC. But with just that, you'd also need a capacitor and resistor to prevent the crackle/pop.
They can make their own IC for about the same price that they'd pay for a discrete transistor, and they'd both be the same tiny package. But their own IC would have its own capacitors and resistors built in; it would cost less and use less space!
If it was DIY or a small business, you'd be right; no special chip is "needed." Or, in a DIY setting you might still have extra CPU pins for GPIO, and you can just turn it off with that.
Another advantage would be that you could connect it to a bus, and control it from software without using extra pins on another chip. I have no idea how robust that brand is when it comes to dealing with broken lid sensors, but on my Thinkpad I can stop it from responding to lid close events if I wanted to. So I'd absolutely insist that something like that be known to the OS as a peripheral.
Don't be such a dunce, I just told you that I'm a mouse, and I'm stirring. There is no sense in making a direct reply just to say that no mice are stirring.
As for abortion, a simple centrist solution: * Access to chemical abortion protected up to 5th month * Abortion banned after 5th month except with exceptions controlled by doctors * Surgical abortion banned except with exceptions controlled by doctors
That would satisfy 70% of Americans on the topic, because there are lots of people who aren't absolutely opposed to it, there are just lots of edge cases they think are horrible or that they falsely believe are common, like "partial-birth abortion." The whole trick in getting the haters down to 20% would be education; it would be easy to get strong opposition among people who understand the details of the policy down to under 20%. The general population, I doubt you'd get it under 25%. But that's still a quality compromise compared to the current situation.
Death penalty, there are lots of people who are against it because of the way it has been wielded disproportionately based on race and economic class. And there are lots of people who would oppose it if they had confidence that people on death row could be punished effectively with a life term without risking that they get released, or spending more money than the current system spends on executing them. I'm against the death penalty, and I think we have too many people in prison, and yet I'd support increased sentences for violent crimes, and I'd support increased penalties on career criminals if I thought that a policy to achieve it wasn't going to be abused by prosecutors. There is lots of room to be "tougher on crime" while reducing the number of executions, so compromise is certainly possible here.
Probably what happened is that before they didn't want to sell, and some major stockholders decided that the time is right, so they made the call.
It isn't likely just "for" one thing. IBM is a professional services company these days, linux support is a big part of their business. When a big company buys their biggest direct competitor, it isn't going to be "for" some trinket. It is a strategic acquisition that not only helps them consolidate their niche, but it also gives IBM a lot of quality engineers.
Cloud is not the reason. It is just one of the many things they're competing in.
If RedHat was a failing company that was being bought cheap, then the buyer might have only wanted one part. Here, IBM is paying a premium price, because the whole thing has value.
Slashdot Mirror
RH is re-inventing the wheel by dropping Btrfs in favor of their own fs, called Stratis.
Nope. Stratis composes existing software so that it adds up to having features people wanted from Btrfs/ZFS.
It uses XFS and the linux devicemapper subsystem to export a D-Bus API. This way, there is new implementation code, but there aren't any new core technologies, and the base filesystem code is already stable.
Btrfs is not going to be a significant part of the future, because it has failed to show itself as worthy. You can't lose data, and you can't change features all the time, if you want to be a filesystem. You have to know what your features are supposed to be, and also how to implement them. People who want to futz with the features and API to get it perfect even when it doesn't yet work that well, they should write browsers or something where that type of practice is accepted. That's the simple fact; considering how long it took Btrfs to get to its current state of development, few people would be willing to use it even if they were told it was finally stable.
I don't doubt there are more than one of you who choose it, I just doubt that if RedHat or IBM goes out collecting technical opinions from professionals that Btrfs will be found to be popular, or to have a use case that it was ever actually stable enough to be assigned to. Stratis already has a more-stable, more-highly-regarded filesystem under the hood.
It is hard to get the code finished when the prototype keeps slipping through your fingers like Btr.
Nonsense, they already support a lot of different stuff.
This is high end support, they have to hire actual smart people who understand *nix to do that work. Because, the employee actually calling from the client company is often a sysadmin or engineer.
The support they do is mostly professional services, not helpdesk. Though they certainly do offer desktop support too. Actual professionals have no problem at all supporting two things, because they're thinking about real problems that are being presented, they're not reading a script or something, or memorizing a checklist.
Most of the speculation as to the reason centers on the cloud technologies RedHat has, and the associated programmers, and IBMs desire to improve the hybrid public/private cloud ecosystem. So that's exactly the opposite direction as wanting to consolidate offerings.
Mostly IBM has been driving consolidation by discontinuing their famously expensive proprietary offerings, and moving customers onto open source instead! They believe in their ability to add value through high quality professional services, they aren't really doing the lock-in thing these days.
I say they would, because the only way that that scenario comes about is if it is IBM buying Microsoft, and you know they would do it.
Once upon a time, most email was delivered by sendmail. Sysadmins joked that you only need a BS to be a sysadmin, but you need a PhD to configure sendmail! Yes, the config was that bad. So then MS put out Exchange, and lots of people started switching. Horrified, IBM wrote Postfix and open sourced it so that there was a non-MS alternative to sendmail. Because obviously everybody using sendmail wanted to switch, as long as there was something with the needed features.
If IBM bought MS, they would totally open source Explorer. Just for giggles. (You can tell they're giggling when their ties oscillate slightly)
That's perhaps the most absurd reason for changing OSes I've ever heard; "somebody random blamed a software bug in a free OS on having low class hardware."
Lets just say, I do believe you that your computer was locking up, but I don't believe you that you know what caused it, or that it was the "IO scheduler."
Gnome 2 was awesome.
I miss the clean thin top and bottom panels with some nice effects from Compiz.
Gnome 2 was awesome, but I don't miss it because Xfce has everything it had, and doesn't bother trying to add new stuff to trip me up.
At first I went to Mate, but they were using language that implied that they were not against the forced changing of paradigms, they just didn't like specific details of the changes. So I don't trust them to stay with the Gnome 2 awesome at all.
Probably he doesn't want to look like an old fart with a dozen xterms.
$ ps ax | grep xterm | wc -l
14
I tried to try the tabbed ones long enough to decide if I liked the paradigm, but all of them that I tried crashed at least one time over a few months of use. xterm has never crashed on me in decades! Also, no weird bugs, because no new features.
I like Qt and use the "konsole" terminal app. Guess I need to find a new tabbed terminal app that I like as much as konsole.
Considering qt and konsole are a lib and an app, respectively, and not the desktop environment, no, you don't.
I run xfce and I can link to Qt and run applications starting with the letter k just fine.
Plus, you'll still be able to install it normally from whatever third party repo you get media codecs from. :)
OS/2 wasn't bad, outside of multitasking.
If they hated us they'd go back to AIX, instead of putting more skin into the linux.
You're a bit confused about a particular cultural detail.
See, it is the hobby distros that are principled and ideological, and interested in a "pure" environment.
Business culture, OTOH, is pragmatic and practical.
Nobody knows how many people involved in the "creativity business" use CentOS because the distro doesn't have a culture of virtue-signaling; nobody is impressed by the choice, even other people who made the same choice don't care if you use it or not. Generally though people in the "creativity business" are surrounded by very snobby non-linux users, so they probably avoid talking about it at all, and stick to talking about application-level stuff.
And why would a distro need to target a user group like that? The whole point of RHEL is that they try to make all the stuff work, instead of focusing on doing a bunch of shit differently than everybody else, or appealing to specific demographics. Whatever business you're in, RHEL is about getting work done, not about being pure.
The best speculation I've seen as to why IBM bought RedHat (other than the developer base) is to improve their integration of public and private crowds. That's all about making all the different weird shit work together! Purity is what the existing offerings have. And silos and lock-in suck. IBM gets it.
They didn't say anything about getting rid of xfce, so business users are safe. ;)
Your a dick shit
"You're"
That's what he said. He fucked his prey in a storefront because there were too many signs out on the street. And he hates God, because reCAPTCHA v2.
Gets yous some English!
Some days I find myself pining for altavista.
These days shouting at them on twitter might be more effective.
Apple laptops are expensive devices with high profit margins, so Apple could afford to spend the money on such a connector.
You don't seem to have really internalized what having a high profit margin means.
You're saying, "Gosh, they could just have a low profit margin on their device instead." While true, it isn't a useful point.
As a consumer I certainly don't want to buy a device with a high profit margin! I certainly wouldn't go around expecting such devices to have quality parts. That would be insane.
Oh thatâ(TM)s right, you didnâ(TM)t think of it.
Thanks Jar-Jar. What do they call switches on your planet?
It isn't as easy as you think; unplug a microphone without turning the gain down. Hear anything?! If you didn't hear a bunch of noise, it means whatever you were unplugging it from didn't use a dumb switch, it either used some silicon or some external passive components to prevent the pop. Anything like a laptop where space is at a premium and it is being manufactured in large quantities, then throwing silicon at the problem is cheaper than the passives. Better results, too.
I don't know about the coward, but I know when I tried to touch an Apple ][e in an inappropriate way it hurt for a whole week.
Right there, on the doll. That's where it hurt.
Once you figure out what "entire computer" means, you'll realize that you're not being sarcastic or insightful, and that it is most certainly "an entire separate computer" but also there are already probably 50 "entire separate computers" on their motherboard.
Even what appears to be an analog op-amp is actually an "entire (digital) computer."
It is a real thing in general contract law, but that isn't going to touch something with specific legislation like DMCA.
More likely, the argument is simply wrong on its claims due to being overly-broad when the statute is actually much narrower. Just because a code is needed doesn't actually make it an access control; many of the ICs in a circuit have to be sent a code during startup, so that the chip can make sure you were trying to start it up before it starts up. And to give you time to get everything ready before telling it to start. Being required at boot doesn't automatically make those things into access controls.
Don't worry, they're iterating as fast as they can.
Is the T2 chip really needed to implement a simple hardware disconnect?
You can't just disconnect the wire, you have to gently reduce the signal or you'll generate a bunch of crackle and pop.
Plus, you'd at least need a transistor. So, an IC. But with just that, you'd also need a capacitor and resistor to prevent the crackle/pop.
They can make their own IC for about the same price that they'd pay for a discrete transistor, and they'd both be the same tiny package. But their own IC would have its own capacitors and resistors built in; it would cost less and use less space!
If it was DIY or a small business, you'd be right; no special chip is "needed." Or, in a DIY setting you might still have extra CPU pins for GPIO, and you can just turn it off with that.
Another advantage would be that you could connect it to a bus, and control it from software without using extra pins on another chip. I have no idea how robust that brand is when it comes to dealing with broken lid sensors, but on my Thinkpad I can stop it from responding to lid close events if I wanted to. So I'd absolutely insist that something like that be known to the OS as a peripheral.
Don't be such a dunce, I just told you that I'm a mouse, and I'm stirring. There is no sense in making a direct reply just to say that no mice are stirring.
As for abortion, a simple centrist solution:
* Access to chemical abortion protected up to 5th month
* Abortion banned after 5th month except with exceptions controlled by doctors
* Surgical abortion banned except with exceptions controlled by doctors
That would satisfy 70% of Americans on the topic, because there are lots of people who aren't absolutely opposed to it, there are just lots of edge cases they think are horrible or that they falsely believe are common, like "partial-birth abortion." The whole trick in getting the haters down to 20% would be education; it would be easy to get strong opposition among people who understand the details of the policy down to under 20%. The general population, I doubt you'd get it under 25%. But that's still a quality compromise compared to the current situation.
Death penalty, there are lots of people who are against it because of the way it has been wielded disproportionately based on race and economic class. And there are lots of people who would oppose it if they had confidence that people on death row could be punished effectively with a life term without risking that they get released, or spending more money than the current system spends on executing them. I'm against the death penalty, and I think we have too many people in prison, and yet I'd support increased sentences for violent crimes, and I'd support increased penalties on career criminals if I thought that a policy to achieve it wasn't going to be abused by prosecutors. There is lots of room to be "tougher on crime" while reducing the number of executions, so compromise is certainly possible here.
Probably what happened is that before they didn't want to sell, and some major stockholders decided that the time is right, so they made the call.
It isn't likely just "for" one thing. IBM is a professional services company these days, linux support is a big part of their business. When a big company buys their biggest direct competitor, it isn't going to be "for" some trinket. It is a strategic acquisition that not only helps them consolidate their niche, but it also gives IBM a lot of quality engineers.
Cloud is not the reason. It is just one of the many things they're competing in.
If RedHat was a failing company that was being bought cheap, then the buyer might have only wanted one part. Here, IBM is paying a premium price, because the whole thing has value.