On the other hand, when I call Rush Limbaugh a fat liar, I do not express any hostility or derision toward fat people in general. Being fat is ugly and unhealthy, and specifically in a wealthy person it may indicate idleness, lack of moderation and inability to apply common sense to the choice of food. However by itself being fat means very little as far as political discussions and propaganda are concerned.
I used VariCAD for all my 3D modeling of sheet metal (bent/riveted/welded) parts, when the metal shop had SolidWorks-based internal workflow. Actual operations specific to sheet metal were handled by the shop based on my models, so I only had to deal with the shape of the part after it was already cut (on a press or with a laser cutter) and bent. It should work fine for general-purpose 3D, it supports export to formats now used for 3D printing, runs on Linux, and is relatively cheap.
If only someone thought about that. *cough* SIGTERM *cough*. Current update design is plain broken. You're too stupid to understand that because youre a fanboy.
You don't even know, what currently implemented update design IS. Web browsers ask the user to restart. Text editors do not. One can be pretty sure, a text editor already editing a document won't get any less secure from continuing running, no matter what security bugs its libraries contain, if it was exploited it's too late, and editing another text file will start a new copy of the editor with the updated library. Only if the user will somehow decide to keep opening documents in the old copy, he will be exposed to any threats, but that would be a monumentally stupid behavior after seeing a list of updates.
You think only libs are dependencies? What about after installation. The package metadata should also contain list of locations of all config files created outside of home directory.
They are a part of either the package itself, or referred from it as configuration files, they are not dependencies. Dependencies are other packages -- libraries, servers, utilities -- that are necessary for a given package to work.
Also apart from settings, there is a dependency on data. Package metadata should have included in design..
It already is.
option for application to specify its default data store directory.
Non-user-specific data is always in the same directory. If the admin does not like it, he can make a symlink, mount point, or maintain it by himself. Package manager will still do the right thing -- it will move or update the data if any conversion is required between versions, but it will not destroy data without asking, or touch custom configuration that the admin maintains by himself. User-specific data is handled by the user in his home directory. Package manager does not touch it, however applications are free to provide their own configuration and update mechanisms whenever (almost never) it may be necessary.
What is the point of having unified system if you cant migrate from one pc to another without any problems.
You don't migrate. You move your hard drive, or copy it -- then new computer *IS* your old computer, just with all hardware replaced. I am currently sitting in front of the Linux box that is essentially the same configuration I have first installed in 1994, except over the years it had at least four complete updates of hardware, at one point the base system was switched from Slackware to Gentoo, and then Gentoo switched from 32-bit to 64-bit with the same set of packages. Home directory stayed the same, and set of packages varied more through the life of hardware or distribution than across the transition. And that's Slackware and Gentoo, distributions not known for easy transitions or maintenance. Debian or Ubuntu wouldn reduce any of such moves to running a simple script.
You're talking about downloads... pause/resume on downloads is something that a highschool student can implement.
Actually he can not because he does not know yet how to verify the integrity of the result.
And indeed some of them have implemented such horrible download managers which are fucked up and dont work on many linux dvd ISOs because file is greater than 4GB and they use 32bit int for file size...
What download managers? What DVDs? The only two programs used for downloading updates are wget and curl, and protocol implementations in package managers. They always work.
I'm talking about the entire update transaction.
DVD images are never downloaded as a part of update.
As it is currently implemented transactional model for all package managers in all distributions is broken and fragile (aka shit). I have to "clean" the state when I lo
Those are applications for viewing remote desktops, not implementations of toolkits that export UI to remote users. You don't know what this whole discussion is about.
X server is a _display server_, and all applications are its clients. It runs on the user's computer and handles all clients -- local, remote, remote forwarded over ssh, even ones running inside locally running virtual machines or containers if the user wishes to use them that way. For the user, they are all just applications with GUI, they can be integrated into any desktop, and they can use the most efficient mechanism among multiple available. Local application will use shared memory and direct access to 3D and video decoder, while remote application would have to tunnel its access through serialized network protocols, however both will benefit from local compositing when their windows overlap, as compositing is handled by a window manager. This is how things work now.
Remote desktop servers are GUI image streaming and input servers, their clients are applications that show remote desktop in a window. At best, they can be hacked to make the desktop background transparent, so remote applications will seem to be integrated into the local desktop. It's a worthless toy compared to a modern networked display architecture.
You are reading my comment backwards. In those situations where applications need to control buffering they can't under X11. Obviously applications that don't care about buffering won't have to worry about it under Wayland or X11.
Again, you have no idea what are you talking about. Applications use interfaces that hide or expose details of display mechanism as the application programmer requires. Most GUI applications' programmers need little more than a widget set and maybe a layout engine. Some have bitmaps and drawing primitives. Some need direct access to 3D primitives, video decoding, etc. Over their history, X11 and libraries that use it, solved all this by creating a stack of libraries, protocols, protocol extensions and data formats. It's not easy to implement and maintain, but it exists to allow developers to work on whatever layers and levels their application is supposed to be. People who do not understand it, get an impression that it all can be reduced to re-implementation of a small subset that they believe to be the only one needed, but this is not true for the whole range of applications and use scenarios that they don't know about.
I don't think so. I think RDP works fine as long as the client (usual meaning, server for X11) has a toolkit capable of responding to high level instructions. So a Qt clients needs to have Qt but not.NET, a Gnome client needs to have GTK...
Again, you have no idea what you are talking about.
RDP is a secure protocol. Encryption is handled like any other network protocol. Authentication is handled using the security system on the server. Session management is handled by the server.
And it's a shit mechanism compared to X11 over SSH, where X11 has local cookies (application access to display server), SSH handles encryption and session authentication (user's access to applications). User is in control of how independent SSH sessions let remote applications on multiple hosts display their interface on his display, while his desktop session can be anything on its own.
KDE, Gnome and most (hopefully) most other toolkits are going to support RDP
First and foremost, even if it was true, this is not something that can be done on the level of toolkit. Authentication, encryption, session management, support for absolutely everything that can be displayed, has to be done once, on one level. With X, it's done that way -- either through X, or ssh that wraps it, and it uses single implementation for all applications that the user can run.
Second, no GUI toolkit supports such a thing, or ever announced any intention to do so, so you are talking out of your ass.
Third, there would be a reason for a toolkit to support its own remote functionality, but that would make any sense only if it was an optional feature that optimized toolkit-specific behavior. RDP is designed for Windows, and would be suboptimal for any modern GUI toolkit other than Windows. Maybe you meant RFB, but RFB is not optimized for anything at all, it's worse than X, NX, and anything else.
Most distributions will include RDP server software
Surely you meant RFB, as RDP is a proprietary Microsoft protocol, and "most distributions" only include a client for it. As I have mentioned before, RFB is inadequate.
So Wayland apps run remotely.
No, they don't. Wayland applications (ones that use Wayland server) absolutely definitely can't run remotely. Qt anf GTK applications intended for Wayland potentially can be hacked with currently-nonexistent versions of Qt and GTK that talk RDP, RFB or some other protocol instead of Wayland interface, but that has absolutely nothing to do with Wayland, relies on other people doing something that those people don't plan doing, and does not cover any application that actually tries to use Wayland interface in its implemented form.
In other words, your claim is false, and unrelated to that, completely idiotic.
It isn't part of the graphics system because part of the design of Wayland is that remote use changes things like buffering strategies that applications needs to know about.
That's bullshit. Application programmer does not have to know details of X protocol, either, yet X11 libraries implement it, and applications use it. Wayland developers merely pretend that they can ignore huge pieces of design because they expect someone to create workarounds for their system's deficiencies. Not even Microsoft dares to do such a thing in GUI implementation, and Microsoft is a notorious bunch of lazy, talentless and bigoted developers.
The graphics system is too far down the chain, far better is something like the toolkit.
GUI toolkit is application specific and optional. Even though it may be a good decision to let it have its own optional protocol extension, it is a completely wrong place to implement a communications mechanism that is a core of remote UI/display support. What you pretend to propose, amounts to little more than insisting on application developers writing their UI as AJAX web pages, as this is the only "toolkit" that currently support remote operation on its own.
Translation:- You continue to be exposed to security vulnerabilities despite a patch already installed.
The alternative is to lose user's data currently held by the running processes.
Windows decided to do the sane thing and prevent users from getting exploited. Linux will catch up some time.
Servers are restarted automatically by package managers, and web/mail clients have restart notification procedure as a part of the update. Your arrogance on the part of your overlords is misplaced.
You can't even do simple shit like backup 'programs and settings' when migrating on Linux computers.
What the Hell are you talking about? User's home directory contains all his settings.
The entire fucking software repository is under control of distribution and still you cant. No.. you have to keep downloading hundreds of megabytes every single time some stupid OSS dev adds some shit in the repo. But surely.. you can pause and resume updates later right? No? hahahaha...dont worry Linux will catch up in some time.
You have no idea what are you talking about. Distribution updates are indexes, lists of packages. User can choose what to upgrade (UI usually insists on at least all security updates, but it's up to the user to reject or postpone the update), however indexes are always updated, so package manager always knows how dependencies graph looks like.
But surely.. you can pause and resume updates later right? No? hahahaha...dont worry Linux will catch up in some time.
Package downloads can be interrupted and restarted, this is what/var/cache/apt/archives/partial/ is for.
Whats awesome is crappy plugins like "fastest mirror" uses mirror choice based on stupid criteria like ping or geo ip, rather than ACTUAL SPEED.
That's because measuring speed precisely enough to choose among hundreds of universities and ISPs would create more traffic than actual update, you dumbass.
Valve has already figured it out for steam updates... also Microsoft... Oh well..
Except both Microsoft and Valve have nearly glacial download speeds no matter where you are. With their resources they could've paid Akamai to handle that.
Linux only works in places where its locked down and completely out of users hands. e.g. servers,phones and embedded devices.
Do your managers give you those idiotic statements to paste into your postings, or does your software automatically add them?
It's worthless -- it's just a server that supports someone else's clients. Any graphics system that intends to replace X, must be able to run any application as a client, remotely. Otherwise it's shit.
Without specification and design documents or working implementation it's all hot air, and judging by how Wayland implementation is going in general, there is no expectation that they will implement things they actuallt agreed upon, leave alone ones they did not.
While many other "rights" of organizations are pure bullshit, property rights of corporations are valid, as in that particular aspect the corporation acts as a proxy for its owners' property rights.
Windows is "certified Unix". And there is Emacs port to it.
The problem is, it's shit, its design is shit, and the only way to do anything with it is in its shitty Windows way. Same applies, at lesser degree, to OSX, iOS and Android, but iOS and Android have an advantage of being used only on phones and tablets, devices that were never intended to do general-purpose computing.
On the other hand, my Maemo-based N900 works remarkably well for its microscopic by modern standards CPU, and its GUI runs entirely on X11.
As someone above said, they're implementing X on top of Wayland, so you'll still be able to run your 1993 Motif applications.
The problem is not with running X applications, it's with running new applications on complex display systems with local and remote displays, screens, window management, handling of input and dealing with misbehaving applications. Any application written for Wayland, Mir, etc. is worthless unless you have exactly the same display configuration, shell and everything, those systems' developers intended to support.
Mir, on the other hand, seems to be related to Ubuntu Mobile, which is probably dead-on-arrival.
I don't know and don't care. If those people wanted to do something useful, they would improve X protocol (what some people already done, better than them), or implement optional extensions that optimize the support for GUI features that people actually want (what some people also done, including some of them). Dropping features and design decisions just because they aren't useful for smartphones, is a stupid and harmful direction of general-purpose systems' development.
And none can coexist on the same system. And none, save for X11, have any usable remote protocol (if copying framebuffers is a usable protocol, then HDMI cable connection is a network).
Food that is intended to be eaten while hot, over an unpredictable and possibly significant amount of time, is served in a condition when it's ready to eat, and will remain so even over the course of a relatively long dinner. This is easier to achieve with the lowest rate of cooling over time and initial temperature close to the upper boundary of the acceptable range. The less is the expected rate of cooling, the farther from the upper boundary this starting temperature can be set, and the food will be the closest to the optimal temperature over the whole time.
To be fair, Mark Zuckerberg and his investors are exactly the same kind of assholes, therefore it's arrogant of him to act as if he is a productive member of society while they are not.
Computational biology graduate student Asa Adadey said the free meal was a draw and said he learned in one mini-course not to cut up all his meat at once before eating it.
Anyone with a brain capable of dealing with science, engineering and math would know that cutting all food before eating it increases the surface area while keeping the total mass and volume unchanged, thus causing the food to cool and dry faster, relative to its original, supposedly optimal for consumption, state. Anyone who is surprised by this, is probably not good at recognizing reasons behind other decisions and rules. He may be is a "trade school" kind of student that collects assorted morsels of prescriptive knowledge and expects it to provide him an easy, comfortable job. Real geeks hate those people, because they pass themselves as competent, cause enormous messes, and a real engineer has to clean up after them instead of doing actual work.
Actually being able to use synaptics touchpad on laptop.
Works great on my laptop (plain vanilla installation of kubuntu).
At the moment, I have CentOS 6.3 installed on my sony laptop
No, you don't, unless you are a professional Microsoft astroturfer. CentOS is a Red Hat Enterprise Linux replacement for people who want to run server software specifically configured for Red Hat, without having to pay Red Hat licensing fees. It's a workaround. It's not supposed to be used by people who have any other choice. It's not even used by most people who have to use commercial software that is specifically configured for Red Hat, because at this point Fedora (also Red Hat product), Debian (the best Linux distribution overall) and Ubuntu (the best user-friendly adaptation of Debian for desktops and laptops, as long as you use Kubuntu or Xubuntu flavors) run that software better than Red Hat does.
So no, you don't have CentOS on your laptop. And if you do, you still have time to replace it with Ubuntu and deny that you ever have it there.
Slashdot Mirror
No, hairyfeet. You are the demons^H^H^H^H^H^Htrolls.
On the other hand, when I call Rush Limbaugh a fat liar, I do not express any hostility or derision toward fat people in general. Being fat is ugly and unhealthy, and specifically in a wealthy person it may indicate idleness, lack of moderation and inability to apply common sense to the choice of food. However by itself being fat means very little as far as political discussions and propaganda are concerned.
lol hairyfeet is talking about trolls.
I used VariCAD for all my 3D modeling of sheet metal (bent/riveted/welded) parts, when the metal shop had SolidWorks-based internal workflow. Actual operations specific to sheet metal were handled by the shop based on my models, so I only had to deal with the shape of the part after it was already cut (on a press or with a laser cutter) and bent. It should work fine for general-purpose 3D, it supports export to formats now used for 3D printing, runs on Linux, and is relatively cheap.
If only someone thought about that. *cough* SIGTERM *cough*. Current update design is plain broken. You're too stupid to understand that because youre a fanboy.
You don't even know, what currently implemented update design IS. Web browsers ask the user to restart. Text editors do not. One can be pretty sure, a text editor already editing a document won't get any less secure from continuing running, no matter what security bugs its libraries contain, if it was exploited it's too late, and editing another text file will start a new copy of the editor with the updated library. Only if the user will somehow decide to keep opening documents in the old copy, he will be exposed to any threats, but that would be a monumentally stupid behavior after seeing a list of updates.
You think only libs are dependencies? What about after installation. The package metadata should also contain list of locations of all config files created outside of home directory.
They are a part of either the package itself, or referred from it as configuration files, they are not dependencies. Dependencies are other packages -- libraries, servers, utilities -- that are necessary for a given package to work.
Also apart from settings, there is a dependency on data. Package metadata should have included in design..
It already is.
option for application to specify its default data store directory.
Non-user-specific data is always in the same directory. If the admin does not like it, he can make a symlink, mount point, or maintain it by himself. Package manager will still do the right thing -- it will move or update the data if any conversion is required between versions, but it will not destroy data without asking, or touch custom configuration that the admin maintains by himself. User-specific data is handled by the user in his home directory. Package manager does not touch it, however applications are free to provide their own configuration and update mechanisms whenever (almost never) it may be necessary.
What is the point of having unified system if you cant migrate from one pc to another without any problems.
You don't migrate. You move your hard drive, or copy it -- then new computer *IS* your old computer, just with all hardware replaced. I am currently sitting in front of the Linux box that is essentially the same configuration I have first installed in 1994, except over the years it had at least four complete updates of hardware, at one point the base system was switched from Slackware to Gentoo, and then Gentoo switched from 32-bit to 64-bit with the same set of packages. Home directory stayed the same, and set of packages varied more through the life of hardware or distribution than across the transition. And that's Slackware and Gentoo, distributions not known for easy transitions or maintenance. Debian or Ubuntu wouldn reduce any of such moves to running a simple script.
You're talking about downloads... pause/resume on downloads is something that a highschool student can implement.
Actually he can not because he does not know yet how to verify the integrity of the result.
And indeed some of them have implemented such horrible download managers which are fucked up and dont work on many linux dvd ISOs because file is greater than 4GB and they use 32bit int for file size...
What download managers? What DVDs? The only two programs used for downloading updates are wget and curl, and protocol implementations in package managers. They always work.
I'm talking about the entire update transaction.
DVD images are never downloaded as a part of update.
As it is currently implemented transactional model for all package managers in all distributions is broken and fragile (aka shit). I have to "clean" the state when I lo
http://www.kde.org/applications/internet/krdc/ [kde.org]
http://sourceforge.net/projects/gnome-rdp/ [sourceforge.net]
Those are applications for viewing remote desktops, not implementations of toolkits that export UI to remote users. You don't know what this whole discussion is about.
X server is a _display server_, and all applications are its clients. It runs on the user's computer and handles all clients -- local, remote, remote forwarded over ssh, even ones running inside locally running virtual machines or containers if the user wishes to use them that way. For the user, they are all just applications with GUI, they can be integrated into any desktop, and they can use the most efficient mechanism among multiple available. Local application will use shared memory and direct access to 3D and video decoder, while remote application would have to tunnel its access through serialized network protocols, however both will benefit from local compositing when their windows overlap, as compositing is handled by a window manager. This is how things work now.
Remote desktop servers are GUI image streaming and input servers, their clients are applications that show remote desktop in a window. At best, they can be hacked to make the desktop background transparent, so remote applications will seem to be integrated into the local desktop. It's a worthless toy compared to a modern networked display architecture.
You are reading my comment backwards. In those situations where applications need to control buffering they can't under X11. Obviously applications that don't care about buffering won't have to worry about it under Wayland or X11.
Again, you have no idea what are you talking about. Applications use interfaces that hide or expose details of display mechanism as the application programmer requires. Most GUI applications' programmers need little more than a widget set and maybe a layout engine. Some have bitmaps and drawing primitives. Some need direct access to 3D primitives, video decoding, etc. Over their history, X11 and libraries that use it, solved all this by creating a stack of libraries, protocols, protocol extensions and data formats. It's not easy to implement and maintain, but it exists to allow developers to work on whatever layers and levels their application is supposed to be. People who do not understand it, get an impression that it all can be reduced to re-implementation of a small subset that they believe to be the only one needed, but this is not true for the whole range of applications and use scenarios that they don't know about.
I don't think so. I think RDP works fine as long as the client (usual meaning, server for X11) has a toolkit capable of responding to high level instructions. So a Qt clients needs to have Qt but not .NET, a Gnome client needs to have GTK...
Again, you have no idea what you are talking about.
RDP is a secure protocol. Encryption is handled like any other network protocol. Authentication is handled using the security system on the server. Session management is handled by the server.
And it's a shit mechanism compared to X11 over SSH, where X11 has local cookies (application access to display server), SSH handles encryption and session authentication (user's access to applications). User is in control of how independent SSH sessions let remote applications on multiple hosts display their interface on his display, while his desktop session can be anything on its own.
No you are missing the point.
KDE, Gnome and most (hopefully) most other toolkits are going to support RDP
First and foremost, even if it was true, this is not something that can be done on the level of toolkit. Authentication, encryption, session management, support for absolutely everything that can be displayed, has to be done once, on one level. With X, it's done that way -- either through X, or ssh that wraps it, and it uses single implementation for all applications that the user can run.
Second, no GUI toolkit supports such a thing, or ever announced any intention to do so, so you are talking out of your ass.
Third, there would be a reason for a toolkit to support its own remote functionality, but that would make any sense only if it was an optional feature that optimized toolkit-specific behavior. RDP is designed for Windows, and would be suboptimal for any modern GUI toolkit other than Windows. Maybe you meant RFB, but RFB is not optimized for anything at all, it's worse than X, NX, and anything else.
Most distributions will include RDP server software
Surely you meant RFB, as RDP is a proprietary Microsoft protocol, and "most distributions" only include a client for it. As I have mentioned before, RFB is inadequate.
So Wayland apps run remotely.
No, they don't. Wayland applications (ones that use Wayland server) absolutely definitely can't run remotely. Qt anf GTK applications intended for Wayland potentially can be hacked with currently-nonexistent versions of Qt and GTK that talk RDP, RFB or some other protocol instead of Wayland interface, but that has absolutely nothing to do with Wayland, relies on other people doing something that those people don't plan doing, and does not cover any application that actually tries to use Wayland interface in its implemented form.
In other words, your claim is false, and unrelated to that, completely idiotic.
It isn't part of the graphics system because part of the design of Wayland is that remote use changes things like buffering strategies that applications needs to know about.
That's bullshit. Application programmer does not have to know details of X protocol, either, yet X11 libraries implement it, and applications use it. Wayland developers merely pretend that they can ignore huge pieces of design because they expect someone to create workarounds for their system's deficiencies. Not even Microsoft dares to do such a thing in GUI implementation, and Microsoft is a notorious bunch of lazy, talentless and bigoted developers.
The graphics system is too far down the chain, far better is something like the toolkit.
GUI toolkit is application specific and optional. Even though it may be a good decision to let it have its own optional protocol extension, it is a completely wrong place to implement a communications mechanism that is a core of remote UI/display support. What you pretend to propose, amounts to little more than insisting on application developers writing their UI as AJAX web pages, as this is the only "toolkit" that currently support remote operation on its own.
Translation :- You continue to be exposed to security vulnerabilities despite a patch already installed.
The alternative is to lose user's data currently held by the running processes.
Windows decided to do the sane thing and prevent users from getting exploited. Linux will catch up some time.
Servers are restarted automatically by package managers, and web/mail clients have restart notification procedure as a part of the update. Your arrogance on the part of your overlords is misplaced.
You can't even do simple shit like backup 'programs and settings' when migrating on Linux computers.
What the Hell are you talking about? User's home directory contains all his settings.
The entire fucking software repository is under control of distribution and still you cant. No.. you have to keep downloading hundreds of megabytes every single time some stupid OSS dev adds some shit in the repo. But surely.. you can pause and resume updates later right? No? hahahaha...dont worry Linux will catch up in some time.
You have no idea what are you talking about. Distribution updates are indexes, lists of packages. User can choose what to upgrade (UI usually insists on at least all security updates, but it's up to the user to reject or postpone the update), however indexes are always updated, so package manager always knows how dependencies graph looks like.
But surely.. you can pause and resume updates later right? No? hahahaha...dont worry Linux will catch up in some time.
Package downloads can be interrupted and restarted, this is what /var/cache/apt/archives/partial/ is for.
Whats awesome is crappy plugins like "fastest mirror" uses mirror choice based on stupid criteria like ping or geo ip, rather than ACTUAL SPEED.
That's because measuring speed precisely enough to choose among hundreds of universities and ISPs would create more traffic than actual update, you dumbass.
Valve has already figured it out for steam updates... also Microsoft... Oh well..
Except both Microsoft and Valve have nearly glacial download speeds no matter where you are. With their resources they could've paid Akamai to handle that.
Linux only works in places where its locked down and completely out of users hands. e.g. servers,phones and embedded devices.
Do your managers give you those idiotic statements to paste into your postings, or does your software automatically add them?
No, it is not, it's an act of war while espionage is not.
Unless applied to playing TF2, of course.
It's worthless -- it's just a server that supports someone else's clients. Any graphics system that intends to replace X, must be able to run any application as a client, remotely. Otherwise it's shit.
Where is the specification?
Without specification and design documents or working implementation it's all hot air, and judging by how Wayland implementation is going in general, there is no expectation that they will implement things they actuallt agreed upon, leave alone ones they did not.
What administration?
People run their applications on multiple hosts.
You have to reboot because the files are in use. What happens to shared libraries that are in use when you update them on Linux?
A copy stays in memory until the last process that uses it, exits. All newly started executables use the updated version.
What, there is another behavior that is not broken?
Don't run it in VMWare under Windows.
While many other "rights" of organizations are pure bullshit, property rights of corporations are valid, as in that particular aspect the corporation acts as a proxy for its owners' property rights.
It's crap, and does not integrate remote applications nearly as well as X.
(much better than X over WAN links in fact)
No one cares about WAN when you have a huge multi-gigabit network with hundreds of hosts.
Windows is "certified Unix". And there is Emacs port to it.
The problem is, it's shit, its design is shit, and the only way to do anything with it is in its shitty Windows way. Same applies, at lesser degree, to OSX, iOS and Android, but iOS and Android have an advantage of being used only on phones and tablets, devices that were never intended to do general-purpose computing.
On the other hand, my Maemo-based N900 works remarkably well for its microscopic by modern standards CPU, and its GUI runs entirely on X11.
As someone above said, they're implementing X on top of Wayland, so you'll still be able to run your 1993 Motif applications.
The problem is not with running X applications, it's with running new applications on complex display systems with local and remote displays, screens, window management, handling of input and dealing with misbehaving applications. Any application written for Wayland, Mir, etc. is worthless unless you have exactly the same display configuration, shell and everything, those systems' developers intended to support.
Mir, on the other hand, seems to be related to Ubuntu Mobile, which is probably dead-on-arrival.
I don't know and don't care. If those people wanted to do something useful, they would improve X protocol (what some people already done, better than them), or implement optional extensions that optimize the support for GUI features that people actually want (what some people also done, including some of them). Dropping features and design decisions just because they aren't useful for smartphones, is a stupid and harmful direction of general-purpose systems' development.
And none can coexist on the same system. And none, save for X11, have any usable remote protocol (if copying framebuffers is a usable protocol, then HDMI cable connection is a network).
Because they suck at design, not at implementation?
Food that is intended to be eaten while hot, over an unpredictable and possibly significant amount of time, is served in a condition when it's ready to eat, and will remain so even over the course of a relatively long dinner. This is easier to achieve with the lowest rate of cooling over time and initial temperature close to the upper boundary of the acceptable range. The less is the expected rate of cooling, the farther from the upper boundary this starting temperature can be set, and the food will be the closest to the optimal temperature over the whole time.
To be fair, Mark Zuckerberg and his investors are exactly the same kind of assholes, therefore it's arrogant of him to act as if he is a productive member of society while they are not.
Computational biology graduate student Asa Adadey said the free meal was a draw and said he learned in one mini-course not to cut up all his meat at once before eating it.
Anyone with a brain capable of dealing with science, engineering and math would know that cutting all food before eating it increases the surface area while keeping the total mass and volume unchanged, thus causing the food to cool and dry faster, relative to its original, supposedly optimal for consumption, state. Anyone who is surprised by this, is probably not good at recognizing reasons behind other decisions and rules. He may be is a "trade school" kind of student that collects assorted morsels of prescriptive knowledge and expects it to provide him an easy, comfortable job. Real geeks hate those people, because they pass themselves as competent, cause enormous messes, and a real engineer has to clean up after them instead of doing actual work.
All supported on Linux. Don't even try hastily googling for "linux bug midi" to dig up some "evidence" posted by 12 years old on Ubuntu forums.
Actually being able to use synaptics touchpad on laptop.
Works great on my laptop (plain vanilla installation of kubuntu).
At the moment, I have CentOS 6.3 installed on my sony laptop
No, you don't, unless you are a professional Microsoft astroturfer. CentOS is a Red Hat Enterprise Linux replacement for people who want to run server software specifically configured for Red Hat, without having to pay Red Hat licensing fees. It's a workaround. It's not supposed to be used by people who have any other choice. It's not even used by most people who have to use commercial software that is specifically configured for Red Hat, because at this point Fedora (also Red Hat product), Debian (the best Linux distribution overall) and Ubuntu (the best user-friendly adaptation of Debian for desktops and laptops, as long as you use Kubuntu or Xubuntu flavors) run that software better than Red Hat does.
So no, you don't have CentOS on your laptop. And if you do, you still have time to replace it with Ubuntu and deny that you ever have it there.