It is somewhat concerning that there is no mention of other toolkits here. I know RedHat is a GTK/GNOME shop, but when they're talking about next generation X, they shouldn't exclude other tookits and frameworks from that discussion.
It seems like they aren't looking at QT/KDE, XUL, OpenOffice.org, Wine, Java (SWT & Swing), and such. Though many other toolkits seem to be moving towards using GTK as a backend on Linux, even a brief mention of KDE or QT would've been nice.
Not that I'm too worried. It seems that even though RedHat and freedesktop.org are dominated by GNOME people, there has been good involvement from other camps. The work being done should benefit everyone, regardless of toolkit or desktop environment (well maybe CDE and Motif will still be butt-ugly, but what can you expect).
The problem is that most of IBM's coders are Windows guys. Many of their configuration utilities will be written only for Win32. For example, if you want to use their rack configuration utitlity, you need Windows.
Much of their in house stuff was written for Windows, and some of it for IE only. I think IBM is ashamed of this, but wouldn't it be intersting if they instead used their own bad experience to sell Linux. Something along the lines of "You see, this is what happens when you lock yourselves to one platform." Use it as an argument to free yourself from the MS platform, and write anything new to be cross platform. It is not like it is that hard nowadays, with XUL, Java, QT, and many other tools available to developers (heck even compiling against Cygwin or Wine libraries would be better than not trying).
Yes, but the cost of sticking with the platform can also be quite large. People just don't notice it because it is status quo. They just see this huge new transition cost and get scared away, without really realizing that it could save them some money.
In IBM's case, maybe it wouldn't even save them any money (their transition costs would be enormous as you say, and they have a lot of in-house Windows expertise built up over the years). But if IBM wants to get serious about moving their customers to Linux, they will need to do so themselves to be taken seriously. So even though some parts of the company would lose a lot of money paying for the switch to Linux, other divisions would later make more money in sales.
So our scientific clusters for bio-informatics, physics, and chemistry are a hobby? Our networking research lab is a hobby? Our webserves, fileservers, print servers, application servers, computation engines, workstations, desktops, and laptops aren't for real work?
That is all news to me. Sorry to say, but around here, Windows is the hobby OS. Windows XP inside is a translation for "lets play some games on this thing" - nothing more than a glorified and insecure gaming console.
Not that Windows can't be used for anything serious either, but you're really out to lunch if you think that Linux is still some pet project that is just a toy for geeks.
Please note that there are legal Linux DVD players available (or if you don't live in DMCA-land, every decss DVD player is also legal). I believe Linspire has one in their click-and-run warehouse, and I'm sure there are others you can buy.
But really, do you think that there will ever be a lawsuit filed against someone watching DVDs using decss? Not bloodly likely.
Well since the hacks were actually on the XFree86 code, and XFree wasn't exactly good about accepting patches, I'd say probably not.
But it is more than patches to XFree86 - they also have a complicated cross-compiling build system that recently got support for the X11 sets. Before they could build just about everything but X11 with this method - now they can build the entire NetBSD release for most every architecture on a single build box from a single source tree.
If they ever do switch to x.org (which is very likely), I'd hope they'd be able to work with the x.org folks to make this an easier task.
I'm not sure if I agree with your assessment of how she got her job, though I do not know enough about about her career previous to HP to say for sure.
However, as a CEO she was a complete disgrace. As one of the most visible and powerful female CEOs in North America, she probably did a lot more harm to her gender than good by setting such a poor example.
I'm sure there are plenty of "she's" that now wish that she'd never become CEO of HP, whether she deserved a shot at the job or not. I think it was a good day for women today - no longer will the people equate female CEO and Carly Fiorina. She was always the first one to come to mind for me, and it wasn't exactly a positive thought.
This is fine, as long as you're willing to support Linux. If not, you'll have people angry at this OS you told them to install, but doesn't work like they expect. But if you're available to help for the first little while though, it can cut significantly cut down on the problems, and the requests for warez (which I will not honour, especially since I have no need for Win32 software myself).
I've done the Linux conversion successfully for a couple people, but I don't think I'd try it for relatives living in a different city than me, or those that are afraid of their computer. So far I've limited it to people who were fed up with Windows, and really did want to try something else.
For those I don't convert to Linux, I could try to encourage them to by a Mac, but I think that'll be futile too. Probably I'll just burn them the OpenCD or gnuWin, and tell them to use the antivirus software and FireFox.
Though you joke, I do think that Ogg Vorbis is quite possibly a format they could offer. I mean why not, it is DRM free and popular in the Linux crowd.
What is more interesting to me is whether they'll offer it in FLAC format. That way you can convert it to any format you want without losing vast amounts of quality. Since they are pushing the fact that being DRM-less allows device portability, why not give format portability too?
Good point. The usage of CBs is much more conducive to good driving. People are not expecting a quick and fluid conversation - long pauses between responses are typical and don't bother the other person. It is okay to either say "just a minute" or just quit talking to focus on the road for a bit.
The etiquette around CBs was build around the fact you use them in vehicles, and are expected to have to drive as well. The etiquette of phone conversations evolved in houses on land lines, and people expect to be able to give and receive the same amount of attention as on a home phone.
People get mad at you if you're only half paying attention to them on the phone (my parents get annoyed at my brother being on the computer and running MSN conversations while on the phone, and only half listening the them). It is just expected that if you bothered to call someone (or pick up the phone), that you're giving the person on the other side your nearly undivided attention. So we have a tendency to do so, whether it is safe or not.
It probably does help that CBs are primarily used by people who spend a lot of time on the road, so they're likely better drivers to begin with. They're also often used in the country or on the highways, where there are generally less hazards.
I'm quite surprised the terms were made public. The terms of the Coke deal here were sealed, much to the displeasure of those outside the union. Often they'll make it public that they're paying for a scoreboard or some other visible piece of hardware, but otherwise keep the deal secret. I imagine it was a similar deal to what you describe though.
I think the biggest reason they used to seal those deals is so that they could offer different deals to different campuses, and not have them realize they're being ripped off. That, and they probably wanted to prevent a bidding war with the competitor - Pepsi and Coke might start driving up the contract amounts through a bidding war, which really is not good for either of them. Negotiate and sign in secret, so that your opponent can't one-up you.
The university here also does not allow Pepsi machines. I wonder if Coca-Cola pwns us too?
* actually they do have an exclusive beverage contract with the campus, so our lousy union gets paid to allow Coke to price gouge us. Quite common in colleges and high schools now.
I was going to say, I'm not sure I'd consider Cygwin the best of Linux.
It is nice to have if you need it, but there are far better Unixes out there than Cygwin on top of Win32. Heck, there are even better Unixes on top of win32, such as CoLinux.
Though others have given their reasons for Mandrake, I'll chime in anyhow. There are two big reasons why I'd standardize on Mandrake Linux across an entire operation:
Package Management and configuration:
- urpmi is a great package management tool, with a good GUI available, but also a very powerful command line version (which is what I use) - updates available for at least 18 months, longer for the Enterprise Server - due to using RPM, 3rd party packages are easier to find (although urpmi contains most everything you'd need, reducing the depedancy on 3rd party packages) - most of their GUI utilities also work on the command line, due to their use of a perl library that can render to GTK or console
Flexibility:
Mandrake's main download edition, as well as the PowerPacks, can be configured on a wide array of hardware and for a wide range of uses. If you want it as a Desktop, specific-use Workstation, or a Server, you can do it. If you want something more specialized, you have these options: - MandrakeMove: LiveCD Desktop oriented version of Mandrake - Discovery: beginner Desktop oriented version - Enterprise Server: similar to RHEL - Multi-Network Firewall: firewall, VPN, router, etc - PowerPack: the works, including commercial packages like Java, and some server GUI wizards if you want that convenience - Download Edition: only OSS software, free to download and use - Clustering: I think this is obvious - versions for i586 and AMD64 for many of the above (PPC versions are sometimes available for the download edition)
These different options give you great flexibility. Here we have a site license, so I put PowerPack on everything from laptops to servers, and have the Clustering edition for the cluster. People who wish to use the same software at home as they do at work can either go with the free download edition, or the fairly cheap and more newbie friendly Discovery or MandrakeMove editions. If you don't have a site license, it is quite possible to put the Download edition on any machines that do not require the commercial software or huge amounts of support. A Club membership an be bought to deal with some support issues if you are only deploying the download edition. A site license will often include a certain number of club memberships, so if you had a few home users running it, they could get some support as well.
They are quite good for support if you ever need it, since they aren't so big a company that they'd ignore their smaller customers (and are small enough that they really pay attention to the big site licenses). But they are also large enough to be profitable as a company, and to have fairly wide usage and support throughout the community.
Everyone thinks of Mandrake as a desktop distro, but it does just as well on the server as Red Hat or Suse in my opinion. Other distros may offer better pure desktop systems (Lindows and Xandros perhaps?), and Debian or the BSDs may be better for some server usage, but I can't think of any Linux or BSD that does many tasks as well as Mandrake does. This is what makes Mandrake so ideal for company/department wide standardization.
They also provide a fair bit more for the power user than most people realize. The easy GUI stuff is there, but there is a lot underneath it for those who don't like point and click. The assumption that you have to graduate to something else less newbie oriented is quite false.
Created due to licensing, but heavily adopted due to people being fed up with XFree86.
The Cygwin folks already had to fork XFree because of the orgainization's refusal to accept patches. And Cygwin was far from being the only ones annoyed with XFree. It was just easier for distros to stick with XFree instead of maintaining their own, and causing a political mess.
The license change was merely the last straw, and was very indicative of how XFree operated. By unilaterally changing the license, then refusing to work with the people who ship their product on fixing it, they showed an even higher level of elitism than before. By this time there was a large enough group X11 developers that were doing great work, but not part of XFree (mainly Keith Packard), that the distros had somewhere else to turn.
So it was more than a simple license change I'm afraid. They kept some of the best developers doing the most innovative work outside of the group, and alienated the very people who distributed their product. Their own elitism made them completely irrelivant in the development of X11, which was supposed to be the entire purpose for XFree's existance.
It is not impossible that this could happen to Firefox too, but right now they are the main drivers in the browser market, and generally are keeping their use base happy. No reason to worry quite yet.
Last I heard it was stable under normal workloads, but could be thrown off by extreme conditions, such as edge conditions when working with 99% full drives and very high workloads.
So in filesystem terms it is considered unstable, but it does indeed work right now. It just doesn't provide assurances and reliability necessary to be part of a mature OS.
Indeed, if anyone can help with the LFS work on NetBSD, it'd be greatly appreciated. Right now there are basically 2 people who look at that code, and it hasn't really seen any work since the summer.
From what I hear, LFS combined with kernel RaidFrame is a killer combination. But until LFS is stablized, nobody will get to see the benefits of this.
Slashdot Mirror
Oh jeez, another easterner complaining about the balmy -20. Come to the prairies to experience the real cold, to the tune of about -40.
But it is a dry cold, so it ain't so bad (as if there can be much difference between a dry -40 and a humid -40).
It is somewhat concerning that there is no mention of other toolkits here. I know RedHat is a GTK/GNOME shop, but when they're talking about next generation X, they shouldn't exclude other tookits and frameworks from that discussion.
It seems like they aren't looking at QT/KDE, XUL, OpenOffice.org, Wine, Java (SWT & Swing), and such. Though many other toolkits seem to be moving towards using GTK as a backend on Linux, even a brief mention of KDE or QT would've been nice.
Not that I'm too worried. It seems that even though RedHat and freedesktop.org are dominated by GNOME people, there has been good involvement from other camps. The work being done should benefit everyone, regardless of toolkit or desktop environment (well maybe CDE and Motif will still be butt-ugly, but what can you expect).
The problem is that most of IBM's coders are Windows guys. Many of their configuration utilities will be written only for Win32. For example, if you want to use their rack configuration utitlity, you need Windows.
Much of their in house stuff was written for Windows, and some of it for IE only. I think IBM is ashamed of this, but wouldn't it be intersting if they instead used their own bad experience to sell Linux. Something along the lines of "You see, this is what happens when you lock yourselves to one platform." Use it as an argument to free yourself from the MS platform, and write anything new to be cross platform. It is not like it is that hard nowadays, with XUL, Java, QT, and many other tools available to developers (heck even compiling against Cygwin or Wine libraries would be better than not trying).
There aint yet no good Lotus Notes client.
Yes, but the cost of sticking with the platform can also be quite large. People just don't notice it because it is status quo. They just see this huge new transition cost and get scared away, without really realizing that it could save them some money.
In IBM's case, maybe it wouldn't even save them any money (their transition costs would be enormous as you say, and they have a lot of in-house Windows expertise built up over the years). But if IBM wants to get serious about moving their customers to Linux, they will need to do so themselves to be taken seriously. So even though some parts of the company would lose a lot of money paying for the switch to Linux, other divisions would later make more money in sales.
Yes, the words "I don't do Windows" do save one from being expected to be the law breaking warez dude for all their friends and family.
And when suddenly they have to pay for their OS (or find someone else they don't know as well to steal it), then Linux starts to look a lot better.
Wow, too bad this story is buried at a 0 moderation right now. If I hadn't already posted here, I'd mod this up.
Linux a hobby?
So our scientific clusters for bio-informatics, physics, and chemistry are a hobby? Our networking research lab is a hobby? Our webserves, fileservers, print servers, application servers, computation engines, workstations, desktops, and laptops aren't for real work?
That is all news to me. Sorry to say, but around here, Windows is the hobby OS. Windows XP inside is a translation for "lets play some games on this thing" - nothing more than a glorified and insecure gaming console.
Not that Windows can't be used for anything serious either, but you're really out to lunch if you think that Linux is still some pet project that is just a toy for geeks.
Please note that there are legal Linux DVD players available (or if you don't live in DMCA-land, every decss DVD player is also legal). I believe Linspire has one in their click-and-run warehouse, and I'm sure there are others you can buy.
But really, do you think that there will ever be a lawsuit filed against someone watching DVDs using decss? Not bloodly likely.
I wonder if this can already be done with the NX protocol or not?
They cross compile XFree86, but it took a bit of work for that to become possible.
I imagine it isn't possible to cross-compile to very many architectures with the current x.org tree, though I could be wrong.
Well since the hacks were actually on the XFree86 code, and XFree wasn't exactly good about accepting patches, I'd say probably not.
But it is more than patches to XFree86 - they also have a complicated cross-compiling build system that recently got support for the X11 sets. Before they could build just about everything but X11 with this method - now they can build the entire NetBSD release for most every architecture on a single build box from a single source tree.
If they ever do switch to x.org (which is very likely), I'd hope they'd be able to work with the x.org folks to make this an easier task.
That is what I've heard - I just never looked it up myself.
At least with her not being employed, her name will slowly fade away, even if it takes quite a long time due to the "legacy" she's left.
I'm not sure if I agree with your assessment of how she got her job, though I do not know enough about about her career previous to HP to say for sure.
However, as a CEO she was a complete disgrace. As one of the most visible and powerful female CEOs in North America, she probably did a lot more harm to her gender than good by setting such a poor example.
I'm sure there are plenty of "she's" that now wish that she'd never become CEO of HP, whether she deserved a shot at the job or not. I think it was a good day for women today - no longer will the people equate female CEO and Carly Fiorina. She was always the first one to come to mind for me, and it wasn't exactly a positive thought.
This is fine, as long as you're willing to support Linux. If not, you'll have people angry at this OS you told them to install, but doesn't work like they expect. But if you're available to help for the first little while though, it can cut significantly cut down on the problems, and the requests for warez (which I will not honour, especially since I have no need for Win32 software myself).
I've done the Linux conversion successfully for a couple people, but I don't think I'd try it for relatives living in a different city than me, or those that are afraid of their computer. So far I've limited it to people who were fed up with Windows, and really did want to try something else.
For those I don't convert to Linux, I could try to encourage them to by a Mac, but I think that'll be futile too. Probably I'll just burn them the OpenCD or gnuWin, and tell them to use the antivirus software and FireFox.
Though you joke, I do think that Ogg Vorbis is quite possibly a format they could offer. I mean why not, it is DRM free and popular in the Linux crowd.
What is more interesting to me is whether they'll offer it in FLAC format. That way you can convert it to any format you want without losing vast amounts of quality. Since they are pushing the fact that being DRM-less allows device portability, why not give format portability too?
Good point. The usage of CBs is much more conducive to good driving. People are not expecting a quick and fluid conversation - long pauses between responses are typical and don't bother the other person. It is okay to either say "just a minute" or just quit talking to focus on the road for a bit.
The etiquette around CBs was build around the fact you use them in vehicles, and are expected to have to drive as well. The etiquette of phone conversations evolved in houses on land lines, and people expect to be able to give and receive the same amount of attention as on a home phone.
People get mad at you if you're only half paying attention to them on the phone (my parents get annoyed at my brother being on the computer and running MSN conversations while on the phone, and only half listening the them). It is just expected that if you bothered to call someone (or pick up the phone), that you're giving the person on the other side your nearly undivided attention. So we have a tendency to do so, whether it is safe or not.
It probably does help that CBs are primarily used by people who spend a lot of time on the road, so they're likely better drivers to begin with. They're also often used in the country or on the highways, where there are generally less hazards.
I'm quite surprised the terms were made public. The terms of the Coke deal here were sealed, much to the displeasure of those outside the union. Often they'll make it public that they're paying for a scoreboard or some other visible piece of hardware, but otherwise keep the deal secret. I imagine it was a similar deal to what you describe though.
I think the biggest reason they used to seal those deals is so that they could offer different deals to different campuses, and not have them realize they're being ripped off. That, and they probably wanted to prevent a bidding war with the competitor - Pepsi and Coke might start driving up the contract amounts through a bidding war, which really is not good for either of them. Negotiate and sign in secret, so that your opponent can't one-up you.
So like anyone else in IT, you're seeing many of them avoid IE like the plague? Those who have a clue anyhow.
Internal IE-only apps may foil that though, especially at Microsoft.
The university here also does not allow Pepsi machines. I wonder if Coca-Cola pwns us too?
* actually they do have an exclusive beverage contract with the campus, so our lousy union gets paid to allow Coke to price gouge us. Quite common in colleges and high schools now.
I was going to say, I'm not sure I'd consider Cygwin the best of Linux.
It is nice to have if you need it, but there are far better Unixes out there than Cygwin on top of Win32. Heck, there are even better Unixes on top of win32, such as CoLinux.
Though others have given their reasons for Mandrake, I'll chime in anyhow. There are two big reasons why I'd standardize on Mandrake Linux across an entire operation:
Package Management and configuration:
- urpmi is a great package management tool, with a good GUI available, but also a very powerful command line version (which is what I use)
- updates available for at least 18 months, longer for the Enterprise Server
- due to using RPM, 3rd party packages are easier to find (although urpmi contains most everything you'd need, reducing the depedancy on 3rd party packages)
- most of their GUI utilities also work on the command line, due to their use of a perl library that can render to GTK or console
Flexibility:
Mandrake's main download edition, as well as the PowerPacks, can be configured on a wide array of hardware and for a wide range of uses. If you want it as a Desktop, specific-use Workstation, or a Server, you can do it. If you want something more specialized, you have these options:
- MandrakeMove: LiveCD Desktop oriented version of Mandrake
- Discovery: beginner Desktop oriented version
- Enterprise Server: similar to RHEL
- Multi-Network Firewall: firewall, VPN, router, etc
- PowerPack: the works, including commercial packages like Java, and some server GUI wizards if you want that convenience
- Download Edition: only OSS software, free to download and use
- Clustering: I think this is obvious
- versions for i586 and AMD64 for many of the above (PPC versions are sometimes available for the download edition)
These different options give you great flexibility. Here we have a site license, so I put PowerPack on everything from laptops to servers, and have the Clustering edition for the cluster. People who wish to use the same software at home as they do at work can either go with the free download edition, or the fairly cheap and more newbie friendly Discovery or MandrakeMove editions. If you don't have a site license, it is quite possible to put the Download edition on any machines that do not require the commercial software or huge amounts of support. A Club membership an be bought to deal with some support issues if you are only deploying the download edition. A site license will often include a certain number of club memberships, so if you had a few home users running it, they could get some support as well.
They are quite good for support if you ever need it, since they aren't so big a company that they'd ignore their smaller customers (and are small enough that they really pay attention to the big site licenses). But they are also large enough to be profitable as a company, and to have fairly wide usage and support throughout the community.
Everyone thinks of Mandrake as a desktop distro, but it does just as well on the server as Red Hat or Suse in my opinion. Other distros may offer better pure desktop systems (Lindows and Xandros perhaps?), and Debian or the BSDs may be better for some server usage, but I can't think of any Linux or BSD that does many tasks as well as Mandrake does. This is what makes Mandrake so ideal for company/department wide standardization.
They also provide a fair bit more for the power user than most people realize. The easy GUI stuff is there, but there is a lot underneath it for those who don't like point and click. The assumption that you have to graduate to something else less newbie oriented is quite false.
Created due to licensing, but heavily adopted due to people being fed up with XFree86.
The Cygwin folks already had to fork XFree because of the orgainization's refusal to accept patches. And Cygwin was far from being the only ones annoyed with XFree. It was just easier for distros to stick with XFree instead of maintaining their own, and causing a political mess.
The license change was merely the last straw, and was very indicative of how XFree operated. By unilaterally changing the license, then refusing to work with the people who ship their product on fixing it, they showed an even higher level of elitism than before. By this time there was a large enough group X11 developers that were doing great work, but not part of XFree (mainly Keith Packard), that the distros had somewhere else to turn.
So it was more than a simple license change I'm afraid. They kept some of the best developers doing the most innovative work outside of the group, and alienated the very people who distributed their product. Their own elitism made them completely irrelivant in the development of X11, which was supposed to be the entire purpose for XFree's existance.
It is not impossible that this could happen to Firefox too, but right now they are the main drivers in the browser market, and generally are keeping their use base happy. No reason to worry quite yet.
Last I heard it was stable under normal workloads, but could be thrown off by extreme conditions, such as edge conditions when working with 99% full drives and very high workloads.
So in filesystem terms it is considered unstable, but it does indeed work right now. It just doesn't provide assurances and reliability necessary to be part of a mature OS.
Indeed, if anyone can help with the LFS work on NetBSD, it'd be greatly appreciated. Right now there are basically 2 people who look at that code, and it hasn't really seen any work since the summer.
From what I hear, LFS combined with kernel RaidFrame is a killer combination. But until LFS is stablized, nobody will get to see the benefits of this.