Debian lets me get stuff done without worrying about source code or compilation at all.
You do know, don't you, that BSD ports is also the basis for BSD packages? You don't need to compile anything at all. Just use the packages and they'll fit into the same system!
Now that I've thought a bit on this problem, I wouldn't even worry about the FHS. Few actually adhere to it, and those that do liberally reinterpret to their advantage.
Instead, we need to make sure that packages do not contain any dependencies on the filesystem layout. Simple. Do you install KDE under/opt/kde,/usr,/usr/local, or/usr/local/kde? To the package it doesn't matter. The contents of the package are going to be marked as ${PREFIX}/share, ${PREFIX}/lib, etc. It's up to the system as to what PREFIX is. If there's an init script that must be placed in/etc or equivalent, mark as an init script, and the system can decide where to put it (/etc/rc.d instead of/etc/init.d). It doesn't matter if my system installs KDE to/opt/kde while another installs to/usr/local, as long as it remains consistant within my system.
The real bugaboo from the beginning of time has been how to identify dependencies. The traditional RPM way has been to use artifical labels, which breaks down if different systems use different labels, since "kdelibs-3.2_2.rpm is different from "kdelibs-3.2.0.rpm", even though the contents might be identical. This can be a real thorny problem when you're trying to make a package work across distros.
No they can't, and that is a good thing. Documents should be documents, not executables. In the real world, very few people embed VB scripts in documents. You may have a need to write a script, but you won't have a need to embed that script in a document.
Linux and BSD are ready now for the corporate enterprise. Not in ten years, not in ten months, but today!
We have open source group ware, open source office productivity tools, open source infrastructure, and open source just about everything else you want. And professional level support for everything above.
Is there any compelling reason to use ISS instead of Apache? Any compelling reason to use Exchange instead of *mail? Does OpenLDAP somehow not meet your needs? The GUI doesn't count, because you're an enterprise, and you have intelligent professional IT administrators. If they bitch about the lack of GUI, replace them with competent personnel.
Is there any compelling reason to stick with Microsoft Office? How many of your users really need the functionality to embed executable applications in their documents? How many times have you ever run across a real-world document that OO.org won't open? There's no reason to put Word on every desktop just because two people in marketing have a "genuine" need for it.
I've played around a bit with KDE's new Kontact. Why someone would want to use Outlook instead of Kontact is beyond me. Ditto for Evolution. These are applications that work with widely used standards. Eliminate the proprietary Microsoft Exchange standards and you eliminate the need for Outlook. The KDE and GNOME desktops are certainly ready to replace Windows in the enterprise. Maybe they're not quite ready for your grandma, but they're more than ready for the corporate desktop.
And software installation? Anyone wh's had to manage the software on half a dozen or more desktops will realize that package managers are clearly superior to the Windows way.
it's not so easy when you have thousands of users...
Oooh, suddenly the situation changes from a single user sitting at home wringing his hands, to the head of IT at a major corporation. Guess what? You have even MORE responsibility to secure those systems! I have no sympathy for you. You're a professional, start acting like one.
Where do you people come from! Is it time for another application of the ClueStick(tm)!
If you're not using a specific port, close it up. That includes 3127. And everything below 3127, and everything above 3127. Close them ALL off except the ones you are specifically using.
Now I realize that this is extremely difficult to do in Windows, but do it anyway. Repeat, do it anyway. This is your responsibility as the owner of a node on the network. And don't think you're done just because you're secured the firewall. Secure all of your client systems as well. My company got hit hard by Blaster because someone walked into the lab with a laptop.
I'm just about set to purchase a 21" Sony Trinitron. I've got one at work and it's the best display I have ever used. Then I go look at LCD's and they suck. So I want to get the Trinitron before CRTs go the way of component modems and floppy drives.
The grandparent post postulated that evolution is a demonstrable fact. I am not arguing the converse, but merely want to see the evidence of factuality. There's plenty of evidence to support the theory of evolution, but they are not the same thing as demonstrable proof.
And what exactly is this "fact" of evolution? Is it the "fact" that every species evolved from another species, all the way back to a single original species? Does this fact preclude an external "intelligent" mechanism, or is that another element of this "fact"?
That evolution occurs is a fact which can be demonstrated.
To play devil's advocate here, please demonstrate this. Please demonstrate one species evolving from another. I don't want you showing fossils or intra-specie variations, but an actual demonstration an one species deriving from another.
I think you are about as likely to get a pony as you are to get a change in the standard employment agreement.
Ah, cynicism. How refreshing.
But it doesn't reflect the reality. You never got your employment agreement modified, but that's because you never tried. I have. And I did. All you have to do is ask.
If they were setting you up for failure during your initial few weeks, then they wouldn't have hired you to start with. Remove your tinfoil hat and realize that most of the world's stupidity can be explained by mere stupidity.
I would still have gotten out of the company, however. Not giving a new employee the employee handbook is just stupid.
The following is a true story. Only the names of projects and managers have been changed to protect the guilty...
Third Sextant project estimation meeting this week. Previous two meeting have not gotten the engineering man months below eighty, the goal is twenty. It looks impossible. The basic problem is that Sextant is a hard realtime embedded medical imaging system using QNX. Marketing is demanding that we use a new Windows-only workflow management solution from a third party, called "eSynerge". Just the massive redesign to fit Windows into the hardware and software architectures is going to take sixty engineer months.
The directory says, "Let's play make-believe. If you had a magic wand and could change anything about the project, would could we do to lower the resource estimate?"
Me: "We could make a requirement for the eSynerge package to perform all of the software installation."
Director: "Good idea. Unrealistic, since we have no control over eSynerge, but a good idea. Any others?"
Bob: "Require eSynerge to provide a hard realtime NT kernel."
Susan: "Require eSynerge to perform all image processing in realtime."
Mark: "We could make eSynerge do everything. All we would need to do is sit back and write process documents."
Director: "Good, good. Now we're getting creative. Let's see what the magic spreadsheet says. Wow! We got the estimate to under eighteen months! I'm going to go take these number into the vice president."
Us: "What?!?! You said to play make believe! Those are all impossible requests! We were joking!"
Director: "I know, but I'll include all of your 'assumptions' as footnotes. The vice president is going to be so happy to see these. Geez, you guys were really productive today."
You're talking about additional software for Windows. I'm talking about the default out-of-the-box window manager, AKA progman.exe. It is the default "wm" that people claim is the epitome of usability. Frankly, they only consider it usable because they don't know any better.
You mean you don't just drag a folder somewhere and call it done?
Hah! Reminds me of a UI epiphany I had a few years ago. I was trying to get a Windows application bundled up ready to go. I was struggling with InstallShield. This was supposed to be the professional version, but I guess I still had to pay extra for the full functionality. So I was planning to just put everything in a zip archive and distribute it like that. Unzip-and-run, in other words. But I was worried, because that was NOT the Windows way.
So I bitched about it to a Mac friend of mine. "What's wrong with that?" he asked. "That's pretty much what we Mac users do." He was right. So I ended up distributing it as a self-extracting zip file. To date I have not had one complaint.
It's curious that Gnome and KDE based their GUI design template on Windows and not the Mac. Clearly, they're basing their design decisions on bringing a Free Windows to the masses, not a Free MacOS.
Big heavy sigh... Gnome is trying to bring Gnome to the masses, and KDE is trying to bring KDE to the masses. Neither is much interested in a free version of the Windows or Mac desktops.
There is one reason, though, why both KDE and Gnome resemble Windows in behavior and feel: users want them to. I just noticed this recently after the KDE 3.2 release, and the flurry of new "bug" reports. KDE and Gnome users want their desktops to look like Windows. So they file bug reports that saying that some minor thing should be more like Windows. They don't say it explicitly in those words, but they do say it. And the developers listen. Both desktops are evolving towards what the majority of users are most comfortable with, and since the majority of users are currently newly reformed Windows users, the direction is towards a Windows feel.
But we're getting a lot of things that aren't even on Microsoft's horizon. Like tabbed pages on every browser in the world except Internet Explorer. Before Mozilla introduced this, no one requested it. No one. But suddenly a new idea came forth and everyone had to have it. In a similar way, no one out there is lobbying to get rid of window shading in favor of the Microsoft way. They've tried it, liked it, and want to keep it. Or look at either Kicker or the Gnome panel. Both are light years ahead of the Windows taskbar, and no one is asking to go back.
This is the evolution of the desktop in action. For a while we're going to have to endure the Windows default feel. But once we get more intermediate and advanced users than newbies, this will change. We already have the changes waiting in the wings, so to speak, ready for user demand to call them forth.
You know, trying to learn Windows from a UNIX POV, I've found some Windows things that are still archaic as well. This creates a significant impact on the user experience.
There are dozens that I could mention, but the biggest is the window manager. Whatever the name is for the Windows window manager, it does not have snapto or window shading. This is a major annoyance when you have multiple windows up on the screen. Neither does it have easily controlled z-ordering. It is not an easy to use window manager. The look may have improved, but the behavior has not changed from Windows 3.0.
The only reason the public has stuck with Windows as long as it has is simply because they are familiar with it. No other reason.
And what, exactly, is wrong with that? [Windows-only applications]
Oh man, you have a lot to learn! The problem with Windows-only applications is that they only run on Windows.
Maybe in your fantasy world you think utopia will only come about when everyone is forced to use Windows, but in the reality it will never happen. What will happen instead is a polarization if increased incompatibilities.
Portability should be your number one concern. Always. If you can't be portable, clearly document why you can't, and where the unportable parts are. As it now stands,.NET is not a portable framework.
Don't blame this on capitalism. The ABA is about as anti-capitalism as you can get.
Re:Why do big companies want pseudo-compiled langs
on
How C# Was Made
·
· Score: 1
But not _much_ slower.
Says the person with the dual CPU server. From the first day of Java, the proselytizers kept telling me to get faster hardware and wait for the next version. I now have hardware 30 times faster, and four major versions later, and Java STILL crawls on my system. Even purely interpreted languages like Python and Ruby are faster!
Re:"Co-opt Java"
on
How C# Was Made
·
· Score: 3, Insightful
Your comment is a fascinating insight into a fanatical mind.
Fanatical no. Cynicism spawned actions of Microsoft? Maybe.
Despite whatever wonderful attributes C# and.NET have, they do not override the fact that the language and framework are under the control of Microsoft. All the people bitching at Sun's control of Java seem to just look the other way when.NET is mentioned. Who gives a rip that they've submitted it to some standards committee? Do you think Microsoft can't "embrace and extend".NET? Do you think they don't have several dozen submarine patents ready to go?
I'll believe the hype when I see a workable, usable, and complete implementation of.NET on any other platform besides Windows. I can take a Java app developed on Windows and run it on Linux, FreeBSD and Solaris. I cannot do the same for *ANY*.NET application.
Microsoft wants developer writing Windows-only applications..NET is one tool they are using to accomplish this.
I didn't take Spanish in high school, but it was obvious talking to those who did what it was like. The class taught "Spanish", as in the language from Spain. Sixty percent of our school were hispanic who spoke a pidgin "Chicana" as a native language. The never did well in Spanish class.
"nuclear power will give us clean limitless energy, and don't worry, we'll deal with the byproducts later because we'll have the tools by then."
And we did it too! We now have ways to safely dispose of nuclear waste. Unfortunately, the politics of the situation means that we are forced to continue storing it in leaking metal drums...
Screw that! I haven't even gotten past the first step of getting the system to even recognize the USB device on bootup. The BIOS says bootable USB is enabled, but there's nowhere to specify it as the first boot device, and it's never accessed.
All these people saying the floppy is dead because you can boot from your USB thumb drive are blowing smoke out their ass. I seriously don't think it can be done.
Things already work out of the box with Windows Explorer. With KDE, I would have to delve into obscure config utilities to make it usable.
Install Windows XP, and two hours and three reboots (four if you want a decent video driver), Windows Explorer works out of the box. Install SuSE Linux, and twenty minutes later with one reboot, Konqueror works out of the box. What's the problem?
Slashdot Mirror
Debian lets me get stuff done without worrying about source code or compilation at all.
You do know, don't you, that BSD ports is also the basis for BSD packages? You don't need to compile anything at all. Just use the packages and they'll fit into the same system!
Now that I've thought a bit on this problem, I wouldn't even worry about the FHS. Few actually adhere to it, and those that do liberally reinterpret to their advantage.
/opt/kde, /usr, /usr/local, or /usr/local/kde? To the package it doesn't matter. The contents of the package are going to be marked as ${PREFIX}/share, ${PREFIX}/lib, etc. It's up to the system as to what PREFIX is. If there's an init script that must be placed in /etc or equivalent, mark as an init script, and the system can decide where to put it (/etc/rc.d instead of /etc/init.d). It doesn't matter if my system installs KDE to /opt/kde while another installs to /usr/local, as long as it remains consistant within my system.
Instead, we need to make sure that packages do not contain any dependencies on the filesystem layout. Simple. Do you install KDE under
The real bugaboo from the beginning of time has been how to identify dependencies. The traditional RPM way has been to use artifical labels, which breaks down if different systems use different labels, since "kdelibs-3.2_2.rpm is different from "kdelibs-3.2.0.rpm", even though the contents might be identical. This can be a real thorny problem when you're trying to make a package work across distros.
No they can't, and that is a good thing. Documents should be documents, not executables. In the real world, very few people embed VB scripts in documents. You may have a need to write a script, but you won't have a need to embed that script in a document.
Linux and BSD are ready now for the corporate enterprise. Not in ten years, not in ten months, but today!
We have open source group ware, open source office productivity tools, open source infrastructure, and open source just about everything else you want. And professional level support for everything above.
Is there any compelling reason to use ISS instead of Apache? Any compelling reason to use Exchange instead of *mail? Does OpenLDAP somehow not meet your needs? The GUI doesn't count, because you're an enterprise, and you have intelligent professional IT administrators. If they bitch about the lack of GUI, replace them with competent personnel.
Is there any compelling reason to stick with Microsoft Office? How many of your users really need the functionality to embed executable applications in their documents? How many times have you ever run across a real-world document that OO.org won't open? There's no reason to put Word on every desktop just because two people in marketing have a "genuine" need for it.
I've played around a bit with KDE's new Kontact. Why someone would want to use Outlook instead of Kontact is beyond me. Ditto for Evolution. These are applications that work with widely used standards. Eliminate the proprietary Microsoft Exchange standards and you eliminate the need for Outlook. The KDE and GNOME desktops are certainly ready to replace Windows in the enterprise. Maybe they're not quite ready for your grandma, but they're more than ready for the corporate desktop.
And software installation? Anyone wh's had to manage the software on half a dozen or more desktops will realize that package managers are clearly superior to the Windows way.
So start migrating!
it's not so easy when you have thousands of users...
Oooh, suddenly the situation changes from a single user sitting at home wringing his hands, to the head of IT at a major corporation. Guess what? You have even MORE responsibility to secure those systems! I have no sympathy for you. You're a professional, start acting like one.
Where do you people come from! Is it time for another application of the ClueStick(tm)!
If you're not using a specific port, close it up. That includes 3127. And everything below 3127, and everything above 3127. Close them ALL off except the ones you are specifically using.
Now I realize that this is extremely difficult to do in Windows, but do it anyway. Repeat, do it anyway. This is your responsibility as the owner of a node on the network. And don't think you're done just because you're secured the firewall. Secure all of your client systems as well. My company got hit hard by Blaster because someone walked into the lab with a laptop.
I'm just about set to purchase a 21" Sony Trinitron. I've got one at work and it's the best display I have ever used. Then I go look at LCD's and they suck. So I want to get the Trinitron before CRTs go the way of component modems and floppy drives.
The grandparent post postulated that evolution is a demonstrable fact. I am not arguing the converse, but merely want to see the evidence of factuality. There's plenty of evidence to support the theory of evolution, but they are not the same thing as demonstrable proof.
And what exactly is this "fact" of evolution? Is it the "fact" that every species evolved from another species, all the way back to a single original species? Does this fact preclude an external "intelligent" mechanism, or is that another element of this "fact"?
From your article: "60% responded...", "half replied...". In other words, a self-selecting survey. Demonstrates nothing.
That evolution occurs is a fact which can be demonstrated.
To play devil's advocate here, please demonstrate this. Please demonstrate one species evolving from another. I don't want you showing fossils or intra-specie variations, but an actual demonstration an one species deriving from another.
I think you are about as likely to get a pony as you are to get a change in the standard employment agreement.
Ah, cynicism. How refreshing.
But it doesn't reflect the reality. You never got your employment agreement modified, but that's because you never tried. I have. And I did. All you have to do is ask.
If they were setting you up for failure during your initial few weeks, then they wouldn't have hired you to start with. Remove your tinfoil hat and realize that most of the world's stupidity can be explained by mere stupidity.
I would still have gotten out of the company, however. Not giving a new employee the employee handbook is just stupid.
The following is a true story. Only the names of projects and managers have been changed to protect the guilty...
Third Sextant project estimation meeting this week. Previous two meeting have not gotten the engineering man months below eighty, the goal is twenty. It looks impossible. The basic problem is that Sextant is a hard realtime embedded medical imaging system using QNX. Marketing is demanding that we use a new Windows-only workflow management solution from a third party, called "eSynerge". Just the massive redesign to fit Windows into the hardware and software architectures is going to take sixty engineer months.
The directory says, "Let's play make-believe. If you had a magic wand and could change anything about the project, would could we do to lower the resource estimate?"
Me: "We could make a requirement for the eSynerge package to perform all of the software installation."
Director: "Good idea. Unrealistic, since we have no control over eSynerge, but a good idea. Any others?"
Bob: "Require eSynerge to provide a hard realtime NT kernel."
Susan: "Require eSynerge to perform all image processing in realtime."
Mark: "We could make eSynerge do everything. All we would need to do is sit back and write process documents."
Director: "Good, good. Now we're getting creative. Let's see what the magic spreadsheet says. Wow! We got the estimate to under eighteen months! I'm going to go take these number into the vice president."
Us: "What?!?! You said to play make believe! Those are all impossible requests! We were joking!"
Director: "I know, but I'll include all of your 'assumptions' as footnotes. The vice president is going to be so happy to see these. Geez, you guys were really productive today."
You're talking about additional software for Windows. I'm talking about the default out-of-the-box window manager, AKA progman.exe. It is the default "wm" that people claim is the epitome of usability. Frankly, they only consider it usable because they don't know any better.
You mean you don't just drag a folder somewhere and call it done?
Hah! Reminds me of a UI epiphany I had a few years ago. I was trying to get a Windows application bundled up ready to go. I was struggling with InstallShield. This was supposed to be the professional version, but I guess I still had to pay extra for the full functionality. So I was planning to just put everything in a zip archive and distribute it like that. Unzip-and-run, in other words. But I was worried, because that was NOT the Windows way.
So I bitched about it to a Mac friend of mine. "What's wrong with that?" he asked. "That's pretty much what we Mac users do." He was right. So I ended up distributing it as a self-extracting zip file. To date I have not had one complaint.
It's curious that Gnome and KDE based their GUI design template on Windows and not the Mac. Clearly, they're basing their design decisions on bringing a Free Windows to the masses, not a Free MacOS.
Big heavy sigh... Gnome is trying to bring Gnome to the masses, and KDE is trying to bring KDE to the masses. Neither is much interested in a free version of the Windows or Mac desktops.
There is one reason, though, why both KDE and Gnome resemble Windows in behavior and feel: users want them to. I just noticed this recently after the KDE 3.2 release, and the flurry of new "bug" reports. KDE and Gnome users want their desktops to look like Windows. So they file bug reports that saying that some minor thing should be more like Windows. They don't say it explicitly in those words, but they do say it. And the developers listen. Both desktops are evolving towards what the majority of users are most comfortable with, and since the majority of users are currently newly reformed Windows users, the direction is towards a Windows feel.
But we're getting a lot of things that aren't even on Microsoft's horizon. Like tabbed pages on every browser in the world except Internet Explorer. Before Mozilla introduced this, no one requested it. No one. But suddenly a new idea came forth and everyone had to have it. In a similar way, no one out there is lobbying to get rid of window shading in favor of the Microsoft way. They've tried it, liked it, and want to keep it. Or look at either Kicker or the Gnome panel. Both are light years ahead of the Windows taskbar, and no one is asking to go back.
This is the evolution of the desktop in action. For a while we're going to have to endure the Windows default feel. But once we get more intermediate and advanced users than newbies, this will change. We already have the changes waiting in the wings, so to speak, ready for user demand to call them forth.
You know, trying to learn Windows from a UNIX POV, I've found some Windows things that are still archaic as well. This creates a significant impact on the user experience.
There are dozens that I could mention, but the biggest is the window manager. Whatever the name is for the Windows window manager, it does not have snapto or window shading. This is a major annoyance when you have multiple windows up on the screen. Neither does it have easily controlled z-ordering. It is not an easy to use window manager. The look may have improved, but the behavior has not changed from Windows 3.0.
The only reason the public has stuck with Windows as long as it has is simply because they are familiar with it. No other reason.
And what, exactly, is wrong with that? [Windows-only applications]
.NET is not a portable framework.
Oh man, you have a lot to learn! The problem with Windows-only applications is that they only run on Windows.
Maybe in your fantasy world you think utopia will only come about when everyone is forced to use Windows, but in the reality it will never happen. What will happen instead is a polarization if increased incompatibilities.
Portability should be your number one concern. Always. If you can't be portable, clearly document why you can't, and where the unportable parts are. As it now stands,
Don't blame this on capitalism. The ABA is about as anti-capitalism as you can get.
But not _much_ slower.
Says the person with the dual CPU server. From the first day of Java, the proselytizers kept telling me to get faster hardware and wait for the next version. I now have hardware 30 times faster, and four major versions later, and Java STILL crawls on my system. Even purely interpreted languages like Python and Ruby are faster!
Your comment is a fascinating insight into a fanatical mind.
.NET have, they do not override the fact that the language and framework are under the control of Microsoft. All the people bitching at Sun's control of Java seem to just look the other way when .NET is mentioned. Who gives a rip that they've submitted it to some standards committee? Do you think Microsoft can't "embrace and extend" .NET? Do you think they don't have several dozen submarine patents ready to go?
.NET on any other platform besides Windows. I can take a Java app developed on Windows and run it on Linux, FreeBSD and Solaris. I cannot do the same for *ANY* .NET application.
.NET is one tool they are using to accomplish this.
Fanatical no. Cynicism spawned actions of Microsoft? Maybe.
Despite whatever wonderful attributes C# and
I'll believe the hype when I see a workable, usable, and complete implementation of
Microsoft wants developer writing Windows-only applications.
I didn't take Spanish in high school, but it was obvious talking to those who did what it was like. The class taught "Spanish", as in the language from Spain. Sixty percent of our school were hispanic who spoke a pidgin "Chicana" as a native language. The never did well in Spanish class.
"nuclear power will give us clean limitless energy, and don't worry, we'll deal with the byproducts later because we'll have the tools by then."
And we did it too! We now have ways to safely dispose of nuclear waste. Unfortunately, the politics of the situation means that we are forced to continue storing it in leaking metal drums...
Screw that! I haven't even gotten past the first step of getting the system to even recognize the USB device on bootup. The BIOS says bootable USB is enabled, but there's nowhere to specify it as the first boot device, and it's never accessed.
All these people saying the floppy is dead because you can boot from your USB thumb drive are blowing smoke out their ass. I seriously don't think it can be done.
Things already work out of the box with Windows Explorer. With KDE, I would have to delve into obscure config utilities to make it usable.
Install Windows XP, and two hours and three reboots (four if you want a decent video driver), Windows Explorer works out of the box. Install SuSE Linux, and twenty minutes later with one reboot, Konqueror works out of the box. What's the problem?
IT JUST WORKS!