Re:"It's not Microsoft"
on
LinuxWorld
·
· Score: 2
if we can't do anything more than mimmic the new features that Microsoft is coming out with,
Although we know that this is a rash and incorrect generalization, I'll address it as if it were correct:
What are we adding? In short, commoditization. Even if Linux were just another Windows (which it isn't), it's free. It commoditizes the operating system, just as TCP/IP commoditized wire protocol and Ethernet commoditized the wire itself. Unseating the monopoly and putting control of the operating system back into the hands of the industry and community is a worthy goal, in and of itself.
Besides, who knows how many slick new innovations could have happened over the last decade if Microsoft hadn't been there to destroy them? We're creating a level playing field, one on which true innovation can happen, whether or not it agrees with the One Microsoft Way. --
I wonder what the percentage of NetWare servers is these days in corporate networks.
Higher than you think. Generally speaking, the bigger the corporation, the more likely they are to still be using Netware. Netware scales beautifully, and in the big shops, there's less of a knee-jerk tendency to rip out working systems and replace them with whatever is trendy.
It's in the small to midrange installations where you see the techno cowboy wannabes who know how to double-click and therefore consider themselves network experts, that Windows NT replaces Netware rapidly. --
If the rumor is true, IBM may be looking to do what we all know Novell needs to do to stay alive, yet doesn't have the guts to do: open-source the Novell Directory. What have they got to lose? Market share? That's trickling away already: Windows NT and Linux are absolutely clobbering Netware in the marketplace. Novell's crown jewels right now are NDS. NDS, a robust, scalable, portable directory service. NDS, a directory that's proven in Fortune 500 installations. NDS, a directory that'll get clobbered once Microsoft Active Directory begins to take hold.
Novell needs to open-source NDS now, before it becomes irrelevant. IBM has the guts to do this, and the bucks to maintain the software as a loss leader to continue to sell services around it. If NDS is open sourced, it will become ubiquitous almost overnight, especially on Linux systems, where it will become the standard directory service. If they choose the GPL, even the Debian folks will adopt it. Imagine being able to write directory-aware applications for Linux, knowing that NDS will always be there.
That's my $0.02, anyway. Personally I think the rumor is bogus. --
The beauty of open source is that you always have the ability to yank pieces out and replace them with whatever you prefer. For the uninitiated, it is actually quite helpful for the 'vanilla' interface to be completely standardized. Newbies like the fact that they can sit down in front of any Windows or Macintosh computer and expect that it'll look like all the rest. Philosophize all you want about choice, but newbies don't see it as choice, they see it as inconsistency and it usually scares them away.
Bring the newbies in with a standardized interface, then let the power users customize their systems. --
Things are moving in the right direction, and in a way that I had imagined a while ago. Didn't Sun announce that they were going to Bonobize the new version of StarOffice? That accomplishes two things:
It provides mature office suite components for the desktop
It throws a lot of weight behind the GNOME desktop and component model
I'm personally quite neutral in the imaginary 'desktop war' but I think it's possible that all this weight from corporate backers might be what it takes to finally put it to rest. --
Microsoft wants everything to be Win32. Pure Win32, no legacy Win16 or MS-DOS. Do we have a problem with that? Of course not! If everything is pure Win32, then the WINE developers can focus on Win32 instead of having to figure out and implement a bunch of legacy cruft.
WinME is still a 32/16 hybrid, and we know it; they've merely put heavier drapes over the DOS layer. Again, so what? You can dual-boot using a 'real' boot manager, the same way you'd do if you had Windows NT and Linux on the same computer. Or you can use WinME as an opportunity to take the plunge and ditch Windows altogether. --
I'd like to congratulate these fine developers. Not because of the technical merit of the project (though it does have merit), but because instead of whining about "X sucks" or "X needs to be replaced" or "Berlin Berlin Berlin Berlin Berlin", they're actually doing something to improve X. Kudos to them. This is how results are delivered in the open source world.
"Just do it." -- Linus Torvalds (paraphrasing Nike)
Seriously, where were all these naysayers when Microsoft was more than two years late shipping Windows NT 5.0 (nee Windows 2000)?
Feature creep, delays, more delays, more feature creep. The product eventually shipped, and people are installing it. Mozilla deserves at least the same level of patience. --
I used to get a lot of use out of my Sniffer (the original one from Network General). I solved a lot of network problems on both ethernet and token ring with that tool. Unfortunately, I don't get to use it much, anymore. Ethernet switches are cheap nowadays, and as a result, networks are more finely segmented, sometimes down to the single node level. Switches, working as designed, filter out all the traffic that's not explicitly unicast towards the sniffer's MAC address (or broadcast, of course). As a result, you don't get the whole picture of what's going on with your network.
Some switches can be programmed to put a port into 'diagnostic mode' (forward all packets to this port because there's a sniffer there) but it's usually more trouble than it's worth, especially when you have a large building with a dozen or more switches.
That said, I'll probably still try out Ethereal. For the times that I still can make use of a sniffer, it'll be nice to get that DOS partition off my laptop. --
Assuming this boycott begins to become successful, the RIAA will simply turn around and say "See what we told you? Record sales are down because of Napster users." --
Personally I like KDE because I can forget I'm using it.
Exactly. In software, as with many other fields, good engineering is whatever gets the job done without calling attention to itself. KDE does just that. It's good-looking, it's fast, it's stable, it's functional, and it doesn't get in your way. It's a tool that allows the computer to work for you instead of vice-versa. Even though I'm a hacker type, I still despise having to type half a dozen shell commands when there's a program out there that'll get the same task done in two mouse clicks.
I applaud the KDE developers and wish them every success. They deserve it. --
Every single flaw in the Mozilla project can be blamed on Jamie Zawinski (JWZ). In fact, JWZ is actually a shill, secretly paid by Microsoft to sabotage the Mozilla effort. --
But using Gecko to render the UI was an incredibly X Window-centric decision. Users on other platforms will not put up with it. Mozilla is dead on non-Unix platforms unless some insane person wraps a new app shell around it.
Yeah, just like that crappy old WinAmp MP3 player. It didn't use native Win32 widgets, and as a result, nobody ever used it, right? --
For a moment, let's put the window manager MDI issue aside. It'd take no more than a few hours for someone to simply shut off the 'taskbar' and let StarOffice run in a more conventional MDI model, like ClarisWorks and Opera do. I want to talk about something else.
The real advantage of StarOffice is its tight integration with Java. Despite the fact that it's not written in Java, it can utilize Java in clever ways. And it uses whatever VM you have on your system rather than carrying its own around, which is of course another advantage.
Recently, I was faced with the task of building a small database for a boring corporate type task. While this could have been done in MS Access or whatever, I wanted to go cross-platform, client/server, and 100 percent Microsoft-free. So here's what I ended up doing: MySQL on the back end, and StarOffice Database on the front end. But there's no MySQL support in StarOffice? True, but there is JDBC support. I located a JDBC driver for MySQL, plugged it in, and everything started working. This may not impress anyone until you come to the realization that no platform-dependent code was written!
And therein, I believe, lies the real power of an office suite that is tightly integrated with Java. Java becomes the 'glue' that pulls various pieces of architecture together. Java becomes the scripting language. Java becomes the language to write StarOffice plugins. This is all good stuff, because all third-party StarOffice stuff is automatically platform-independent.
A couple of side notes: I think that two things would benefit StarOffice in the short term: first, the built-in web browser should be an embedded Mozilla 1.0 (when it's eventually released); and second, Sun should get super aggressive about bundling free copies of StarOffice with new PC's -- not just Linux machines, but Windows machines as well. With a free 'good enough' office suite in their hands, many users wouldn't bother spending the 500 bucks on another suite.
Yes, this is the standard response to the litany of "X sucks" rants, and it is the appropriate one. Users flocked to Windows 3.1 despite its horrifying lack of technical sophistication, because there were lots of applications which ran on it. Application support is the #1 impediment to quick adoption of Linux on the desktop. You've heard this before. You've heard it too much. The bottom line is that it's true. Get the apps out there and the architecture will follow. --
Linux as the new Unix, officially?
on
Endgame For SCO
·
· Score: 1
Doesn't SCO currently own the rights to the 'official' Unix, as in the main source tree descended from Bell Labs?
If so, perhaps it's time to take that code, GPL it, and merge the useful bits into Linux (both the kernel and the pieces that make up a complete OS). If this were done, Linux would become the new Unix(tm), not only in practice but titularly as well. This closes the story in a positive light. Instead of the history books recounting that 'Linux killed the Unix market' we would read that 'Linux was such a strong proposition that Unix was merged into it'. Interesting thoughts. --
A friend of mine observed many years ago that the primary appeal of alternative media is that it bypasses the traditional hierarchy. At the time he was speaking of 'zines, and also of BBS's (in fact, he made this comment on UNCENSORED! BBS, a small online community which I still operate today). This sentiment holds just as true today for the Internet -- perhaps even moreso. Everyone is a potential publisher. When more bandwidth becomes available, everyone will be a potential broadcaster as well.
Bypassing the traditional hierarchy. That's what it's all about. Undoubtedly this scares the daylights out of those who control the traditional hierarchy. --
Self-install will become a far more widespread option when the so-called 'Splitterless' ADSL becomes more widespread. It is called so because, as the name suggests, no splitter is necessary at the demarc. The 'modem' extracts the necessary frequencies for its own use and ignores the POTS bandwidth. The practical upshot of all this is that you can plug the 'modem' into any phone jack.
Splitterless wiring, combined with a USB interface to the computer, results in an installation that even a brain-dead Windroid can complete. Look for this to become the norm. --
This kind of thing would have been useful a decade ago. At this point, multicast technology is maturing to the point where we'll soon be able to have nearly infinite digital media capability without the ugly overhead of having to transmit a separate stream to each listener/viewer. At that point, everyone is a potential broadcaster, and the only difference between NBC and Joe Sixpak with a video camera is that the former has a bigger advertising budget. --
So... what happens if, say, half of the Linux webservers switch to Tux over the next year or so? Do these webservers report as Tux servers, or Apache servers with a Tux kernel accelerator installed? The former could be a problem for purely stupid reasons: if 'Apache' held 30% market share, and 'Tux' held 30% market share, Microsoft would immediately claim victory in the 'web server war' -- as a result, all of the 'go with the market leader' people would begin installing IIS. --
It would be interesting to see these benchmarks run twice, once with single-NIC servers and once with multiple-NIC servers. Windows NT2K aced the Mindcraft benchmarks by effectively coupling CPU's and NIC's. (One may wonder whether the Mindcraft tests were also run with single NIC machines, and the benchmark's sponsor [MS] elected not to have those results published.) Linux's IP stack has been continually improved since Mindcraft, so it would be nice to see how far the kernel developers have come since then. --
Slashdot Mirror
What are we adding? In short, commoditization. Even if Linux were just another Windows (which it isn't), it's free. It commoditizes the operating system, just as TCP/IP commoditized wire protocol and Ethernet commoditized the wire itself. Unseating the monopoly and putting control of the operating system back into the hands of the industry and community is a worthy goal, in and of itself.
Besides, who knows how many slick new innovations could have happened over the last decade if Microsoft hadn't been there to destroy them? We're creating a level playing field, one on which true innovation can happen, whether or not it agrees with the One Microsoft Way.
--
Hmm. I wonder ... if you install AOL for Linux, does it 'accidentally' un-configure your eth0 and ppp0 interfaces, leaving only 'aol0' active?
--
It's in the small to midrange installations where you see the techno cowboy wannabes who know how to double-click and therefore consider themselves network experts, that Windows NT replaces Netware rapidly.
--
If the rumor is true, IBM may be looking to do what we all know Novell needs to do to stay alive, yet doesn't have the guts to do: open-source the Novell Directory. What have they got to lose? Market share? That's trickling away already: Windows NT and Linux are absolutely clobbering Netware in the marketplace. Novell's crown jewels right now are NDS. NDS, a robust, scalable, portable directory service. NDS, a directory that's proven in Fortune 500 installations. NDS, a directory that'll get clobbered once Microsoft Active Directory begins to take hold.
Novell needs to open-source NDS now, before it becomes irrelevant. IBM has the guts to do this, and the bucks to maintain the software as a loss leader to continue to sell services around it. If NDS is open sourced, it will become ubiquitous almost overnight, especially on Linux systems, where it will become the standard directory service. If they choose the GPL, even the Debian folks will adopt it. Imagine being able to write directory-aware applications for Linux, knowing that NDS will always be there.
That's my $0.02, anyway. Personally I think the rumor is bogus.
--
The beauty of open source is that you always have the ability to yank pieces out and replace them with whatever you prefer. For the uninitiated, it is actually quite helpful for the 'vanilla' interface to be completely standardized. Newbies like the fact that they can sit down in front of any Windows or Macintosh computer and expect that it'll look like all the rest. Philosophize all you want about choice, but newbies don't see it as choice, they see it as inconsistency and it usually scares them away.
Bring the newbies in with a standardized interface, then let the power users customize their systems.
--
I'm personally quite neutral in the imaginary 'desktop war' but I think it's possible that all this weight from corporate backers might be what it takes to finally put it to rest.
--
Microsoft wants everything to be Win32. Pure Win32, no legacy Win16 or MS-DOS. Do we have a problem with that? Of course not! If everything is pure Win32, then the WINE developers can focus on Win32 instead of having to figure out and implement a bunch of legacy cruft.
WinME is still a 32/16 hybrid, and we know it; they've merely put heavier drapes over the DOS layer. Again, so what? You can dual-boot using a 'real' boot manager, the same way you'd do if you had Windows NT and Linux on the same computer. Or you can use WinME as an opportunity to take the plunge and ditch Windows altogether.
--
I'd like to congratulate these fine developers. Not because of the technical merit of the project (though it does have merit), but because instead of whining about "X sucks" or "X needs to be replaced" or "Berlin Berlin Berlin Berlin Berlin", they're actually doing something to improve X. Kudos to them. This is how results are delivered in the open source world.
"Just do it." -- Linus Torvalds (paraphrasing Nike)
"Shut up and show me the code." -- Eric Raymond
--
Seriously, where were all these naysayers when Microsoft was more than two years late shipping Windows NT 5.0 (nee Windows 2000)?
Feature creep, delays, more delays, more feature creep. The product eventually shipped, and people are installing it. Mozilla deserves at least the same level of patience.
--
I used to get a lot of use out of my Sniffer (the original one from Network General). I solved a lot of network problems on both ethernet and token ring with that tool. Unfortunately, I don't get to use it much, anymore. Ethernet switches are cheap nowadays, and as a result, networks are more finely segmented, sometimes down to the single node level. Switches, working as designed, filter out all the traffic that's not explicitly unicast towards the sniffer's MAC address (or broadcast, of course). As a result, you don't get the whole picture of what's going on with your network.
Some switches can be programmed to put a port into 'diagnostic mode' (forward all packets to this port because there's a sniffer there) but it's usually more trouble than it's worth, especially when you have a large building with a dozen or more switches.
That said, I'll probably still try out Ethereal. For the times that I still can make use of a sniffer, it'll be nice to get that DOS partition off my laptop.
--
The above message was moderated down in error. It should not be "-1, Troll" but rather "2, Funny"
Thank you.
--
Assuming this boycott begins to become successful, the RIAA will simply turn around and say "See what we told you? Record sales are down because of Napster users."
--
The word "Genome" sounds too much like "GNOME" to me. Therefore I detest Hunkapiller because KDE is much, much better.
--
I applaud the KDE developers and wish them every success. They deserve it.
--
How about we just kill JWZ? That would be a big help to Mozilla.
--
Every single flaw in the Mozilla project can be blamed on Jamie Zawinski (JWZ). In fact, JWZ is actually a shill, secretly paid by Microsoft to sabotage the Mozilla effort.
--
But using Gecko to render the UI was an incredibly X Window-centric decision. Users on other platforms will not put up with it. Mozilla is dead on non-Unix platforms unless some insane person wraps a new app shell around it.
Yeah, just like that crappy old WinAmp MP3 player. It didn't use native Win32 widgets, and as a result, nobody ever used it, right?
--
For a moment, let's put the window manager MDI issue aside. It'd take no more than a few hours for someone to simply shut off the 'taskbar' and let StarOffice run in a more conventional MDI model, like ClarisWorks and Opera do. I want to talk about something else.
The real advantage of StarOffice is its tight integration with Java. Despite the fact that it's not written in Java, it can utilize Java in clever ways. And it uses whatever VM you have on your system rather than carrying its own around, which is of course another advantage.
Recently, I was faced with the task of building a small database for a boring corporate type task. While this could have been done in MS Access or whatever, I wanted to go cross-platform, client/server, and 100 percent Microsoft-free. So here's what I ended up doing: MySQL on the back end, and StarOffice Database on the front end. But there's no MySQL support in StarOffice? True, but there is JDBC support. I located a JDBC driver for MySQL, plugged it in, and everything started working. This may not impress anyone until you come to the realization that no platform-dependent code was written!
And therein, I believe, lies the real power of an office suite that is tightly integrated with Java. Java becomes the 'glue' that pulls various pieces of architecture together. Java becomes the scripting language. Java becomes the language to write StarOffice plugins. This is all good stuff, because all third-party StarOffice stuff is automatically platform-independent.
A couple of side notes: I think that two things would benefit StarOffice in the short term: first, the built-in web browser should be an embedded Mozilla 1.0 (when it's eventually released); and second, Sun should get super aggressive about bundling free copies of StarOffice with new PC's -- not just Linux machines, but Windows machines as well. With a free 'good enough' office suite in their hands, many users wouldn't bother spending the 500 bucks on another suite.
--
Yes, this is the standard response to the litany of "X sucks" rants, and it is the appropriate one. Users flocked to Windows 3.1 despite its horrifying lack of technical sophistication, because there were lots of applications which ran on it. Application support is the #1 impediment to quick adoption of Linux on the desktop. You've heard this before. You've heard it too much. The bottom line is that it's true. Get the apps out there and the architecture will follow.
--
Doesn't SCO currently own the rights to the 'official' Unix, as in the main source tree descended from Bell Labs?
If so, perhaps it's time to take that code, GPL it, and merge the useful bits into Linux (both the kernel and the pieces that make up a complete OS). If this were done, Linux would become the new Unix(tm), not only in practice but titularly as well. This closes the story in a positive light. Instead of the history books recounting that 'Linux killed the Unix market' we would read that 'Linux was such a strong proposition that Unix was merged into it'. Interesting thoughts.
--
A friend of mine observed many years ago that the primary appeal of alternative media is that it bypasses the traditional hierarchy . At the time he was speaking of 'zines, and also of BBS's (in fact, he made this comment on UNCENSORED! BBS, a small online community which I still operate today). This sentiment holds just as true today for the Internet -- perhaps even moreso. Everyone is a potential publisher. When more bandwidth becomes available, everyone will be a potential broadcaster as well.
Bypassing the traditional hierarchy. That's what it's all about. Undoubtedly this scares the daylights out of those who control the traditional hierarchy.
--
Self-install will become a far more widespread option when the so-called 'Splitterless' ADSL becomes more widespread. It is called so because, as the name suggests, no splitter is necessary at the demarc. The 'modem' extracts the necessary frequencies for its own use and ignores the POTS bandwidth. The practical upshot of all this is that you can plug the 'modem' into any phone jack.
Splitterless wiring, combined with a USB interface to the computer, results in an installation that even a brain-dead Windroid can complete. Look for this to become the norm.
--
This kind of thing would have been useful a decade ago. At this point, multicast technology is maturing to the point where we'll soon be able to have nearly infinite digital media capability without the ugly overhead of having to transmit a separate stream to each listener/viewer. At that point, everyone is a potential broadcaster, and the only difference between NBC and Joe Sixpak with a video camera is that the former has a bigger advertising budget.
--
So ... what happens if, say, half of the Linux webservers switch to Tux over the next year or so? Do these webservers report as Tux servers, or Apache servers with a Tux kernel accelerator installed? The former could be a problem for purely stupid reasons: if 'Apache' held 30% market share, and 'Tux' held 30% market share, Microsoft would immediately claim victory in the 'web server war' -- as a result, all of the 'go with the market leader' people would begin installing IIS.
--
It would be interesting to see these benchmarks run twice, once with single-NIC servers and once with multiple-NIC servers. Windows NT2K aced the Mindcraft benchmarks by effectively coupling CPU's and NIC's. (One may wonder whether the Mindcraft tests were also run with single NIC machines, and the benchmark's sponsor [MS] elected not to have those results published.) Linux's IP stack has been continually improved since Mindcraft, so it would be nice to see how far the kernel developers have come since then.
--