Here's one of my biggest reasons for using Linux. Control. You are the master of your domain, not M$. In fact, if you want to tinker with the OS, go for it.
I'd also be wondering about a company that sends back data (who knows what) when you register the OS. There was a big brouhaha about this a while back. Also, and I haven't verified this, but a colleague running XP claims that while he is surfing the net, the OS is sending data to m$.com. Now, he hasn't run a dump on the connection, but I'd be interested if anyone has any details on this, could be just a config issue, or just totally wrong, but (and here's the rub) I wouldn't put it past M$ to track every surfer's habits similar to doubleclick - this is valuable information.
Yeah, I think Windows has improved a bit, especially now with the competition from Linux. Look at their divulging the source code to the Chinese gov. and you get an idea of the tremendous influence Linux has had. Also, it's interesting to see the new Server 2003 ads the expound on the money saved. How you save money when you are spending not only for the license, new hardware required, cost to test and to upgrade your network etc etc - but hey that's marketing, always a schmooze. By the time you even begin to recoup your investment, there's Server 2005, after all that's the business model. I recall a while back when M$ was trumpeting some feature which was essentially something that had always been in Unix (ln -s I think, but according to M$ this was their extraordinary innovation).
So if you're a Windows Developer in a Windows World, but interested in peering into the skies at the other planets rest assured you can dabble with since, unlike M$, many popular tools are available for Linux are just as available for Windows since as open source they run happily on NT as on Unix. Furthermore, you can try cygwin or even Knoppix. As for VS studio, dunno, although you can check out the MONO project. There have been some threads recently of running VisualFoxPro on Linux using Wine, so maybe VB too... Better yet, take a look at the Eclipse project from which you could write any sort of scripts, PHP or whatever. KDevelop is amazing also - very familiar to the windows developer, don't fret, the Linux world is not requiring you to learn vi or Emacs..
Also, the new Linux distros are *very* cool in that much is already installed and configured for you. For example, I went to install mysql on my Mandrake box, when I realized I already had a version installed and running. One of the nice features of Mandrake which allows you to select the type of server you are running, i.e. database, web, games, file server, and install the appropriate packages.
Therefore, as far as installing Linux and being productive, you'd be surprised at how easy it is. In fact, check some comparisons at cnet, you might find that Windows config is actually more of a pain than Linux, my take is that this is because Linux is almost always installed rather than pre-installed like Windows. So by nature it has to be good, if not better than Windows. My personal experience with this has been that Windows (at least older versions, not sure about XP) is that it paints you into a corner and can be twice as frustrating as Linux upgrades.
Games. Linux could use some improvement here, but then you could get a PS2 or xbox, too right?
Anyway, if you're interested in the vast expanding worlds of open source, grab a copy of SuSE, RedHat or Mandrake. You won't be disappointed.
Yeah, C# is out, there's a flicker of interest perhaps, but fleeting, who wants to go backwards? Too little, too late. I used to browse MSDN, eagerly await the latest copy of VS, get wrapped up in COM, and well, years have past and I haven't looked back since.
Gads, you're right. He gave the software out for free, I think it was even open source. Sold his soul like Anders Hellsberj of Delphi fame?
I'm really wondering what folks like Miquel de Icaza are thinking spending the time writing MONO. Perhaps useful as a C# clone, but I'm sure Microsoft will use a similar argument for running C# bytecode on non-Microsoft platforms such as MONO.
Here's an excert from The Register:
"Prior to the demonstration, Hentzen received a call from Ken Levy, Microsoft's Visual FoxPro marketing manager, telling him that he would be in violation of the EULA (End User Licensing Agreement) for VFP if he demonstrated (or ran) the development tool on Linux."
I mean, which is more preposterous, going from Dos->Windows->NT, or going from Minix->Linux?
My copy of Silberschatz & Galvin's "Operating Systems Concepts" isn't too detailed, but I do recall that in addition to the Vax/VMS heritage (Dave Cutler being the architect) that there was a Mach kernel influence...hm...
Anyway, SCO, listen up. There's potential here, at least as solidly founded as your case against IBM/RedHat/SuSE. Or maybe HP vs. M$?
Completely agree, and I would clarify that to be "ban all Windows executables *and* byte code " (or whatever the M$ term is for their Java clone's p-code). Yeah, I think M$ couldn't care less if a few hobbyists run VFP on their Linux box, but if they start running.exe's, or byte code look out!
Does anyone remember the spoof (written in the form of a tv show script) in which the Star Trek Enterprise is confronted with a Klingon warrior star ship, their cannons having failed, out comes a bunch of...lawyers... with brief cases! Piles upon piles of documents! Unreal hourly rates! Micorsoft was not founded on technical quality, but it's skill in the courtroom. Interesting how those founding traits permeate and define an organization.
Let's not forget, MSFT purchased FP. Fox Software was a small company from Ohio I think that made a well designed, elegant xbase engine. It ran on Xenix/Unix. Our company ran a multi-user business system on Fox that was comparable to a mini-computer system. This was when everybody else would be happy runing Wordperfect with their single user DOS/Windows session.
It was/is? extraordinarly flexible, for example there were some extensions written by a programmer at the Jet Propulsion Lab in Pasadena that were simply amazing. I think Java/Python/PHP have somewhat taken the limelight, but this illustrates that once you have a great peice of software, people will write other great peices of software that works with it.
One programmer who wrote a retail system in Fox called it the COBOL of the 90s.
And I agree with you point - the software simply gets the job done. Often consultants/managers get lost in the "we can do this in JAVA (or.NET or whatever) it'll be cool", where the real requirement can be met in good ol FoxPro. So everybody gets distracted by FUD, while the FPers keep writing code that works. But point also taken that RAD tools!= great design.
The previous poster was not talking about whether the x86 64 bit was an improvement over the 32 bit, they were talking about the x86 architecture as a whole being a 'hack' vs. chips like the PowerPC, Alpha, etc. That of course is another argument.
I have a copy of the previous edition of this book, written by Doug Tidwell, James Snell and Pavel Kulchenko. In addition to perl it covers Java and C#. Concise, well written, simple examples. A thoroughly good book. So I would tend to disagree - while I haven't been impressed with *all* of the O'Reilly books (XML in a Nutshell indeed is essentially a Javadoc of the API), buyer beware.
If we take his basic question, "XML is too hard for beginners", well you're correct there are lots of tools & libraries. Here are a few:
http://castor.exolab.org/
From the website:
"It's basically the shortest path between Java objects, XML documents and SQL tables. Castor provides Java to XML binding, Java to SQL persistence, and then some more. "
Pretty amazing I say.
Sun's JAXB is similar.
As for Pull parsers, Dennis Sosnoski has some interesting articles at IBM Developer Works:
http://www-106.ibm.com/developerworks/xml/library/ x-injava/#6
Here's an example (real-world):
Company running Net.Commerce wanted to write a fancy 'configurator'. Java was all the rage then (still is), so it was proclaimed that the configurator be written in Java, in spite of the fact that:
We had many experienced Net.Data, and Net.Commerce (C++ based). NC has some nice utilities which make it easy to setup a commerce site, but also has basic utilities for retrieving data.
But no, they wrote it in Java. Instead of using Net.Data (since they didn't have experienced Java prog), and even instead of using JSPs, the Java code *built the pages* in Java! Yeah, it looked like page one of the Java servlet tutorial from 8 years ago. Now how easy is that to read and debug?
Furthermore, database connectivity was so badly handled, so that the system could handle about 1 user at most without locking up. (Their solution was to synchronize the servlet! yowee!)
Granted, with experience Java programmers, with good tools (open source database pooling utilities or WebSphere etc), and knowledge of how JSPs, XMLC, XSL, work they could've done a better job, perhaps better than NC.
But for just getting the work done and out the door, and the users happy, Net.Commerce would've fit the fill fine. Most of the time was spent in design - the design was great, the followthrough (the actual app) sucked
That was a lesson for all.
Er...actually OS/2 is still supported...
Not sure how your theory fits in with IBM's support of Linux (I can run Db2, WebSphere, WSAD/Eclipse, MQSeries, Tivoli, Notes (Wine) etc etc on Linux.
Nor do I see it fitting in with their support of Java (some 80% or so of the spec for J2EE was contributed by IBM)
Nor their contributions to Apache and other open source projects (Eclipse)
Certainly supporting M$ is a source of income for IBM just as many other firms, but I find it pretty impressive that they support such a wide variety of platforms. That surely makes life hard for them, and easy for companies like M$ (then again, Win95vsWin98vsNTvs2000vsXP you get the picture..)
Sure it's good - it was based on OS/2...interesting the article glosses over that a little...
Cringley said the same about IBM 10 yrs ago
on
The Faded Sun
·
· Score: 1
My take is he is trying to 'wake-up' Sun, shock them out of their malaise. Read his book - accidental empires, in which he predicts the doom of IBM in a few years. Now that didn't happen, and I don't think his ideas were all that good either. What Sun really needs is someone like Gerstner to focus Sun's energies.
Cringley is a great storyteller, and alos very good at summarizing, in a humorous way, the past. I don't think he has the same ability with the present. But who does?
Hopefully Sun will get their act together. I believe they have great technology, but they need to move, embrace Linux, rather than fight it, and also make sure their crown jewel, Java, runs well on Solaris, and they need to... Hmm - hard isn't it?
As is usual in these lists, no one is really answering the main posters question: what companies are successfully making a business with open source?
IBM is. Not only is IBM contributing huge amounts of software ($40 million for Eclipse, XML4J, etc) to the open source world, they also use it in their projects or regular systems (for example, Ant is used to simplify WebSphere Commerce Suite build and deploy).
Even better, they *make* software that runs on Linux, such as WebSphere, DB2, MQSeries, etc.
Slashdot Mirror
I'd also be wondering about a company that sends back data (who knows what) when you register the OS. There was a big brouhaha about this a while back. Also, and I haven't verified this, but a colleague running XP claims that while he is surfing the net, the OS is sending data to m$.com. Now, he hasn't run a dump on the connection, but I'd be interested if anyone has any details on this, could be just a config issue, or just totally wrong, but (and here's the rub) I wouldn't put it past M$ to track every surfer's habits similar to doubleclick - this is valuable information.
Yeah, I think Windows has improved a bit, especially now with the competition from Linux. Look at their divulging the source code to the Chinese gov. and you get an idea of the tremendous influence Linux has had. Also, it's interesting to see the new Server 2003 ads the expound on the money saved. How you save money when you are spending not only for the license, new hardware required, cost to test and to upgrade your network etc etc - but hey that's marketing, always a schmooze. By the time you even begin to recoup your investment, there's Server 2005, after all that's the business model. I recall a while back when M$ was trumpeting some feature which was essentially something that had always been in Unix (ln -s I think, but according to M$ this was their extraordinary innovation).
So if you're a Windows Developer in a Windows World, but interested in peering into the skies at the other planets rest assured you can dabble with since, unlike M$, many popular tools are available for Linux are just as available for Windows since as open source they run happily on NT as on Unix. Furthermore, you can try cygwin or even Knoppix. As for VS studio, dunno, although you can check out the MONO project. There have been some threads recently of running VisualFoxPro on Linux using Wine, so maybe VB too... Better yet, take a look at the Eclipse project from which you could write any sort of scripts, PHP or whatever. KDevelop is amazing also - very familiar to the windows developer, don't fret, the Linux world is not requiring you to learn vi or Emacs..
Also, the new Linux distros are *very* cool in that much is already installed and configured for you. For example, I went to install mysql on my Mandrake box, when I realized I already had a version installed and running. One of the nice features of Mandrake which allows you to select the type of server you are running, i.e. database, web, games, file server, and install the appropriate packages.Therefore, as far as installing Linux and being productive, you'd be surprised at how easy it is. In fact, check some comparisons at cnet, you might find that Windows config is actually more of a pain than Linux, my take is that this is because Linux is almost always installed rather than pre-installed like Windows. So by nature it has to be good, if not better than Windows. My personal experience with this has been that Windows (at least older versions, not sure about XP) is that it paints you into a corner and can be twice as frustrating as Linux upgrades.
Games. Linux could use some improvement here, but then you could get a PS2 or xbox, too right? Anyway, if you're interested in the vast expanding worlds of open source, grab a copy of SuSE, RedHat or Mandrake. You won't be disappointed.
Yeah, C# is out, there's a flicker of interest perhaps, but fleeting, who wants to go backwards? Too little, too late. I used to browse MSDN, eagerly await the latest copy of VS, get wrapped up in COM, and well, years have past and I haven't looked back since.
Gads, you're right. He gave the software out for free, I think it was even open source. Sold his soul like Anders Hellsberj of Delphi fame? I'm really wondering what folks like Miquel de Icaza are thinking spending the time writing MONO. Perhaps useful as a C# clone, but I'm sure Microsoft will use a similar argument for running C# bytecode on non-Microsoft platforms such as MONO. Here's an excert from The Register: "Prior to the demonstration, Hentzen received a call from Ken Levy, Microsoft's Visual FoxPro marketing manager, telling him that he would be in violation of the EULA (End User Licensing Agreement) for VFP if he demonstrated (or ran) the development tool on Linux."
I mean, which is more preposterous, going from Dos->Windows->NT, or going from Minix->Linux? My copy of Silberschatz & Galvin's "Operating Systems Concepts" isn't too detailed, but I do recall that in addition to the Vax/VMS heritage (Dave Cutler being the architect) that there was a Mach kernel influence...hm... Anyway, SCO, listen up. There's potential here, at least as solidly founded as your case against IBM/RedHat/SuSE. Or maybe HP vs. M$?
Completely agree, and I would clarify that to be "ban all Windows executables *and* byte code " (or whatever the M$ term is for their Java clone's p-code). Yeah, I think M$ couldn't care less if a few hobbyists run VFP on their Linux box, but if they start running .exe's, or byte code look out!
Does anyone remember the spoof (written in the form of a tv show script) in which the Star Trek Enterprise is confronted with a Klingon warrior star ship, their cannons having failed, out comes a bunch of ...lawyers ... with brief cases! Piles upon piles of documents! Unreal hourly rates! Micorsoft was not founded on technical quality, but it's skill in the courtroom. Interesting how those founding traits permeate and define an organization.
Let's not forget, MSFT purchased FP. Fox Software was a small company from Ohio I think that made a well designed, elegant xbase engine. It ran on Xenix/Unix. Our company ran a multi-user business system on Fox that was comparable to a mini-computer system. This was when everybody else would be happy runing Wordperfect with their single user DOS/Windows session. It was/is? extraordinarly flexible, for example there were some extensions written by a programmer at the Jet Propulsion Lab in Pasadena that were simply amazing. I think Java/Python/PHP have somewhat taken the limelight, but this illustrates that once you have a great peice of software, people will write other great peices of software that works with it. One programmer who wrote a retail system in Fox called it the COBOL of the 90s. And I agree with you point - the software simply gets the job done. Often consultants/managers get lost in the "we can do this in JAVA (or .NET or whatever) it'll be cool", where the real requirement can be met in good ol FoxPro. So everybody gets distracted by FUD, while the FPers keep writing code that works. But point also taken that RAD tools!= great design.
The previous poster was not talking about whether the x86 64 bit was an improvement over the 32 bit, they were talking about the x86 architecture as a whole being a 'hack' vs. chips like the PowerPC, Alpha, etc. That of course is another argument.
I have a copy of the previous edition of this book, written by Doug Tidwell, James Snell and Pavel Kulchenko. In addition to perl it covers Java and C#. Concise, well written, simple examples. A thoroughly good book. So I would tend to disagree - while I haven't been impressed with *all* of the O'Reilly books (XML in a Nutshell indeed is essentially a Javadoc of the API), buyer beware.
If we take his basic question, "XML is too hard for beginners", well you're correct there are lots of tools & libraries. Here are a few: http://castor.exolab.org/ From the website: "It's basically the shortest path between Java objects, XML documents and SQL tables. Castor provides Java to XML binding, Java to SQL persistence, and then some more. " Pretty amazing I say. Sun's JAXB is similar. As for Pull parsers, Dennis Sosnoski has some interesting articles at IBM Developer Works: http://www-106.ibm.com/developerworks/xml/library/ x-injava/#6
Take a look at Sun's JAXB or The Castor Project (http://castor.exolab.org/), which are better solutions to this problem.
Here's an example (real-world): Company running Net.Commerce wanted to write a fancy 'configurator'. Java was all the rage then (still is), so it was proclaimed that the configurator be written in Java, in spite of the fact that: We had many experienced Net.Data, and Net.Commerce (C++ based). NC has some nice utilities which make it easy to setup a commerce site, but also has basic utilities for retrieving data. But no, they wrote it in Java. Instead of using Net.Data (since they didn't have experienced Java prog), and even instead of using JSPs, the Java code *built the pages* in Java! Yeah, it looked like page one of the Java servlet tutorial from 8 years ago. Now how easy is that to read and debug? Furthermore, database connectivity was so badly handled, so that the system could handle about 1 user at most without locking up. (Their solution was to synchronize the servlet! yowee!) Granted, with experience Java programmers, with good tools (open source database pooling utilities or WebSphere etc), and knowledge of how JSPs, XMLC, XSL, work they could've done a better job, perhaps better than NC. But for just getting the work done and out the door, and the users happy, Net.Commerce would've fit the fill fine. Most of the time was spent in design - the design was great, the followthrough (the actual app) sucked That was a lesson for all.
Both France and Germany have USD Billions of contracts with Iraq ... that's about to be null and void, eh?
So what does that makes us? Super power wimps?
Er...actually OS/2 is still supported... Not sure how your theory fits in with IBM's support of Linux (I can run Db2, WebSphere, WSAD/Eclipse, MQSeries, Tivoli, Notes (Wine) etc etc on Linux. Nor do I see it fitting in with their support of Java (some 80% or so of the spec for J2EE was contributed by IBM) Nor their contributions to Apache and other open source projects (Eclipse) Certainly supporting M$ is a source of income for IBM just as many other firms, but I find it pretty impressive that they support such a wide variety of platforms. That surely makes life hard for them, and easy for companies like M$ (then again, Win95vsWin98vsNTvs2000vsXP you get the picture..)
Sure it's good - it was based on OS/2...interesting the article glosses over that a little...
My take is he is trying to 'wake-up' Sun, shock them out of their malaise. Read his book - accidental empires, in which he predicts the doom of IBM in a few years. Now that didn't happen, and I don't think his ideas were all that good either. What Sun really needs is someone like Gerstner to focus Sun's energies. Cringley is a great storyteller, and alos very good at summarizing, in a humorous way, the past. I don't think he has the same ability with the present. But who does? Hopefully Sun will get their act together. I believe they have great technology, but they need to move, embrace Linux, rather than fight it, and also make sure their crown jewel, Java, runs well on Solaris, and they need to... Hmm - hard isn't it?
As is usual in these lists, no one is really answering the main posters question: what companies are successfully making a business with open source? IBM is. Not only is IBM contributing huge amounts of software ($40 million for Eclipse, XML4J, etc) to the open source world, they also use it in their projects or regular systems (for example, Ant is used to simplify WebSphere Commerce Suite build and deploy). Even better, they *make* software that runs on Linux, such as WebSphere, DB2, MQSeries, etc.