That's a tempest in a teapot. MS Office files still need to be serializable somehow, and that format cannot possibly be worse than it is right now. If anything, it may get better with Microsoft's push for XML.
Functionally, database-based file systems are an old hat. If they were the magic bullet Microsoft and BeOS think they are, they would have caught on long ago. To me, it looks more like a bunch of college hackers getting mightily excited about a whizbang feature of little real value. (Database based file systems have worked well in some niche markets--IBM is selling some systems with such file systems.)
Something needs to be done about indexing and search, but putting a database into the kernel is not the right thing.
When I visit their site with Mozilla 0.9.9, I get an error message that they only support IE and Netscape4. And for their handheld, they chose a window system that excludes most open source toolkits from co-existing on the same screen.
Looks to me Sharp is trying to milk open source for its advertising value, but somehow they just don't get it.
Thanks, but I'll have a look at the new Sony instead. Commercial developers don't have to pay money to develop for the Sony, and it's a cool piece of hardware anyway.
Making the file system a database didn't start with Be or Microsoft, it goes back several decades to IBM.
There are brute force ways of doing it: you build some kind of ad hoc database system and dump it into kernel space. You may be able to engineer such a system reasonably well, but to me, it is in bad taste: indexing is such a complex and application dependent area that nobody can guess ahead of time what kind of indexing people will want a few years from now. The Be file system looks like it's too complicated to interoperate well, and too simplistic to be of much use for anything rather than fairly primitive indexing operations.
A better way of doing this is to figure out a protocol for notification and updates between a traditional file system and user-space database indexing services. Yes, that's harder, but that's what software engineers get paid to figure out. And, as far as I'm concerned, if you can't figure out how to do it right, it's better not to do it at all rather than doing something half-baked.
It doesn't matter how well it works. The main thing that matters to most people in an MS Office replacement is how well it reads and writes MS Office files. And that's, unfortunately, a moving target.
Java, as a matter of fact, isn't a product of a company, but of the entire industry, with the especial exception of microsoft:)
If only that were true. Unfortunately, it isn't. Java, like Microsoft Windows, is really the product of only one company, Sun, although a number of other companies (IBM, Apple) are reselling it with some modifications.
Microsoft can't spend more money in.NET and C# than the community spends in Java.
What makes you think Microsoft is going it alone? The usual suspects are investing in C# and.NET. And C#/.NET isn't even going for the same people as Java--it will find acceptance quickly among Microsoft's current VC++ and VB programmers, because it's a whole lot better than what they have right now.
They could have been more careful about things like the memory model.
There is nothing wrong with the C#/CLR "memory model". By default, it is safe, just as in Java. If you write an unsafe model, the memory model is unsafe, just like it is in Java. Oh, you say, Java doesn't have unsafe modules. But it does. They are called "JNI". The only difference to C#'s unsafe modules is that JNI is less efficient and harder to program. (Both Java's and C#'s security models label unsafe code as such.)
I guess one of my pet areas is scientific computation. They might have done something creative to make that easier.
This is adding insult to injury. C# has value classes, operators, multidimensional arrays, and easy and efficient interfaces to native code. Sun and Gosling have been promising some of those features for years and failed to deliver on even the simplest of them. The best we are getting is a cumbersome proposal from IBM for multidimensional arrays that most implementations will probably not even bother to optimize.
And, I mean, the fact that the syntax [of C#] is so much -- is like exactly the same, or just about exactly the same [as that of Java].
Well, gee, what a coincidence. Microsoft thought Java was a great idea, but they wanted to have their own libraries. Sun sues them. So, they did the next best thing: they cloned Java as much as they could, fixed a bunch of small things Sun has been promising to fix for years, and called it C#. What does Gosling expect Microsoft to do? Just roll over and die? And Sun really has a double standard there: when Apple exposes all their native platform APIs to Java, that's fine. It's just not fine when Microsoft does it. Who's going to get sued next? What can open source developers do with Java before Sun is going to try and sue them?
I am no friend of Microsoft, and I won't use a Microsoft-only platform. But I am really getting tired of the marketing fluff coming out of Sun. When Java originally came out, Sun was promising a well-defined, open, standardized, and efficient platform. Today, it's a huge system with incompletely specified APIs, lousy support for high-performance computations, and no independent third party implementations (all compliant Java2 implementations depend to a large degree on Sun's source code). Sun has dropped out of every standardization process around, and they have been threatening others with lawsuits left and right.
I don't want to be tied to either a litigious Sun Java monopoly nor to a bundling Microsoft.NET monopoly. If Sun doesn't clean up its act quickly, after seven years of lobbying for Java and using it for lots of software, I'm dropping it. And I suspect others are getting similarly annoyed with Sun.
Sorry, pages that use Nuscrapisms like FONT tags instead of CSS are not "designed with a clue"
The guy claimed that Netscape 4 can render well designed pages. How on earth is your comment relevant to that?
Personally, I happen to use Mozilla. But I would prefer if all pages were kept simple enough so that they do render correctly in Netscape 4. In particular, that means I would greatly prefer not to see either CSS or FONT tags on the web.
Re:.NET is overall a good thing.
on
Java v. .Net?
·
· Score: 2
This is good. Amateur radio is regulated as a non-commercial space. If amateur radio operators manage to create a non-commercial, high-speed packet switching network, then maybe we can reclaim some space and create a network that's much closer to the original spirit of the Internet.
That's just too bad for Bluetooth. Amateur radio is licensed for those bands. Unlicensed users are secondary and have to accept interference. If they didn't like the deal, they didn't have to build devices that used amateur radio bands.
People don't exist to give companies a level, competitive playing field. Rather, companies exist in order to satisfy the needs of people. High speed Internet access companies have failed to do so, and that's why municipal goverments have stepped in.
In any case, in a democracy, it is up to the people to decide how public rights-of-way and public airwaves are allocated. We have made a decision in many places to have public utilities, and we can do the same thing with Internet access if we think it serves our needs better.
Now that all those media giants with their "creaative talent" and "amazing content" are on-line, that ads and Flash applets blink at us from everywhere, how could the web possibly be less fun than when it was mostly pages created by amateurs for fun? Maybe Americans have more taste and community spirit than the media giants and politicians give them credit for after all. Let's hope politicians will take that into account when drawing the line between the rights of Disney and the rights of grandma.
.NET is overall a good thing.
on
Java v. .Net?
·
· Score: 3, Insightful
Love to learn the bomb, and realize that.NET is overall actually a pretty good thing. First of all, it moves the Microsoft world away from Visual C++, and VB. This will result in Windows programmers that are used to working and thinking much more like the rest of the world. It will also result in Windows programs that are more reliable and easier to port to other platforms.
The C# language and core libraries underlying.NET are also well-documented and standardized. This means that free and third party implementations are going to happen (in fact, Mono is almost there), and a lot of free code can move more easily between the Windows and non-Windows worlds than was possible before.
And Sun needs the competition. The threat to their business from C#/.NET may finally get Sun to open up Java more. And it may get Sun to finally address some serious limitations of Java and the JVM that they have been promising to address for years but failed to do anything about.
In the long run, I think the two platforms will just merge. Runtimes will simultaneously support JVM and CLR, and Java and C# compilers will target each others runtimes.
All this is good as far as I'm concerned. I'm using Java for a lot of my work right now, but I may give Mono a try once it is fully self-hosting on Linux.
While there are some packages in Matlab that are doubtlessly an added value, those are not usually the packages that people use in education.
What has made Matlab so entrenched and valuable is the network effect: some people are using it, and therefore other people have to use it, too. This does demonstrate something about the "commercial viability" of free software: it can be highly profitable to establish a standard by giving away free software and eventually making the project proprietary and start charging huge amounts of money for it. That's a lesson everybody should learn--before they start using "free" software that is somehow enmeshed with commercial interests or comes with non open-source compliant licenses.
While Sun and Microsoft fight it out for the minds of computer science majors, another company has pretty much won the battle when it comes to engineering: MathWorks's Matlab has become the de-facto standard for computing in engineering and some areas of science and applied math. You can't exchange code with many others in the field unless you buy their software. Many research results are built on it and only reproducible using it. Oh, sure, it's cheap as long as you are a student or professor, but once you graduate, expect to pay many thousands of dollars even for a basic license, and many students graduating from top engineering and research labs are largely incapable of programming in anything else. The Matlab success story is a monopolist's dream.
I don't see that much of a difference between C#/.NET and the Java2 platform in terms of how closely they are tied to one company or the other (while several Java2 systems are available, they are all derived from Sun's code). Both.NET and Java2 have incomplete subsets that are available in open source form (Java 1.1, Mono), but, ultimately, both are proprietary platforms.
In fact, source access to the Java2 platform under the SCSL has onerous "contamination provisions" and I think using it in a computer science course is irresponsible because it may contaminate students for the rest of their professional lives.
What we really need is better open source, non-proprietary implementations of either language that colleges can use. These then give students access to tools they can use after they graduate wherever they work, and they can work with the full source code without selling their souls. And, besides, colleges shouldn't focus so much on just one language anyway.
It really doesn't matter whether a non-Microsoft office suite is better than MS Office. I would argue that there are already plenty of such office suites. Most people who are forced to run MS Office need to do so in order to read the MS Word and MS PowerPoint files that people send them. And, unfortunately, because many such files seem to include weird, undocumented formatting and executable code, non-MS Office suites have a really hard time reading them.
This is mostly fluff from IBM, the company that failed with OS/2 on the desktop and whose UNIX server line is slowly being replaced by Linux. They just don't get it.
While some of the Gnome and KDE users may suffer from Windows-envy, most traditional UNIX and Linux users use the UNIX and Linux desktop because they prefer it. We aren't quietly suffering with some inferior system, we are using the desktops we are using because we like them. After all, it's not like we haven't already paid for Windows, usually many times, anyway. The only reason I use non-Linux desktops from time to time is because of a deluge of MS Word and MS PPT junk that ends up in my inbox and can't always be ignored.
On the server side and issues of scalability and support, these folks also just don't get it. Shared memory approaches are inherently and intrinsically non-scalable: adding processors to a shared memory machine has quickly diminishing returns and escalating costs. You need to make that bargain if you buy an old-technology database like Oracle and think of databases in a 1960's big-box sort of way, and vendors will happily sell you the premium hardware to implement such outdated systems, but that doesn't make it reasonable or cost effective. If a vendor mentions "shared memory" and "scalable" in the same breath, hold on to your wallet and run.
Linux is both scalable and robust in the way that matters: you can put together very cheaply hundreds of boxes of commodity hardware. If a few fail and the overall distributed system is designed right, nobody cares. You don't need 24x7 service. You don't even need premium hardware support contracts.
Doubtlessly, Linux is not the right system for large crowds of desktop users, and it's not the right system for large crowds of ex-mainframe and ex-AIX hackers. And, frankly, I hope it will never be.
The only place where BeOS still has the advantage is in userspace, where BeOS totally whips GNOME and KDE in terms of speed, ease-of-use, simplicity, consistancy, etc. Apparently, this fork tries to take the best ideas from both sides.
Can you be more specific? What interesting ideas does BeOS actually contribute? When I looked at the documentation, I didn't see anything other than a fairly run-of-the-mill C++-based toolkit.
Merely being fast, low latency, and simple isn't exactly a new idea, and it isn't very hard for a system that is only a few years old. X11 has been up against plenty of fast, low latency, and simple GUIs before, and it has always come out ahead, by user choice.
Troff and Scribe were in common use, but they were commercial-only. Because of that, fewer people used them and even fewer people hacked on them, and they didn't evolve and eventually fell into disuse. TeX took over because it was free and it had excellent support for math, even if the rest of it was pretty iffy.
I use TeX/LaTeX for all my papers and presentations, but I have to say, TeX really has a lot of design problems, problems that even a herculean effort like LaTeX can't completely fix.
I think the credit for TeX's success should got to Lamport, who managed to create a usable type setting system on top of TeX. TeX itself is an oddball collection of some nice algorithms, a good math typesetting sub-language, but also lots of weird scoping rules and lots of global variables. Other things that have been essential to TeX's survival have been the addition of direct PostScript font support and pdftex by a number of volunteers.
At this point, I think TeX only holds on to life because LaTeX is an entrenched standard. If you designed something from scratch, it might look like LaTeX in most cases, but the underlying mechanisms for macro expansion, layout, and fonts would be completely different from, and much more modern than, TeX's.
Of course: if it's not broken, don't fix it. But most software is completely broken, and, if you avoid the second system effect, can actually get significantly better if you rewrite using your better understanding of the problem, as well as more modern tools.
Of course, if you rewrite using the same old style and tools and add 2x features while you are at it, rewriting doesn't make sense.
Slashdot Mirror
Functionally, database-based file systems are an old hat. If they were the magic bullet Microsoft and BeOS think they are, they would have caught on long ago. To me, it looks more like a bunch of college hackers getting mightily excited about a whizbang feature of little real value. (Database based file systems have worked well in some niche markets--IBM is selling some systems with such file systems.)
Something needs to be done about indexing and search, but putting a database into the kernel is not the right thing.
Looks to me Sharp is trying to milk open source for its advertising value, but somehow they just don't get it.
Thanks, but I'll have a look at the new Sony instead. Commercial developers don't have to pay money to develop for the Sony, and it's a cool piece of hardware anyway.
There are brute force ways of doing it: you build some kind of ad hoc database system and dump it into kernel space. You may be able to engineer such a system reasonably well, but to me, it is in bad taste: indexing is such a complex and application dependent area that nobody can guess ahead of time what kind of indexing people will want a few years from now. The Be file system looks like it's too complicated to interoperate well, and too simplistic to be of much use for anything rather than fairly primitive indexing operations.
A better way of doing this is to figure out a protocol for notification and updates between a traditional file system and user-space database indexing services. Yes, that's harder, but that's what software engineers get paid to figure out. And, as far as I'm concerned, if you can't figure out how to do it right, it's better not to do it at all rather than doing something half-baked.
It doesn't matter how well it works. The main thing that matters to most people in an MS Office replacement is how well it reads and writes MS Office files. And that's, unfortunately, a moving target.
If only that were true. Unfortunately, it isn't. Java, like Microsoft Windows, is really the product of only one company, Sun, although a number of other companies (IBM, Apple) are reselling it with some modifications.
Microsoft can't spend more money in .NET and C# than the community spends in Java.
What makes you think Microsoft is going it alone? The usual suspects are investing in C# and .NET. And C#/.NET isn't even going for the same people as Java--it will find acceptance quickly among Microsoft's current VC++ and VB programmers, because it's a whole lot better than what they have right now.
There is nothing wrong with the C#/CLR "memory model". By default, it is safe, just as in Java. If you write an unsafe model, the memory model is unsafe, just like it is in Java. Oh, you say, Java doesn't have unsafe modules. But it does. They are called "JNI". The only difference to C#'s unsafe modules is that JNI is less efficient and harder to program. (Both Java's and C#'s security models label unsafe code as such.)
I guess one of my pet areas is scientific computation. They might have done something creative to make that easier.
This is adding insult to injury. C# has value classes, operators, multidimensional arrays, and easy and efficient interfaces to native code. Sun and Gosling have been promising some of those features for years and failed to deliver on even the simplest of them. The best we are getting is a cumbersome proposal from IBM for multidimensional arrays that most implementations will probably not even bother to optimize.
And, I mean, the fact that the syntax [of C#] is so much -- is like exactly the same, or just about exactly the same [as that of Java].
Well, gee, what a coincidence. Microsoft thought Java was a great idea, but they wanted to have their own libraries. Sun sues them. So, they did the next best thing: they cloned Java as much as they could, fixed a bunch of small things Sun has been promising to fix for years, and called it C#. What does Gosling expect Microsoft to do? Just roll over and die? And Sun really has a double standard there: when Apple exposes all their native platform APIs to Java, that's fine. It's just not fine when Microsoft does it. Who's going to get sued next? What can open source developers do with Java before Sun is going to try and sue them?
I am no friend of Microsoft, and I won't use a Microsoft-only platform. But I am really getting tired of the marketing fluff coming out of Sun. When Java originally came out, Sun was promising a well-defined, open, standardized, and efficient platform. Today, it's a huge system with incompletely specified APIs, lousy support for high-performance computations, and no independent third party implementations (all compliant Java2 implementations depend to a large degree on Sun's source code). Sun has dropped out of every standardization process around, and they have been threatening others with lawsuits left and right.
I don't want to be tied to either a litigious Sun Java monopoly nor to a bundling Microsoft .NET monopoly. If Sun doesn't clean up its act quickly, after seven years of lobbying for Java and using it for lots of software, I'm dropping it. And I suspect others are getting similarly annoyed with Sun.
The difference is that Windows really does suck :-)
The guy claimed that Netscape 4 can render well designed pages. How on earth is your comment relevant to that?
Personally, I happen to use Mozilla. But I would prefer if all pages were kept simple enough so that they do render correctly in Netscape 4. In particular, that means I would greatly prefer not to see either CSS or FONT tags on the web.
Close, but not just quite yet...
This is good. Amateur radio is regulated as a non-commercial space. If amateur radio operators manage to create a non-commercial, high-speed packet switching network, then maybe we can reclaim some space and create a network that's much closer to the original spirit of the Internet.
That's just too bad for Bluetooth. Amateur radio is licensed for those bands. Unlicensed users are secondary and have to accept interference. If they didn't like the deal, they didn't have to build devices that used amateur radio bands.
It is for that reason that amateur radio is licensed. RF safety is a big part of the amateur radio license.
In any case, in a democracy, it is up to the people to decide how public rights-of-way and public airwaves are allocated. We have made a decision in many places to have public utilities, and we can do the same thing with Internet access if we think it serves our needs better.
Now that all those media giants with their "creaative talent" and "amazing content" are on-line, that ads and Flash applets blink at us from everywhere, how could the web possibly be less fun than when it was mostly pages created by amateurs for fun? Maybe Americans have more taste and community spirit than the media giants and politicians give them credit for after all. Let's hope politicians will take that into account when drawing the line between the rights of Disney and the rights of grandma.
The C# language and core libraries underlying .NET are also well-documented and standardized. This means that free and third party implementations are going to happen (in fact, Mono is almost there), and a lot of free code can move more easily between the Windows and non-Windows worlds than was possible before.
And Sun needs the competition. The threat to their business from C#/.NET may finally get Sun to open up Java more. And it may get Sun to finally address some serious limitations of Java and the JVM that they have been promising to address for years but failed to do anything about.
In the long run, I think the two platforms will just merge. Runtimes will simultaneously support JVM and CLR, and Java and C# compilers will target each others runtimes.
All this is good as far as I'm concerned. I'm using Java for a lot of my work right now, but I may give Mono a try once it is fully self-hosting on Linux.
What has made Matlab so entrenched and valuable is the network effect: some people are using it, and therefore other people have to use it, too. This does demonstrate something about the "commercial viability" of free software: it can be highly profitable to establish a standard by giving away free software and eventually making the project proprietary and start charging huge amounts of money for it. That's a lesson everybody should learn--before they start using "free" software that is somehow enmeshed with commercial interests or comes with non open-source compliant licenses.
While Sun and Microsoft fight it out for the minds of computer science majors, another company has pretty much won the battle when it comes to engineering: MathWorks's Matlab has become the de-facto standard for computing in engineering and some areas of science and applied math. You can't exchange code with many others in the field unless you buy their software. Many research results are built on it and only reproducible using it. Oh, sure, it's cheap as long as you are a student or professor, but once you graduate, expect to pay many thousands of dollars even for a basic license, and many students graduating from top engineering and research labs are largely incapable of programming in anything else. The Matlab success story is a monopolist's dream.
In fact, source access to the Java2 platform under the SCSL has onerous "contamination provisions" and I think using it in a computer science course is irresponsible because it may contaminate students for the rest of their professional lives.
What we really need is better open source, non-proprietary implementations of either language that colleges can use. These then give students access to tools they can use after they graduate wherever they work, and they can work with the full source code without selling their souls. And, besides, colleges shouldn't focus so much on just one language anyway.
It really doesn't matter whether a non-Microsoft office suite is better than MS Office. I would argue that there are already plenty of such office suites. Most people who are forced to run MS Office need to do so in order to read the MS Word and MS PowerPoint files that people send them. And, unfortunately, because many such files seem to include weird, undocumented formatting and executable code, non-MS Office suites have a really hard time reading them.
While some of the Gnome and KDE users may suffer from Windows-envy, most traditional UNIX and Linux users use the UNIX and Linux desktop because they prefer it. We aren't quietly suffering with some inferior system, we are using the desktops we are using because we like them. After all, it's not like we haven't already paid for Windows, usually many times, anyway. The only reason I use non-Linux desktops from time to time is because of a deluge of MS Word and MS PPT junk that ends up in my inbox and can't always be ignored.
On the server side and issues of scalability and support, these folks also just don't get it. Shared memory approaches are inherently and intrinsically non-scalable: adding processors to a shared memory machine has quickly diminishing returns and escalating costs. You need to make that bargain if you buy an old-technology database like Oracle and think of databases in a 1960's big-box sort of way, and vendors will happily sell you the premium hardware to implement such outdated systems, but that doesn't make it reasonable or cost effective. If a vendor mentions "shared memory" and "scalable" in the same breath, hold on to your wallet and run.
Linux is both scalable and robust in the way that matters: you can put together very cheaply hundreds of boxes of commodity hardware. If a few fail and the overall distributed system is designed right, nobody cares. You don't need 24x7 service. You don't even need premium hardware support contracts.
Doubtlessly, Linux is not the right system for large crowds of desktop users, and it's not the right system for large crowds of ex-mainframe and ex-AIX hackers. And, frankly, I hope it will never be.
Can you be more specific? What interesting ideas does BeOS actually contribute? When I looked at the documentation, I didn't see anything other than a fairly run-of-the-mill C++-based toolkit.
Merely being fast, low latency, and simple isn't exactly a new idea, and it isn't very hard for a system that is only a few years old. X11 has been up against plenty of fast, low latency, and simple GUIs before, and it has always come out ahead, by user choice.
I use TeX/LaTeX for all my papers and presentations, but I have to say, TeX really has a lot of design problems, problems that even a herculean effort like LaTeX can't completely fix.
At this point, I think TeX only holds on to life because LaTeX is an entrenched standard. If you designed something from scratch, it might look like LaTeX in most cases, but the underlying mechanisms for macro expansion, layout, and fonts would be completely different from, and much more modern than, TeX's.
in short, a duplicate story, patented, and an old hat in the amateur radio community.
Of course, if you rewrite using the same old style and tools and add 2x features while you are at it, rewriting doesn't make sense.