Secondly, wildly broad claims normally start, and increasingly narrow claims are made as one works down the ordered list. Therefore, the first thing you claim is going to be ridiculously broad. Generally speaking, the first few claims in a patent are not serious attempts to patent something. The last few claims are the ones of importance.
Yes, the first few claims are "overly broad", and that is the problem. The patent office has allowed those claims. That means they are in force until someone goes through the trouble and expense of trying to have them struck down in court. And they may not succeed at that, because contrary to your inane suggestions, there is no law saying that "the first 10 claims of a patent are always invalid".
The fact of the matter is, patents are structured with a series of claims, as you can see in the linked article. Claims are the only thing with meaning in a patent. The rest is provided for reference, clarification, and simple defence. So the only thing of true importance in this document are the claims, near the top. [...] Nothing like patents to bring out the righteous indignance.
Ah, well, you speak like someone who likes to pretend to know something about patent law on Slashdot. Congratulations, you have successfully parrotted the two or three key phrases to fool the high school students on Slashdot. Of course, to the rest of us, you are stating the obvious and your comments are merely redundant noise.
In reality, this patent should outrage any engineer and software developer. It is particularly outrageous because the patent examiners allowed the first 10 claims, but the remaining 10 claims are almost as bad.
However, as far as this being a Microsoft patent, the only thing that is unusual about that is that Microsoft didn't use to write a lot of patents at all. Now they have merely caught up with common industry practice in the creation of such stupid patents. We shouldn't hold that against Microsoft any more than we hold it against any other company that engages in such practices, but we should certainly criticize Microsoft for this patent as much as we would criticize IBM or Apple. Companies like that cannot unilaterally stop writing such patents because they need them for trading. But they can become politically active and have the patent system changed. In fact, they are probably pretty much the only players who can ultimately get patent law changed.
I explained why the perception existed. It's due entirely to the way the software has been positioned by its authors. Try to keep up.
Well, I suppose if you jump to conclusions after reading the first few lines on a web page, you can have all sorts of perceptions. Anybody who actually bothers to look at the screen shots,read the FAQ, read the progress reports, or read the mailing list will see that Gtk# is where most of the GUI action is.
Publishing files on the web requires nothing more than dragging them into a folder. If people want to add text, they can choose among hundreds of programs to help them, many far easier to use than P2P software.
Keep in mind that the web was designed for scientists to share information in the first place; even a sufficient knowledge of raw HTML is so trivial that anybody working in education should be able to grasp it with little training.
How can an Open Standards UNIX company with its roots in BSD and contributions of millions of lines of code to the Open Source community (just look at OpenOffice.org) be a threat to Linux?
Sun's contributions to the open source community are useful and appreciated, but that doesn't change the fact that Java is not open and presents a threat.
By analogy, would you sign over ownership of your house to me if I gave you my old car for free? Probably not.
How can a company who is comitted to playing fair and abiding by Open Standards, with public documentation be a threat to Linux?
The Java documentation may be "public", but it comes with a several page long license attached to it that imposes strong constraints on anybody using the specification.
Why would Sun hate Linux when it keeps closed, proporietary, buggy, virus and trojan-infested, expensive Windows off PCs?
Your argument makes no sense. Sun is selling SPARC-based systems. Why would they want to keep buggy software off PCs? The worse PCs are technically, the easier it is for Sun to compete with them. Linux makes PCs into a very viable, low-cost alternative to Sun's products and that is destroying Sun's market. Of course, if Sun could wave a magic wand and make Linux go away, they would do so. However, since there is no way in hell they could succeed at that, they are living with an uneasy truce.
In any case, we don't have to guess what Sun management is thinking about Linux and Gnome, they are telling us. For example, Schwartz called it "open source crap" and the "Linux mistake". And McNealy is busy spreading FUD about Linux, comparing it to illegal MP3 file sharing and raising questions about copyright liabilities.
Some of you slashbots really need your heads examining!
Some of you Java zealots really need to get over this "we think Sun is a nice company and therefore we trust them" thing and look at actual licenses.
"Manual modification" is the simple installation of an Infiniband networking card.
Well, that is a big deal, given that on the x86 side, many shops will sell you ready-made units (or whole racks) at little additional cost. Installing custom cards also plays havoc with repair and maintenance because it causes finger pointing among vendors.
If you think the air conditioning and maintenance costs are being ignored, you're ignorant to the entire process going on here in Blacksburg.
Desktop enclosures take up much more space than 1U rack mounts, which means they are much more expensive to install per unit than rack mounts, no matter what your per square foot cost is. Furthermore, desktops on shelves are significantly more time consuming to service, further increasing costs.
You'll also come to see that while using SPEC, the cluster is still going to destroy most clusters at ten times the price.
Oh? Care to put some side-by-side numbers behind that claim? When I calculate it, even in purchase costs along, the G5 boxes come out behind Xeon or Opteron. And the lack of rack mounting makes the costs even higher in the long run.
A 1U dual G5 Xserve may be a while since Apple seems to have trouble cooling the processors in 1U enclosures...
What's the advantage of this over a course web page and personal web pages? I mean, isn't that what web pages were designed for, file sharing and information sharing among individuals? And we even have URLs we can include in mail messages to point people at things. And while for MP3 sharing web pages may not work and not everybody has web hosting, evey university student should have web hosting and the skill to put up a web page somehow.
If Point A is true, then the "migration" theme in Point B has to be prompted by something other than Java since Java makes the underlying OS irrelevant.
You'll have to take that up with Turgid--it was his argument. He just applied it to the wrong language. He claimed that Mono enabled people to migrate from Linux to Windows easily and that that was a threat to Linux. But that is false as far as Mono is concerned because Mono programs rely so much on Linux-specific APIs. If Turgid's concern is valid, however, it does apply to Java because Java really does make it easy for people to migrate off of Linux.
Is Turgid's concern valid? Well, since Sun actually supplies key parts of all Java 2 implementations for Linux (e.g., Swing), Sun itself might engineer a migration off of Linux. They certainly have lots of motivation to drive people to Solaris from Linux. It certainly is a danger, and it's a danger that simply does not exist with Mono.
Furthermore, every programmer that uses the Java APIs is one less programmer that uses the APIs of open libraries and toolkits.
To make my point clearer, here's the same argument made using HTML:
The real danger is Microsoft. It is Microsoft's stated goal with HTML to make the underlying operating system and browser irrelevant. That is Microsoft is really trying to migrate people off of Linux by getting programmers to develop in HTML.
Your analogy doesn't work. Java is a set of programming APIs that encapsulates all of the operating system. If you write an application in Java, you don't depend on the underlying OS anymore. On the other hand, HTML isn't an encapsulation of an operating system. You can't write an application "in" HTML, you can only write it using HTML. The relationship between Java and the OS is not comparable to the relationship between HTML and the OS.
The argument does apply to other cross-platform toolkits, however. So, I think most Linux GUIs should not be written in Tcl/Tk or FLTK (wxWindows has platform-specific hooks, so it's less bad).
$38 million works out to $63000 for each one of the 600 Dell nodes. But the 1750 and 2650 are your run-of-the-mill $3500-$4000 Xeon rack mounts. Where are the other $59000 per machine going? Someone is comparing apples and oranges (no pun intended).
Also, "flops" is a pretty meaningless measure of performance. If you want to use benchmarks, at least use SPEC, and on those, the Macs are not particularly outstanding in terms of price/performance.
Finally, the Macs aren't rack mounts and they require manual modification for use in clusters; that alone makes them an iffy choice. VT can get cheap student labor to do that sort of thing and may not care about bulk, air conditional, or maintenance costs, but all of those matter big time in the real world.
Sorry, but someone is dreaming if they think G5's are the epitome of cluster computing. They may be better than previous Macs, and they may have entered the realm of plausibility, but you are still going to be far better off with rackmounted 1U dual-processor Linux machines in every respect.
Since there are many more applications being written that are targeted to the Windows platform than any other platform, such a move on Microsoft's part would render Mono an unusable platform - developers could not trust that applications targeting.NET could run under Mono.
We are in complete agreement there: nobody will be able to build a Microsoft.NET clone that runs applications written for Windows reliably. But, so what? An attempt at.NET compatibility by Mono still makes it much easier for Windows programmers to switch to Linux. It also makes porting to Linux much easier and I would expect a lot of C# library code (non-GUI code) can be written to work easily on both Windows and Linux. Both of those seem like good things to me.
Yes, Mono would still remain an interesting runtime environment, but do we really need another?
Yes, I think we do. The only other choice in this space is the Java language and Java platform, and it has numerous encumbrances and problems. In contrast to Java, ECMA C# is a free standard and Mono provides a large, complete, open source platform with easy access to existing open source libraries like Gtk+.
Mono and C# aren't perfect, but they seem sufficiently different from Java to have a reason for existing.
and if MS decides to use its patents against competing implimentations of the.net apis?
If you look on the go-mono.com web site, you'll see that most screenshots are from Gtk# applications. Mono has a rich set of non-.NET APIs, and those seem to form the backbone of most Mono applications. So, even if Microsoft's patent is enforcable, it wouldn't affect most users of Mono since they are not using.NET.
Now, let's say that Microsoft introduced a new Java API, or changed a Java API when Java was successful, did that change the success of Java?
Well, it did change things. Java has lots of problems on the "run anywhere" side of things as it is, and when major java programs were written with MS-only APIs, cross-platform dreams were totally over.
First of all, that is just what Apple has been doing with the OSX version of Java.
And in any case, why is cross-platform support supposed to be good? I care about Linux and Linux APIs. Why would I want to sink to the lowest common denominator of Windows, Solaris, and Linux (which is what a cross-platform environment means) for being able to run software on Windows that I never want to run on Windows in the first place?
I suspect that.NET will also have major Win32-only parts. If a goal of mono is just to be a development platform, fine. If anyone thinks that apps written for MS.NET will be cross-platform, then they haven't been reading their recent history.
You got it: Mono is turning into a good environment for Linux developers to develop Linux applications. Its.NET compatibility is merely there to make it easy to also attract Windows programmers to the platform.
And if Mono is just about a dev environment, then why bother? I can't really see why I should switch to C#.
That depends on what you are switching from.
If you are switching from C++, then one reason would be that C# is much easier to learn and debug while still given you easy and fast access to pointers and native libraries.
If you are switching from Java, there are a number of reasons. First, C# with Gtk# gives you much better Gnome desktop integration than Java. Also, Mono is open source while there are no open source implementations of the Java 2 SE platform. C# has much easier access to existing C/C++ libraries than Java. And C# fixes a number of serious limitations of the Java language (although it introduces some new problems of its own).
So, yes, Mono is primarily "just" a development platform. But it is a free, open source development platform that is quite good and powerful..NET compatibility and the possibility of Windows/Linux cross-platform applications are a nice additional feature, but if they went away, it wouldn't be the end of the world.
You are, for free as in beer, providing Microsoft with a well-made tool for helping them to get people to migrate off of UNIX and onto Windows.
Mono's.NET compatibility, such as it is, is a way for Windows programmers to migrate to UNIX/Linux. That seems like a good thing to me.
Migration in the other direction is not a real concern when using Mono: most Mono applications will simply not easily work on Microsoft.NET because they usually have lots of dependencies on Linux libraries. For example, of all the screen shots on the Mono home site, almost all of them are Gtk# based.
Of course, if you work at it, you can use Mono to write cross-platform applications that also run on Windows, but, then, you can do the same thing with wxWindows, FLTK, Python, Perl and lots of other toolkits and environments. Mono doesn't keep you from writing something that will work on Windows, but it certainly doesn't encourage it.
The real danger is Sun. It is Sun's stated goal with Java to make the underlying operating system or native toolkit irrelevant. That is Sun really is trying to migrate people off of Linux by getting programmers to develop in and for Java APIs. Of course, they would like to migrate them to Solaris, but they might more readily migrate to Windows. Even without all the legal b.s. from Sun, that alone would constitute a big threat to Linux.
Without SUN UNIX would not have been more successful. I was a PhD student around 1990, and at this time it was SUN, HP and IBM fighting for the academic world (giving away cheap workstations etc).
So, you are saying Sun tried to compete on price, and I agree. But how did that help UNIX in the long run? Sun cut prices but, like all other UNIX vendors, they failed to create a thriving market in commercial or open source end user applications.
SUN was never alone in the market, they had fierce competition from other UNIX vendors. If SUN was really so bad w.r.t. GUI and API's, then why did they gain most of the marketshare?
As you point out yourself, they were cheap and they came out of a university environment and marketed to a university environment.
It was a free market with open competition, so I cannot see how a single company would have been able to do damage to UNIX.
Sun's war on X11 and Motif did quite a bit of damage and in the end was futile. And bad standards like NFS also hurt the UNIX world quite a bit. So, yes, a single company can do quite a bit of damage to UNIX.
And Sun is gearing up to repeat history: they are trying again to replace UNIX and POSIX with a set of what amounts to Sun-proprietary APIs. As before, they will lose in the end, but they will do some damage in the process.
If any, it was the collective vendors that spoilt it, not SUN alone. By the way, do you remember the X11-GUI war between OpenLook (sun) with a good looking and easy to program toolkit and interesting windowing system (News, which is very much like Apples display postscript of today) against the rest of the UNIX world pushing for g*dawful Motif?
I remember it very well: both OpenWindows and NeWS were poorly conceived, poorly executed, and proprietary. They lost in the market because they were bad. The fact that they lost to X11 and Motif tells you how bad they really were. Keep in mind that this really was a user choice: people got OpenWindows and NeWS preinstalled and they often still installed MIT X11.
If SUN would have had their way, UNIX on the desktop would have been in a much better shape.
No, it would merely have been Sun-proprietary. But, if at first you don't succeed, try again. So, Sun is trying again to create a proprietary UI by replacing X11 APIs with Java APIs. And they are about as successful as they were with NeWS, which is to say, not very much.
You see, Sun really wants to be just like Microsoft: they want to control the APIs and see the money rolling in. The part Sun didn't get back then and the part Sun still isn't getting is that creating a good API is neither necessary nor sufficient for getting lots of people to adopt it.
I think the point was that if you think it is an inaccurate view, maybe you should get the mono people to rewrite their documentation to reflect more accurately where the project is and where it's going.
Well, the fact that almost all the screenshots are from Gtk# applications and that there is an entire Gtk# wiki should make this pretty obvious. And the FAQ talks at length about how to deal with the possibility that Microsoft might try to shut down the project using their single, unissued patent (provided they get it, of course).
Still, I agree that it would be nice for the Mono project to emphasize its widespread use of open source APIs more in the mission statement. On the other hand, they may have their own reasons for emphasizing that they are striving for.NET interoperability--keep in mind that there are businesses associated with the Mono project, and they do want to use it to allow Windows programmers to migrate to Linux. Or they may simply not have caught up with reality--that.NET compatibility for Mono has turned out to be far less important and that Gtk#/Gnome# are far more important than they originally were.
Please, don't forget who your enemy is and who your friend is, take just a bit of historical perspective!
Corporate history is no guarantee whatsoever for future actions. Look at Caldera/SCO: one day, they are a big open source Linux company, then they turn around and start taking people to court over code they themselves contributed.
Without SUN UNIX would have died a long time ago (in the real corporate world that is) and thus Linux and FreeBSD would have been much less relevant as well.
Or maybe UNIX would actually be more successful. Sun's incompetent and unsuccessful attempts at giving a good UI to UNIX, their insecure, unreliable, and unscalable network file system, and their painful and misguided attempts to switch to SysV APIs did a lot of damage to UNIX. Sun could have had the entire academic market and from there conquered the desktop with UNIX; instead, they drove their academic customers into the arms of Microsoft and Apple by delivering machines that can only be described as user-hostile to anybody but hardcore UNIX users.
Do you know who is our enemy and who is our friend?
No, and the concept doesn't make sense when applied to companies. Companies make short-term, profit maximizing decisions. Sun does that as much as Microsoft. It just happens to be the case that Sun is in a position where some open source releases are useful for them right now.
Therefore, I care about specifics: licenses, contracts, standards. The licenses and agreements under which Sun distributes Java are very dangerous for open source developers, as is Sun's extensive patent portfolio on Java-related inventions.
Fact is, SUN has NEVER misused its control over API's, but instead has given away many products and standards without ever playing tricks or misusing its position (think about virtually any UNIX network protocol and influence in most UNIX API's).
Well, and for most of their history, Sun was fairly rich and successful. Now they aren't anymore, and that's why there is a serious risk that their policies will change. In fact, as far as I am concerned, Sun is already "misusing" their control over certain APIs, namely the Java APIs.
Yes, Ximian launched mono to create an open source implementation of.NET. How does that contradict what I said? How does the reason for which Ximian launched Mono have anything to do with where the project is going, how people are using it, or what control Microsoft has over it?
The fact remains that a free and open subset of.NET, namely ECMA C#, is supported by Mono and makes a great programming platform when combined with open source toolkits like Gtk+ and Gnome. Microsoft has no claims against that combination, and, in fact, I doubt they really mind that it exists. And it is that combination that successful open source Mono projects will be written in, not the.NET APIs.
In contrast, the full.NET APIs of Mono are cumbersome and inefficient; they will be useful to have around for compatibility, but if they went away, it wouldn't matter much at this point. Their function is to allow Microsoft programmers to migrate more easily to Linux. There is no risk in trying to offer that capability: at worst, Microsoft might succeed in shutting down full.NET compatibility, which would leave us no worse off than we are now.
Whether any other claim by Microsoft on.NET is even valid or enforceable remains to be seen; so far, they don't even have a patent.
Microsoft has several means at their disposal to effectively shut down Mono if it should ever gain critical mass.
Those claims are based on the inaccurate perception that the success of Mono depends on.NET compatibility and that Mono applications are.NET applications. That's, in fact, just false. Most current uses of Mono are based on ECMA C# and Gtk#, not.NET. In fact, one of the big strengths of C# is that, unlike Java, C# makes it easy to reuse existing C and C++ libraries; in that, it is much like the relationship of C++ to C. If you already know Gnome, you can start using C# to develop Gnome applications much more easily than picking up Java and Swing (and the Mono/Gtk# applications will work better, too).
The company to worry about is Sun: open source Java applications do use all-Sun APIs; interfacing with native libraries is just too much hassle, and that's no accident: Sun wants you to use their APIs and give up on the free, open source APIs. And, despite all the JCP mumbo-jumbo, Sun has a lot of control over the Java platform, through numerous patents, through owning key parts of the actual implementation of key parts of the Java platform (e.g., Swing), and through their ownership of the specification and the certification process.
So, if you are worried about Microsoft's ownership of.NET, just don't use.NET. In fact, I wouldn't touch.NET simply because I think it's technically not very good. But you can still use Mono, which is shaping up to be a great, general-purpose programming platform. And because existing open source libraries, like Gtk+, Gnome, expat, X11, etc., is so easily accessible, it's very easy to start using Mono--it's just a nicer version of C++.
There are plenty of solar-powered calculators that work in normal light. They make a decent backup for when you are out of batteries or your plane is going down...
That would be a major change for KDE, given that Qt is such an integral part. However, it might actually be a good way for KDE to eliminate their dependency on TrollTech and replace Qt with something LGPL'ed--the Qt license has been a major obstacle to KDE's acceptance I believe.
Lastly, rendering speed improvements could be realized. Aside from dedicated HW doing the rendering, if the OS did it in a trusted manner then it could be faster than many libraries and/or programmer hacks.
Yeah, right, let's put everything into the kernel. You can see how well that worked for Windows.
Both Linux and Macintosh have client/server GUI systems that do rendering not in the OS but in user mode processes. That's the right way to do it. Something as complex as a GUI does not belong into the kernel (the "OS"). However, display-side SVG rendering in X11 would be a welcome addition, and that is being planned. That will give people the same kind of features that Apple gets with DisplayPDF, but with a somewhat cleaner and far more standards-compliant graphics language (SVG vs. PDF).
The smaller reasons include much less data - if a graphics card rendered SVG, then the connection between the computer and the card could be slow and very long distance. Hard drives space isn't an issue, except in power-conscious embedded areas where smaller graphics files could make a difference.
Pretty much all graphics cards these days have 2D acceleration. Therefore, computers already communicate them using vector graphics operations.
Objects can be clearly rendered in SVG regardless of the output device resolution.
You're confusing representation with graphics model. SVG gives you a text-based representation of an object-based graphics model. You don't need that for geometric scaling. Postscript, DisplayPostscript, OpenGL, GDI+, and XRENDER all give you geometric scalability without anything like SVG.
Furthermore, rescaling GUI objects is not simply scaling them geometrically. Support for scaling graphics in the the window system will help somewhat with scaling over a limited range, but it won't magically make a UI or even an icon scale from PDA to huge screen sizes.
SVG is useful and it will come. But it isn't the panacea that you seem to think it is, and it isn't all that new either.
Seriously, for a single home-use desktop, how much of the extra stuff is absolutely necessary?
Most of it. That "extra stuff" isn't functionality, it's all the infrastructure for making X11 run fast on modern desktop machines with large screens and lots of windows.
Perhaps if someone would release a similar slimmed-down version for desktops, then people would quit bitching about it.
The functionality is in X11 because people want it in their desktops. And why slim it down when X11 is already slimmer and faster than Windows or MacOS?
But dont let them compete against others and pretend that they're accomplishing anything.
And what are olympic athletes "accomplishing" right now?
The drama of sports is seeing what the human body can achieve through disciplined training and talent.
So, if someone's performance happens to be the product of a well-chosen mating, it's a "drama", but if it's the product of gene therapy, it's not? Sorry, I don't see it. I think the olympics are a carnival side show, with or without doping or gene therapy.
The US doesn't have have, and really never had, a true free market.
Quite right. And the US attained its greatest strength was when its markets were actually less free than they are now.
Look for totalitarian states and you will find abysmal economies. Look for strong socialistic countries and you will find better, but still poor economies. Your strong economies are found in nations with moderate to light socialism. In other words, where the government doesn't try to manage all aspects of the market.
Quite true again. But in order to get such markets and such a society, a lot of things need to come together. For the US to be able to do this, a lot of factors had to come together.
But it has been relatively freer than those in many other nations. This isn't a US thing. Europe and Australia also have markets relatively freer than most other nations.
And as I was saying, I think the US is finding out that merely having a "freer" market is not sufficient.
X11 worked fine on 20MHz 68k workstations with 4Mbytes of RAM. In the days of 400MHz PDAs with 64M+ of RAM, there really is no need to "replace" X11, in particular given their stamp-sized screens. And because X11 is neutral on policy, unlike other window systems, you can build a great handheld or embedded environment on top of it. In practice, of course, X11 is already more efficient in both memory and speed than either Qt/Embedded or the Windows GDI, so it is also hard to see why people think they can do better.
Besides, there is another problem with GPL'ed toolkits: the competition lets commercial developers develop for free. GPL'ed toolkits and window systems are no good: any Linux-based PDA is only going to succeed widely if commercial developers can develop for it for free. That's a bargain free software has always been making (e.g., you can develop commercial software with gcc for Linux and not open source it). Toolkits need to have licenses like LGPL, X11, or BSD.
So, please forget about all this "GPL'ed embedded toolkit" nonsense. Use X11 and one of several embedded toolkits for handheld development: wxWindows, FLTK, or GPE.
The default XFree86 port to handhelds, of course, has a lot of useless overhead in it because it is just a recompilation of a desktop server. That can be reduced, but you don't need to invent a new, incompatible window system to do it. An embedded X11 server can probably be stripped down to 100-200k. That is where the effort of people looking to create "efficient" embedded servers should be directed.
Slashdot Mirror
Secondly, wildly broad claims normally start, and increasingly narrow claims are made as one works down the ordered list. Therefore, the first thing you claim is going to be ridiculously broad. Generally speaking, the first few claims in a patent are not serious attempts to patent something. The last few claims are the ones of importance.
Yes, the first few claims are "overly broad", and that is the problem. The patent office has allowed those claims. That means they are in force until someone goes through the trouble and expense of trying to have them struck down in court. And they may not succeed at that, because contrary to your inane suggestions, there is no law saying that "the first 10 claims of a patent are always invalid".
The fact of the matter is, patents are structured with a series of claims, as you can see in the linked article. Claims are the only thing with meaning in a patent. The rest is provided for reference, clarification, and simple defence. So the only thing of true importance in this document are the claims, near the top. [...] Nothing like patents to bring out the righteous indignance.
Ah, well, you speak like someone who likes to pretend to know something about patent law on Slashdot. Congratulations, you have successfully parrotted the two or three key phrases to fool the high school students on Slashdot. Of course, to the rest of us, you are stating the obvious and your comments are merely redundant noise.
In reality, this patent should outrage any engineer and software developer. It is particularly outrageous because the patent examiners allowed the first 10 claims, but the remaining 10 claims are almost as bad.
However, as far as this being a Microsoft patent, the only thing that is unusual about that is that Microsoft didn't use to write a lot of patents at all. Now they have merely caught up with common industry practice in the creation of such stupid patents. We shouldn't hold that against Microsoft any more than we hold it against any other company that engages in such practices, but we should certainly criticize Microsoft for this patent as much as we would criticize IBM or Apple. Companies like that cannot unilaterally stop writing such patents because they need them for trading. But they can become politically active and have the patent system changed. In fact, they are probably pretty much the only players who can ultimately get patent law changed.
I explained why the perception existed. It's due entirely to the way the software has been positioned by its authors. Try to keep up.
Well, I suppose if you jump to conclusions after reading the first few lines on a web page, you can have all sorts of perceptions. Anybody who actually bothers to look at the screen shots,read the FAQ, read the progress reports, or read the mailing list will see that Gtk# is where most of the GUI action is.
Publishing files on the web requires nothing more than dragging them into a folder. If people want to add text, they can choose among hundreds of programs to help them, many far easier to use than P2P software.
Keep in mind that the web was designed for scientists to share information in the first place; even a sufficient knowledge of raw HTML is so trivial that anybody working in education should be able to grasp it with little training.
How can an Open Standards UNIX company with its roots in BSD and contributions of millions of lines of code to the Open Source community (just look at OpenOffice.org) be a threat to Linux?
Sun's contributions to the open source community are useful and appreciated, but that doesn't change the fact that Java is not open and presents a threat.
By analogy, would you sign over ownership of your house to me if I gave you my old car for free? Probably not.
How can a company who is comitted to playing fair and abiding by Open Standards, with public documentation be a threat to Linux?
The Java documentation may be "public", but it comes with a several page long license attached to it that imposes strong constraints on anybody using the specification.
Why would Sun hate Linux when it keeps closed, proporietary, buggy, virus and trojan-infested, expensive Windows off PCs?
Your argument makes no sense. Sun is selling SPARC-based systems. Why would they want to keep buggy software off PCs? The worse PCs are technically, the easier it is for Sun to compete with them. Linux makes PCs into a very viable, low-cost alternative to Sun's products and that is destroying Sun's market. Of course, if Sun could wave a magic wand and make Linux go away, they would do so. However, since there is no way in hell they could succeed at that, they are living with an uneasy truce.
In any case, we don't have to guess what Sun management is thinking about Linux and Gnome, they are telling us. For example, Schwartz called it "open source crap" and the "Linux mistake". And McNealy is busy spreading FUD about Linux, comparing it to illegal MP3 file sharing and raising questions about copyright liabilities.
Some of you slashbots really need your heads examining!
Some of you Java zealots really need to get over this "we think Sun is a nice company and therefore we trust them" thing and look at actual licenses.
"Manual modification" is the simple installation of an Infiniband networking card.
Well, that is a big deal, given that on the x86 side, many shops will sell you ready-made units (or whole racks) at little additional cost. Installing custom cards also plays havoc with repair and maintenance because it causes finger pointing among vendors.
If you think the air conditioning and maintenance costs are being ignored, you're ignorant to the entire process going on here in Blacksburg.
Desktop enclosures take up much more space than 1U rack mounts, which means they are much more expensive to install per unit than rack mounts, no matter what your per square foot cost is. Furthermore, desktops on shelves are significantly more time consuming to service, further increasing costs.
You'll also come to see that while using SPEC, the cluster is still going to destroy most clusters at ten times the price.
Oh? Care to put some side-by-side numbers behind that claim? When I calculate it, even in purchase costs along, the G5 boxes come out behind Xeon or Opteron. And the lack of rack mounting makes the costs even higher in the long run.
A 1U dual G5 Xserve may be a while since Apple seems to have trouble cooling the processors in 1U enclosures...
What's the advantage of this over a course web page and personal web pages? I mean, isn't that what web pages were designed for, file sharing and information sharing among individuals? And we even have URLs we can include in mail messages to point people at things. And while for MP3 sharing web pages may not work and not everybody has web hosting, evey university student should have web hosting and the skill to put up a web page somehow.
If Point A is true, then the "migration" theme in Point B has to be prompted by something other than Java since Java makes the underlying OS irrelevant.
You'll have to take that up with Turgid--it was his argument. He just applied it to the wrong language. He claimed that Mono enabled people to migrate from Linux to Windows easily and that that was a threat to Linux. But that is false as far as Mono is concerned because Mono programs rely so much on Linux-specific APIs. If Turgid's concern is valid, however, it does apply to Java because Java really does make it easy for people to migrate off of Linux.
Is Turgid's concern valid? Well, since Sun actually supplies key parts of all Java 2 implementations for Linux (e.g., Swing), Sun itself might engineer a migration off of Linux. They certainly have lots of motivation to drive people to Solaris from Linux. It certainly is a danger, and it's a danger that simply does not exist with Mono.
Furthermore, every programmer that uses the Java APIs is one less programmer that uses the APIs of open libraries and toolkits.
To make my point clearer, here's the same argument made using HTML:
The real danger is Microsoft. It is Microsoft's stated goal with HTML to make the underlying operating system and browser irrelevant. That is Microsoft is really trying to migrate people off of Linux by getting programmers to develop in HTML.
Your analogy doesn't work. Java is a set of programming APIs that encapsulates all of the operating system. If you write an application in Java, you don't depend on the underlying OS anymore. On the other hand, HTML isn't an encapsulation of an operating system. You can't write an application "in" HTML, you can only write it using HTML. The relationship between Java and the OS is not comparable to the relationship between HTML and the OS.
The argument does apply to other cross-platform toolkits, however. So, I think most Linux GUIs should not be written in Tcl/Tk or FLTK (wxWindows has platform-specific hooks, so it's less bad).
$38 million works out to $63000 for each one of the 600 Dell nodes. But the 1750 and 2650 are your run-of-the-mill $3500-$4000 Xeon rack mounts. Where are the other $59000 per machine going? Someone is comparing apples and oranges (no pun intended).
Also, "flops" is a pretty meaningless measure of performance. If you want to use benchmarks, at least use SPEC, and on those, the Macs are not particularly outstanding in terms of price/performance.
Finally, the Macs aren't rack mounts and they require manual modification for use in clusters; that alone makes them an iffy choice. VT can get cheap student labor to do that sort of thing and may not care about bulk, air conditional, or maintenance costs, but all of those matter big time in the real world.
Sorry, but someone is dreaming if they think G5's are the epitome of cluster computing. They may be better than previous Macs, and they may have entered the realm of plausibility, but you are still going to be far better off with rackmounted 1U dual-processor Linux machines in every respect.
Since there are many more applications being written that are targeted to the Windows platform than any other platform, such a move on Microsoft's part would render Mono an unusable platform - developers could not trust that applications targeting .NET could run under Mono.
.NET clone that runs applications written for Windows reliably. But, so what? An attempt at .NET compatibility by Mono still makes it much easier for Windows programmers to switch to Linux. It also makes porting to Linux much easier and I would expect a lot of C# library code (non-GUI code) can be written to work easily on both Windows and Linux. Both of those seem like good things to me.
We are in complete agreement there: nobody will be able to build a Microsoft
Yes, Mono would still remain an interesting runtime environment, but do we really need another?
Yes, I think we do. The only other choice in this space is the Java language and Java platform, and it has numerous encumbrances and problems. In contrast to Java, ECMA C# is a free standard and Mono provides a large, complete, open source platform with easy access to existing open source libraries like Gtk+.
Mono and C# aren't perfect, but they seem sufficiently different from Java to have a reason for existing.
and if MS decides to use its patents against competing implimentations of the .net apis?
.NET.
If you look on the go-mono.com web site, you'll see that most screenshots are from Gtk# applications. Mono has a rich set of non-.NET APIs, and those seem to form the backbone of most Mono applications. So, even if Microsoft's patent is enforcable, it wouldn't affect most users of Mono since they are not using
First of all, that is just what Apple has been doing with the OSX version of Java.
And in any case, why is cross-platform support supposed to be good? I care about Linux and Linux APIs. Why would I want to sink to the lowest common denominator of Windows, Solaris, and Linux (which is what a cross-platform environment means) for being able to run software on Windows that I never want to run on Windows in the first place?
I suspect that
You got it: Mono is turning into a good environment for Linux developers to develop Linux applications. Its
And if Mono is just about a dev environment, then why bother? I can't really see why I should switch to C#.
That depends on what you are switching from.
If you are switching from C++, then one reason would be that C# is much easier to learn and debug while still given you easy and fast access to pointers and native libraries.
If you are switching from Java, there are a number of reasons. First, C# with Gtk# gives you much better Gnome desktop integration than Java. Also, Mono is open source while there are no open source implementations of the Java 2 SE platform. C# has much easier access to existing C/C++ libraries than Java. And C# fixes a number of serious limitations of the Java language (although it introduces some new problems of its own).
So, yes, Mono is primarily "just" a development platform. But it is a free, open source development platform that is quite good and powerful.
You are, for free as in beer, providing Microsoft with a well-made tool for helping them to get people to migrate off of UNIX and onto Windows.
.NET compatibility, such as it is, is a way for Windows programmers to migrate to UNIX/Linux. That seems like a good thing to me.
.NET because they usually have lots of dependencies on Linux libraries. For example, of all the screen shots on the Mono home site, almost all of them are Gtk# based.
Mono's
Migration in the other direction is not a real concern when using Mono: most Mono applications will simply not easily work on Microsoft
Of course, if you work at it, you can use Mono to write cross-platform applications that also run on Windows, but, then, you can do the same thing with wxWindows, FLTK, Python, Perl and lots of other toolkits and environments. Mono doesn't keep you from writing something that will work on Windows, but it certainly doesn't encourage it.
The real danger is Sun. It is Sun's stated goal with Java to make the underlying operating system or native toolkit irrelevant. That is Sun really is trying to migrate people off of Linux by getting programmers to develop in and for Java APIs. Of course, they would like to migrate them to Solaris, but they might more readily migrate to Windows. Even without all the legal b.s. from Sun, that alone would constitute a big threat to Linux.
Without SUN UNIX would not have been more successful. I was a PhD student around 1990, and at this time it was SUN, HP and IBM fighting for the academic world (giving away cheap workstations etc).
So, you are saying Sun tried to compete on price, and I agree. But how did that help UNIX in the long run? Sun cut prices but, like all other UNIX vendors, they failed to create a thriving market in commercial or open source end user applications.
SUN was never alone in the market, they had fierce competition from other UNIX vendors. If SUN was really so bad w.r.t. GUI and API's, then why did they gain most of the marketshare?
As you point out yourself, they were cheap and they came out of a university environment and marketed to a university environment.
It was a free market with open competition, so I cannot see how a single company would have been able to do damage to UNIX.
Sun's war on X11 and Motif did quite a bit of damage and in the end was futile. And bad standards like NFS also hurt the UNIX world quite a bit. So, yes, a single company can do quite a bit of damage to UNIX.
And Sun is gearing up to repeat history: they are trying again to replace UNIX and POSIX with a set of what amounts to Sun-proprietary APIs. As before, they will lose in the end, but they will do some damage in the process.
If any, it was the collective vendors that spoilt it, not SUN alone. By the way, do you remember the X11-GUI war between OpenLook (sun) with a good looking and easy to program toolkit and interesting windowing system (News, which is very much like Apples display postscript of today) against the rest of the UNIX world pushing for g*dawful Motif?
I remember it very well: both OpenWindows and NeWS were poorly conceived, poorly executed, and proprietary. They lost in the market because they were bad. The fact that they lost to X11 and Motif tells you how bad they really were. Keep in mind that this really was a user choice: people got OpenWindows and NeWS preinstalled and they often still installed MIT X11.
If SUN would have had their way, UNIX on the desktop would have been in a much better shape.
No, it would merely have been Sun-proprietary. But, if at first you don't succeed, try again. So, Sun is trying again to create a proprietary UI by replacing X11 APIs with Java APIs. And they are about as successful as they were with NeWS, which is to say, not very much.
You see, Sun really wants to be just like Microsoft: they want to control the APIs and see the money rolling in. The part Sun didn't get back then and the part Sun still isn't getting is that creating a good API is neither necessary nor sufficient for getting lots of people to adopt it.
I think the point was that if you think it is an inaccurate view, maybe you should get the mono people to rewrite their documentation to reflect more accurately where the project is and where it's going.
.NET interoperability--keep in mind that there are businesses associated with the Mono project, and they do want to use it to allow Windows programmers to migrate to Linux. Or they may simply not have caught up with reality--that .NET compatibility for Mono has turned out to be far less important and that Gtk#/Gnome# are far more important than they originally were.
Well, the fact that almost all the screenshots are from Gtk# applications and that there is an entire Gtk# wiki should make this pretty obvious. And the FAQ talks at length about how to deal with the possibility that Microsoft might try to shut down the project using their single, unissued patent (provided they get it, of course).
Still, I agree that it would be nice for the Mono project to emphasize its widespread use of open source APIs more in the mission statement. On the other hand, they may have their own reasons for emphasizing that they are striving for
Please, don't forget who your enemy is and who your friend is, take just a bit of historical perspective!
Corporate history is no guarantee whatsoever for future actions. Look at Caldera/SCO: one day, they are a big open source Linux company, then they turn around and start taking people to court over code they themselves contributed.
Without SUN UNIX would have died a long time ago (in the real corporate world that is) and thus Linux and FreeBSD would have been much less relevant as well.
Or maybe UNIX would actually be more successful. Sun's incompetent and unsuccessful attempts at giving a good UI to UNIX, their insecure, unreliable, and unscalable network file system, and their painful and misguided attempts to switch to SysV APIs did a lot of damage to UNIX. Sun could have had the entire academic market and from there conquered the desktop with UNIX; instead, they drove their academic customers into the arms of Microsoft and Apple by delivering machines that can only be described as user-hostile to anybody but hardcore UNIX users.
Do you know who is our enemy and who is our friend?
No, and the concept doesn't make sense when applied to companies. Companies make short-term, profit maximizing decisions. Sun does that as much as Microsoft. It just happens to be the case that Sun is in a position where some open source releases are useful for them right now.
Therefore, I care about specifics: licenses, contracts, standards. The licenses and agreements under which Sun distributes Java are very dangerous for open source developers, as is Sun's extensive patent portfolio on Java-related inventions.
Fact is, SUN has NEVER misused its control over API's, but instead has given away many products and standards without ever playing tricks or misusing its position (think about virtually any UNIX network protocol and influence in most UNIX API's).
Well, and for most of their history, Sun was fairly rich and successful. Now they aren't anymore, and that's why there is a serious risk that their policies will change. In fact, as far as I am concerned, Sun is already "misusing" their control over certain APIs, namely the Java APIs.
Yes, Ximian launched mono to create an open source implementation of .NET. How does that contradict what I said? How does the reason for which Ximian launched Mono have anything to do with where the project is going, how people are using it, or what control Microsoft has over it?
.NET, namely ECMA C#, is supported by Mono and makes a great programming platform when combined with open source toolkits like Gtk+ and Gnome. Microsoft has no claims against that combination, and, in fact, I doubt they really mind that it exists. And it is that combination that successful open source Mono projects will be written in, not the .NET APIs.
.NET APIs of Mono are cumbersome and inefficient; they will be useful to have around for compatibility, but if they went away, it wouldn't matter much at this point. Their function is to allow Microsoft programmers to migrate more easily to Linux. There is no risk in trying to offer that capability: at worst, Microsoft might succeed in shutting down full .NET compatibility, which would leave us no worse off than we are now.
.NET is even valid or enforceable remains to be seen; so far, they don't even have a patent.
The fact remains that a free and open subset of
In contrast, the full
Whether any other claim by Microsoft on
Microsoft has several means at their disposal to effectively shut down Mono if it should ever gain critical mass.
.NET compatibility and that Mono applications are .NET applications. That's, in fact, just false. Most current uses of Mono are based on ECMA C# and Gtk#, not .NET. In fact, one of the big strengths of C# is that, unlike Java, C# makes it easy to reuse existing C and C++ libraries; in that, it is much like the relationship of C++ to C. If you already know Gnome, you can start using C# to develop Gnome applications much more easily than picking up Java and Swing (and the Mono/Gtk# applications will work better, too).
.NET, just don't use .NET. In fact, I wouldn't touch .NET simply because I think it's technically not very good. But you can still use Mono, which is shaping up to be a great, general-purpose programming platform. And because existing open source libraries, like Gtk+, Gnome, expat, X11, etc., is so easily accessible, it's very easy to start using Mono--it's just a nicer version of C++.
Those claims are based on the inaccurate perception that the success of Mono depends on
The company to worry about is Sun: open source Java applications do use all-Sun APIs; interfacing with native libraries is just too much hassle, and that's no accident: Sun wants you to use their APIs and give up on the free, open source APIs. And, despite all the JCP mumbo-jumbo, Sun has a lot of control over the Java platform, through numerous patents, through owning key parts of the actual implementation of key parts of the Java platform (e.g., Swing), and through their ownership of the specification and the certification process.
So, if you are worried about Microsoft's ownership of
There are plenty of solar-powered calculators that work in normal light. They make a decent backup for when you are out of batteries or your plane is going down...
That would be a major change for KDE, given that Qt is such an integral part. However, it might actually be a good way for KDE to eliminate their dependency on TrollTech and replace Qt with something LGPL'ed--the Qt license has been a major obstacle to KDE's acceptance I believe.
SVG is already widely used and supported in Gnome.
Lastly, rendering speed improvements could be realized. Aside from dedicated HW doing the rendering, if the OS did it in a trusted manner then it could be faster than many libraries and/or programmer hacks.
Yeah, right, let's put everything into the kernel. You can see how well that worked for Windows.
Both Linux and Macintosh have client/server GUI systems that do rendering not in the OS but in user mode processes. That's the right way to do it. Something as complex as a GUI does not belong into the kernel (the "OS"). However, display-side SVG rendering in X11 would be a welcome addition, and that is being planned. That will give people the same kind of features that Apple gets with DisplayPDF, but with a somewhat cleaner and far more standards-compliant graphics language (SVG vs. PDF).
The smaller reasons include much less data - if a graphics card rendered SVG, then the connection between the computer and the card could be slow and very long distance. Hard drives space isn't an issue, except in power-conscious embedded areas where smaller graphics files could make a difference.
Pretty much all graphics cards these days have 2D acceleration. Therefore, computers already communicate them using vector graphics operations.
Objects can be clearly rendered in SVG regardless of the output device resolution.
You're confusing representation with graphics model. SVG gives you a text-based representation of an object-based graphics model. You don't need that for geometric scaling. Postscript, DisplayPostscript, OpenGL, GDI+, and XRENDER all give you geometric scalability without anything like SVG.
Furthermore, rescaling GUI objects is not simply scaling them geometrically. Support for scaling graphics in the the window system will help somewhat with scaling over a limited range, but it won't magically make a UI or even an icon scale from PDA to huge screen sizes.
SVG is useful and it will come. But it isn't the panacea that you seem to think it is, and it isn't all that new either.
Seriously, for a single home-use desktop, how much of the extra stuff is absolutely necessary?
Most of it. That "extra stuff" isn't functionality, it's all the infrastructure for making X11 run fast on modern desktop machines with large screens and lots of windows.
Perhaps if someone would release a similar slimmed-down version for desktops, then people would quit bitching about it.
The functionality is in X11 because people want it in their desktops. And why slim it down when X11 is already slimmer and faster than Windows or MacOS?
But dont let them compete against others and pretend that they're accomplishing anything.
And what are olympic athletes "accomplishing" right now?
The drama of sports is seeing what the human body can achieve through disciplined training and talent.
So, if someone's performance happens to be the product of a well-chosen mating, it's a "drama", but if it's the product of gene therapy, it's not? Sorry, I don't see it. I think the olympics are a carnival side show, with or without doping or gene therapy.
The US doesn't have have, and really never had, a true free market.
Quite right. And the US attained its greatest strength was when its markets were actually less free than they are now.
Look for totalitarian states and you will find abysmal economies. Look for strong socialistic countries and you will find better, but still poor economies. Your strong economies are found in nations with moderate to light socialism. In other words, where the government doesn't try to manage all aspects of the market.
Quite true again. But in order to get such markets and such a society, a lot of things need to come together. For the US to be able to do this, a lot of factors had to come together.
But it has been relatively freer than those in many other nations. This isn't a US thing. Europe and Australia also have markets relatively freer than most other nations.
And as I was saying, I think the US is finding out that merely having a "freer" market is not sufficient.
X11 worked fine on 20MHz 68k workstations with 4Mbytes of RAM. In the days of 400MHz PDAs with 64M+ of RAM, there really is no need to "replace" X11, in particular given their stamp-sized screens. And because X11 is neutral on policy, unlike other window systems, you can build a great handheld or embedded environment on top of it. In practice, of course, X11 is already more efficient in both memory and speed than either Qt/Embedded or the Windows GDI, so it is also hard to see why people think they can do better.
Besides, there is another problem with GPL'ed toolkits: the competition lets commercial developers develop for free. GPL'ed toolkits and window systems are no good: any Linux-based PDA is only going to succeed widely if commercial developers can develop for it for free. That's a bargain free software has always been making (e.g., you can develop commercial software with gcc for Linux and not open source it). Toolkits need to have licenses like LGPL, X11, or BSD.
So, please forget about all this "GPL'ed embedded toolkit" nonsense. Use X11 and one of several embedded toolkits for handheld development: wxWindows, FLTK, or GPE.
The default XFree86 port to handhelds, of course, has a lot of useless overhead in it because it is just a recompilation of a desktop server. That can be reduced, but you don't need to invent a new, incompatible window system to do it. An embedded X11 server can probably be stripped down to 100-200k. That is where the effort of people looking to create "efficient" embedded servers should be directed.