.Net idea of being JIT only is a problem.

What is a buffer overflow. http://en.wikipedia.org/wiki/Buffer_overflow_protection That right running a binary built by a incompetent twit is the only reason that happens who did not turn on protection against it with modern binaries. Modern day C and C++ are just as resistant to buffer overflow as .net and java if you build with the protection options on.

“write once, execute everywhere” Does not require your code to remain JIT for unsuitable environments like LiveCD’s. The means take the byte code of Java and spit out a native binary is a good feature. Gcj is highly useful at times.

You will find us Linux users more tolerant to java than .net because its possible for us to use Java without JIT.

iLia Java also does not give us legal migraines. So Java is a workable Language. Of course people fail to note if there Java programs work converted to native binaries or not. Would be useful to know.

iLia Linux users are complex beasts if you have not worked it out yet. We have issue with Java JIT but we don’t have a major issue with Java itself because we can use it to make native binaries that will work properly with the OS kernel when we need to.

With .net if we could have it spit out native binaries that would leave the legal issue with the developer who made the binary if something goes wrong not us. So we could have it build in a country where patents don’t apply. Problem solved.

Note that you took because I said Java JIT was a problem that we could not still use Java. Limited understanding of this problem as well.

So what do you propose? Write applications in C or C++ using Win32 API or GTK+? Or maybe you forgot about buffer overflow? And what about “write once, execute everywhere?” even with there drawbacks JIT based systems are very good for Linux, simply because they allow developers to create applications for Linux even without caring about this OS, all these Java based IDEs and other tools they are a kind of popular on Linux, aren’t they? Java brings new application to Linux, and these applications are same on all platforms where Java works, so no need to learn new tools in order to be able to work on Linux.

So asking who has been sued over Mono yet its unlikely due to how few use it in production.

Wine due to its development methods is very hard to prove the patent has been infringed. The development method blackbox wine uses does cause problems trying to bring to patent case. Patent can be voided if proven the invention is not unique or creative enough. This has happened with BIOS patents.

Mono is different because its built from a Microsoft Spec so all Microsoft has to manage to prove is the patent grant is voided to attack. The lack of trust of Mono is based on how Microsoft treated FAT and other specs they put out there. Once bitten twice shy. You cannot expect us to be any other way. It is really up to Microsoft to earn the trust back.

To be correct they can because its written in the patent grant so you make a deviates in particular ways and you void the patent grant.

TM Repository
–That’s now how a standard works. Adobe can’t sue me if I create a PDF writer that implements a content format that isn’t in the official PDF spec.–
To be correct Adobe can. In the patent grant Adobe clearly states that all copy protection stuff cannot be broken.

If you create a tool with the express goal of breaking Adobe copy protection on PDF. Adobe is in the legal right to sue you with every patent they have over PDF that you use. So you create system for store the password as plain text inside the pdf as extra content you have now broken PDF standard and now can be sued.

Adobe PDF patent grant is fairly open has a few teeth. With Adobe PDF extra content better not be the password to the PDF file. Other than that they are fair allowing.

TM Repository Microsoft is always very clear in there patent grants that if you implement anything that is not in the documentation they have the right to sue you over what you implement does not matter if there implementation did it or not because they never declare their implementation the reference implementation. Failure to do that cause legal problem when trying to make something compatible.

Adobe PDF grant is different you have the right to go by the reference implementation and request and update to the PDF standard. The reference implementation being Adobe Reader. So if Adobe Reader does something and the written standard says otherwise you can implement as per the Adobe reader but have to place a request for update to standard. Doing that you are legally free and clear.

Its all about the wording of the grant. Every grant MS does they are very tight on the wording. So yes you can be sued because of it.

TM Repository basically you don’t know the legal of this. Pulling in PDF to protect yourself shows how stupid you are. It never cross you mind that adding password of the file in plain format to the PDF file is forbin by PDF rules. Yes it technically possible to breach the rules if you write up a PDF document and part of the content is the password to the file.

Every patent grant has conditions you have to obey them.

“And the most popular Mono applications don’t use them. And many other non standardized namespaces can be easily replaced. So no major problems here.”

Its not something that simple to replace. Build you binary as per standard MS .Net engine will fail to run it at times. Because the alignments are wrong for particular things in the standard. These are things MS has changed to increase performance of .net but they failed to update the standard.

http://opensimulator.org ran into this first hand. Mono project has got around this problem by breaking the standard itself. MS patent grant is only valid if you follow the standard to the letter.

iLia
–Are you afraid of competition from Java and .Net developers?–
There is a simple technical issue at the core of it.

The Linux kernel under memory pressure can drop normal executable code from memory. Linux kernel knows if you attempt to change a binary on the filesystem. So if the binary on the file system is not changed the Linux kernel does not send the executable sections and the non changed data sections of a binary to swap. Instead can just drop it out of memory and use that memory space for something else latter read it back from disc.

As so as you get to JIT you have a problem. You get under memory pressure and the JIT converted code ends up shoved to swap.

gnote vs tomboy is a very good compare. It can show the clear difference they are basically the same program gnote is the C# of tomboy converted to C++. Java compliers exist for a reason. If are running Java programs were you might run out memory building a native binary out of it can be a good thing then the OS kernel work effectively.

The issue comes down to that JIT in Java and .Net have never ever design hooks to work with the OS core effectively when memory pressure happens. It also becomes problematic trying to.

Like if you delete the generated native code from JIT completely you have to waste cpu time regenerating it. If you are on a livecd with no writeable swap sending to swap is not an option. So now you are stuffed because you placed a .net or java binary on a livecd because they take there full memory foot print. People say about JIT can clean up after itself.

Problem is the Linux OOM Killer and windows equal for when the OS is out of memory there is no way for it to ask the JIT applications hey can you please release some stuff I need the memory back.

Basically its a split brain problem. You have two things managing memory not talking properly to each other.

iLia as normal for a MS troll us Linux people have to be some form of bigot there cannot be a major technical issue. MS .net cannot have a major design flaw. MS would never hurt you. Right. Reality does not agree with these statement ment.

Until someone designs a solution to the technical at the core of the JIT memory problem you cannot expect the Linux people to perfectly like java or .net.

You will find Linux people liking vala its like .net syntax we don’t hate the syntax we hate the result.

The reality here you are the bigot iLia. You are being insulting you not willing to believe there is a issue. You believe everything is good. So you never considered the other side might have a point.

Sorry to quote myself, but funny enough, Miguel de Icaza posted this:

http://linuxfonts.narod.ru/files/Miguel.de.Icaza-What.Killed.the.Linux.Desktop.html

He does an honest reflection on why he thinks the Linux desktop has had a rough time and the reasons why he uses OSX now. That’s right, your “Microsoft MVP” runs OSX.

You’ve already begun distancing yourself from him; quick, finish him off! But first ask yourself one question, “is this good for the company?”

http://www.corporatelifesucks.org/images/officespace.jpg

Yes you do. There are plenty of Shell replacements for Windows:

http://en.wikipedia.org/wiki/List_of_alternative_shells_for_Windows

There are explorer replacements:

http://www.techrepublic.com/blog/five-apps/five-free-replacements-for-windows-explorer/1103

There are taskbar replacements and docks:

http://rocketdock.com/
http://www.stardock.com/products/objectdock/

Is your Google broken?

That’s now how a standard works. Adobe can’t sue me if I create a PDF writer that implements a content format that isn’t in the official PDF spec.

Likewise, Apple can’t be sued for making a 3-channel headphone jack that deviates form the 1/8th inch headphone jack spec.

Similarly, Microsoft can’t sue me if I create a .NET implementation that deviates from the spec.

All it means is that I can’t gain full certification that my software is compatible with the spec from standards board.

All you’re doing is inventing hypothetical scenarios based on nonsense.

Wait. You’re upset about loyalty and evangelism? That’s rich. You and your cheer squad here are some of the most desperately devout Linux enthusiasts I’ve seen. You evangelize it with every post! You’ve picked your team/religion and you’ll stand by it no matter what, because being pragmatic would mean having to admit you were wrong some of the time.

You think this makes your righteous, but it really just makes you a fool who refuses to acknowledge the holes in his sinking ship. Were you wise enough to admit the holes were there, then perhaps you could actually save it from sinking. Instead, you hope for divine intervention from a mythical army of developers that you aren’t even a part of.

And if you mean loyalty from their users, hah, good luck. Most users aren’t loyal to Windows or Microsoft in the slightest. Their desktop may run Windows, but they often own Macbooks, iOS devices, android devices, PS3s and Wiis, etc. If you mean their employees, people come and go all the time. There are only a handful of people who’ve actually stuck with the company for very long.