Isn't it a bit late to be worrying about the consequences of MS asserting its IPR?
Or, to turn the problem around, what led you to believe that an exact clone of the CLR and Dotnet APIs would not be challenged by MS?
I'm afraid I doubt if PR concerns will count for much when half the company's revenue is at risk from OS clones, after all, bad press has hardly deterred MS from fighting the antitrust case with everything it had.
Go ahead and try your luck, but I regret that I for one shall not be reaching into my pocket to support the right-to-slavishly-mimic-MS legal fund.
I don't see anybody stating that Linux VMs are a novel idea, far from it.
One of the reasons that Mono got peoples' backs up was that high-quality efforts such as Kawa and Parrot were already well established, and both of these targeted a number of different languages.
There was no need to copy Microsoft verbatim, since the goal of complete portability and interoperability (e.g. Photoshop for Linux) was very unlikely to be achieved. Therefore, development should have continued with one of the existing projects, incorporating any new ideas as appropriate.
One aspect that you gloss over is the utility of having the intermediate language ("bytecode") close or equivalent to the source language. The clear trend of modern development tools is to rely on introspecting components and representing them as source, even if only as an API. These kind of advances are precluded if you have a large gap between the two.
Yes, you should stop work on Mono now as you may be making users liable to lawsuits from the owner(s) of Dotnet patents.
No, SAMBA should not be halted, since protocols are much harder to patent than code, although Andrew Tridgell has certainly recommended inventing a new CIFS.
Perhaps you can point me to the formal definition of a scripting language (Python) vs. a general purpose programming language (LISP)?
My argument never mentioned obfuscation, the point is about semantic equivalence of distributed code vs. manipulated code.
Lastly, for Java-class languages, efficiency is no longer the concern of source-to-bytecode translation, but primarily of bytecode-to-machine-code translation. The existence of decompilers is proof of this - the decompiled code does not differ substantially from the original source, therefore no significant optimization has taken place.
Maybe now an OSS equivalent [...].NET will be developed
Indeed.
One interesting angle is that an OSS VM can very naturally enforce open source. While Java bytecode can be turned back into source code, minus the comments and some stylistic info, an OSS-centric VM could have a bytecode that was exactly equivalent to the source code. This way, it would be effectively impossible to ship anything other than the source.
It's been obvious for 40 years (LISP) or maybe 55 (Turing's ACE Report) that programs-are-data, and tools today like IBM Eclipse go as far as they reasonably can to treat Java this way.
OSS has the opportunity to steal a march on Java and Dotnet and converge the worlds of users and developers. This is a natural evolutionary step, but OSS is only model that has a strong reason to promote it.
Do you have benchmark figures indicating that Java is slower than CLR code? My limited experience is that the reverse is true by a considerable margin, plus there's the theoretical argument that run-time optimization (Java JIT) will always beat compile-time optimization (IL compiler).
I've also taken a PC-developed Java application and deployed it on a 64-way machine, achiving near linear scalability. Has anyone achieved similar results with Dotnet?
One might have entertained the hope that this application would bring Mono advocates back into the world occupied by the rest of us.
But based on past experience, hack first and ask questions afterwards is likely to remain the general response, the absurd naivety of De Icaza continuing to infect a whole group of people who might otherwise be able to contribute a lot to open source developments.
Developers and especially users of Mono are now obliged to navigate a complex morass of ownership claims and associated policies. Unlike the carefree, here today, gone tomorrow leet hackers, 'legaleeze' will apply to those of us investing real resources into Mono in considerable force and with potentially unlimited extent.
You aren't, or you weren't - AFAIR the RDB project at IBM - System R*? - originally called the language SEQUEL (hence Ingres QUEL, which might be the commercial product someone else remembers).
However the IBM TLA police were called in (they turned a number of products into TLAs for some reason) and officially renamed it S.Q.L., so it's an SQL database these days.
But against this prospect, some way off at least for Europarl, if not the Commission, must be balanced the fact that WTO and related bodies are past masters of the art of 'divide and conquer'.
For example, when seeking to prevent poorer countries producing their own AIDS drugs, the US (acting on behalf of the large pharms) threatened the likes of the Dominican Republic, India and Brazil with sanctions unless they supported the WTO's TRIPS agreements.
Only an economic unit the equal of the USA can face down such threats, and the EU could well develop something of a spine and do so on occasion. Sometimes it pays to be part of a big club!
[OT rant: This is an example of the/. effect where the same debate gets less rather than more informed at each/. appearance. I see many of the usual suspects are missing from the Java vs. Dotnet match this time, so maybe they got bored and this is the cause of the relatively fact-free exposition this time. Groundhog days might be less frequent if there was a better search engine, or/. had a structure beyond what time of day it was when the editor put finger to keyboard.]
Anyway, back to the already well-rehearsed point at issue:
Dotnet does not consist only of C# and the CLI.
Even excluding other languages, the APIs included in the standard are about a tenth of those currently comprising Dotnet (about 120 out of 1200 classes). The bulk of the APIs are for Dotnet components such as Windows Forms, ADO.NET, ASP.NET, Web Services and Web Forms.
Not only are these not standardized, there are no announced plans to standardize them and there are reasons to believe that they are protected by patents (hence the do-your-worst nature of Miguel's note on patent implications).
(why make it.NET compatible at all?) You really have to go out of your way to make things incompatible (after all, the ECMA specs are fairly thorough...)
Whoops! Straight into trap number 1 (Dotnet is not an ECMA standard). And you're an early Mono developer... hate to think how many less well informed people are going to follow you.
...we predict a resurgence of interest in CSV files among diabolical masterminds everywhere.
Just wondering whether this really has anything to do with XML... if so, should it? Not sure when a data structure last had its own universal global framework for supporting rapid discovery and sharing of suspected criminal and terrorist evidence by law enforcement agencies but I suppose this counts as progress of a kind, though personally I think I'd sleep better if the authorities kept an eye on all manifestations of Directed Acyclic Graphs.
But since, according to cladistics, birds are dinosaurs, Milner's statement is correct. There's no evidence that flight evolved more than once in bird ancestors.
One trusts that this isn't representative of the current crop of American science graduates...
If you study science, you may notice that industrial countries do not progress in line.
The UK, and other countries outside the USA, have been responsible for a huge number of advancements since the industrial revolution. The jet engine, cavity magnetron (microwaves), radar, LCDs...
And for the record:
1. Alexander Graham Bell was Scottish
2. Joseph Swann demonstrated the electric light in England some months before Edison
Just finished developing a large app on Linux, demoing it on Windows and deploying it to a huge 64 way Sun box. Yes, it really works... amazing. OK, there's not much of a GUI, but Swing and SWT do work.
Cool. But apparently it's OK to just bend its clauses far enough so that they are effectively abolished, without rousing the average citizen to action.
Sure, but then you can abolish the constitution and make it 'unwritten' (i.e. evolutionary) like here in the UK.
Seems like now you get to have the worst of both worlds - haunted by clearly fossilized policies for things like firearms while enlightened and public-spirited principles are distorted and sidelined.
If only this were true. You'll find that it's the majority verdict, not the dissenters, that attempts to decide what parts of the constitution are still relevant to the modern era, based on whether they are compatibile with international treaties, applicable to digital media etc.
A more brazen attempt as policy-making on the fly would be hard to envisage - certainly Stevens never sullies his judgement with such extraneous factors.
Boy, this is funny. So you're trying to tell us that the majority verdict isn't riddled with appeals to pragmatism, references towards practices elsewhere, best guesses and other forms of 'new millenium handwaving'?
You really need to re-read Stevens to understand how a logical and therefore constitutional case is made. His argument only refers to constitution and case-law, it's the other guys who mix in the latter-day mumbo-jumbo.
To take just one example, a key plank of the majority position is the constitutionality of the CTEA, but, almost incredibly, here's the justification they quote from the CTEA decision:
"[I]n an era of multinational publishers and instantaneous electronic transmission," the court said, "harmonization in this regard has obvious practical benefits" [Ginsbury p7]
I'm sure that in the Hollywood version brave revolutionaries will be asserting their rights to fair use in the face of corporate redcoat henchmen, however in this case history is not on the side of a good drama.
The 1710 Statute of Anne, rather than ushering in "one of the most corrupt periods in English History", was designed to fix the abuses of the previous century. In addition to limited term copyrights, it heralded a freer press, as advocated by John Milton's "Speech for the liberty of unlicensed printing" back in 1644. (Strictly speaking, the royal license requirement, as seen on the front of reproduction Shakespeare folios, had already lapsed in 1688 with the 'glorious revolution').
The Statute of Anne was seen as good practice by the framers, and indeed both this and later findings by Parliament are quoted approvingly in the dissenting finding by Stevens. That's why the clause in the Constitution resembles this statute and not some reformed version.
Next up: decide royalties due to the heirs of John Locke for his contributions to the Constitution:-)
Slashdot Mirror
You wouldn't be confusing the C Sharp language and the CLR with the whole of Dotnet would you?
If not, I'd appreciate a reference where MS states the intention of making ASP.NET, Windows Forms, ADO.NET etc. ECMA standards.
Isn't it a bit late to be worrying about the consequences of MS asserting its IPR?
Or, to turn the problem around, what led you to believe that an exact clone of the CLR and Dotnet APIs would not be challenged by MS?
I'm afraid I doubt if PR concerns will count for much when half the company's revenue is at risk from OS clones, after all, bad press has hardly deterred MS from fighting the antitrust case with everything it had.
Go ahead and try your luck, but I regret that I for one shall not be reaching into my pocket to support the right-to-slavishly-mimic-MS legal fund.
I don't see anybody stating that Linux VMs are a novel idea, far from it.
One of the reasons that Mono got peoples' backs up was that high-quality efforts such as Kawa and Parrot were already well established, and both of these targeted a number of different languages.
There was no need to copy Microsoft verbatim, since the goal of complete portability and interoperability (e.g. Photoshop for Linux) was very unlikely to be achieved. Therefore, development should have continued with one of the existing projects, incorporating any new ideas as appropriate.
One aspect that you gloss over is the utility of having the intermediate language ("bytecode") close or equivalent to the source language. The clear trend of modern development tools is to rely on introspecting components and representing them as source, even if only as an API. These kind of advances are precluded if you have a large gap between the two.
Yes, you should stop work on Mono now as you may be making users liable to lawsuits from the owner(s) of Dotnet patents.
No, SAMBA should not be halted, since protocols are much harder to patent than code, although Andrew Tridgell has certainly recommended inventing a new CIFS.
Perhaps you can point me to the formal definition of a scripting language (Python) vs. a general purpose programming language (LISP)?
My argument never mentioned obfuscation, the point is about semantic equivalence of distributed code vs. manipulated code.
Lastly, for Java-class languages, efficiency is no longer the concern of source-to-bytecode translation, but primarily of bytecode-to-machine-code translation. The existence of decompilers is proof of this - the decompiled code does not differ substantially from the original source, therefore no significant optimization has taken place.
Maybe now an OSS equivalent [...] .NET will be developed
Indeed.
One interesting angle is that an OSS VM can very naturally enforce open source. While Java bytecode can be turned back into source code, minus the comments and some stylistic info, an OSS-centric VM could have a bytecode that was exactly equivalent to the source code. This way, it would be effectively impossible to ship anything other than the source.
It's been obvious for 40 years (LISP) or maybe 55 (Turing's ACE Report) that programs-are-data, and tools today like IBM Eclipse go as far as they reasonably can to treat Java this way.
OSS has the opportunity to steal a march on Java and Dotnet and converge the worlds of users and developers. This is a natural evolutionary step, but OSS is only model that has a strong reason to promote it.
Do you have benchmark figures indicating that Java is slower than CLR code? My limited experience is that the reverse is true by a considerable margin, plus there's the theoretical argument that run-time optimization (Java JIT) will always beat compile-time optimization (IL compiler).
I've also taken a PC-developed Java application and deployed it on a 64-way machine, achiving near linear scalability. Has anyone achieved similar results with Dotnet?
One might have entertained the hope that this application would bring Mono advocates back into the world occupied by the rest of us.
But based on past experience, hack first and ask questions afterwards is likely to remain the general response, the absurd naivety of De Icaza continuing to infect a whole group of people who might otherwise be able to contribute a lot to open source developments.
Developers and especially users of Mono are now obliged to navigate a complex morass of ownership claims and associated policies. Unlike the carefree, here today, gone tomorrow leet hackers, 'legaleeze' will apply to those of us investing real resources into Mono in considerable force and with potentially unlimited extent.
You aren't, or you weren't - AFAIR the RDB project at IBM - System R*? - originally called the language SEQUEL (hence Ingres QUEL, which might be the commercial product someone else remembers).
However the IBM TLA police were called in (they turned a number of products into TLAs for some reason) and officially renamed it S.Q.L., so it's an SQL database these days.
1) Last I heard, only C# and the CLR were ECMA standards - the bulk of the Dotnet APIs remain proprietary.
2) If the ties to Win32 are minimal, why are the Mono guys relying on WINE for their GUI APIs?
3) Why do these tendendious statements appear unmodified every time Dotnet comes up?
But against this prospect, some way off at least for Europarl, if not the Commission, must be balanced the fact that WTO and related bodies are past masters of the art of 'divide and conquer'.
For example, when seeking to prevent poorer countries producing their own AIDS drugs, the US (acting on behalf of the large pharms) threatened the likes of the Dominican Republic, India and Brazil with sanctions unless they supported the WTO's TRIPS agreements.
Only an economic unit the equal of the USA can face down such threats, and the EU could well develop something of a spine and do so on occasion. Sometimes it pays to be part of a big club!
Actually that was trap number 2... never mind.
/. effect where the same debate gets less rather than more informed at each /. appearance. I see many of the usual suspects are missing from the Java vs. Dotnet match this time, so maybe they got bored and this is the cause of the relatively fact-free exposition this time. Groundhog days might be less frequent if there was a better search engine, or /. had a structure beyond what time of day it was when the editor put finger to keyboard.]
[OT rant: This is an example of the
Anyway, back to the already well-rehearsed point at issue:
Dotnet does not consist only of C# and the CLI.
Even excluding other languages, the APIs included in the standard are about a tenth of those currently comprising Dotnet (about 120 out of 1200 classes). The bulk of the APIs are for Dotnet components such as Windows Forms, ADO.NET, ASP.NET, Web Services and Web Forms.
Not only are these not standardized, there are no announced plans to standardize them and there are reasons to believe that they are protected by patents (hence the do-your-worst nature of Miguel's note on patent implications).
(why make it
You really have to go out of your way to make things incompatible (after all, the ECMA specs are fairly thorough...)
Whoops!
Straight into trap number 1 (Dotnet is not an ECMA standard). And you're an early Mono developer... hate to think how many less well informed people are going to follow you.
...we predict a resurgence of interest in CSV files among diabolical masterminds everywhere.
Just wondering whether this really has anything to do with XML... if so, should it? Not sure when a data structure last had its own universal global framework for supporting rapid discovery and sharing of suspected criminal and terrorist evidence by law enforcement agencies but I suppose this counts as progress of a kind, though personally I think I'd sleep better if the authorities kept an eye on all manifestations of Directed Acyclic Graphs.
The reference to fluid dynamics was a bit of a giveaway, wasn't it?
But since, according to cladistics, birds are dinosaurs, Milner's statement is correct. There's no evidence that flight evolved more than once in bird ancestors.
One trusts that this isn't representative of the current crop of American science graduates...
If you study science, you may notice that industrial countries do not progress in line.
The UK, and other countries outside the USA, have been responsible for a huge number of advancements since the industrial revolution. The jet engine, cavity magnetron (microwaves), radar, LCDs...
And for the record:
1. Alexander Graham Bell was Scottish
2. Joseph Swann demonstrated the electric light in England some months before Edison
In case anyone was wondering, that building is still around, as are some historic prints.
FWIW, there used to be a Guildhall University too, but it just merged and changed its name.
me too - Java (as no one else has mentioned it)
Just finished developing a large app on Linux, demoing it on Windows and deploying it to a huge 64 way Sun box. Yes, it really works... amazing. OK, there's not much of a GUI, but Swing and SWT do work.
I took a couple of runs at this, and your other post - I encourage everyone else to do the same!
Very cogent.
Cool. But apparently it's OK to just bend its clauses far enough so that they are effectively abolished, without rousing the average citizen to action.
Sure, but then you can abolish the constitution and make it 'unwritten' (i.e. evolutionary) like here in the UK.
Seems like now you get to have the worst of both worlds - haunted by clearly fossilized policies for things like firearms while enlightened and public-spirited principles are distorted and sidelined.
If only this were true. You'll find that it's the majority verdict, not the dissenters, that attempts to decide what parts of the constitution are still relevant to the modern era, based on whether they are compatibile with international treaties, applicable to digital media etc.
A more brazen attempt as policy-making on the fly would be hard to envisage - certainly Stevens never sullies his judgement with such extraneous factors.
As a European, can I just say thanks a lot?
This means that we can look forward to corresponding legislation on climate change, the international criminal court, gun control etc. right?
Or is it maybe that big-business sponsored WTO programs are the real driving force here, and not a desire to match European practice at all?
Boy, this is funny. So you're trying to tell us that the majority verdict isn't riddled with appeals to pragmatism, references towards practices elsewhere, best guesses and other forms of 'new millenium handwaving'?
You really need to re-read Stevens to understand how a logical and therefore constitutional case is made. His argument only refers to constitution and case-law, it's the other guys who mix in the latter-day mumbo-jumbo.
To take just one example, a key plank of the majority position is the constitutionality of the CTEA, but, almost incredibly, here's the justification they quote from the CTEA decision:
"[I]n an era of multinational publishers and instantaneous electronic transmission," the court said, "harmonization in this
regard has obvious practical benefits" [Ginsbury p7]
So now who's making up policy as they go along?
I'm sure that in the Hollywood version brave revolutionaries will be asserting their rights to fair use in the face of corporate redcoat henchmen, however in this case history is not on the side of a good drama.
:-)
The 1710 Statute of Anne, rather than ushering in "one of the most corrupt periods in English History", was designed to fix the abuses of the previous century. In addition to limited term copyrights, it heralded a freer press, as advocated by John Milton's "Speech for the liberty of unlicensed printing" back in 1644. (Strictly speaking, the royal license requirement, as seen on the front of reproduction Shakespeare folios, had already lapsed in 1688 with the 'glorious revolution').
The Statute of Anne was seen as good practice by the framers, and indeed both this and later findings by Parliament are quoted approvingly in the dissenting finding by Stevens. That's why the clause in the Constitution resembles this statute and not some reformed version.
Next up: decide royalties due to the heirs of John Locke for his contributions to the Constitution