QuoteOriginally posted by: ellenQuoteOriginally posted by: CuchulainnOne of the shortcomings I find with C++ in finance apps (a lot of which is numerical) is the lack of good support for functional programming.So, I think the language is not the problem but the software paradigm is more important. Unfortunately, C#, Java and C++ are used too one-sided (OO paradigm) and others are just as important (generic, functional,...)I think having a peep at Haskell and F#... //I hardly ever hear people talking about design and pattern here, just language which is a non-issue. It's like building a house without the drawings. A good design can be implemented in any language.Hallelujah!!! You just answered your own question.Story of my life just kiddingNow I have some questionsThree Questions

QuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuote Seriously, you are asking more of a economic question....i.e. technology will always evolve ...Nope, I am not asking an economic question; this is a software language thread and I was referring to very specific issues, namely the pervasiveness of C/C++ in critical systems.Why won't C++ evolve in the same way that COBOL did? COBOL was in all manner of critical enterprise systems prior to 2000. But the move from roll-your-own enterprise IT systems to off-the-shelf ERP system meant that COBOL became marginalized.Very good prospective.C++ will not evolve to the fullest extend because most software developers do not write code in notepad or vi anymore. (well, maybe Cuchulainn does ). 90% of C++ IDEs production are controlled by a single monopoly and tied to a single OS and Framework. If you believe on the contrary, your should put your money on MSFT and go 'buy and hold' This is actually not a joke, I know people who put their money on J++ 10 years back.

QuoteOriginally posted by: ellenQuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: CuchulainnQuote Seriously, you are asking more of a economic question....i.e. technology will always evolve ...Nope, I am not asking an economic question; this is a software language thread and I was referring to very specific issues, namely the pervasiveness of C/C++ in critical systems.Why won't C++ evolve in the same way that COBOL did? COBOL was in all manner of critical enterprise systems prior to 2000. But the move from roll-your-own enterprise IT systems to off-the-shelf ERP system meant that COBOL became marginalized.Very good prospective.C++ will not evolve to the fullest extend because most software developers do not write code in notepad or vi anymore. (well, maybe Cuchulainn does ). 90% of C++ IDEs production are controlled by a single monopoly and tied to a single OS and Framework. If you believe on the contrary, your should put your money on MSFT and go 'buy and hold' This is actually not a joke, I know people who put their money on J++ 10 years back.Eclipse + CDT? J++ died a miserable death because people realised MS tried to release a proprietary form of the Java language which wasn't portable -> defeating the whole purpose of the language. The IDE had nothing to do with it. What IDE do people developing C++ on Linux (KDE etc) use?Languages survive because they're suited to a particular task, not because they have a fancy IDE.

Quotemost software developers do not write code in notepad or vi anymore. (well, maybe Cuchulainn does ).Nah, I use a line editor, vi is for wimps Edlin is a great editor, secure and fast. Just type 'edlin' <filename> at the DOS prompt and see what happens. here

QuoteOriginally posted by: dirtydroogQuoteOriginally posted by: ellenC++ will not evolve to the fullest extend because most software developers do not write code in notepad or vi anymore. (well, maybe Cuchulainn does ). 90% of C++ IDEs production are controlled by a single monopoly and tied to a single OS and Framework. If you believe on the contrary, your should put your money on MSFT and go 'buy and hold' This is actually not a joke, I know people who put their money on J++ 10 years back.Eclipse + CDT? J++ died a miserable death because people realised MS tried to release a proprietary form of the Java language which wasn't portable -> defeating the whole purpose of the language. The IDE had nothing to do with it. What IDE do people developing C++ on Linux (KDE etc) use?Languages survive because they're suited to a particular task, not because they have a fancy IDE.But doesn't the IDE play a role in this? A good IDE can broaden the applicability of a language or help correct for deficiencies in language. To the extent that the editor aids in writing correct syntax or the compiler/debugger is smart at pinpointing faults, is the extent that a deficient language can become sufficient.

QuoteOriginally posted by: dirtydroogQuoteOriginally posted by: ellenEclipse + CDT? J++ died a miserable death because people realised MS tried to release a proprietary form of the Java language which wasn't portable -> defeating the whole purpose of the language. The IDE had nothing to do with it. What IDE do people developing C++ on Linux (KDE etc) use?Languages survive because they're suited to a particular task, not because they have a fancy IDE.This is right, but what is the market share for Eclipse + CDT users? Do financial engineers use it? We know (just by browsing this forum) that many use some form of combination of .Net, VC+, C#, Excel provided by monopolized entity. We also know that some computer science students and Cuchulainn use Edlin to write C++. What about others? Will be nice to know what people use for C++ environment today... that will give us a better take on its future. Ho: more then 80% of software engineers use Microsoft IDE to write C++H1: less then 80% use Microsoft IDE to write C++p < 0.05Test: Reject Ho if favor of H1

QuoteWill be nice to know what people use for C++ environment today... that will give us a better take on its future. Most of the people I know use VS2005 and Excel 2003. But with new developments in h/w I reckon that Linux interoperability will be important. Some sav vy folks use C++/CLI in combination with C#/.NET. Now that is an idea If you work on a green field project, C# might be an idea. The big question is if it would evolve into another language. ////BTW I was kidding about edlin; Hollerith is safer. Easy to backup. And then paper tape is easy to debug.

QuoteOriginally posted by: CuchulainnQuoteWill be nice to know what people use for C++ environment today... that will give us a better take on its future. Most of the people I know use VS2005 and Excel 2003. But with new developments in h/w I reckon that Linux interoperability will be important. Some sav vy folks use C++/CLI in combination with C#/.NET. Now that is an idea If you work on a green field project, C# might be an idea. The big question is if it would evolve into another language. ////BTW I was kidding about edlin; Hollerith is safer. Easy to backup. And then paper tape is easy to debug. Thats right, take the C# out of .Net and you'll get Java on steroids... that will definitely evolve for some time. Unfortunately monopolies have different prospective.

QuoteOriginally posted by: ellenHallelujah!!! You just answered your own question.But just because a language does not have all the features in my ideal wish-list does not imply that it will not be used in industry. What happens is that developers either use workarounds (e.g. function objects in C++) or do not support the feature. That's a kind of law of physics in IT land.And if change is needed, then so be it. Moving from C++ to any other language is much easier than moving from language X to C++.//A hardware analogy is VHS versus Betamax or Apple II versus IBM PC...

What would be nice would be to have support in C++ for FORmula TRANslation, for example applications that involve algorithms whose output is input for other algorithms. This should be as generic and as extendible as possible. Like doing maths.

I would just like to put my pimp hat on here...Arguing about the merits and problems of C++ is a perfectly rational debate.However, some newbies may choose to read this as some endorsement of the failure to learn C++, and stick with Java and VBA.From a careers perspective that is a mistake.My view is C++ is currently the best option for coding up finance apps, but if I were to be wrong in that, it will still take 5 years at a minimum from now for Java or C# to become languages worth learning for this line of work.

Dominic,My posts were meant to give a balanced account and not fall into the _hammer_and_nail_ trap. The power of C++ is its malleability ... it can be extended to do anyhing. What I hear from people who choose other languages is that they hit a stone way, sooner or later. QuoteMy view is C++ is currently the best option for coding up finance apps, but if I were to be wrong in that, it will still take 5 years at a minimum from now for Java or C# to become languages worth learning for this line of work.I agree. My feeling is that Java has had 10 years the time to oust C++ for this (very) technical area and it has not succeeded, so it probably will not happen. C# is another story, but I agree with the excellent remarks by Athletico on this one.

Java may not be the option in QF.If some people view that C++ is a superset of C# (I may be wrong) in some aspects, why can't C#?If C# is not good enough, what is the main issue to be an alternative?I really want to know what C# cannot do until many think that C++ is the only language to learn for QF application? Bang my head, please.QuoteOriginally posted by: DCFCI would just like to put my pimp hat on here...Arguing about the merits and problems of C++ is a perfectly rational debate.However, some newbies may choose to read this as some endorsement of the failure to learn C++, and stick with Java and VBA.From a careers perspective that is a mistake.My view is C++ is currently the best option for coding up finance apps, but if I were to be wrong in that, it will still take 5 years at a minimum from now for Java or C# to become languages worth learning for this line of work.

QuoteOriginally posted by: CuchulainnDominic,My posts were meant to give a balanced account and not fall into the _hammer_and_nail_ trap. The power of C++ is its malleability ... it can be extended to do anyhing. What I hear from people who choose other languages is that they hit a stone way, sooner or later. QuoteMy view is C++ is currently the best option for coding up finance apps, but if I were to be wrong in that, it will still take 5 years at a minimum from now for Java or C# to become languages worth learning for this line of work.I agree. My feeling is that Java has had 10 years the time to oust C++ for this (very) technical area and it has not succeeded, so it probably will not happen. C# is another story, but I agree with the excellent remarks by Athletico on this one.Why are C# and Java different? The former is practically a clone of the latter.

QuoteWhy are C# and Java different? The former is practically a clone of the latter. They are very similar, sure with the qualifications. C# learned from Java's mistakes. C# has Winforms which beats AWT/Swing hands-down. C# works with Excel. C# has built-in generics; Java generics is just a wrapper. Java does not support operators (important for maths folk). C# and C++/CLI can be integrated into one application. MS is king of the desktop. Java is better than C# for distributed/concurrent computing . I don't like the way Java was designed; deriving everything from object is heresy IMO. C# has the same problem, it's a Smalltalk legacy. Stroustrup was careful to avoid this trap.That's the way I see it Apart from these minor differences, they are indeed very similar.