On Mon, 2013-09-09 at 11:07 +0200, Ramon wrote: […] > Is it? Why compete? The only way to attracts large numbers of C++ > developers is to become more and more like C++ (incl. of course, > massive amounts of libraries and tools) and to end up as some > kind of C+++.
The "space of the game" is native code applications. The players currently are Fortran, C, C++, D, Go, Rust, Haskell, OCaml. There are others but they are second division rather than first division. Thus, almost by definition D is competing with C++ for use statistics. > Python is similar to - nothing (commonly used) - and yet it grew > wildly. There are so many to complain about Python's weird > indentation syntax. And yet they come and use it. Because it > promises something tangible and it delivers. Because there is > "the Python way". Because there excellent docs. And because there > is no real competitor. I agree, currently, and quite bizarrely, Python is unique amongst programming languages in that it is seen as the natural partner of native code components. > Had van Rossum tried to please the perl crowd, he might have > attracted some more and quicker but today Python would be a small > niche thingy nobody'd care much about. Python and Perl did compete but they did so head on. It was a philosophical "head on" so compromise was never an issue! > I feel we should largely ignore C++. I feel that D is grossly > inconsequent in a) - very smartly - aiming to be what C++ wanted > to be and b) - not at all smartly - trying to please the C++ > crowd and to mimick C++ up to the point of at least seriously > considering mimicking leper and plague of C++, too. > > D already *is* what C++ wanted to be, namely a more modern C with > OO. D shouldn't measure itself against C++ but rather against > what C++ wanted to be. > > And there is another immensely important factor: reliability and > safety. > > This world gets ever more dependent on software - and software is > ever more recognized as unreliable and insecure; hell, there is > even an industry living from that (anti virus, anti-malware, etc, > etc). > > THAT's the sweet spot. To be what C++ wanted to be - plus - a > strong focus on reliability and safety. C++11 has revitalized C++ in ways that are only just showing themselves. This is a threat to D gaining traction. I am confident D can win the battle for the hearts and minds of native code programmers over C++, but it remains a "head to head" and C++ is established and accepted. D is the newcomer and has to dislodge entrenched position. There will be an interesting analogy with Java 8 in JVM land. > The Ada people are not stupid. There is a good reason for them to > ponder a year or longer over a new keyword. Bertrand Meyer may > have it implemented in a way that looks strange to many but that > man isn't stupid at all. The lesson to learn from those two > languages known for reliability? Have a tight definition and > think long and hard before you make the slightest changes. And > *always* keep your "guiding principles" in mind. Ada and Eiffel are niche languages, Eiffel more so than Ada. Whether good, bad, or doesn't matter this is the case: they are languages used in a very small domain, and even there C++ is allowed. -- Russel. ============================================================================= Dr Russel Winder t: +44 20 7585 2200 voip: sip:russel.win...@ekiga.net 41 Buckmaster Road m: +44 7770 465 077 xmpp: rus...@winder.org.uk London SW11 1EN, UK w: www.russel.org.uk skype: russel_winder
signature.asc
Description: This is a digitally signed message part