On 12/23/11, David Christensen <dpchr...@holgerdanske.com> wrote: > Someone wrote: >> I am like you and wrote most of my C++ during the >> early years of the language. I used the AT&T Cfront version 1.2 >> compiler for years. Always on Unix machines and never on Windows. >> >> I have become disillusioned with the new C++ that has the kitchen sink >> in it. It has become the new ADA.
Is ADA evil? >> I believe it makes easy programs >> easy and hard programs harder. So now I program in C and haven't >> written much C++ in years. > > I learned Perl in 1998, and have been using it as my primary programming > language/ system since ~2001. Easy stuff is easy, and I can manage > moderate stuff. Brilliant/ workaholic people can make Perl do anything, > but I'm not there. Me neither, but I find that, for certain classes of problems, I'm definitely more productive in perl than in C. > The core language and library are outstanding for a > non-corporate, non-standards body product. But the rest of the library > (CPAN) is a volunteer community effort. Scope and quality of individual > CPAN modules range from good to garbage. And, as they say, "a > programming language is only as good as its library". The stream of > available books has fallen off, although a new Camel book is due shortly. > > > Perl's motto "There's more than one way to do it" (TIMTOWTDI) is great > for "duct tape that holds the Internet together". But the dark side of > TIMTOWTDI is that effort grows exponentially as you cobble together Perl > modules with differing designs and interfaces. "Conceptual integrity" > and "systems programming product" appear to be precluded (ref. Brooks, > "The Mythical Man-Month"). Perl is a naturally evolved language. > I've come to the conclusion that I need another language/ system for > advanced applications; one that meets Brook's ideals. So, I > periodically consider other languages/ systems. Join the crowd. > I learned C++ in 1995 and used it professionally until ~2000. While I > loved C, I have mixed memories of C++. I have positive memories of STL > and mixed memories of Rational Rose. Boost seems to be upcoming and C++ > is fully supported by Umbrello. C and C++ seem to occupy significant > positions in the Linux ecosystem, and are broadly supported on many > platforms. C and C++ are currently about the closest thing you can have to true cross-platform for large-scale apps. The issues you are fighting with are less the language and more the support tools. Underlying that is the variety of disciplines. When the rubber meets the road, there's, ahem, more than one way to do it. (cough) > Java is a corporate product, and doesn't feel truly FOSS (especially > since Oracle). I am leery about its apparent requirement for IDE's -- I > prefer the understanding and control afforded by the shell. And, I've > heard one too many Java factory-factory-factory-factory jokes. I decided something like a decade ago that Java is the new CoBOL. Haven't changed my opinion there. > C# by itself looks appealing, but C#/.NET is corporate and and I can't > stand Windows or the Microsoft ecosystem. Mono seems to be doomed to > playing catch-up. I'd rather invest my efforts into a FOSS language/ > system that has first-class support on all the common platforms and the > possibility for embedded systems. Indeed. > Python has been upcoming for years. Python's "Benevolent Dictator for > Life" and "there should be one -- and preferably only one -- obvious way > to do it" philosophies would seem more likely to achieve Brook's ideals. > But, Python's whitespace sensitivity has always put me off, and the > version 2 => 3 language change still seems to be propagating through the > standard library. Python looks to me like it ought to be a fun replacement for BASIC. I mean that in a good way. > Go is intriguing. It's corporate, but feels FOSS. The designers appear > to have very specific goals in mind (e.g. conceptual integrity). The > language incorporates many desirable advanced features, while eschewing > certain others commonly found in other languages. Code is compact and > regular. Go compiles and runs fast, and is being used "for real stuff". > But, the language is still evolving (version 1 due next year?), > documentation is terse, and the library has limited breadth and depth. > I'm still hoping/ waiting for Google to "pull the cork out" and develop > Go to its full potential. > > > Any other comments/ suggestions regarding programming languages/ systems > for advanced applications on Linux? > > > David Well, I'd tell you about the beautiful language I am working on. Except, it's less working on and more on the back burner. And, since Go is too much in a state of flux for you, I'm sure you don't want to mess with the state of flux my mind is in. My language is based on Forth and it should be ready sometime in the next century. Assuming I can make some sort of deal with God or with my children. (Sigh. My son hates programming. He likes mucking around with the source code for his customized Three Kingdoms MUD. Perl. I've tried to teach him C and it's like he thinks I'm trying to brainwash him for even mentioning it. Heh.) The general advice I've heard is not to limit oneself by language choices. I myself have a hard time finding time to take that advice, but, it seems like good advice. (Should I hit the send button or the discard button?) -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/caar43ipzmkoxo-nftd1t0k_c-+0cszdspit-yw7jwnmd4nt...@mail.gmail.com
