Operations Technical Procedure Template: Fill & Download for Free

Your question brought an analogy to my mind.“Are old spoken languages like English, French, and Spanish still used? If so, why aren’t the documents modernized?”Absurd? Yes, of course. Why? Human spoken and written languages like English, French, and Spanish are still heavily used and are up-to-date. In fact, this answer is one such example; it’s in English!Similarly, computer programming languages such as Fortran and Cobol are still heavily used and are up-to-date. Thus, the languages in question are already modernized.Your question contains a very important implicit assumption that is false: Old = obsolete. This is simply not true.To see why, let’s look at the case of human spoken and written languages. There are documents written in some form of English that go back as far as 500 AD. That’s over 1500 years of use as a distinct language. During that time, English has kept up with the times by changing its vocabulary and grammatical structure to suit the needs of its current generations of users. That is true also for French, Spanish, and most other Indo-European languages. Also true for currently used non-Indo-European languages. Heck, Chinese has been around for thousands of years and it is still heavily used and up-to-date, in spite of the fact that its primary writing system uses 70,000 ideographs.The same principles apply to computer programming languages. Like human languages, computer programming languages die when they no longer fulfill the needs of the current generations of users. Computer programming languages survive and thrive when they continue to fulfill the needs of users.I don’t know much about Cobol. I took one course in it in 1979 and have never used it since. However, I can speak a great deal about Fortran. I learned it in 1975 and have used it continuously since then. It is still my preferred programming language for new development. I also served for 9 years on the ISO/IEC/ANSI Fortran Standards Technical Committee, the committee that develops new versions of the Fortran International Standard for the ISO.Fortran is very much a modern, up-to-date programming language. As of today (September 30, 2018), there are six (6) published International Standards for Fortran: FORTRAN 66, FORTRAN 77, Fortran 90, Fortran 95, Fortran 2003, and Fortran 2008. Within a few months there will be a 7th: Fortran 2018. Plans are already underway for developing the feature set for Fortran 202x. So the ISO is keeping the Fortran programming language up to date. BTW, procedure templates are at the top of the feature list for Fortran 202x.Fortran has many features that are found in most modern programming languages. It also has quite a few features that greatly aid in software development but are in few, if any, programming languages. Here is a brief and far from exhaustive list:Modules. Data and procedures can be encapsulated in a coherent programing unit.Derived Types, aka data structures.PointersObject-Oriented Programming (OOP): Inheritance, Type-bound procedures, encapsulation, polymorphism, pointers to procedures.Explicit interfacesOptional procedure argumentsUser-defined generic proceduresUser-defined operators and assignmentKeyword-based procedure argumentsWhole-array operationsArray-slice operationsAllocatable arrays (at run time)Character strings are first-class data typesComplex numbers are first-class data typesMultiple representation methods (kinds) of all 5 intrinsic data types (real, complex, integer, character, and logical (Boolean))User-defined parameterized derived typesAsynchronous I/OBoth record-based and stream-based I/OUser-defined derived-type I/O proceduresLong variable names (up to 63 characters)Interoperability with CIEEE floating-point exception handlingParallel processing (via Co-Arrays)Submodules, i.e., separation of module interface from module implementationHope this helps.

What should people stop doing when writing C++ code?Stop using macrosLeave text preprocessing in the 1980’s (where it should have rightfully died).If you’re using a macro and you can’t formally explain why you’re not instead using another language feature, then there’s at least a 90% chance you don’t know what you’re doing.Stop using inheritance to solve every problemComposition is better. The first thing the average CS sophomore should learn to re-contextualize is inheritance. It is the best solution to 10% of your problems, and the worst solution to 90% of your problems. Stop sprinkling it everywhere like it’s your favorite sriracha.Stop reinventing the wheel to solve common problemsThere’s almost an STL library for everything these days. Don’t shim in your own pthreads wrapper when std::thread will do just fine. Don’t typedef some arcane C-function pointer and write your own templates when you can just throw std::function in to get the job done.Stop copying and pasting1000 SLOC where 100 SLOC would do isn’t 10-times the technical value, it is 10-times the technical debt.Keep your code DRY. Throw in a template or two, use std::function and std::transform to iterate structures and operate procedures upon their members. If you need 4 nested loops then there’s no reason to spread that unreadable bug-ridden mess to a dozen locations.If you’re doing the same in 5 places and a template, inheritance, or another C++ language construct won’t consolidate these then this is the rare occasion when reluctant macro use might be needed, but at the back of your mind those 5 lines where you’ve sprinkled macros to consolidate should always haunt you like a long lost first love turned bitter. At least until you wake from a dead sleep in cold sweats, unable to rest until you find out what you did wrong and refactor.Stop adding using directives all over the place, and NEVER use them in a headerYou’re littering your filthy namespace all over the place, and potentially changing linking and program behavior in ways you cannot always anticipate.Worse yet, if this goes into a header then you don’t know how this will impact user code. A user might add your header to their file to implement a new feature, and simply because you wanted to throw in a using namespace X; somewhere in your header their driver code now executes X:read() and X::write() instead of their UserNS::read() and UserNS::write() functions and they have no idea why their code suddenly has new bugs in old working features based on nothing directly to do with the new features they were implementing.There’s some wiggle room on that never, but the only exception is when your using directive rests inside a closed-scope within a template header. Even this should make you uncomfortable, and the proper solution would still be to fully qualify the namespaces you use in that header at their points of usage.

No.the ternary operator is technically an expression with a return value. An if-then-else is a procedural structure (a statement)The are not semantically equivalent: to construct the effect of a ternary-op you need to at least associate a return-like variable to an if.To create an if using a ternary op, you have to transform into expressions the “if” branches (so they have to return something: not all statements do that)The ternary operator is also the default construct of of the “common type” concept of the standard library represented by std::common_type.See std::common_type - cppreference.comIt is vital for template meta-programming as well.