Question

I wanted to know

Python is suited for what kind of applications?

I am new to Python world but I know it’s a scripting language like Perl but I was not sure about the kind of applications which one would build using Python and would certainly appreciate if someone can provide some useful information.

Answer 1

It’s hard to think of kinds of general applications where Python would be unsuitable, but there are several kinds where, like just about all higher-level languages akin to it, it might be considered a peculiar and probably inferior choice.

In “hard real time” applications, all dynamic memory allocation and freeing, and especially garbage collection, are quite understandably frowned upon; this rules out almost all modern languages (including Python, but also Java, C#, etc, etc), since almost all of them rely on dynamic memory handling and garbage collection of some kind or other.

If you’re programming for an “embedded device” which you expect to be produced and sold in huge numbers, every bit of ROM may add measurably to the overall costs, so you want a language focused on squeezing the application down to the last possible bit — any language that relies on a rich supporting runtime environment or operating system (including Python, and, again, also Java, C#, etc, etc) would no doubt force you to spend extra on many more bits of ROM (consider threaded-interpretive languages like good old Forth: they can make a substantial application’s code be measurably more compact than straightforward machine code would!).

There many be other niches that share similar constraints (mostly focused on MEMORY: focus on using as few bits as possible and/or strictly confining execution within precisely predefined limits — no dynamism, no allocation, no garbage collection, etc, etc), and basically the case would once again incline in similar ways (for example, there are server applications, intended to run on myriads of servers, which can save many megabytes per server if coded in C++ [especially if without “allegedly-smart” pointers;-)] rather than Java, Python, C#, and so on).

Of course there are excellent reasons most modern languages (Python, Java, C#, etc) choose to do dynamic memory allocation, garbage collection, and so forth, despite the importance of application niches where those techniques are a negative aspect: essentially, if you can possibly afford such nice memory handling, writing applications becomes MUCH, MUCH easier, and a whole class of problems and bugs connected with the need to carefully manage memory if you lack such support can go away — programmer productivity really soars… IF garbage collection and the like can be afforded at all, that is. For example, if an application was going to run on a few hundreds or thousands of servers, I probably wouldn’t bother coding it in C++ with manual memory management in order to save memory; it’s only at tens and hundreds of thousands of servers, that the economics of all those extra megabytes really kicks in.

Note that, despite the common misconception that “interpreted languages” (ones with a rich underlying runtime or VM, like Java, C#, Python, etc) “are slow”, in fact for most CPU-intensive applications (such as scientific computation), Python is perfectly suitable, as long as the “rich supporting runtime environment” (e.g. numpy) is factored in. So, that is not really a factor — though memory consumption and garbage collection CAN be, in some niches.