What is the difference between runtime and compile-time? [closed]

Question

So what is a runtime? Is it a virtual machine that executes half-compiled code that cannot run on a specific processor. If so, then what's a virtual machine? Is it another software that further translates the half-compiled code to machine specific code? So what if we are talking about one of those languages that don't compile to intermediate code but rather translate/compile directly to machine code. What's a runtime in that situation? is it the hardware (CPU and RAM)?

Also, what's the difference between compile-time and runtime? Are they stages of a software lifecycle. I mean a program is originally a bunch of text files, right? So you compile or translate those to a form of data that then either can be loaded to memory and executed by the processor or if it's a "managed" language, then it would need further compilation before it can run on hardware. What exactly is a managed language?

Lastly, is there such a thing as debug-time and what is it?

I'm in my first term studying computer science, and it really confuses me how illogically things are taught. "Information" is being shoved down my throat, but whenever I try to make sense out of everything by organizing everything related into a single system of well defined components and relations, I get stuck.

Thanks in advance, Garrett

Answer 1

The kind of code suitable for reasoning by human beings (let's call it "source code") needs to pass through several stages of translation before it can be physically executed by the underlying hardware (such as CPU or GPU):

1. Source code.
2. [Optionally] intermediate code (such as .NET MSIL or Java bytecode).
3. Machine code conformant to the target instruction set architecture.
4. The microcode that actually flips the logical gates in silicon.

These translations can be done in various phases of the program's "lifecycle". For example, a particular programming language or tool might choose to translate from 1 to 2 when the developer "builds" the program and translate from 2 to 3 when the user "runs" it (which is typically done by a piece of software called "virtual machine"¹ that needs to be pre-installed on user's computer). This scenario is typical for "managed" languages such as C# and Java.

Or it could translate from 1 to 3 directly at build time, as common for "native" languages such as C and C++.

The translation between 3 and 4 is almost always done by the underlying hardware. It's technically a part of the "run time" but is typically abstracted away and largely invisible to the developer.

The term "compile time" typically denotes the translation from 1 to 2 (or 3). There are certain checks that can be done at compile time before the program is actually run, such as making sure the types of arguments passed to a method match the declared types of method parameters (assuming the language is "statically typed"). The earlier the error is caught, the easier it is to fix, but this has to be balanced with the flexibility, which is why some "scripting" languages lack comprehensive compile-time checks.

The term "run-time" typically denotes the translation from 2 (or 3) all the way down to 4. It is even possible to translate directly from 1 at run-time, as done by so called "interpreted languages".

There are certain kinds of problems that can't be caught at compile time, and you'll have to use appropriate debugging techniques (such debuggers, logging, profilers etc...) to identify them at run-time. The typical example of a run-time error is when you try to access an element of a collection that is not there, which could then manifest at run-time as an exception and is a consequence of the flow of execution too complex for the compiler to "predict" at compile time.

The "debug time" is simply a run-time while the debugger is attached to the running program (or you are monitoring the debug log etc.).

---

¹ Don't confuse this with virtual machines that are designed to run native code, such as VMware or Oracle VirtualBox.