For a university course I have to write a http server which is supposed to run on both Linux and Windows.
I have got a humble Linux machine which I don't think can handle any kind of heavy virtual environment, neither I'm willing to go through the hassle of installing it.
This is the first project of mine complex enough (I estimate ~1.5 months to develop) to require an environment sufficiently comfortable to alternate rapidly between short coding and testing sessions (the latter on both platforms, of course).
So, I was wondering what could be the best set up for this situation. I think testing it on Wine would be ok (it is not a real-world thing, after all), and I installed MinGW for the Windows-targeting part.
Basically, a simple well-written makefile could solve my problem... It should build both the Linux and Windows binaries and place them in the respective folders (the Windows one in the Wine sub-tree) and I'm all done! But I feel very inexperienced in this thing and I really don't know where to start. Maybe the make manual, ahah!:)
Thoughts, suggestions, anything I didn't think/know!
Thank you!
(PS. I'm planning to use emacs as editor, or maybe learn vim. Unless eclipse provide some kind of skynet-like plugin that entirely solve this problem...:)
You're on the right track. It's not that complicated, really, thanks to MinGW. You basically need two things:
The code has to be portable across the OSes. MinGW has some POSIX support, but you'll probably need to either use Cygwin in order to be able to use the POSIX interface or have your own compatibility layer for interfacing with the OS. I'd probably go for Cygwin as then you can code only against POSIX and won't have to test and debug your compatibility layer. Also, make sure you won't use any external libraries that are OS specific. Non-portable code often results in a compile error, but make sure you test the application thoroughly anyway.
The toolchains for targeting Linux and Windows. You already have them, you just need to use them correctly. Normally you'd use a variable like $(CROSS_COMPILE) as a prefix when calling the toolchain during cross compilation. So when compiling for Linux, you call gcc, ld, etc. (having the CROSS_COMPILE variable empty), and when compiling for Windows you call e.g. i486-mingw32-gcc, i486-mingw32-ld etc., i.e. CROSS_COMPILE=i486-mingw32-. Or just just define CC, LD etc. depending on the target.
I wrote a small game on Linux and made it run on Windows as well. If you browse the code, you can see the code has next to no #ifdef jungle (basically just some extra debugging features enabled for Linux), and the Makefile is simple as well, with no complicated handling for cross-compilation, just the possibility to override CC etc. like it should be. As lots of important open source software is written this way (especially software that's used by the desktop and embedded devices), you should also be able to find lots of other examples on how to set up the build environment correctly.
As for testing the application on Windows, I think the best option is if you can find a real Windows machine somehow. If you do everything correctly, it should run the same as on Linux and you won't need to continuously test your application on both OSes. If testing on a Windows machine is not possible, a VM would be the next best choice, though it would probably be more difficult to set it up. Wine is a good backup plan, but I don't think you can be sure your application works well on Windows if you only tested it on Wine.
Related
I'm starting to work on my master thesis at the moment and I have a (maybe) specific question...
I want to stay on windows OS and run a Linux VM via VirtualBox combined with Vagrant. No Problem. I like the feature to reset the VM via vagrant easily.
The next target is using features like auto-completing or similar while developing in C++. This would help me to work with unknown includes/libraries.
Is it possible to access the filesystem/compiler of the VM while using an IDE (like clion) installed on windows? Without explicit loading of the gui und running the IDE on it? Kinda like working with cygwin? I don't want to use cygwin because it doesn't support c++11 standard (or is there a way???)
Maybe you know an alternative way. I would be glad for all hints solving my problem.
I don't know much about cygwin, though I would be surprised if they cannot get recent versions of gcc. But for certain, you can use MSYS2 to get very recent versions of gcc and many other linux packages, which will support C++11.
It's a matter of opinion how best to do cross-platform development, but an alternative worth mentioning is to use cmake for your project. When you want to code in windows, it can make MSVC 2015 project files for you -- when you want to compile in linux, it can find the dependencies and generate a makefile for you to use. IIUC, cmake is the most widely used cross-platform build system right now, besides gnu make itself. (I'm pretty sure it's more popular than "autotools" nowadays, and its definitely more popular than scons.) The advantage is that you avoid the need to maintain multiple platform-specific project files that essentially say the same thing with different formatting.
Programming languages are platform independent, so why is it that we can't write a program that will run on both a PC and a Mac?
I want to develop a software and I'm on a mac, but I want it to run on a PC also, is it possible to develop such a software without having to require the user to download a special program that will make my program compatible with their computer?
The problem with this is that most software is dependent on the OS to handle some tasks. Yes, most programming languages are compatible with many platforms, but the OS provides a lot of support. When software uses the OS, it is sometimes called making a system call. If you want here is some more information.
Theoretically if you write your program in a 'high level language' it should be portable between two operating systems.
Practically however, the differences start from the very beginning - the API of choice, which works on one and does not on another(Such as, Mac's BSD API is incompatible with Win32 API) and boils down to the very last, which is, executable format, linker and loader. Each operating system has its own quirks.
Then comes the difference between the underlying architecture. Previously Macs ran on PowerPC architecture and Motorola architectures, while PCs used Intel. Since Macs have switched to Intel, there have been attempts at making cross platform executables inside Apple. Most attempts have failed.
There is however a way around your problem. You can use a very high level language such as Python to code and then distribute your python code to your PC friends.(But remember remember, you need a Python interpreter in your PC friends' computers for your program to run). I have successfully ported Python programs from Mac to PC with 0 code changes, and sometimes requiring only 2-4% code changes.
Simple answer: because language per se is not enough to make an application cross-platform. Also the framework it uses must be cross-platform too, frameworks are required for everything: handling data, displaying things, communicate with the hardware, multi threading, etc
This can usually be done:
by choosing a complete solution like Java, which will actually run on both platforms seamlessly and even with the same binary
by using C/C++ and cross platform libraries so that the same program can be compiled for both platforms (keep in mind that you can't distribute the real same binary, you need to compile two in any case)
by writing the logic of your program just using standard libraries and a standard language and then attach whatever you need for a specific platform just to build two different libraries. Of course you will have to wrap as much as you need so that the cross-platform part of your program doesn't know it
Mind that developing cross-platform applications which are not trivial examples like a game (for which there are plenty of cross-platform APIs) without using a complete solution like Java is not an easy task at all. Especially because most of the GUI you can build are strictly platform specific and relies on their own frameworks.
If you want an application to run "anywhere" your best option is a JIT type language which means that it compiles as it runs (Just In Time) for the platform that it's on. Really the language that stands out in my mind is Java (there's others and personally I don't like java). However, it's not quite that simple. For example a Window on a Mac computer has pieces and functions that a Window on a PC doesn't have and vice versa. And other operating systems don't even have windows or anything equivilant yet still run Java like Android or iOS for example or countless Linux Distros. And that's just a very basic example it gets MUCH MUCH harrier. Really the best way to build an application that can be used by anyone on just about any device is going web based.
The lesson is that if it was that simple a lot of people wouldn't have jobs and it never will be that simple, things will always progress and change and not everyone is going to want to do the same thing with their OS as someone else. There's a million ways to skin a cat and there's many more ways to implement something in an OS.
Yes, it is possible. But it is quite tricky. You need to:
Use a cross platform language (this is the easy part, many languages run on different plaftforms)
Avoid using any platform-specific features (usually not too hard, but needs testing)
Ensure you have cross platform libraries for all your dependencies (hard!)
Because of the library issue in particular, there are very few options that work across platforms. Your best options are probably:
A JVM language (like Java, Scala or Clojure) - because the JVM abstracts away from platform specific features, pure Java applications and libraries will run on any platform. Java probably has the best ecosystem of cross platform libraries and tools as a result.
JavaScript - quite a good option if you don't mind running in a browser. There are lots of quirks to deal with, but JavaScript is one of the best cross-platform options because of it's ubiquity.
I was wondering is there a free Windows version of GCC.
I know there is minigw and something else but I don't know how to use them.
Sorry if this should be on SU.
The main choices are either MinGW or CygWin.
CygWin is a more complete UNIX-like environment than MinGW as it offers quite a lot of tools over and above development stuff. Even to the point of a full X-Windows server so you can develop software that'll run on both UNIX-like systems and Windows.
The installer is good but I would suggest installing everything even if you think you don't need it. Disk space is cheap and I've had problems in the past trying to get stuff going on partial installs (whether 1.7, or even earlier, fixes this, I don't know - I always do full installs).
However, it relies on the CygWin UNIX emulation DLL which, if I remember rightly, has restrictions for non-free software.
MinGW is more concentrated on the development tools. It generates native Windows applications rather than running under a emulation DLL like CygWin.
It used to be difficult to install with having to do MinGW, MSYS and others separately but it's come a long since then and has an easy graphical installer.
I believe it can do graphical applications using native Windows calls rather than via X-Windows, since it just links to the normal Windows runtimes.
If you want to know how to use either of them, you really have to look through the docs found at those links I provided - check the Documentation link on the left for MinGW (particularly Getting started) or the CygWin FAQ.
As for GUIs for development, I've never used one for CygWin - I'm old enough that I remember mark sense cards so I'm not scared of the command line interface :-).
I used Eclipse with CDT running over MinGW and wasn't that impressed although admittedly that was an early version. Don't get me wrong Eclipse is a brilliant tool and we use it for both Java and Linux/C development, I just had a lot of troubles with Eclipse/CDT under Windows.
Code::Blocks, on the other hand, was absolutely brilliant but you should check them all out to see which one suits you best. As I said, the last time I looked was about five years ago, an eternity in the IT world.
I am a bit late, but since the question may still arise...
gcc for Windows (including Ada, C, C++ and Fortran compilers) can be obtained from MinGW-builds on SourceForge: http://sourceforge.net/projects/mingwbuilds/files/host-windows/releases/
As of august 2013, there are 32 and 64 bits versions of gcc-4.8.1.
I'm certain this has been asked before, with with nearly 900k questions, it's hard to find things :)
We are starting a project where we want our C++ and Python to run both under a Unix environment but also under Windows. We want to make our project easy to contribute to as developers, and so we wish to have what each side feels is "natural."
In this case, under Unix, it's automake/autoconf/etc. Under Windows, anything but visual studio project files seems like a mistake. This is mostly because we want windows-native hackers to have a way to easily develop in their natural environment.
How are people doing this sort of thing today? Are Windows coders comfortable in a non-GUI world, so we are trying too hard?
Another goal is that we need automated testing of our project, ideally using the same test code (google-test) on all platforms, etc. This may be matter for another question, though.
Multi-platform development can be a tricky situation. Here are some things we do to ease the process:
Source Control (this one is obvious, a solid code-base that works with both Windows and Linux, we use SubVersion)
Create standards (All of our Makefiles look identical and are easy to implement as are our .sln files for Visual Studio)
Build the application on all platforms, frequently
Be aware of cross library use like QT and any quirks that might contain between Windows and Linux
Sometimes you can't avoid it: #ifndef's to execute some code specific to a platform
We develop in Windows, Sun and Linux (RedHat) with multiple applications both visual and command line based. I hope that helps!
Is there any difference between makefile in Windows and Linux?
If I know how to use it in Linux, is it necessary to learn
something new when in Windows system?
The fine folks over at GnuWin32 have Gnu make built for Windows as a native binary. It works well, but does have a few minor quirks due to a well-intentioned hack that tries to equate a target named foo with the file foo.exe. The GnuWin32 package does include a nice PDF of the manual, IIRC.
I use it on XP alongside GCC from the MinGW project, as well as with several different embedded systems cross compiler toolchains. MinGW also provides a native build of Gnu Make.
We do use Cygwin for a couple of projects, in particular one where the system-on-a-chip vendor provided a complete, working toolchain and build environment hosted in Cygwin. Moving it to a native Windows shaped build would be prohibitively expensive, and would make it harder to accept updates from the vendor.
If you want to use mostly native Windows tools, but need to start by running configure on an existing source kit, then take a look at MSYS. This started as a fork of Cygwin by the MinGW developers, and it intends to provide a minimal set of unix-like tools with a bash shell so that configure can be used on Windows. It isn't as complete as Cygwin, but it is a lot lighter weight.
You will find that the manual for Gnu Make is mostly true. However, you have to keep in mind that the content of a Makefile is not written in a single language. The stuff that defines variables and lays out the dependancy tree is one language. Each line of an action is a miniature shell script that is in the language of the default shell (which might be CMD.EXE rather than /bin/sh on Windows), and each command has options and arguments.
If you are comfortable with your *nix build environment, then Cygwin might be the path of least pain. But MSYS is often sufficient, and there is something to be said for adopting the native tools of the target platform so that you have a sense of how your users see Windows, in which case MinGW and GnuWin32 are handy resources to know about.
In addition to what iftrue note that Microsoft has a Makefile based build tool called nmake with some changes in syntax/semantics from the traditional *nix based make tool. This page lists some of the differences.
Well, if you download "cygwin", you can run the "make" tool directly on windows, but you may want to look into a true cross-platform build tool such as "cmake". It's like make, but it's a bit more spiffy and is inherently cross-platform.