It's been my understanding the OS sits on top of the hardware. Is it more or less the same to run windows from a macbook? When installing SQL on a windows partition, does this install similar to an all Windows setup?
I've heard the kernel is the main connector between hardware and basic OS, so would the mac kernel cause potential differences in operation?
Would installing the linux OS also adhere to these rules?
Thanks, and sorry for the simple question.
Generally, you are correct to say that, installing different operating systems on the same hardware would be the same. You will be able to both install Windows and Linux on the same laptop (whether that would be an Asus brand laptop, or HP, or whatever). Once you install an OS on some hardware, and the OS is able to recognize the hardware, and is able to utilize it, then you are in the clear. What's important is to install on OS that is compatible with the architecture of the computer. So if you get a Linux distro that supports x86 architecture, then you would have to install it on hardware that is with a x86 architecture.
Side note: Modern OS's are very smart and they have a wide range of architecture support (list of Linux architecture support, Windows support for ARM, Apple also has a wide range of architecture support).
Since you are asking about a macbook and Windows, then the short answer is: there won't be a problem for you to install Windows on your mac. Apple even gives you Boot Camp to easily do this (there are also quite a bit of recent tutorials on this topic as well).
So the end experience would be almost the same as having Windows on any other machine.
I've heard the kernel is the main connector between hardware and basic OS, so would the mac kernel cause potential differences in operation?
This is true. The kernel is the heart of any OS, but once you have your Windows running, it would be using its own heart and it won't touch the mac kernel. So if you remove your macOS and install only Windows, then only the Windows kernel would be taking control of the Mac hardware. But if you load your macOS, then the Mac kernel would be running and operating on the hardware.
Will Windows run faster on Mac hardware than macOS on its hardware? It's debatable and I would assume not a lot of studies have been made in that sphere. But, at least, it will run.
But what about dual-booting your macbook with Linux? Technically, it is possible (and the principle is the same), but Apple have made restrictions to their firmware, limiting the option of having both a macOS and a Linux distro at the same time. What's so different here than the case with Windows? Well mainly that the firmware of the macbook (the software embedded in the hardware of the laptop) doesn't allow for Linux to be installed. Maybe things have changed, but these are the (not so recent anymore) news I know about (I guess there are still ways of installing Linux on mac hardware).
I am New to VxWorks. First of all Can I use VxWorks OS as Normal OS on my PC? I mean Can I run my application software on VxWorks OS?
While VxWorks can run on PC hardware, it is not a general-purpose OS for running independent executables. VxWorks is an RTOS library; you statically link it to your application and the whole runs as a monolithic executable.
It does support a command line interface (intended primarily for development and debug), and from that it is possible to dynamically load and link object files, but these are not independent executables in the sense they are in a GPOS; they essentially become part of the monolithic application.
An RTOS such as OS/9 or QNX would be more suited as these can operate more like a GPOS in the sense of loading and executing independently linked executables.
In any event, application software must be specifically built for these targets.
For versions of VxWorks prior to VxWorks 6, the answer by clifford provides a good explanation of why this is not really possible.
VxWorks 6 introduced Real Time Processes (RTPs). These are independant, user mode applications, running on top of the VxWorks OS. Dependant on how the VxWorks OS has been configured and built, these RTP applications may have access to POSIX libraries, and so you may be able to run POSIX applications (eg linux programs) with little modification.
However, these must still be built for VxWorks, ideally linked against your own VxWorks Source Build.
You cannot, however, just pick up any old application and expect it to run. You are never going to get Word or Excel to run.
If I install a bare-metal hypervisor (say, ESXi), would it allow me to run Windows 7 concurrently with Linux?
Would it allow me to run multiple instances of Windows 7?
When I'm sitting at the PC that's running Win7 and Linux on a hypervisor, which OS do I see when I look at the screen? (I'm suspecting that the only way to access either OS is to do a remote login.)
Assuming the answer to #2 is yes, how do you manage multiple installs of Win7 on the same hard drive?
Thanks in advance!
If you simply want to run Linux and Windows in parallel you may of course do this on e.g. ESXi. Still, the OSes would run with virtualized (or emulated) hardware available to them, i.e. you would not be able to easily access all the hardware directly and the hypervisor itself not only introduces an overhead but this overhead is not deterministic.
If you want to run an RTOS (like Real-Time Linux) or any other RTOS, then you need a "real-time hypervisor".
You can google for such hypervisors - there are a few out there.
(I dont want to recommend one here as we are a vendor of such a solution our selves)
Regards
GFL
I'm reading an article about Xen, a virtual machine monitor. They say that an operating system requires some modification in order to be able to act as a guest OS on top of Xen. Now, for an OS like Linux, I can understand what a "modification" might mean but in the case of an OS like say, Windows XP, what does it mean? I mean, XP is closed source proprietary OS right?
It means exactly the same thing. It's just harder because the source is not widely available. Note that the modifications are no longer required when Xen is used in conjunction with hardware virtualization.
Xen was originally a university project. As a researcher (or as a government agency tasked with infrastructure security), you can get the Windows sourcecode from Microsoft. You're just not allowed to distribute your own version of Windows.
This is exactly what they did: they had the sourcecode of Windows XP and then they ported XP to Xen, just to show that if Microsoft were ever to officially support Xen, it would be doable.
On the other hand, although para-virtualization cannot provide an end-to-end virtualization solution, it can improve performance.
Does anyone know of a good software development framework or similar that has the following properties?
Cross platform: it should be runnable on XP, Vista, OSX and common versions of Linux (such as Ubuntu and Kubuntu).
No installation: Be able to run the software from a USB stick without having to copy anything to the host machine.
Have good GUI support (this is why this question doesn't give a suitable answer, as far as I can tell).
Permissive licensing such as LGPL or BSD or such.
Among the softer requirements are having a set of abstractions for the most common backend functionality, such as sockets, file IO, and so on (There is usually some platform specific adaptations necessary), and supporting a good language such as Python or C++, though it is usually fun to learn a new one (i.e. not perl).
I think possible candidates are Qt 4.5 or above (but IFAIK Qt software will not run on Vista without any installation(?)), some wxWidgets or maybe wxPython solution, perhaps gtkmm. The examples I have found have failed on one or another of the requirements. This does not mean that no such examples exist, it just means that I have not found any. So I was wondering if anyone out there know of any existing solutions to this?
Some clarifications;
By "framework" I mean something like Qt or gtkmm or python with a widget package.
This is about being able to run the finished product on multiple platforms, from a stick, without installation, it is not about having a portable development environment.
It is not a boot stick.
It is ok to have to build the software specifically for the different targets, if necessary.
The use case I am seeing is that you have some software that you rely on (such as project planning, administration of information, analysis tools or similar) that:
does not rely on having an internet connection being available.
is run on different host machines where it is not really ok to install anything.
is moved by a user via a physical medium (such as a USB stick).
is run on different operating systems, such as Windows, Vista, Ubuntu, OSX.
works on the same data on these different hosts (the data can be stored on the host or on the stick).
is not really restricted in how big the bundled framework is (unless it is several gigabytes, which is not really realistic).
It is also ok to have parallel installations on the stick as long as the software behaves the same and can work on the same data when run on the different targets.
A different view on the use case would be that I have five newly installed machines with Vista, XP, OSX, Ubuntu and Kubuntu respectively in front of me. I would like to, without having to install anything new on the machines, be able to run the same software from a single USB stick (meeting the above GUI requirements and so on) on each of these five machines (though, if necessary from different bundles on the stick).
Is this possible?
Edit:
I have experimented a little with a Qt app that uses some widgets and a sqlite database. It was easy to get it to work on an ubuntu dist and on osx. For windows xp and vista I had to copy QtCored4.dll, QtGuid4.dll, QtSqld4.dll and mingwm10.dll to distribution directory (this was debug code) and I copied the qsqlited4.dll to a folder named "sqldrivers" in the distribution directory.
You mention wxWidgets but dismiss it as failing at least one of the requirements.
I don't know what your requirements are and in what way wxWidgets wouldn't work for you, but IMO it does fulfill them:
Cross platform: it should be runnable on XP, Vista, OSX and common versions of Linux.
It does run on those platforms, but "common versions of Linux" isn't good enough, as you can never be sure that the necessary GUI libraries for wxGTK (which should not be linked to statically) will be installed. This is however a problem for other solutions as well, unless you plan to put everything onto the stick.
No installation: Be able to run the software from a USB stick without having to copy anything to the host machine.
See the previous point, you would need to specify which libraries are needed on Linux. Also you could specify at build time not to use some of the system-provided libraries (for example for graphics, compression, regexes) but to use the wxWidgets-internal libraries instead.
Have good GUI support
Check.
Permissive licensing such as LGPL or BSD or such.
Check. You can statically link wxWidgets into your application too.
supporting a good language such as Python or C++
Supports both, and there are bindings to other languages as well.
having a set of abstractions for the most common backend functionality, such as sockets, file IO, and so on
It does have some abstractions like that, but you can link to other cross-platform libraries as well.
We use wxWidgets for FlameRobin, a graphical administration program for the Firebird SQL server. It has active ports to Windows, Linux and Mac OS X, and has been compiled for at least some BSD variant and Solaris as well. It definitely runs from a stick on Windows, I haven't tried with Linux or Mac OS X, but I don't see why it shouldn't there too.
Java.
It has GUI support.
It provides your network/file/etc. abstractions.
It is cross-platform. Most platforms you can think of have a JRE available.
No need to install a JRE. Most users probably already have one, and if not, you can run the appropriate JRE right off the stick.
You can provide several startup scripts for various platforms to run the app under the appropriate JRE.
Something else to consider is HTML+Javascript. :D
You can look at Mono it cross platform, has GUI (GTK+, or Winforms 2.0) and I can execute code without installing.
This might not be crossplatform, but is maybe even better, it dont even use the platform : linux on a stick :-)
The subtitle is
Take your Java workspaces wherever you go on a USB key
Here with java and eclipse, but nothing stops you there of course.
http://knol.google.com/k/inderjeet-singh/installing-a-ubuntu-hardy-heron-java/1j9pj7d01g86i/2#
Well, it depends on what you mean by 'package'. Kylix came close to being such a thing. It was QT based, and it allowed you to write once and compile for Windows + Linux. However, it was not an open source solution.
I asked a similar question in this link
http://www.24hsoftware.com/DevelopersForum/CrossPlatform-C-Library.html
and the best asnswer seems to be QT.
I have started using QT, but it is not as easy as I expected mainly due to deployment problems due to the DLL hell, Winsxs hell and manifest hell.
Tclkit is a single-file, self-contained Tcl/Tk system. The mac version I have is about 3.8 megs. You can get a version for just about any modern OS. I carry around a thumb drive that has mac, windows and linux binaries so I can run my scripts on any platform. No install is required, just copy one file wherever you want.
The most recent versions of tcklit use native, themed widgets (though, on *nix there really isn't a single "native" set of widgets...)