binary compatibility support for .NET assembly

[someone]

Suppose that I have a class in an assembly that is delivered to the user,
what can I do to change the class so that it doesn't break the
binary compatibility? That is, user application can run with recompiling
and relinking.

I know that if I define an interface, and only expose the interface but not
the class which implments the interface, I can add a data member to the
class without breaking the binary compatibility. If the class itself,
rather than the interface is exposed to user, can I still add a data member
to the class without breaking the binary compatibility? What are other
restrictions that won't break the binary compatibility?

In pure C++, adding a data member to a class that is exposed to the user
will break compatibility.

Thanks.

 

[someone]

Suppose that I have a class in an assembly that is delivered to the user,
what can I do to change the class so that it doesn't break the
binary compatibility? That is, user application can run with recompiling
and relinking.

A mistake: 'with' should be 'without'.

 

[Niki Estner]

Suppose that I have a class in an assembly that is delivered to the user,
what can I do to change the class so that it doesn't break the
binary compatibility? That is, user application can run with recompiling
and relinking.

..net applications are just-in-time compiled; There is no "linker". The kind
of tasks that used to be done by the linker are done at runtime (by the JIT)
anyway.

I know that if I define an interface, and only expose the interface but not
the class which implments the interface, I can add a data member to the
class without breaking the binary compatibility. If the class itself,
rather than the interface is exposed to user, can I still add a data member
to the class without breaking the binary compatibility?
Yes.

What are other restrictions that won't break the binary compatibility?

Don't remove methods, make them private or change their signature if they
are used by the client. You'll get a "MissingMethodException" if the JIT
can't find a method that's called in your code.

In pure C++, adding a data member to a class that is exposed to the user
will break compatibility.

Yep. There are ways around this, but they usually require some work.
That's one of the reasons why .net was invented.

Niki

 

[Niki Estner]

.net applications are just-in-time compiled; There is no "linker". The kind
of tasks that used to be done by the linker are done at runtime (by the JIT)
anyway.

Sorry, that wasn't 100% correct: There is a linker (al.exe) that fuses
modules and resources, but it is rarely used, at least not in a usual
VS-build cycle.

Niki

 

[someone]

What about enumeration constants defined? if their value is changed, I
assume that the binary compatibility is broken. I know that this is true in
C++.

 

[Niki Estner]

I'm sorry, my first reply was definitely too short; Unfortunately I didn't
find any good article on the subject; So, here's a short list of things I
that came to my mind right now; In general
- You may safely change default values of data members
- You may safely change all private/internal fields of a class/struct
- You may usually add properties or methods to an existing class or struct
(this only hurts the client if it derives a class from that class, and
happens to add a property/method with the same signature)
- You may usually add data members and events to a class or struct, unless
you use binary serialization in your clients (the binary layout changes with
new data members)
- Removing, renaming or changing the type/signature of public/protected
properties/data members/events/methods can cause clients to crash (however
unlike C++ you'll get something like a "MethodNotFoundException", and no
memory will be overwritten) if (and only if) the client uses that member
- Removing an interface or changing the base class may hurt clients, if the
client casts to that class/interface
- Changing static constants/enums causes clients to break (didn't think of
that before your post)

Of course this list isn't complete, and there are may still be some cases
where even these rules don't work (e.g. if the client tries to invoke a
non-existing method using late-binding; If you add that method, the client
will behave differently). But I hope that this at least gives you a rough
sketch of what is done at compile-time, and what is done at run-time.

In general, if you expect an assembly to change often, using interfaces for
communication is not a bad idea.

Niki