Related
I'm designing an API with a specific pattern in mind, but don't know if this pattern has a name. It's similar to the Command pattern in GoF (Gang of Four) but not exactly.
One simple example of it I can find is in Eclipse where you manipulate a project (IProject), not by calling methods on the project that change its state, but by this 3 step process:
extracting its state into a descriptor object (IProjectDescription) with getDescription
setting properties on the descriptor. E.g. setName
applying the descriptor back to the original project with setDescription
The general principle seems to be that you have a complex object as part of a framework with many potentially interdependent properties, and rather than working directly on that object, one property at a time, you extract the properties into a simple data object, manipulate that, and apply it back.
It has some of the attributes of the Command pattern, in that the data object encapsulates all of the changes like a Command would - but it's not really a Command, because you don't execute it on the object, it's simply a representation of the state of the object.
It also has some attributes of a Transactional API, in that, by making the changes all in one hit with the set... call, you allow for the entire modification to effectively "roll back" if any one property changes fails. But while that's an advantage of the approach, it's not really the main purpose of it. And what's more, you can achieve the transactional nature without this approach, by simply adding transactional methods to the API (like commit and rollback)
There are two advantages in this pattern that I do want to exploit - although I don't see them being exploited by the eclipse example above:
You can represent the meaningful state of the underlying object while its implementation changes. This is useful for upgrading, or copying state from different types of representations. Say I release a new version of my API where I create an object Foo2 which is a totally new form of my old Foo1, but both have the same basic properties. To upgrade a Foo1 to a Foo2, I can extract those properties as a FooState. foo2.setFooState(foo1.getFooState) as simply as that. The way in which the properties are interpreted and represented is encapsulated in the Foos and can be totally different.
I can persist and transmit the state of the underlying object with my simple data object, where persisting the object itself would be much more complex. So I can extract the state of Foo as a FooState, and persist it as a simple XML document then later apply it to some new object by "loading" it and applying it. Or I can transmit the FooState simply to a webservice as a JSON object whereas the Foo itself is too big and complex to transmit. (Or the objects on each end of the service call are entirely different, like Foo1 and Foo2)
Anyway, I can't find an name or example of this pattern anywhere, neither in the Gang of Four design patterns, nor even in Martin Fowler's comprehensive "bliki"
Data Transfer Object(DTO) that Martin Fowler describes in his book Principles of Enterprise Application Architecture seems to be for the purpose you describe in point 2.
A DTO is a fairly simple extraction of the more complex Domain Model that it represents.
Fowler describes that the usage of a DTO in combination with an assembler can be used to keep the DTO independent from the actual Domain Object(or Objects) that it is supposed to represent. The assembler knows how to create a DTO from the Domain Object and vice versa. Also he mentions that the DTO needs to be serializable to persist/transmit its state. What you describe in point 2 seems to match this description.
What you've described in point 1 though does not seem to be an intended purpose, but definitely seems achievable using this pattern.
I'm not sure if you went through the Pattern catalog of his book or the book itself. The book itself describes this in much greater detail.
You may also want to have a look at Transfer Object definition from Oracle which Fowler says here is what he describes as DTO.
Not every design is documented as a single Design Pattern, in fact most system designs are combinations of multiple patterns.
However one part of what you're doing, with IProjectDescription is using a Memento, however yours seems to be a Polymorphic variation. Consider Patterns as they appear in Pattern Catalogues to be the pared down to the essential starting point not the end result. Patterns are by there very nature supposed to be extended and combined.
The Command pattern can give you Commit and RollBack (Do/Undo) and combining it with Memento in that way is a quite common approach. The same thing is seen in the Java Servlet API with HttpRequest & HttpResponse.
When do you encourage programming against an interface and not directly to a concrete class?
A guideline that I follow is to create abstractions whenever code requires to cross a logical/physical boundary, most especially when infrastructure-related concerns are involved.
Another checkpoint would be if a dependency will likely change in the future, due to possible additional concerns code (such as caching, transactional awareness, invoking a webservice instead of in-process execution) or if such dependencies have direct references to infrastructure integration points.
If code depends on something that does not require control to cross a logical/physical boundary, I more or less don't create abstractions to interact with those.
Am I missing anything?
Also, use interfaces when
Multiple objects will need to be acted upon in a particular fashion, but are not fundamentally related. Perhaps many of your business objects access a particular utility object, and when they do they need to give a reference of themselves to that utility object so the utility object can call a particular method. Have that method in an interface and pass that interface to that utility object.
Passing around interfaces as parameters can be very helpful in unit testing. Even if you have just one type of object that sports a particular interface, and hence don't really need a defined interface, you might define/implement an interface solely to "fake" that object in unit tests.
related to the first 2 bullets, check out the Observer pattern and the Dependency Injection. I'm not saying to implement these patterns, but they illustrate types of places where interfaces are really helpful.
Another twist on this is for implementing a couple of the SOLID Principals, Open Closed principal and the Interface Segregation principle. Like the previous bullet, don't get stressed about strictly implementing these principals everywhere (right away at least), but use these concepts to help move your thinking away from just what objects go where to thinking more about contracts and dependency
In the end, let's not make it too complicated: we're in a strongly typed world in .NET. If you need to call a method or set a property but the object you're passing/using could be fundamentally different, use an interface.
I would add that if your code is not going to be referenced by another library (for a while at least), then the decision of whether to use an interface in a particular situation is one that you can responsibly put off. The "extract interface" refactoring is easy to do these days. In my current project, I've got an object being passed around that I'm thinking maybe I should switch to an interface; I'm not stressing about it.
Interfaces abstraction are convenient when doing unit test. It helps for mocking test objects. It very useful in TDD for developing without actually using data from your database.
If you don't need any features of the class that aren't found in the Interface...then why not always prefer the Interface implementation?
It will make your code easier to modify in the future and easier to test (mocking).
you have the right idea, already. i would only add a couple of notes to this...
first, abstraction does not mean 'interface'. for example, a "connection string" is an abstraction, even though it's just a string... it's not about the 'type' of the thing in question, it's about the intention of use for that thing.
and secondly, if you are doing test automation of any kind, look for the pain and friction that are exposed by writing the tests. if you find yourself having to set up too many external conditions for a test, it's a sign that you need a better abstraction between the thing your testing and the things it interacts with.
I think you've said it pretty well. Much of this will be a stylistic thing. There are open source projects I've looked at where everything has an interface and an implementation, and it's kind of frustrating, but it might make iterative development a little easier, since any objects implementation can break but dummies will still work. But honestly, I can dummy any class that doesn't overuse the final keyword by inheritance.
I would add to your list this: anything which can be thought of as a black box should be abstracted. This includes some of the things you've mentioned, but it also includes hairy algorithms, which are likely to have multiple useful implementations with different advantages for different situation.
Additionally, interfaces come in handy very often with composite objects. That's the only way something like java's swing library gets anything done, but it can also be useful for more mundane objects. (I personally like having an interface like ValidityChecker with ways to and-compose or or-compose subordinate ValidityCheckers.)
Most of the useful things that come with the Interface passing have been already said. However I would add:
implementing an interface to an object, or later multiple objects, FORCES all the implementers to follow an IDENTICAL pattern to implement contract with the object. This can be useful in case you have not so OOP-experienced-programmers actually writing the implementation code.
in some languages you can add attributes on the interface itself, which can be different from the actual object implementation attribute as sense and intent
This is probably a newbie question since I'm new to design patterns but I was looking at the Template Method and Strategy DP's and they seem very similar. I can read the definitions, examine the UML's and check out code examples but to me it seem like the Strategy pattern is just using the Template Method pattern but you just happen to passing it into and object (i.e. composition).
And for that matter the Template Method seems like that is just basic OO inheritance.
Am I missing some key aspect to their differences? Am I missing something about the Template Method that makes it more that just basic inheritance?
Note: There is a previous post on this (672083) but its more on when to use it, which kind of helps me get it a bit more but I want valid my thoughts on the patterns themselves.
It basically all comes down to semantics. The strategy pattern allows you to pass in a particular algorithm/procedure (the strategy) to another object and that will use it. The template method allows you to override particular aspects of an algorithm while still keeping certain aspects of it the same (keep the order the same, and have things that are always done at the start and end for example... the 'template') while inheritance is a way of modelling 'IS-A' relationships in data models.
Certainly, template methods are most easily implemented using inheritance (although you could just as easily use composition, especially once you have functors), and strategy patterns are frequently also template methods but where the syntax is similar the meanings are vastly different.
The Strategy design pattern
provides a way to exchange the algorithm of an object
dynamically at run-time
(via object composition).
For example, calculating prices in an order processing system.
To calculate prices in different ways,
different pricing algorithms can be supported
so that which algorithm to use can be selected (injected) and exchanged dynamically at run-time.
The Template Method
design pattern
provides a way to
redefine some parts of the behavior of a class statically at compile-time
(via subclassing).
For example, designing reusable applications (frameworks).
The application implements the common (invariant) parts of the behavior
so that users of the application can write subclasses to redefine
the variant parts to suit their needs.
But subclass writers should neither be able to change the invariant parts of
the behavior nor the behavior's structure
(the structure of invariant and variant parts).
By putting functionality into a function, does that alone constitute an example of encapsulation or do you need to use objects to have encapsulation?
I'm trying to understand the concept of encapsulation. What I thought was if I go from something like this:
n = n + 1
which is executed out in the wild as part of a big body of code and then I take that, and put it in a function such as this one, then I have encapsulated that addition logic in a method:
addOne(n)
n = n + 1
return n
Or is it more the case that it is only encapsulation if I am hiding the details of addOne from the outside world - like if it is an object method and I use an access modifier of private/protected?
I will be the first to disagree with what seems to be the answer trend. Yes, a function encapsulates some amount of implementation. You don't need an object (which I think you use to mean a class).
See Meyers too.
Perhaps you are confusing abstraction with encapsulation, which is understood in the broader context of object orientation.
Encapsulation properly includes all three of the following:
Abstraction
Implementation Hiding
Division of Responsibility
Abstraction is only one component of encapsulation. In your example you have abstracted the adding functionality from the main body of code in which it once resided. You do this by identifying some commonality in the code - recognizing a concept (addition) over a specific case (adding the number one to the variable n). Because of this ability, abstraction makes an encapsulated component - a method or an object - reusable.
Equally important to the notion of encapsulation is the idea of implementation hiding. This is why encapsulation is discussed in the arena of object orientation. Implementation hiding protects an object from its users and vice versa. In OO, you do this by presenting an interface of public methods to the users of your object, while the implementation of the object takes place inside private methods.
This serves two benefits. First, by limiting access to your object, you avoid a situation where users of the object can leave the object in an invalid state. Second, from the user's perspective, when they use your object they are only loosely coupled to it - if you change your implementation later on, they are not impacted.
Finally, division of responsility - in the broader context of an OO design - is something that must be considered to address encapsulation properly. It's no use encapsulating a random collection of functions - responsibility needs to be cleanly and logically defined so that there is as little overlap or ambiguity as possible. For example, if we have a Toilet object we will want to wall off its domain of responsibilities from our Kitchen object.
In a limited sense, though, you are correct that a function, let's say, 'modularizes' some functionality by abstracting it. But, as I've said, 'encapsulation' as a term is understood in the broader context of object orientation to apply to a form of modularization that meets the three criteria listed above.
Sure it is.
For example, a method that operates only on its parameters would be considered "better encapsulated" than a method that operates on global static data.
Encapsulation has been around long before OOP :)
A method is no more an example of encapsulation than a car is an example of good driving. Encapsulation isn't about the synax, it is a logical design issue. Both objects and methods can exhibit good and bad encapsulation.
The simplest way to think about it is whether the code hides/abstracts the details from other parts of the code that don't have a need to know/care about the implementation.
Going back to the car example:
Automatic transmission offers good encapsulation: As a driver you care about forward/back and speed.
Manual Transmission is bad encapsulation: From the driver's perspective the specific gear required for low/high speeds is generally irrelevant to the intent of the driver.
No, objects aren't required for encapsulation. In the very broadest sense, "encapsulation" just means "hiding the details from view" and in that regard a method is encapsulating its implementation details.
That doesn't really mean you can go out and say your code is well-designed just because you divided it up into methods, though. A program consisting of 500 public methods isn't much better than that same program implemented in one 1000-line method.
In building a program, regardless of whether you're using object oriented techniques or not, you need to think about encapsulation at many different places: hiding the implementation details of a method, hiding data from code that doesn't need to know about it, simplifying interfaces to modules, etc.
Update: To answer your updated question, both "putting code in a method" and "using an access modifier" are different ways of encapsulating logic, but each one acts at a different level.
Putting code in a method hides the individual lines of code that make up that method so that callers don't need to care about what those lines are; they only worry about the signature of the method.
Flagging a method on a class as (say) "private" hides that method so that a consumer of the class doesn't need to worry about it; they only worry about the public methods (or properties) of your class.
The abstract concept of encapsulation means that you hide implementation details. Object-orientation is but one example of the use of ecnapsulation. Another example is the language called module-2 that uses (or used) implementation modules and definition modules. The definition modules hid the actual implementation and therefore provided encapsulation.
Encapsulation is used when you can consider something a black box. Objects are a black box. You know the methods they provide, but not how they are implemented.
[EDIT]
As for the example in the updated question: it depends on how narrow or broad you define encapsulation. Your AddOne example does not hide anything I believe. It would be information hiding/encapsulation if your variable would be an array index and you would call your method moveNext and maybe have another function setValue and getValue. This would allow people (together maybe with some other functions) to navigate your structure and setting and getting variables with them being aware of you using an array. If you programming language would support other or richer concepts you could change the implementation of moveNext, setValue and getValue with changing the meaning and the interface. To me that is encapsulation.
It's a component-level thing
Check this out:
In computer science, Encapsulation is the hiding of the internal mechanisms and data structures of a software component behind a defined interface, in such a way that users of the component (other pieces of software) only need to know what the component does, and cannot make themselves dependent on the details of how it does it. The purpose is to achieve potential for change: the internal mechanisms of the component can be improved without impact on other components, or the component can be replaced with a different one that supports the same public interface.
(I don't quite understand your question, let me know if that link doesn't cover your doubts)
Let's simplify this somewhat with an analogy: you turn the key of your car and it starts up. You know that there's more to it than just the key, but you don't have to know what is going on in there. To you, key turn = motor start. The interface of the key (that is, e.g., the function call) hides the implementation of the starter motor spinning the engine, etc... (the implementation). That's encapsulation. You're spared from having to know what's going on under the hood, and you're happy for it.
If you created an artificial hand, say, to turn the key for you, that's not encapsulation. You're turning the key with additional middleman cruft without hiding anything. That's what your example reminds me of - it's not encapsulating implementation details, even though both are accomplished through function calls. In this example, anyone picking up your code will not thank you for it. They will, in fact, be more likely to club you with your artificial hand.
Any method you can think of to hide information (classes, functions, dynamic libraries, macros) can be used for encapsulation.
Encapsulation is a process in which attributes(data member) and behavior(member function) of a objects in combined together as a single entity refer as class.
The Reference Model of Open Distributed Processing - written by the International Organisation for Standardization - defines the following concepts:
Entity: Any concrete or abstract thing of interest.
Object: A model of an entity. An object is characterised by its behaviour and, dually, by its state.
Behaviour (of an object): A collection of actions with a set of constraints on when they may occur.
Interface: An abstraction of the behaviour of an object that consists of a subset of the interactions of that object together with a set of constraints on when they may occur.
Encapsulation: the property that the information contained in an object is accessible only through interactions at the interfaces supported by the object.
These, you will appreciate, are quite broad. Let us see, however, whether putting functionality within a function can logically be considered to constitute towards encapsulation in these terms.
Firstly, a function is clearly a model of a, 'Thing of interest,' in that it represents an algorithm you (presumably) desire executed and that algorithm pertains to some problem you are trying to solve (and thus is a model of it).
Does a function have behaviour? It certainly does: it contains a collection of actions (which could be any number of executable statements) that are executed under the constraint that the function must be called from somewhere before it can execute. A function may not spontaneously be called at any time, without causal factor. Sounds like legalese? You betcha. But let's plough on, nonetheless.
Does a function have an interface? It certainly does: it has a name and a collection of formal parameters, which in turn map to the executable statements contained in the function in that, once a function is called, the name and parameter list are understood to uniquely identify the collection of executable statements to be run without the calling party's specifying those actual statements.
Does a function have the property that the information contained in the function is accessible only through interactions at the interfaces supported by the object? Hmm, well, it can.
As some information is accessible via its interface, some information must be hidden and inaccessible within the function. (The property such information exhibits is called information hiding, which Parnas defined by arguing that modules should be designed to hide both difficult decisions and decisions that are likely to change.) So what information is hidden within a function?
To see this, we should first consider scale. It's easy to claim that, for example, Java classes can be encapsulated within a package: some of the classes will be public (and hence be the package's interface) and some will be package-private (and hence information-hidden within the package). In encapsulation theory, the classes form nodes and the packages form encapsulated regions, with the entirety forming an encapsulated graph; the graph of classes and packages is called the third graph.
It's also easy to claim that functions (or methods) themselves are encapsulated within classes. Again, some functions will be public (and hence be part of the class's interface) and some will be private (and hence information-hidden within the class). The graph of functions and classes is called the second graph.
Now we come to functions. If functions are to be a means of encapsulation themselves they they should contain some information public to other functions and some information that's information-hidden within the function. What could this information be?
One candidate is given to us by McCabe. In his landmark paper on cyclomatic complexity, Thomas McCabe describes source code where, 'Each node in the graph corresponds to a block of code in the program where the flow is sequential and the arcs correspond to branches taken in the program.'
Let us take the McCabian block of sequential execution as the unit of information that may be encapsulated within a function. As the first block within the function is always the first and only guaranteed block to be executed, we can consider the first block to be public, in that it may be called by other functions. All the other blocks within the function, however, cannot be called by other functions (except in languages that allow jumping into functions mid-flow) and so these blocks may be considered information-hidden within the function.
Taking these (perhaps slightly tenuous) definitions, then we may say yes: putting functionality within a function does constitute to encapsulation. The encapsulation of blocks within functions is the first graph.
There is a caveate, however. Would you consider a package whose every class was public to be encapsulated? According to the definitions above, it does pass the test, as you can say that the interface to the package (i.e., all the public classes) do indeed offer a subset of the package's behaviour to other packages. But the subset in this case is the entire package's behaviour, as no classes are information-hidden. So despite regorously satisfying the above definitions, we feel that it does not satisfy the spirit of the definitions, as surely something must be information-hidden for true encapsulation to be claimed.
The same is true for the exampe you give. We can certainly consider n = n + 1 to be a single McCabian block, as it (and the return statement) are a single, sequential flow of executions. But the function into which you put this thus contains only one block, and that block is the only public block of the function, and therefore there are no information-hidden blocks within your proposed function. So it may satisfy the definition of encapsulation, but I would say that it does not satisfy the spirit.
All this, of course, is academic unless you can prove a benefit such encapsulation.
There are two forces that motivate encapsulation: the semantic and the logical.
Semantic encapsulation merely means encapsulation based on the meaning of the nodes (to use the general term) encapsulated. So if I tell you that I have two packages, one called, 'animal,' and one called 'mineral,' and then give you three classes Dog, Cat and Goat and ask into which packages these classes should be encapsulated, then, given no other information, you would be perfectly right to claim that the semantics of the system would suggest that the three classes be encapsulated within the, 'animal,' package, rather than the, 'mineral.'
The other motivation for encapsulation, however, is logic.
The configuration of a system is the precise and exhaustive identification of each node of the system and the encapsulated region in which it resides; a particular configuration of a Java system is - at the third graph - to identify all the classes of the system and specify the package in which each class resides.
To logically encapsulate a system means to identify some mathematical property of the system that depends on its configuration and then to configure that system so that the property is mathematically minimised.
Encapsulation theory proposes that all encapsulated graphs express a maximum potential number of edges (MPE). In a Java system of classes and packages, for example, the MPE is the maximum potential number of source code dependencies that can exist between all the classes of that system. Two classes within the same package cannot be information-hidden from one another and so both may potentially form depdencies on one another. Two package-private classes in separate packages, however, may not form dependencies on one another.
Encapsulation theory tells us how many packages we should have for a given number of classes so that the MPE is minimised. This can be useful because the weak form of the Principle of Burden states that the maximum potential burden of transforming a collection of entities is a function of the maximum potential number of entities transformed - in other words, the more potential source code dependencies you have between your classes, the greater the potential cost of doing any particular update. Minimising the MPE thus minimises the maximum potential cost of updates.
Given n classes and a requirement of p public classes per package, encapsulation theory shows that the number of packages, r, we should have to minimise the MPE is given by the equation: r = sqrt(n/p).
This also applies to the number of functions you should have, given the total number, n, of McCabian blocks in your system. Functions always have just one public block, as we mentioned above, and so the equation for the number of functions, r, to have in your system simplifies to: r = sqrt(n).
Admittedly, few considered the total number of blocks in their system when practicing encapsulation, but it's readily done at the class/package level. And besides, minimising MPE is almost entirely entuitive: it's done by minimising the number of public classes and trying to uniformly distribute classes over packages (or at least avoid have most packages with, say, 30 classes, and one monster pacakge with 500 classes, in which case the internal MPE of the latter can easily overwhelm the MPE of all the others).
Encapsulation thus involves striking a balance between the semantic and the logical.
All great fun.
in strict object-oriented terminology, one might be tempted to say no, a "mere" function is not sufficiently powerful to be called encapsulation...but in the real world the obvious answer is "yes, a function encapsulates some code".
for the OO purists who bristle at this blasphemy, consider a static anonymous class with no state and a single method; if the AddOne() function is not encapsulation, then neither is this class!
and just to be pedantic, encapsulation is a form of abstraction, not vice-versa. ;-)
It's not normally very meaningful to speak of encapsulation without reference to properties rather than solely methods -- you can put access controls on methods, certainly, but it's difficult to see how that's going to be other than nonsensical without any data scoped to the encapsulated method. Probably you could make some argument validating it, but I suspect it would be tortuous.
So no, you're most likely not using encapsulation just because you put a method in a class rather than having it as a global function.
This is a question with many answers - I am interested in knowing what others consider to be "best practice".
Consider the following situation: you have an object-oriented program that contains one or more data structures that are needed by many different classes. How do you make these data structures accessible?
You can explicitly pass references around, for example, in the constructors. This is the "proper" solution, but it means duplicating parameters and instance variables all over the program. This makes changes or additions to the global data difficult.
You can put all of the data structures inside of a single object, and pass around references to this object. This can either be an object created just for this purpose, or it could be the "main" object of your program. This simplifies the problems of (1), but the data structures may or may not have anything to do with one another, and collecting them together in a single object is pretty arbitrary.
You can make the data structures "static". This lets you reference them directly from other classes, without having to pass around references. This entirely avoids the disadvantages of (1), but is clearly not OO. This also means that there can only ever be a single instance of the program.
When there are a lot of data structures, all required by a lot of classes, I tend to use (2). This is a compromise between OO-purity and practicality. What do other folks do? (For what it's worth, I mostly come from the Java world, but this discussion is applicable to any OO language.)
Global data isn't as bad as many OO purists claim!
After all, when implementing OO classes you've usually using an API to your OS. What the heck is this if it isn't a huge pile of global data and services!
If you use some global stuff in your program, you're merely extending this huge environment your class implementation can already see of the OS with a bit of data that is domain specific to your app.
Passing pointers/references everywhere is often taught in OO courses and books, academically it sounds nice. Pragmatically, it is often the thing to do, but it is misguided to follow this rule blindly and absolutely. For a decent sized program, you can end up with a pile of references being passed all over the place and it can result in completely unnecessary drudgery work.
Globally accessible services/data providers (abstracted away behind a nice interface obviously) are pretty much a must in a decent sized app.
I must really really discourage you from using option 3 - making the data static. I've worked on several projects where the early developers made some core data static, only to later realise they did need to run two copies of the program - and incurred a huge amount of work making the data non-static and carefully putting in references into everything.
So in my experience, if you do 3), you will eventually end up doing 1) at twice the cost.
Go for 1, and be fine-grained about what data structures you reference from each object. Don't use "context objects", just pass in precisely the data needed. Yes, it makes the code more complicated, but on the plus side, it makes it clearer - the fact that a FwurzleDigestionListener is holding a reference to both a Fwurzle and a DigestionTract immediately gives the reader an idea about its purpose.
And by definition, if the data format changes, so will the classes that operate on it, so you have to change them anyway.
You might want to think about altering the requirement that lots of objects need to know about the same data structures. One reason there does not seem to be a clean OO way of sharing data is that sharing data is not very object-oriented.
You will need to look at the specifics of your application but the general idea is to have one object responsible for the shared data which provides services to the other objects based on the data encapsulated in it. However these services should not involve giving other objects the data structures - merely giving other objects the pieces of information they need to meet their responsibilites and performing mutations on the data structures internally.
I tend to use 3) and be very careful about the synchronisation and locking across threads. I agree it is less OO, but then you confess to having global data, which is very un-OO in the first place.
Don't get too hung up on whether you are sticking purely to one programming methodology or another, find a solution which fits your problem. I think there are perfectly valid contexts for singletons (Logging for instance).
I use a combination of having one global object and passing interfaces in via constructors.
From the one main global object (usually named after what your program is called or does) you can start up other globals (maybe that have their own threads). This lets you control the setting up of program objects in the main objects constructor and tearing them down again in the right order when the application stops in this main objects destructor. Using static classes directly makes it tricky to initialize/uninitialize any resources these classes use in a controlled manner. This main global object also has properties for getting at the interfaces of different sub-systems of your application that various objects may want to get hold of to do their work.
I also pass references to relevant data-structures into constructors of some objects where I feel it is useful to isolate those objects from the rest of the world within the program when they only need to be concerned with a small part of it.
Whether an object grabs the global object and navigates its properties to get the interfaces it wants or gets passed the interfaces it uses via its constructor is a matter of taste and intuition. Any object you're implementing that you think might be reused in some other project should definately be passed data structures it should use via its constructor. Objects that grab the global object should be more to do with the infrastructure of your application.
Objects that receive interfaces they use via the constructor are probably easier to unit-test because you can feed them a mock interface, and tickle their methods to make sure they return the right arguments or interact with mock interfaces correctly. To test objects that access the main global object, you have to mock up the main global object so that when they request interfaces (I often call these services) from it they get appropriate mock objects and can be tested against them.
I prefer using the singleton pattern as described in the GoF book for these situations. A singleton is not the same as either of the three options described in the question. The constructor is private (or protected) so that it cannot be used just anywhere. You use a get() function (or whatever you prefer to call it) to obtain an instance. However, the architecture of the singleton class guarantees that each call to get() returns the same instance.
We should take care not to confuse Object Oriented Design with Object Oriented Implementation. Al too often, the term OO Design is used to judge an implementation, just as, imho, it is here.
Design
If in your design you see a lot of objects having a reference to exactly the same object, that means a lot of arrows. The designer should feel an itch here. He should verify whether this object is just commonly used, or if it is really a utility (e.g. a COM factory, a registry of some kind, ...).
From the project's requirements, he can see if it really needs to be a singleton (e.g. 'The Internet'), or if the object is shared because it's too general or too expensive or whatsoever.
Implementation
When you are asked to implement an OO Design in an OO language, you face a lot of decisions, like the one you mentioned: how should I implement all the arrows to the oft used object in the design?
That's the point where questions are addressed about 'static member', 'global variable' , 'god class' and 'a-lot-of-function-arguments'.
The Design phase should have clarified if the object needs to be a singleton or not. The implementation phase will decide on how this singleness will be represented in the program.
Option 3) while not purist OO, tends to be the most reasonable solution. But I would not make your class a singleton; and use some other object as a static 'dictionary' to manage those shared resources.
I don't like any of your proposed solutions:
You are passing around a bunch of "context" objects - the things that use them don't specify what fields or pieces of data they are really interested in
See here for a description of the God Object pattern. This is the worst of all worlds
Simply do not ever use Singleton objects for anything. You seem to have identified a few of the potential problems yourself