按键盘上方向键 ← 或 → 可快速上下翻页，按键盘上的 Enter 键可回到本书目录页，按键盘上方向键 ↑ 可回到本页顶部！
————未阅读完？加入书签已便下次继续阅读！
key/value pairs in the section。
…………………………………………………………Page 362……………………………………………………………
340 CH AP T E R 1 2 ■ L E A R N I N G A B OU T A PP L I CA TI O N CO N F I G U R AT IO N AN D D Y N A M I C L O AD I N G
o You can enhance the application configuration file； but that implies implementing
types that specifically outline which XML nodes and attributes are to be present in the XML。
o When processing data； you sometimes need a type （such as ConfigurationInfo） that is
used to store information in a structured manner for a specific reason。
o has plenty of functionality that allows a developer to dynamically execute code。
o One way of dynamically executing code is to load an assembly； instantiate a type； and
then cast the instance to a specific known type。
o You can instantiate private types declared in an assembly。
o You can use the GAC to store assemblies that will be shared by multiple applications。
o Use the GAC only if you must。 If you do use the GAC； you need to create strongly named
assemblies using signing。
o Regardless of how an assembly is versioned； a version number should exist and be used。
o In convention…based coding； your code makes assumptions about its behavior。
Convention…based coding is more pact； but because assumptions are made； the
code is less flexible。 However； do not confuse flexibility with being the best way to do
something。 You should have flexible code when you need it； not just because it can be
written that way。
Some Things for You to Do
The following are some exercises for you to do to apply what you’ve learned in this chapter。
1。 Rewrite the configuration code so that the configuration file specifies a directory where
all assemblies in that directory are loaded。
2。 Having read all of the assemblies with the rewritten configuration； call a predefined
type that will tell the caller which types are available and their appropriate types。
3。 Modify ConfigurationLoader so that it uses the new configuration architecture to
instantiate types。
…………………………………………………………Page 363……………………………………………………………
C H A P T E R 1 3
■ ■ ■
Learning About Multithreading
Your brain allows you to multitask。 For example； you can prepare dinner while talking on the
telephone。 However； this multitasking has limits； and you can do only two or three things
simultaneously。 But suppose you could put down the work； start another piece of work； then
put that down； and then switch to the original work。 How many tasks could you handle at the
same time？ Probably a few hundred； because what you are doing is serializing the multitasking。
Now suppose you and another person are preparing dinner in the kitchen； but you are not
municating with each other。 What is the likelihood that you will run into the other person？
Probably pretty high。 What I am trying to get across here is the difference between multitasking
with a single brain and multitasking with multiple brains。 Multitasking always has a cost； which is
orchestration。 And sometimes doing more multitasking is not going to speed things up。 There
is a limit to how many brains are required to run an efficient kitchen。
The focus of this chapter is how to write code that is capable of multitasking。 The operating
system multitasks； by default； but whether or not your program multitasks depends on how
you write your code。 For example； suppose the lighting controller introduced in Chapter 8 had
to control 3；000 rooms？ To be able to process such a large number of rooms； you would use the
same code； except run it in parallel。 And therein lies the problem of code that needs to multitask。
It’s more difficult to make code run efficiently in parallel。 A mon problem in multitasking
applications is the dreaded deadlock。 A deadlock occurs when two tasks need data from each
other to continue； which means neither can continue。
Running code in parallel requires coding discipline； as you will learn in this chapter。 We
will use only a single project； named JugglingTasks； which is a console application that imple
ments the techniques demonstrated in this chapter。
Understanding Multitasking
One of the biggest jumps in puting history was when the puter went from a machine
that carried out a single task to a multitasking machine。 The old modore 64 and VIC…20
were single…tasking machines。 You started the puter； and then the puter waited for
you to do something。 If you decided to make the puter run a loop saying “hello” millions
of times； that is all the puter would do。 The puter could not do one thing and then do
something else while waiting for an answer from the executing task。 How could anyone get
anything done？ （On the other hand； you could argue that people might be more productive if
they could work in only a single program； and not also check their e…mail； look at the latest blog
entry from their favorite author； and so on。）
341
…………………………………………………………Page 364……………………………………………………………
342 CH AP T E R 1 3 ■ L E A R N I N G A B OU T M U L T I TH R E A DI N G
Things changed quite dramatically when puters could multitask。 Server puters
running operating systems like Unix were multitasking from day one。 Operating systems like
Windows were not 100％ preemptive multitasking until Windows NT and Windows 95。 Notice
that I used the term preemptive； which makes a big difference。
Plain…vanilla multitasking is multitasking where individual tasks cooperate and allocate
resources as a collective。 It is essentially a dead…end idea； and the best example of how it worked is
Windows 3。0 and Windows 3。1 （but these were full of hacks）。 Now let’s see what preemptive
multitasking means。
Preemptive Multitasking
Preemptive multitasking is when the operating system controls which application does what
and when。 Realize that you can run an application because applications are treated as po
nents。 You don’t believe me？ Create a console application and rename the method Main（） to
Mains（）。 See what happens。 The reality is that the Main（） method； as it is declared； is an API
used by the operating system to run your ponent； which masquerades as a program。
So we have this program called the operating system that runs ponents called programs。
The next question is how can multiple programs run at the same time？ The answer is that the
operating system program is no ordinary program。 It is the foundation upon which applica
tions can be launched and managed。 The operating system hooks into special features of the
microprocessor that allow the operating system to time slice the individual programs。 Because
your programs are ponents； they will never need to access the microprocessor directly。
Time Slicing
Time slicing is when an operating system can dictate for how much time a program is allowed
to execute。 Between the times of execution； the program is in a state of deep freeze and does
nothing。 You； as a user； are not aware of the time slices； because a time slic