Errata

"macorlib.dll" should be "mscorlib.dll"

The heading of the dotted line circle reads macorlib.dll.
It should be mscorlib.dll.

Change "We explain assemblies in detail in Chapter 16." to "We explain assemblies in detail in Chapter 17."

Change ?The semicolon is used to terminate a statement. ... The Braces are used to group multiple statements ... The period denotes ... The parentheses are used ... (the equal sign ... ? to ?A semicolon is used to terminate a statement. ... Braces are used to group multiple statements ... A period denotes ... Parentheses are used ... (an equal sign ...?

Although this is very early in the book and in the development of C# 4.0 concepts, perhaps a footnote should be added that refers to pg. 947, first paragraph, 2nd and 3rd sentences, which say about multiple application domains :

"... each domain has separate memory, and objects in one domain cannot interfere with those in another. Furthermore, static members of the same class have independent values in each domain."

The text refers the reader to Chapter 19 for a discussion of the lock statement and the Monitor class but that discussion is in Chapter 21.

The example code demonstrates aliasing a namespace, but the expression references a class, not a namespace and will not compile.

Replace ?The syntax for using indexers is like that for using arrays when the index is an integer type.? with ?The syntax for using indexers is like that for using arrays except that the index parameter(s) can be any type(s).?

"To get to its SharedOwned field [...]" should be replaced with: "To get to its SharesOwned field [...]"

Replace ?? the default accessibility members of a class or struct? with ?? the default accessibility of members of a class or struct?

Add a reference to this subsection to the index.

"nee- / ded" to "need- / ed"

This sentence should say :

"... get back an object, but an STRING subclass will also do:",

instead of :

"... get back an object, but an OBJECT subclass will also do:",

This sentence states :

"If the enumerator implements IDisposable, the (bold font) foreach statement also acts as a (bold font) using statement, ..."

It should say :

"If the enumerator implements IDisposable, the (bold font) using statement also acts as a (bold font) finally statement, ..."

Should be:

... (see "try Statements and Exceptions" on page 138 in Chapter 4 or ...

Change "third" to "fourth". 1) Static, 2) Dynamic, 3) Custom, 4) Language, 5!) Interop. Wait change it to "fifth" or "another"!

* "Among the our main types of binding are: ..." -- Monty Python, Spanish Inquisition sketch.

Replace ?We describe how to write your own attributes in Chapter 17.? With ?We describe how to write your own attributes in Chapter 18.?

Replace ?The types for communicating with WCF are in, and are below, the System.ServiceModel namespace.? with ?The types for communicating with WCF are in, and below, the System.ServiceModel namespace.?

Change "intero- / perbility" to "inter- / operability"

Replace ?The result of the final comparison ?? with ?The result of the third example ??

Array.CreateInstance appears as two separate words (Array.CreateIn stance)

Close up the white space between ?}? and ?Public delegate TOutput Converter ? .?

TryGetValue appears as two different words (TryGet Value).

It looks like there is some typographical problem throughout the book; I sent similar errata for pg. 277, and I suspect there may be more. The typeface used for function names does not display a hyphen if one of these names is spread upon two different lines; this might lead to confusion.

In the last line of the code example, better use " - Len: " instead of just "Len: " to avoid messing the file names.

Page 478, code listing in "Opt-in Disposal":
Code defines a class called "HouseManager" but the following paragraph in pg 479 2nd sentence:
"This, however, relies on the consumer knowing what's inside Demo's Dispose method."

There is no Demo class.

Same for page 479, 2nd code listing:
Constructor called Demo in a class called HouseManager.

This sentence states :

"In effect, the server GC is tuned for throughput rather than latency."

The tradeoff could be more exact by stating :

"In effect, the server GC is tuned for maximizing throughput rather than minimizing incidents of latency."

This sentence says :

"... forcing collection can hurt performance by unnecessarily promoting Gen0 objects to Gen1."

Shouldn't this sentence also state :

"... forcing collection can hurt performance by unnecessarily promoting Gen0 objects to Gen1 (and Gen1 objects to Gen2)."

The body should read

if (message == null) throw new ArgumentNullException("message");

not "message != null".

Everywhere I see "overridden" on this page, and on page 516 in "Postconditions and Overridden Methods", it seems wrong. I believe "overriding" is correct.

"You keep using that word. I do not think it means what you think it means." -- Inigo Montoya

Regarding the absence of the free GUI debugger DbgCLR(.exe) in SDKs past v2.0, pages found in Google about DbgCLR suggest downloading the free Visual Studio Express to get a CLR debugger with a GUI interface. Of course, the Express edition is not stripped down to just the debugger. Also, the lifetime of the Express edition is limited.

The Visual Studio 2010 Integrated Shell (but not the Isolated Shell) also appears to contain the Visual Studio debugger. This could be a useful alternative to downloading the (temporary) instance of Visual Studio Express. The description of it says :

"By using the Visual Studio Shell (integrated mode), you will not need to make Visual Studio a prerequisite on your user machines."

I have verified this by installing the Integrated Shell on a machine that did not have Visual Studio installed. It can be downloaded from here :

http://www.microsoft.com/download/en/details.aspx?displaylang=en&id=115

Should a footnote be added, warning not to confuse the WebClient class in System.Net namespace with the WebClient class in Silverlight's subset of the .NET Framework?

This sentence states :

"...(this allows it to be sited in Visual Studio's designer)."

As with suggestions I made for pages 852 and 870, it would be more explicit to state :

"...(this allows it to be sited in the Visual Studio designer panel's component tray)."

Should "context.response.ContentLength64 = " read

"context.response.ContentLength = " ?

A missing 'a' in assembly
Line says "The compiled code in an ssembly"

Single dispatch is explained in the preceeding paragraphs. The last sentence of the 3rd paragraph from the bottom states :

"..., only the runtime type of the receiver can vary which method gets called."

Shouldn't this sentence say :

"... only the compile-time type of the receiver can vary which method gets called."


The first sentence of the 3rd paragraph from the bottom states :

"The runtime decision of whether of whether to invoke a dog's Walk method or a cat's Walk method ..."

Unfortunately, it is the following, 2nd sentence that develops the fact that the decision is made at compile-time, not runtime. Nonetheless, should the first sentence be more exact and say :

"The compile-time decision of whether of whether to invoke a dog's Walk method or a cat's Walk method ..."

These sentences state :

"The result of Intersect is (italic font) less restrictive when Demanded, requiring (italic font) intersected permissions to be met.
The result of Union is (italic font) more restrictive when Demanded, requiring both permission sets to be met."

According to Visual Studio 2010's Help, the italicized words should be in the other respective sentences. Also, in the 2nd clause of the third sentence should have the opposite meaning. These sentences should read :

"The result of Intersect is (italic font) more restrictive when Demanded, requiring (italic font) intersected permissions to be met.
The result of Union is (italic font) less restrictive when Demanded, requiring (italic font) either one of or both permission sets to be met."

This sentence states :

"To maintain security, methods that call link-demanded methods must themselves perform demands or link demands - ..."

Should this be more precise and say :

".. methods that call link-demanded methods must themselves perform demands or link demands on THEIR immediate caller - ..."

This sentence states :

"Partially trusted assemblies can never call security critical methods in fully trusted assemblies."

It would be more exact (albeit lengthier) to state :

"Methods in partially trusted assemblies can never call security critical methods in fully trusted assemblies."

This sentence states :

"This works well as long as all partaking assemblies do the same thing - ..."

Shouldn't this say :

"This works well as long as all the methods in other assemblies that call of your method do the same thing - ..."

This sentence states :

"Write [SecuritySafeCritical] wrappers around calls to X."

It would be more exact to state :

"Write [SecuritySafeCritical] wrappers around calls (by methods in assembly T) to methods in assembly X."

This sentence states :

"Go fully critical yourself."

Would it be clearer to state :

"Make your assembly T fully critical."

This sentence states :

"Write [SecurityCritical] wrappers around calls to X."

It would be more explicit to state :

"Write [SecurityCritical] wrappers around the methods in assembly T's calls to methods in X."

This sentence states :

"In other words, it's now entirely up to the host as to how permissions should be restricted - ..."

It would be more explicit to include examples, as in this version (if my understanding is correct) :

"In other words, it's now entirely up to the host (such as ASP.NET or an application domain) as to how permissions should be restricted - ..."

On page 954, in the second sentence of the paragraph after the tip, this section is referenced. It might be helpful to add a footnote on pg. 769 that references the "Isolating Types and Assemblies" section on pg. 954.

The second sentence starts with :

"(If you don't call Wait and abandon the task, ..."

This clause is a little ambiguous. It could be taken to mean that if BOTH wait is not called and the task is not abandoned. It would be clearer to say :

"(If you don't call Wait, but instead abandon the task, ..."

The heading "Static methods" should be "Static members".

This sentence starts off with :

"In other words, by making static methods thread-safe, ..."

This should instead be :

"In other words, by making static members thread-safe, ..."

This sentence states :

"..., it pays to be cautious in assuming whether a method is read-only in nature."

It instead state :

"..., it pays to be cautious in assuming whether a member is read-only in nature."

The second sentence states :

"The main thread of a WPF or Windows application, for example, ..."

I believe what is meant is :

"The main thread of a WPF or Windows Forms application, for example, ..."

I assume Windows console apps were not meant to be included.

This bullet point states :

"An implementation of IComponent allowing it to be sited in Visual Studio's designer"

To be clearer, this would be in the Component Tray at the bottom of Visual Studio's designer.

This bullet point states :

"A Component implementation, allowing it to be sited in Visual Studio's designer"

It should state :

"An IComponent implementation, allowing it to be sited in the Visual Studio designer's component tray". See the correction for pg. 852.

This bullet point states :

"If the query starts with Enumerable.Range, replace the latter with ParallelEnumerable.Range."

There is no "former" in the first clause of this sentence, unless it is supposed to be "the query". It would be clearer to say :

"... replace the start with ParallelEnumerable.Range."

To be explicit, an aside or footnote should be added saying that Parallel.Invoke should not be confused with P/Invoke (interop).

The first sentence states :

"This program will handle 1,000 concurrent requests on fewer than 10 pooled threads."

However, the first executable statement in method Serve in class Server on pg. 932 is :

ThreadPool.SetMINThreads (50, 50);

Therefore, shouldn't the first sentence state :

"... on as few as 50 pooled threads."

the signature seemingly should be that of CreateInstanceAndUnwrap, instead of CreateInstanceFromandUnwrap

Add a footnote somewhere warning not to confuse P/Invoke (interop) with Parallel.Invoke.

This should be "Basic Multilingual Plane".

"lambda" typo (as "lambca")

"delgates" => "lambda expressions (see lambca expressions)".

"Extern" is also discussed on pg. 766

Add pages 420 and 510 as an index entry. Of course, the purity discussed on pg. 510 is not about PLINQ.

Immutable objects are also referred to on pg. 253.

Add a reference to pg. 585 for IComponent in the index.

(Ambient) properties are also discussed on pg. 152.

Under index entry for "signaling", the 2nd sub-entry says :

"with Wait and Pulse (see Wait and
Pulse1)"

this should be :

"with Wait and Pulse (see Wait and
Pulse)"

The index entry for "WPF" lists only pg. 188. However, the book also briefly mentions WPF on pages 492, 493, 822, 840, and 851, from what I've seen so far.

Windows Forms are also referred to on pages 493, 666, 670, 756, 791, 799, 814, 851-852, 840, 869-872, and 909.

WPF/Windows Presentation Foundation is also referred to on pages 183, 234, 665-666, 670, 668, 756-757, 791, 799, 814, 852, 869-872, and 909.