HWTSAM

Information and discussion on the MS Press release "How We Test Software at Microsoft"

Buy HWTSAM Now!

The HWTSAM bug list (so far)

Posted by Alan

No book on testing would be complete without a bug list. HWTSAM is no exception! Some "book people" call this stuff errata, but let's just call a bug a bug.

There are two distinct types of bugs: Author bugs, and Production bugs. The Author bugs are those where we, as authors, made mistakes. The Production bugs were introduced after our final sign off. To you, the end user, it doesn't matter - these are bugs. If we ever do a second edition, we'll take care of all of these.

Here's what we know about so far:

1) In the intro, I state that Ken approached me in the fall of 2007. On reflection, it was actually late fall 2006. I began writing the first chapter in January 2007 (chapter 4 to be exact). Minor issue, but an Author Bug nonetheless.

2) The table on p. 19 is missing a line between USA (California) and Hyderabad (Production Bug)

3) There's a bit of weirdness with the code coverage data on page 119. Bj explains it in detail on his blog here. (Author bug)

4) On page 125, the text refers to the Int.Parse() function, but it doesn?t exist in the code fragment (oops - Author Bug)

5) In the code sample on the bottom of page 235, I have mismatched quotes in two of the lines. (definite Author Bug).

6) The story about testing on page 227 has the phrase "as shown in the following graphic" For technical reasons, we pulled that graphic, but missed that bit of referring text (Author Bug)

7) Figure 14-2 on page 327 reads "Browsers (IF, Firefox, Mozilla)".  The IF should be IE or Internet Explorer but not IF (Author Bug)

8) Part IV starts on p. 363, but the section text at the top of the pages changes to part IV at the beginning of chapter 14. (Production Bug)

9 )In  the index (p. 413), Königsberg has garbage characters in place of the ö (Production Bug)

10) We forgot to include a bibliography (or forgot to think thoroughly about including one). Although we reference a lot of books via footnotes in the text, one tester in particular is quite annoyed that we didn't reference the writing that inspired us in the first place. I'll post a bibliography sometime and include one in the 2nd edition.

I'm sure I'm missing a few more. I'll add them as I find or hear about them.

Be the first to rate this post

  • Currently 0/5 Stars.
  • 1
  • 2
  • 3
  • 4
  • 5
Posted on: 4/7/2009 at 8:32 AM
Actions: E-mail | Kick it! | DZone it! | del.icio.us
Post Information: Permalink | Comments (4) | Post RSSRSS comment feed

Related posts

Comments

Ondrej Par cz

Tuesday, April 28, 2009 5:19 AM

Ondrej Par

Hi,

other bugs I've found so far:

p120: I think the postincrement operator in "retval = index++;" is redundant, and the sentence "(...) so we enter this block and increment the value of index by 1 and assign it to the retVal variable." is obviously wrong.

p123: the title of figure 5-6 is wrong (the function is SimpleGetNT5ClientVersion).

p124: "the fourth CFD demonstrates how the control fl ow jumps around A6 to the A7 case statement and returns the string WinXp" should read "the fourth CFD demonstrates how the control fl ow jumps around A4 to the A5 case statement and returns the string WinXp"

Bj Rollison us

Tuesday, April 28, 2009 6:11 AM

Bj Rollison

Hi Ondrej,

Thanks for pointing out these oversights.

The incremental operator "retVal = index++;" is not redundant, it is just not doing anything, or as you mention it is not incrementing the value of index by 1. It should have been "retVal = index + 1;" to indicate the ordinal position of the character in the string (starting at 1).

Ondrej Par cz

Wednesday, April 29, 2009 2:17 AM

Ondrej Par

Hi Bj,

thanks for clarifying that. The description of this example, however, doesn't explicitly mention that the characters are counted from 1.

Another problem in this chapter is the result of basis path testing of Luhn's algorithm. AFAIK the number "0" is valid Mod10 number, am I wrong?

Bj Rollison us

Thursday, April 30, 2009 3:14 AM

Bj Rollison

Hi Ondrej,

You're right, we don't explicity state characters are counted from 1. The implication is that the ordinal value of the first character in a string is 1 because most people don't start counting at 0.

And yes, the number 0 is a valid Mod10 number, which is why I wrote "...assume that the developer validates the argument value as an integer greater than 0..." on page 142.

Comments are closed