Embedded questions normally don't allow Subject-Auxiliary inversion, though it's required for normal questions:
- How can we do this? (Aux can and Su we are inverted)
- *How we can do this? (No inversion; ungrammatical)
However, in spoken English (which is similar to the chatty, informal style used in many computer books), an embedded question complement can optionally invert the subject and the first auxiliary verb, like a regular question does.
- I wonder how we can do this. (No inversion; normal for embedded Q)
- I wonder how can we do this? (Su-Aux inverted; marked pragmatically)
There's a pragmatic distinction between the two forms, with and without the inversion. The presenting construction -- the one that resembles a real question -- is in fact intended to operate like a real question, and thus invite an answer from the addressee, rather than being intended to be interpreted as a simple rhetorical question.
For that reason, it's often encountered punctuated like a question; and sometimes a comma intonation -- or even a semicolon -- dashes in to separate the clauses.
- I wonder, how can we do this?
- I wonder -- how can we do this?
Now, this particular instance is in a piece of writing, where any question the author asks has to be rhetorical, after all, so this usage in this context probably is meant simply to increase the chumminess of the Boost.Interprocess salesman's pitch. If you're interested, there's an extensive literature on such syntactic marking in Pragmatics; the key term is "indirect question".
- I wouldn't worry about this (unless I were being paid to edit it for an ESL textbook).