Week of 12 July 2010
Update: Thursday, 15 July 2010 12:41 -0400
We pretty much finished the build on the extreme system yesterday,
although there are still cables hanging out of it every which way.
That's the sixth and final of the book's project systems. Two of them
are already in use. We gave the budget system to Barbara's sister
Frances to replace her old system, and the appliance system is now
working as the replacement for my old den system. We'll get the media
center system installed next, but that'll probably be after we finish
the book. Eventually, the extreme system will replace my main office
Speaking of the book, our to-production deadline is 31
July, which is a Saturday. In practice, that means the real deadline is
Monday, 2 August. I assume my editor already has a production slot
reserved, so that's an absolute drop-dead deadline. Once the book goes
to the production folks, it's completely out of our hands, other than
answering queries and proofing the PDFs and making minor corrections.
see that B&N is currently listing the book with a 15 November pub
date, but that's really just a guess that O'Reilly made months ago.
Well, more than a guess, because I'm sure they absolutely, positively
wanted the book to be available in time for Christmas sales, which
means a pub date no later than mid-November. Based on past experience
with how fast O'Reilly works, I'd guess the book will actually be
published in October or even late September. A lot depends on what else
is in their queue, what priorities are assigned to the different books,
the printer's schedule, and so on. But I'm confident that the book will
be available well in time for Christmas sales.
We'll just hope
the book sells well. Ten years ago, a lot of computer books would have
a sell-in (pre-purchases by distributors and booksellers) of 10,000 to
15,000 copies, which meant those books essentially earned out their
advances immediately. Nowadays, a typical sell-in is just a few
thousand copies. Bookstores that used to have several aisles devoted to
computer books may now have just one small section. Instead of ordering
a dozen copies of a lot of titles, they'll order two or three copies of
a few titles. So sell-in is still important--to get the books on the
shelves in the first place--but sell-through (re-orders) is now the
important metric in terms of earning out the advance and getting into
In terms of meeting the to-production
deadline, we're behind where I wanted to be, but ahead of what I
actually expected. We have three weeks left, and the bulk of the work
is done. Of course, that's a result of working a seven-day per week
schedule since April. Once I get the extreme system chapter finished
and off to the tech reviewers (Brian Bilbrey and Ron Morse), I'll start
incorporating tech review comments in those chapters and also do
re-write on the early narrative chapters and the preface.
Realistically, we may finish up a few days ahead of schedule, which is
extraordinary given the short deadlines for this book.
Of course, I'll be busier than the proverbial one-armed paper hanger
for the next three weeks, so updates here will be short and sporadic.
Hah! Just did the smoke test on the extreme system, which powered up
normally. Counting the power supply and CPU cooler, this beast has nine
(!) fans, so it's not exactly a silent PC. With the cover off, it looks
like Times Square in there, with seven red-LED fans and a bunch of
blinking blue LEDs on each memory module. Of course, it'll look like
that with the cover on, too, because the left side cover is transparent
It really annoys me when people claim credentials they don't have.
Frankly, it annoys me when some people claim credentials that they do
have. People in non-rigorous disciplines should not be using
the title "doctor", even if they're legally entitled to it. To me,
the title "doctor" means, or should mean, that the person holds an MD
or a Ph.D. or a D.Sc. in a rigorous discipline, such as hard science, mathematics,
or engineering. I think it's dubious, to say the least, to award that
title to someone in a field where correct answers are a matter of
opinion. Someone who is truly accomplished in, say, history, should
hold a masters as a terminal degree. And, in fact, that's the way it used to be, even at the world's best universities.
Diluting the Ph.D is bad enough, but people claiming throwaway Ph.D.'s as real is much worse. So I was pleased to read Gillian McKeith does not have a Ph.D. on Pharygula. That article is baed on Ben Goldacre's Bad Science article
on the subject. As PZ remarks, if someone earned a real Ph.D. it's
easy enough to find out. I did a quick check of two friends who hold
real doctorates. It took me all of 15 seconds to find Mary Chervenak's
doctoral dissertation, and not much longer to find Paul Jones's.
- This is interesting. The journal Neurology has just published an article
that suggests there is an inverse correlation between head size (used
as a proxy for brain size) and the severity of Alzheimer's symptoms.
Apparently, larger brains have more excess capacity that is brought
into play automatically as Alzheimer's damages a brain. I hope it's
true, because I wear an extremely large hat.
But I wonder if it
is true. Women typically have smaller brains than men, so this research
suggests that women should suffer disproportionately from Alzheimer's.
Is that the case? I have limited exposure to Alzheimer's patients, most
from the time when my mother was in the nursing home. There were
certainly a lot more women Alzheimer's patients there, but there were a
lot more women there, period. Women live longer than men.
Of course, this may all be immaterial to my situation. I think my head is shrinking as I get older.
I sent the first draft of the extreme system chapter off for tech
review yesterday. I've already gotten back Ron's comments and I expect
Brian's will follow soon. That gives me 11 of 12 review comments on the
six build chapters, so I'm going to spend today going through them and
incorporating the changes and suggestions. Once I finish that, I'll go
to work on updating the two early narrative chapters.
interesting. Usually, at this stage of a book I'm running flat-out,
convinced that there's no way I can get everything I want to do done in
the time remaining before deadline. This book was on a very short
deadline, and yet it feels as though I have plenty of time to finish it
up and even do a final pass through to make some minor tweaks. I may
even take a day off to do some lab stuff and shoot a YouTube video.
I just finished posting the second drafts of the six system chapters.
Ron and Brian have been turning around tech review comments very fast,
and at this point those chapters are ready for our editor to starting
hacking on. Today and tomorrow, Barbara is off on day trips: Galax, VA
today with her parents, and Lexington, NC tomorrow with her friend
Marcie. I'll be working on the two narrative chapters on fundamentals
and choosing/buying components.
I'm waiting for FedEx to show up with a couple of 32 GB Lexar Echo SE Backup Drives,
which will feature in the book. As it turns out, they're arriving just
in time. I do my daily backups to two internal hard drives on my main
system, as well as two external hard drives. Then, being a
belt-and-suspenders kind of guy, I copy the backup to a USB flash
drive. The problem is, the size of my daily backups grows when I do a
book. Ordinarily, my and Barbara's working data directories (including
our compressed Linux home directories) might be 500 MB. I've been
writing those data sets to a pair of Kingston 4 GB flash drives that I
bought two years ago.
Not only are those Kingston drives small,
they're fairly slow. Realistically, I get at most 4 to 5 MB/s when
writing. With a 500 MB data set, that's only a couple of minutes. But
my daily backup data set is now 4.1 GB (and growing), so it'd take 15
minutes or so to write, assuming the drive had enough capacity. With
flash drives, there's always a trade-off between capacity and speed.
The fastest flash drives are available only in small capacities, mainly
because few people would pay the price for a flash drive that was both
fast and high-capacity. Conversely, most high-capacity flash drives are
quite slow, because (again) few people would pay the price to have both
speed and capacity.
The Lexar Echo SE drives strike a happy
medium. They're available in 16, 32, 64, and 128 GB capacities, and
their write speeds are specified as "up to" 10 MB/s. Of course, that's
almost certainly specified for copying single large files, because
copying many small files incurs a lot of overhead and can slow write
speeds dramatically. Still, I'm hoping that these drives will be
roughly twice as fast as what I have now, not to mention eight times
larger. I'll run them through their paces over the next week or two.
1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010