Dr. Jerry Pournelle

Email Me

Why not subscribe now?

Chaos Manor Subscribe Now

Useful Link(s)...


Hosting by

Powered by Apache

Computing At Chaos Manor:
March 11, 2010

The User's Column, March 2010
Column 356
Jerry Pournelle jerryp@jerrypournelle.com
Copyright 2010 Jerry E. Pournelle, Ph.D.

A Coming New Revolution

The history of mankind is the history of civilizations. There are of course many variations, but a study of successful civilizations everywhere reveals a common - I'd say actually universal - theme. Successful civilizations grow - if they didn't grow we wouldn't consider them successful - and as they grow they devote more and more of their output to structure. Moreover, this isn't a linear expansion; they devote a larger and larger percentage of this growing output to internal structure.

This continues until there is so much structure that growth slows. Alas, the growth of structure doesn't slow despite the lack of growth in output to support it. Eventually there is so much structure that it's very hard for anything to be done, and the civilization collapses. There are a number of ways that this decivilization can take place, and this isn't the place to describe them; we can all think of our favorite examples.

Western civilization shares this history. From time to time, though, there have been spurts of output growth so large that the structure builders can't keep up. These are pivotal events in the history of civilizations. The best known such in Western Civilization were the Discovery of the New World, the combination of the Enlightenment and Industrial Revolution, various other phases of the Industrial Revolution, and then, beginning in about 1980, the Computer Revolution. In every case there were spurts of growth accompanied by periods of freedom; and in every case the structure builders eventually caught up. The Computer Revolution, for example, was at least as profound in its economic effects as some of the earlier phases of the Industrial Revolution - but the regulators and structure builders needed far less time to catch up. As a simple example, the State of California added 40,000 state employees in the past eight years. No one I know believes there was any noticeable improvement in state services during that time. It was just an increase in structure, hardly noticed during the boom times.

I bring this up as background to what I see as another coming revolution.

I'm not entirely certain what to call this new revolution. Various names suggest themselves, but it's not as obvious as The Computer Revolution. (Long time readers will recall that when I was lead columnist for both Byte and Byte's sister magazine Popular Computing, my Popular Computing column title was "The Computer Revolution.")

The Computer Revolution certainly changed the way of life for most people in civilized societies, particularly in the West but also in Japan and the Orient. Employment patterns changed. The very nature of employment changed. Management of enormous enterprises were simplified (or appeared to be), and large organizations became much larger. Manufacturing became so efficiently productive that while manufacturing output increased, employment in manufacturing declined sharply. Note that this had already happened in agriculture half a century before: far fewer farm workers grew far more food. It took a while for the Computer Revolution to do the same in manufacturing. Both trends continue and are not likely to be reversed.

It's not clear - at least to me - what all specific effects the new revolution will have. Some are easy to predict. Others remain obscure. The overall effect, though, is likely to be equally as profound as the Computer Revolution. I admit that's vague, but so is our understanding of the new technologies that make up the next revolution.

Marks of the New Revolution

By the next revolution I do not mean the "singularity" which has been the lead topic of many essays, discussions, and conferences. The "singularity" generally refers to the time when a workable self-reproducing artificial intelligence is developed. It may or may not include nano-technology and nanobots. The singularity is predicted for any time after about 2030, with a general consensus among those interested in the subject that it will happen before the end of the 21st Century, probably well before.

Note that "singularity" is a term of art taken from physics, and refers to a condition beyond which we can no longer make predictions. Science just doesn't work in a singularity. Science fiction writer Vernor Vinge used the term to mean a point in the future at which so many technological developments matured that it is simply impossible to predict what happens next. Ray Kurzweil has run with that concept, and his book The Singularity is Near explains his views better than I can, as does the Singularity web site http://singularity.com/. I will write about the revolution leading to the singularity another time.

The revolution I refer to is something beyond the Computer Revolution, but doesn't go as far as a singularity. For lack of something better I have thought of calling it the Knowledge Delivery Revolution. It began with many manifestations, spawned iPod and iPaq, continued with iPhone and Kindle, and will grow stronger with the iPad and the coming Microsoft iPad rival Courier. Components of this revolution include the rapid development of WiFi and telephone networks, as well as the continued development of technologies growing out of the TabletPC.

Another component of the coming revolution is continuous WiFi connectivity, which requires battery life. Peter Glaskowsky notes "The latest Wi-Fi chipsets consume very little power on average - which you can see yourself, with your iPhone, since you can associate it with your home Wi-Fi yet it runs for a day or two on about 5 watt-hours of energy. Modern Wi-Fi chips consume only about 50 milliwatts on average to keep the connection open, which amounts to an hour or two difference in an iPhone but is entirely negligible in a PC, and they turn off automatically when not actively used. Power consumption while operating is also far lower than it used to be." I expect that technology to improve.

There are some minor signs of this new revolution already. Apple share prices went up on announcement of the April 3 shipping date for iPad. Lots of people want iPads and Apple's main problem may be making enough of them. I'm looking forward to seeing what's really in the iPad, but I'm not quite that eager to get one. It's a Kindle with color and a lot more, but for now the Kindle will have to do for me.

I'm a bit more interested in the Microsoft Courier, but there are fewer solid details of what it will be than there are for iPad. What we see of Courier is less than you get from a peep show, and I know only one person with much Courier experience. From what I've heard I'll like it.

It's Coming

We're at the beginning stages of this revolution. We glimpse the coming technology. Present technology is not quite powerful enough and certainly costs too much. Both those factors will change, certainly over a decade and probably much faster. My first computer, Ezekiel, an S-100 2 megahertz box with 64K (that's Kilobytes) of memory and 128K "mass storage" on 8-inch floppy disks, cost $12,000 in 1979. It didn't do a lot compared to machines that came out a few years later, but it changed my life: my writing output went way up once I learned to use that machine, even though my writing and editing program was primitive by any reasonable standard. (Old Zeke was on display in the Smithsonian Museum of American History in the Hall of Communications and Computing when they closed that hall for refurbishment. I was told he'd be back.) A few years later IBM came out with the first PC. Computer prices fell as computing power increased, and the computer revolution was on. Many of us remember the economic ferments of the time, including booms, a bust, and another boom all coming in rapid succession.

No one regulated the Computer Revolution. It was a wild frontier, and it just grew; no one in government understood it well enough to regulate it, and it was creating new wealth and wealthy people every month. At the same time the Department of Defense was developing the ARPANET. The backbone was much faster, but most users experienced it at 300 Baud (300 bits/second, 1200 if you could afford a really fast modem).

In 1982 I said that by the year 2000 anyone in Western Civilization would be able to get the answer to any question: that knowledge would no longer be hoarded or restricted but freely available. Not long after we had the CDROM, which greatly lowered the price of knowledge distribution. Not long after we saw the beginnings of the Internet, followed by the World Wide Web. Those sparked another economic boom.

I think that's happening again. This time the principal effect of the Computer Revolution will be on publishing, education, news delivery, and entertainment. The effect on each of these important areas of life will be at least as great as the effects of the Computer Revolution have been on the way we live.

For a picture of some of the possible effects of the Knowledge Delivery Revolution, go look at Microsoft's pitches about their Courier which they'll be able to deliver Real Soon Now. You can see new pictures of the coming Courier at this link. There's more on Engadget, and YouTube. Now these are of course teasers (Microsoft calls them concept videos), and in the absence of a shipping product they have to serve as Microsoft's answer to the upcoming Apple iPad. We don't know when Courier will ship, or what it will cost, and it's all speculation on what it can do. As Peter Glaskowsky "notes there hasn't been a machine released in the last 5 years that can do what my Motion LS800 can do, and there hasn't been a machine released since 1997 that can do what the Newton did." All true, but what man can do, man can aspire to. What Microsoft and Apple have done, rapidly improving technology dictates that they or others will surpass. I hope Courier will be able to do what Microsoft shows in concept, but it doesn't have to for us to get some idea of what's in store for us. This puts a lot of new meaning into the concept of a Personal Digital Assistant. Information is available everywhere: this makes it more manageable.

I've frequently said that currently the best research tool I know of is a good TabletPC with Microsoft OneNote. Assuming that Courier has some kind of USB port and thus a way to attach an external keyboard - and if it doesn't, the next iteration will - then we're looking at a netbook TabletPC. I'm told that the projected Courier won't run OneNote, at least not the OneNote I am familiar with, but I'll bet it won't be long before there's a good enough version from someone. Think OneNote with a better interface than the current TabletPC. Now imagine university science courses designed to take advantage of that. Textbooks will have video illustrations including cut scenes from great lecturers; imagine learning quantum mechanics from Dick Feynman, all built into your textbook reader. Imagine being able to work physics problems by simply drawing and animating them - and everyone has that capability.

Note this isn't a pitch for Courier. All I'm doing is a linear projection of what's almost certainly coming in the next year or so. It's easy enough to make the same projection about media consumption, both informative and entertainment. Newspapers and paperback books are already affected. TV, both broadcast and cable, have already been affected and will be even more so. The whole notion of libraries is changing.

As the technology improves and becomes cheaper, the demand for these devices will increase. With larger numbers using them, more people will develop applications. We saw all that in the Computer Revolution. It will happen again, and the way we gather and disseminate information and knowledge will be changed forever.

The immediate effects of Kindle and iPad and Courier won't be so great, of course. The machines cost too much and do too little. Battery life is too short. All that will change.

I can hardly wait to see what happens over the next few years.

Create Your Own

Another important trend is the empowerment of artists. We've seen this one coming for quite a while, and the trend is accelerating. See this article from Australia. At the moment the hardware is slower than we like, and the software takes a bit longer to learn than we care for, but the technical costs of making a good movie are now lower than the credit card credit limit of many would-be directors. We saw that happen with the recording industry. We're seeing it happen with book publishing - the cost of self-publishing has plummeted, and distribution costs have fallen as well.

Now it's cinema's turn.

Installing Windows XP Under Windows 7

I have managed to install Windows XP on my various Windows 7 machines. It was not particularly difficult, but it was very annoying. First comes the fight to find the download page. Use Explorer (when you're about to install something in Windows, it's always best to use Explorer rather than Firefox or some other browser), and keep trying. Use Google to look for ways to install XP under Windows 7. (Explorer defaults to Bing, which also works, but oddly enough I have found Google finds Microsoft download sites better.) You'll get many pages from many places, and the ones high on the search return didn't lead me to a Microsoft download site. I expect they want to sell you something. I never bothered to find out.

There will also be a bunch of Microsoft pages on the subject, and since you want to safely download a change to Microsoft Windows you want to be sure you're dealing with a genuine Microsoft site, those are the ones to pay attention to. Most of the Microsoft sites will give you information on system and software compatibility (e.g., you need Windows 7 Professional, Ultimate, or Enterprise to install and run the Virtual XP) and much of it is worth reading. Then you will be frustrated trying to find out just where to find the programs to download: most of the Microsoft sites give you all the information except where to get what you're looking for.

Eric Pobirs tells me that "The XP Mode components are actually pretty easy to find if you call them by the right name. A search in either Google or Bing takes you almost directly there." I'm sure he's right, but that wasn't my experience, probably because I haven't mastered finding short tags for search engines. I wish you better luck. The important point is that you want to find the correct Microsoft page, and I'm just suspicious enough that I want to get there directly, not through a third party. In any event, keep trying. Eventually you'll find this Microsoft page, and what you need is all there. You fill in the information about your system, and up will come two download windows, one to bring in XP, the other to bring in the Virtual Machine. You download in that order, XP first, then the Virtual software.

Before you can actually get the software Windows Genuine Disadvantage will want to run, and it will probably want to install an add-on. The various security features built into Internet Explorer and Windows 7 will fight you on this. You have to keep on giving it permission to run things. Some of those requests will be obscure because there will be multiple windows, and the right one may not be on top. Keep trying. At some point you will get to download and install the first program. This will be more or less automatic, but stay alert for security authorization questions and agreement acceptance demands. Eventually you will be finished with installing XP. Once that one is done, you can download and install the second program which is the Virtual system program. That one goes much easier.

The system needs to restart now; after which you can go to Start, Windows Virtual PC, and that will have the XP link. You can pin this to the start menu, or the dock oops taskbar. Once that is all done, you'll have a large Windows XP Window. Open "My Computer" and you will see a C Drive, but that's not the C Drive of the parent machine. Keep exploring and you will see and access the parent system C Drive under that systems name, as you might see networked systems. To run programs you have to install them on the virtual machine. Depending on the program that may not require anything more than moving or copying it from the parent C drive to the virtual C drive. None of this is particularly complicated, and the real secret to success is to have faith and don't be intimidated. It will all work.

One more installation detail: in order to run the virtual machine, hardware assisted virtualization must be turned on in your BIOS. If it's not, you'll get a message to that effect. The remedy is to restart the system, use the appropriate key - F2 worked for me - to catch the system before it boots so that it goes into the BIOS, and enable hardware assisted virtualization in the BIOS. If you have no idea what hardware assisted virtualization authorization means - I didn't - you'll find it under the SECURITY settings, and it's easy enough to change it from <disabled> to <enabled>; at least it was on my Intel Extreme motherboard. It was already enabled on the Gigabyte motherboard in Bette, the Core 2 Quad 6600 that serves as the general purpose system here. Once you've done that, if you just exit and save and allow the system to boot, the next time you try to bring up Virtual XP you'll get the same error message, ‘hardware assisted virtualization not enabled'. Shut the system down - power off - and count to twenty. Let it reboot and this time it will work.

The Virtual XP system is a separate machine, more or less. Fool around with it and you'll learn to use it soon enough.

I Ching and other ancient programs

Windows 7 works just fine for most of what I do, and never found a real need to install Virtual XP on 32-bit Windows 7 machines, although once I found how easily it installed I did put it on a couple of those just in case I do need it. It was a different story with 64-bit systems. With 64-bit Windows 7, you can't run a number of older programs. My favorite example is Jim Baen's Electric Dragon, an ancient DOS program that uses the primitive DOS graphics symbols to build and display I Ching hexagrams, then calls up the interpretation of the primary hexagram and the one derived after the moving lines have done their thing. It's primitive but I like it better than the free on-line I Ching programs. Baen examined a number of I Ching translations before licensing the one he chose, and I find it congenial. I also find the rather crude interface a satisfying way to consult the oracle.

As to why bother with the I Ching, I have no real answer; Jim Baen and I used to discuss this fairly often. We both agreed that the ancient compilers of the I Ching oracle had an uncanny understanding of human nature. One doesn't have to believe there's anything beyond that for it to be useful. Of course many do, and there are many theories connecting the I Ching with the universe. I'll leave those speculations to others. I find it useful for focusing thoughts.

Electric Dragon runs nicely in a command window in 32-bit Windows 7, but it doesn't work in 64-bit Windows 7. It works just fine in a command window in Virtual XP under Windows 7. So do a number of old DOS programs that I couldn't get to work in Windows 7 64-bit.

Note also that a number of features in Office that worked just fine in Windows XP and even in Vista need considerable attention to get them running in Windows 7. The most notable is autocomplete, but there are others; if you haven't noticed their absence, you probably don't miss them. You can turn autocomplete on, but you have some work to do to get it right. This is the Microsoft page describing how to manage it in Office 2003. In Word 2003 if you typed in the day of the week and a comma, it offered to bring in the date as well; now it won't do that. There are other changes, all minor.

While I sort of got used to Word 2007 and now use it, I don't like some of those changes. For example, I had got used to Clippy, the crazy animated search assistant, although I suspect I am the only person alive who actually misses him. The remedy to that is to install Office 2003 under Virtual XP. It runs just fine, so now I have Clippy back if I want him, as well as some other features I missed. I have to admit that I've got used to Office 2007 and I haven't used Office 2003 as much as I thought I would, but at least I have the choice. Office 2007 and Office 2003 won't co-exist in Vista or Windows 7. They do in XP, but I have 2007 installed directly in Windows 7 and 2003 in the XP virtual machine.

Eric Pobirs notes that many of his clients miss Clippy, or his doggy or kitty avatar. It was mostly the computer press people who hated Clippy; ordinary users had a different view. Eric says "These users weren't necessarily making use of the function to solve problems but they were very attached to those little features that made the machine more comfortable." I tend to agree: I always had Clippy open with the animation turned on.

Incidentally, those who wonder about Word 2007 autocomplete and why Microsoft changed things as they did, can get the whole story here.

Another ancient DOS program I used to run in a command windows is Alien Names, a neat shareware program by Ralph Roberts that generates odd alien character names using user-selected rules. That runs fine in 32-bit Windows 7, but not in 64-bit. Installing Virtual XP fixed that, and Alien Names runs in an XP command window - but that turns out to be mostly irrelevant, because Roberts has posted Alien Names on line in a more elegant version with a better interface.

System Slowdowns I

I haven't seen them on TV, but I have heard many radio advertisements in which people talk about how their computer system has been running so slowly they contemplated buying a new machine, but then bought a product - generally a registry cleaner - that restored system speed and how their computer now is running as fast and as smoothly as it did when they first bought it.

I don't suppose I need to say it, but for the record that's bunk and you'd be silly to buy any such programs.

There is a grain of truth in it. Windows systems do get cluttered with tag ends of deleted programs. Windows Updates can get confused, or so I am told. The upshot is that Leo Laporte and others recommend that you reinstall your operating system about once a year, a sort of cybernetic spring cleaning. I suppose it is a good idea, but I haven't found that necessary, and the reinstallation of all my applications would be a pain.

Recently, though, I had no choice: Emily's hard drive was hit with some kind of worm that was vicious enough that I decided to remove the hard drive, install a new Seagate terabyte drive, and install Windows 7 from scratch. If there was performance improvement I didn't notice it. Of course Emily is a less than one year old Intel Quad Extreme system, and has lots of resources, so nothing seems to slow her down. Your mileage may differ, but in my case there wasn't much room for improvement to begin with.

If you do decide to do a prophylactic Windows reinstallation, be aware that many third parties want to sell you stuff that in theory will make it easier. I don't recommend any of them.

There is one exception. Laplink PCMover can be very useful in moving applications without having to reinstall them from scratch. Alex used PCMover to move his applications from an XP machine to a 64-bit VMWare virtual Windows 7 machine on a 17 inch MacBook Pro. It wasn't entirely smooth, but all problems were resolved.

System Slowdowns II

Systems do slow down, sometimes noticeably. It generally happens any time after the system is three years or older.

Hard drives have a known pattern of failure. (Of course this is probabilistic, but it's based on a very great deal of experience.) Either they die in their first month of use, or they work quietly and efficiently for about three years. After about three years there is an increasing probability of failure without warning. There is also an increasing probability of warning failure signs, some physical such as chattering, but also performance degradation: the drive slows down. This is often the result of a sector going bad so that the drive has to do multiple reads to get that sector's data. We used to call that "soft errors."

This can have a dramatic effect on your overall system. Most PC's already have more CPU power than they'll need, so they are IO limited anyway. Slowing disk performance slows the whole computing process. The probability of this kind of error goes up more or less linearly after three years. The function isn't linear, it's quite complex, but you don't need to know about incomplete gamma functions; the point is that failure gets more probable with time; and the system slows down as soft errors increase.

There are two ways to fix this. One is simply to replace the drive with a new one. That's a fairly simply process, even for laptops: I had no problem doing it with the ThinkPad a year or so ago, and it's even easier to do it with desktop machines. Symantec's Norton Ghost is a good tool for cloning your current system drive to a new drive; all you need is a good USB drive enclosure. I have found the "Toaster" to work very well for me: it accepts both laptop and desktop drives, and it's very easy to use. The whole process takes less than an afternoon.

The other way to fix your failing disk problem is with SpinRite. To find out how that works, go to Steve Gibson's web site. If you're not familiar with Gibson's site, you ought to be. Steve is a bit of a fanatic on Internet Security, and you can learn a lot at his place.

Spinrite will "fix" your disk by finding failing sectors and marking them as unusable, so the operating system no longer writes anything to them and never tries to read them. That can speed up your system remarkably. Do note, though, that in this age of cheap drives, you may find it simpler just to put in a new one. Gibson's SpinRite can take many hours to run to completion, generally longer than you'd need to clone the drive.

There was a time when SpinRite was a realistic alternative to installing a new disk, and I know a few experts who will say it still is, but I don't know any who actually recommend it. When hard drives get a significant number of failing sectors it's a sign that many more are very likely. It will tell you if there's nothing wrong with your drive, meaning that something else has slowed your system, and that's a good thing to know. I wouldn't recommend using SpinRite as an alternative to replacing a failing drive.

Bob Thompson says "I haven't used Spinrite in 20 years. It was actually useful back in the days of ST-506/412 drives, which could be low-level formatted by the user. The introduction of ATA drives, which are low-level formatted at the factory and can't be low-level formatted by the user, rendered Spinrite pretty much useless.

"A failing drive needs to be replaced, not "refreshed", whatever that means, and it's a truly terrible idea to write recovered data to the problem hard drive. For recovering data, I'd recommend dd_rhelp, which gets as much data off the failing drive as quickly as possible and then goes back and tries to recover data from damaged sectors."

Winding Down

The gadget of the month is the BACtrack S75 Pro breathalyzer. There's not a lot to say about this: it seems to work, it's easy to use, and if you throw parties where alcohol is served it would be a handy thing to have just to warn people. Of course I can't measure this thing's accuracy, which is why I say it "seems" to work; it has not given me any problems, it does everything it says it will, and it's easy to use. My first test said I had zero blood alcohol level, which is true. After swishing my mouth out with Scope it gave a result of over 0.3 which according to the interpretation chart included with the well written and complete accompanying manual means I am probably unconscious. A friend who'd had a glass of wine with dinner about half an hour before testing got 0.02%, which seems about right. I haven't had a chance to try it with anyone who'd been doing some hard drinking. I conclude that if you need a breathalyzer, this will do it.

We'd intended to see something else, but what we saw this month was Dear John. It's a classical romantic drama and I thought it was well done; we enjoyed it. Some of the plot is a bit contrived in that I would have thought the girl intelligent enough to achieve her results in a different way, but if she had there wouldn't have been a story to tell. The other parts of the story were well thought out and well presented. As to performances, Amanda Seyfried is a charming ingénue, a rich young girl who isn't like all her contemporaries; while Channing Tatum is a believable taciturn Green Beret. Both Seyfried and Tatum were models before they turned to acting, so you expect them to look good, and they do. Not a terribly serious movie, but well worth the price to see it at ArcLite.

The light reading of the month is Sarah Hoyt's Darkship Thieves from Baen Books. Hoyt has a knack for story telling, and the ability to blend elements of science fiction and fantasy into believable worlds. Darkship Thieves is the story of Athena Hera Sinistra, a princess of an Earth that has continued to evolve and build structure until there is very little freedom left in human life, who meets a hero from a society that might have been developed in consultation with Ayn Rand. Both societies have developed space travel, but this isn't a technological novel; it's a classic adventure romance, and very readable.

The book of the month is Selling Peace: Inside the Soviet Conspiracy that Transformed the U.S. Space Program by Jeffrey Manber. Manber was CEO of MirCorp, the company that leased the Russian Mir space station and sold tickets, and was thus in a position to know what was going on during the first attempts to commercialize space access. His story of Dennis Tito's adventure rings true enough so far as I can tell. I knew Tito before his space adventure, back when the LA Opera League used to hang out in his hill-top mansion complex (well, we had about five major fund raising parties there, anyway). Manber's prose isn't as exciting as it could be, which is probably just as well, and he tends to give more details than I really wanted to know, but this is a very interesting insider's account of an important period in space commercial development.

The computer book of the month is Real-Time Java Programming with Java RTS by Eric Bruno and Greg Bollella, Prentice Hall. Like many of the books I recommend, this is a combination of introduction and handbook. It's not an introduction to Java, and unless you're already familiar with standard Java programming it's not likely to be of much use; but if you do Java programming and want to learn about Sun Microsystems Java Real-Time System, this will be a good start. Java RTS is used for, among other things, writing games, including games that can be run as apps in small systems. Java has been around for a long time. For an example of what was done with early versions of Java RTS, see YARTS. The program is a great deal more sophisticated now, and there are many phone apps written in Java RTS.

Linux enthusiasts will enjoy the second edition to Mark Sobell's Practical Guide to Linux, Editors, and Shell Programming, also from Prentice Hall. This is a classic handbook with a thousand pages of instructions, examples, lessons, tutorials, and exercises on how to use Linux and the Mac OS X command line terminal including the Mac OS X programming tools. Many consider this the best single volume work on the subject, and working one's way through this book is a good way to go from beginner to near journeyman level in working with Linux.

Practical Digital Photomicrography by Dr. Brian Matsumoto is a good practical introductory handbook that will be of use to biologists and others who need to do microphotography in their research and who don't have the funds to hire an expert. It discusses consumer point and shoot cameras as well as Digital SLR's and dedicated Photomicrography cameras; mounts and accessories; and of course techniques. It also discusses microscopes including older microscopes suitable for digital photomicrography, and the attachment hardware you will need. The book assumes that you know what you want to photograph and concentrates on how to do that with best results.

The Art of Scalability: Scalable Web Architecture, Processes, and Organizations for the Modern Enterprise by Martin L. Abbott and Michael T. Fisher comes with a lot of high praise from an impressive list of managers and business experts. Abbott and Fisher were at PayPal/eBay at a critical time, and are well known consultants. This book is for IT managers, and one would more likely encounter it as a text in an MBA program than in a technical IT course. It's not a book for coders. It is a book for those who want to manage code shops. I would recommend it for code experts who hope to go into management; it would also be appropriate for managers who don't do code, as it does go into some technical introductions. There's a lot of meat in this book.