Dr. Jerry Pournelle

Email Me


Why not subscribe now?

Chaos Manor Subscribe Now



Useful Link(s)...

JerryPournelle.com


Hosting by
  Bluehost


Powered by Apache

Computing At Chaos Manor:
The Mailbag

Jerry Pournelle jerryp@jerrypournelle.com
www.jerrypournelle.com
Copyright 2009 Jerry E. Pournelle, Ph.D.

May 19, 2009

ESCAPE FROM HELL, by Larry Niven and Jerry Pournelle, is now shipping. This is the story of a science fiction writer who finds himself in Dante's Inferno. It might be subtitled Vatican II meets Dante. A number of readers have reported favorably...

The audio book versions of INFERNO, as well as ESCAPE FROM HELL, both read by Tom Weiner, are available from Blackstone. You can find a review of the audio version of Inferno together with a map I drew for the first edition of Inferno at this link. They're good readings, unabridged.


In your Chaos Manor column here, you said:

Open Source attempts to allow free access to various programs - such as PHP and MySQL - including the source code, and impose on those who take advantage of this the same conditions: that is, if you use MySQL and in doing it make useful modifications to the code, you can't copyright your improvements.

MySQL is distributed under a dual license. You can either use the GNU Public License, or use their commercial license if you find the GPL's requirements unacceptable. The GPL does not preclude one from copyrighting improvements to code; in fact it depends on that copyright. The GPL is designed to work with copyright law. The Free Software Foundation's counsel has a simple legal argument for violators: "You are copying my client's copyrighted work without his permission. Stop it, or I will get a judge to make you stop." Because the GPL is the license that makes the copying legal, someone who copies in violation of the GPL is guilty of copyright infringement.

What the GPL (and other licenses subsumed under the Open Source Definition) requires is that when you make those modifications, and convey the modified program to someone else, you must give them the same rights you had in the original program. When you release your modified code under the GPL or a compatible license, you have satisfied that requirement. Now your copyright on the modifications joins the prior authors' copyright in their code and makes you all potential plaintiffs if a downstream developer violates the license under which you all distributed your work.

--
Monty Harder

Thanks. My legal friends seem to disagree on just how effective the Open Source agreements are at enforcing what they intend, but I have neither an opinion on the matter or the right to one. Thanks again.


Chaos Manor Reviews Column: iWork '09

Hi Jerry,

I confess to curiosity: as an owner of iWork '08, I didn't see a single reason to "upgrade" to iWork '09 based on reading Apple's release information, especially as there is no "upgrade" price; you just buy a new copy at full retail.

Granted I'm a minimalist user of any word processing software: the most complex document I maintain at the moment is my CV, and for MS Office documents that iWork has trouble with I can always try using Open Office. So far, I haven't received any document that I can't read, one way or another.

Best regards,

Giles

Thanks. On reflection, I have to say I have not had any real reason to upgrade, and I probably said that you probably should upgrade largely as a reflex. It's not terribly expensive, and it does encourage Apple to continue working on the program. As I understand it, there has been progress on compatibility with other suites, and that may be worth something.

I have used iWork, but I confess that I tend to use Word 2o08 when I work with the MacBook Air. Actually I tend to do much of my actual writing in Office 2007 with a Windows operating system, and just about all of my fiction in Word 2003. They're both Good Enough. I have been slow about changing over to Mac from Windows, and as I have said often, I doubt it much matters now. Both Mac OS X and Windows-Vista operating systems and applications have for years been better than Good Enough for almost everything I do. Macs do try to work with you a bit more than Windows systems, and Macs do seem to be more intuitive with audio and visual editing - but I don't do a lot of either of those, and Windows has been good enough for what I do...


Windows 7 release Candidate

Jerry,

I downloaded the 64 bit version of Windows 7 RC and installed it. Here are a couple of gotchas.

The system I used for the install has a diskette drive. Before the first Install Windows Dialog box would appear I had to put a diskette in the diskette drive. It didn't matter what diskette it was I just had to insert one.

The system has a D-Link DGE-530T gigabit Ethernet adapter. The RC has no driver for this. Had to boot back to XP and download the Vista 64 bit driver. I hope it works. I haven't booted back to Win 7 to install it.

First impressions, a very pretty face, much like Vista, However, there is additional obfuscation over and above what Vista added over XP. I took a casual look to find where to change screen resolution under personalization and didn't find it where expected.

Thank goodness, Microsoft is not too big to fail.

Bob Holmes

Most of the reports I have on Windows 7 have been highly favorable, and I am getting more and more tempted to try it; and having said that I continue to use XP in VMware on the iMac and Vista 64 one my major communications machines. Once a system has enough memory and disk storage and processor speed, the OS doesn't seem to matter so much. Of course I do have the usual 15-second hangup when Outlook goes out to get the mail even on a 64-bit Intel Core 2 Quad system with 8 gigabytes of memory. That happens infrequently enough that I just put up with it; the remedy is always just to wait.


Wolfram Alpha:

Wolfram Alpha Gets Mixed Reviews

So-called computational-knowledge engine Wolfram Alpha went live Friday after weeks of speculation that it would revolutionize search or even challenge Google.

See Full Post: Wall Street Journal blog link.

Rick Hellewell

Dan Spisak says

Okay guys, this is some seriously amazing stuff. Check it out:

http://www.wolframalpha.com/screencast/introducingwolframalpha.html

It's worth the 13 minutes.

-Dan S.

Probably the simplest comment is from Eric Pobirs:

Wow, it's like an Artificial ADD system. A curious person goes in and is never heard from again.

Eric

We have only begun to explore the potential of Wolfram Alpha. I would not be astonished to see it surpass Google, at least among sophisticated users. There is enormous potential.


Vinge on what if the Singularity doesn't happen?

http://www.kurzweilai.net/meme/frame.html?main=/articles/art0696.html

Frankly, while I have been fascinated by his books, I found this article very difficult. I'm not sure I understand him. Perhaps your readers would enjoy your explanation.

My best,

Fred Kruger

"The Singularity" is brought about when we invent truly intelligent Artificial Intelligence systems. There are those who think Wolfram Alpha a significant step toward "the singularity."

Vernor is always worth paying attention to, but I have long had my doubts about discontinuities in human history. They do happen, but they are rare.

My view of what happens if we don't go through a singularity was given in A Step Farther Out, particularly the section called Survival With Style, and depicted in fiction in Exile - and Glory! and in the novel I did with Charles Sheffield, Higher Education. I may yet do another novel that combines the Higher Education and Exile--and Glory! worlds.

Singularity believers have said that it's no good planning for the world after the Singularity because we can't possibly understand those conditions. Waiting for the Singularity seems to have some similarities to those who await the End of Days and the Rapture of the Saints. I think we can plan for the future, and we'd better do it.

I can imagine an Artificial Intelligence that is arrogantly convinced that it knows The Truth. We seem to have managed that with natural intelligence, and the results have not always been good...


Dr. Pournelle,

A hardware review:

Last week I rented a Nikon CoolScan 5000 (link) from a local camera store. Renting (including the slide feeder) cost $150, vs $1200 to buy. The slide feeder can take up to 50 slides at once. Load it up, select the number to scan, tell it to start, walk away for an hour or so. Takes about a minute/slide. Nikon has a system called DigitalICE which automagically removes scratches, dust, and dirt, if they aren't too big. I blew dust off the slides, by blowing on them, before I loaded them. There are all sorts of settings which I didn't mess with as the defaults worked fine. The primary ones I played with were bit depth and film type. A bit depth of 8 bits/color yields very good 6 MB jpegs. 16 bits/color yields 32 MB tiffs. I went with jpeg which is certainly Good Enough. Film types are Negative, Positive, and Kodachrome. Kodachrome has its own setting because the dye layers were very different and therefore have to be scanned differently. Scanned 600 slides over the course of a week.

It works. Very well. Here's an Ektachrome slide from 1983: Flickr link

And an Ektachrome infrared: another Flickr link

Highly recommended if you have lots of slides you want to digitize.

And you may want to. Over the years I used mostly Ektachrome and Kodachrome, but I also shot several rolls in high school of Seattle film Works film. That was a bad idea. Those slides have faded very badly. The Kodak slide films held up beautifully, the SFW ones did not. A few more years and those would've been gone forever. Dad has Kodachrome from 1950 that looks like it was shot yesterday. So if you have old slides, check to see if they need to be digitized. Ones shot on Kodak films and properly cared for are probably OK.

Kit Case

Thanks. I have a stack of slides that need converting, and I never seem to get a round tuit, which is unfortunate. Some of those slides are irreplaceable, so I have hesitated to hire this out.

Alas I don't have a week to put into this just now.


Mail Bag

Jerry,

You replied to someone with the statement "The intention of Wikipedia was to allow correction on the theory that truth will out; but the Global Warming controversy is a counter example. Once the 'consensus' is established, then 'denial' becomes nearly criminal."

Well yeah, but I've found that to be true no matter where or what human beings are debating. Take String Theory for example. No one in physics dares to say it's wrong. And the same is true for any topic in any given group of people. Once a consensus is reached you don't dare go against it. So much for the scientific principle.

Dean

Now imagine an artificial intelligence that knows it is right...


Subject: misguided

You wrote:

"Textbook prices won't come down until the entire education package - tuition, certification, instruction, textbooks - becomes competitive. The means for doing that already exist - the Internet is certainly good enough, and some of the greatest lectures ever given are available on line for free. The problems of distance learning are not primarily technical. There is an educationist oligopoly that controls "accreditation" and other gate-keeping functions, and it works hard to eliminate competition by the usual means, including government monopolies.

Eventually this monopoly will be broken; we see more and more alternative education institutions every year."

Being lectured at does not constitute learning. I've been mulling this over with respect to the physics textbook discussion on the mailbag.

First of all, government does not have a monopoly. This country is filled with private schools, from K through graduate school, and they don't really do any better or cheaper.

Second, to have an informed opinion, one needs to be informed. In this case, informed about the mechanisms of learning and memory. This is one of my areas of research. There is, for example, a reason why it took several thousand years to arrive at Newton's laws. There are hardwired features in the human brain that help us with seat of the pants reasoning in a complex world, but that get in our way when trying to learn physics. There are some fiendishly clever experiments on month old babies that reveal rudimentary understandings of contact interactions, gravity and force and motion relationships. None of that could have come from experience since they have had no experience.

Now there are equally clever ways of enabling people to effectively rewire their brains. But to my knowledge, all of these depend heavily on dialog and interaction, with teachers, with other learners, and most importantly with phenomena. Much of this is difficult to impossible to deliver in a distance learning environment, at least in my bailiwick.

You could do a lot worse than to read this, from the National Academies of Science:

http://www.nap.edu/openbook.php?record_id=6160

Free and all.

If you have more than a passing interest in this issue, I can point you toward some other scientific research on the subject.

--
Dr. Paul J. Camp
Spelman College
Department of Physics

"The beauty of the cosmos derives not only from unity in variety but also from variety in unity"
-- Umberto Eco
The Name of the Rose

I replied:

Not sure who you are writing to. Or what you think I believe, but I did not think I had implied that textbooks were lectures. There are two education matters here: textbooks and their prices, and education itself. Both are overpriced.

As to lectures not being education, most of the universities certainly seem to assume they are. And they assign people to give the lectures who aren't very good at it. Whereas George Mosse at the University of Iowa lectured to the entire freshman class -- required course -- on Western Civilization and he certainly got a lot of kids interested in it; and more to learn something whether they wanted to or not. But I agree, most lectures are not educational. Fortunately a few are. The Internet can make those available to everyone.

But mostly I see no reason why there ought not be national certification in the manner of CPA: it doesn't matter what education you have there is the Board Exam. We could do that for a lot of subjects. Then those who want expensive 4 year liberal arts can have them. Those who don't can try another route.

As to whether private and home schools don't do better, we certainly don't act as if we believed that. Try giving a kid at UCLA a scholarship to Harvard or Yale...

I am glad we know the secrets of teaching, but I fear they aren't very widely applied.

That's true. I wouldn't argue with you about that. The reward structure in higher education is all screwed up, with insignificant dependence on provably effective pedagogy. I recall a story told at a meeting by an award winning lecturer (by Eric Mazur of Harvard, if memory serves, but it may not) who got a lot of kids interested in science and physics in particular. But then he started pre and post testing with the Force Concept Inventory and found that all these excited kids who "really learned a lot" in his class actually exited just as pre-Newtonian as they had come in. Mazur cared. Most faculty don't.

The trouble is that it isn't really a case of monopoly power that can be broken through technology, and the internet in particular (which was the specific point in your column I was taking issue with). A large part of the problem is applying research on learning and memory more broadly in classrooms. But I've come to believe, after 20 years of trying to do exactly that in my classrooms, that an equally large problem is that education is essentially a sort of handicraft industry. This means that productivity improvements are difficult to achieve. Output is limited not so much by the ability of the educational machine to deliver knowledge, but by the intrinsic limitations of the human mind to absorb it and integrate it with existing knowledge. That process can be helped along but it basically doesn't have an accelerator. Where a traditional classroom environment fails is in the assumption that this process is basically the student's problem, not the instructor's.

I'm skeptical whether national board certification would be a cure all. To some extent, it already exists. You can get discipline specific national certification for K-12 education. But research indicates that nationally certified teachers are not significantly better than those who are not certified. And universities are primarily interested in research so certification is a nonstarter there. So I suspect making that sort of certification more broadly required would have a limited effect. It certainly hasn't kept ambulance chasers out of the legal business (or ambulance fillers out of the medical business). Or crooks out of the CPA business.

I've generally been pretty appalled by the output of places like Harvard and Yale. They seem to be more about getting plugged into the network rather than actually learning something. The aforementioned Eric Mazur has to teach physics without labs. Harvard doesn't allow them. You should look up a video from the 1980's by Phil Sadler of the Harvard Smithsonian Center for Astrophysics. He interviewed graduating seniors at Harvard commencement asking them very simple science questions -- why is it hot in Summer? What causes the phases of the Moon? Here's a battery, a bulb and a wire -- make the bulb light. You can find it at Annenberg's site:

http://www.learner.org/resources/series28.html?pop=yes&vodid=199341&pid=9

You'll have to sign up to watch it (free) and it is about 20 minutes long. We may think we want to go to Harvard for the education, but what we are really thinking of is the opportunities it creates, and those come mostly from the social network you get slotted into.

Textbooks and lectures suffer from some similar issues but lectures have been more extensively investigated. Lectures are not intrinsically bad, but they have a serious timing issue associated with them. There is a time for telling, and that time is when a need to know has been generated. A freshman's ability to generate that need is much more rudimentary than, say, a postdoc's, but I wonder how much of that is due to greater learning, greater ability, or simply filtering out people who couldn't.

In any case, learning from textbooks is more mysterious. I actually submitted a research proposal to NSF to investigate this very thing (turned down, but I think I can clean it up). The problem is that we assumed kids learned how to read by fifth grade and this is simply more of the same. It isn't. It is a complex skill that involves the coordination of multiple subskills -- lexical access, proposition assembly, contextual evaluation, integration of prior knowledge, and so on. We teach this using narrative materials, and then expect the same skills will carry over to scientific written materials. A little thought should reveal how crazy that is. Lexical access and proposition assembly are seriously complicated by the fact that the lexicon is not just words and the propositions being assembled are not just sentence clauses, but rather are idiosyncratic amalgams of words, pictures, graphs, numerical data and equations. And the process is significantly complicated by the fact that there are gaps that have to be recognized and filled in. This is obviously true in journal articles, which refer to other articles rather than reproduce their details, but it is also true in textbooks where the process of getting from equation 5.10 to 5.11 may be described (or not) rather than worked out in detail.

Reading a textbook is a learned skill (and what I proposed to the NSF was to study its development) and when I realized that I looked by at my old textbooks and notes on journal articles. I could see that process unfolding in the highlighting and other artifacts I left behind.

Ok, I'm getting longwinded so I'll shut up now. If I've misunderstood or mischaracterized your position, I apologize. My overall point is that the problems of distance learning ARE primarily technical, but not in the sense of delivery mechanisms. The technical issues have to do with cognition and memory. Translating didacticism to an electronic environment will certainly do no worse than was done to those graduating seniors at Harvard, but it will do no better either and it will be missing the network opportunities.

I agree that textbook prices are scandalous, and I have a long story about that too (I saw it happen from the inside) but I'll spare you.

I think we are not discussing the same subject at all. You cannot get education credentials without taking education courses. It is not possible merely to demonstrate that you know what you are doing, or take an examination. You have to subject yourself to the education establishment to get "credentials". Ask any retired military instructor about what you have to go through to become 'credentialed', even though they may have been teaching mathematics for twenty years.

If we know how to teach then we ought to allow those who can teach to do it; which is to say we need to have a different bureaucracy certifying whether one has learned anything from the bureaucracy that teaches how to teach.

I have never thought there is a simple solution to the situation. For one, a great number of people need to be taught skills, not educated.

As to the success of distance learning, I am following several experiments; in all the cases my readers seem to be learning calculus and engineering, and doing so much more cheaply through Internet classes. We will have to see, but I would be willing to bet one can learn more from a good Internet lecture than from a non-English speaking graduate assistant at UCLA - and it will cost a great deal less, too.

Of course the social networking at Harvard is good. Hotel management is taught very well at Cornell. But graduation from a major university is not really a certification nor does it say what you actually learned (given grade inflation everywhere how could it?) Independent certification - the way CPA exams are handled, as an example -- is not the answer to all problems, but it does give potential employers something other than the name of the institution to go on in looking over job candidates. Cornell has a great course. Waybelow Normal hasn't much of a reputation. But if we have two candidates, one from Cornell in the 10th percentile on a national certification examination, and one from Waybelow Normal in the 95th percentile, we at least have something to decide.

I have never suggested that I knew of a cure all. Indeed, I don't think in that kind of terms at all. But we spend a lot on what we call education and we do not seem to be getting our money's worth and I suspect that making educators compete for students will help. Making those who teach the arbiters of their effectiveness seems bull headed. Have someone else certify whether the kids learned or not. Like the old Regents exams in New York which were damned effective and thus emasculated.

I agree that no one should evaluate themselves. I'm skeptical that market forces will fix things. The reason is that markets sell product that is in demand, and in education the product in demand is not so much knowledge as it is credentials and resume items. Simple example: almost every high school, public or private, offers AP courses for resume enhancement. Frankly, I'd rather my students have no physics at all than come into my class with AP credit. I have hard data to show that it didn't do any good and my general impression from dialog with them is that they have learned all sorts of things that aren't true and that actively interfere with what I'm trying to accomplish.

Anyway, the only reason I responded, since you're going to do something with this discussion, is to point out that we do have external evaluative exams mandated under No Child Left Behind. They don't work.

There are a variety of reasons for that. Under the banner of local control, there are no uniform standards. Underperforming schools are defunded, making their job harder, instead of being restaffed.

But it is also very tricky to determine what you really want to know in the format of a standardized exam. It can be done -- the Force and Motion Conceptual Evaluation is a good example. But the FMCE is based on years of research into the development of Newtonian understanding and the identification of consistent intermediate states of knowledge -- you can use it to measure not only what they understand but how close they are to a physicist's understanding. Most such exams are not. This is the reason my classes are driven by anchoring experiences and Socratic dialog thereon, and my exams are heavy on qualitative and semiquantitative explanation. I want to understand the way they model real phenomena. Informative as it may be, that sort of evaluation would be difficult to deploy on a larger scale for obvious reasons. So we give instead the test that can be economically administered and hope that it tells us something.

I fear we continue not so much to disagree as to talk at cross purposes. I have no brief for No Child Left Behind. I do think an achievements examination to show that a high school diploma has value might be useful.

No Child Left Behind assumes that everyone will get some value from what Bill Gates calls a world class university prep education. Of course this not being Lake Wobegon, everyone will not benefit from a world class university prep education; but worse, examinations in algebra and high-g loaded subjects are given to everyone, while effectiveness is judged by the median class score. The way to raise the median class score is for the teacher to spend a lot of time with the marginal students, getting them up a few points to the detriment of those who could have used the attention to learn a lot more. This often drives the bright kids crazy.

What I have contended is that a system of qualification/certification independent of where the student went to school or what courses were taken would reward those institutions that actually teach the students, and while it's not a panacea, it seems preferable to having the professors of education decide who has been educated. That can apply both to college level and high school level skills.

There are two parts to the education dilemma in the US. One is education of the kind of students you teach. As you observe, the high schools don't seem to be doing a great job of preparing them, and you have to do their work for them. Thanks for doing that. The other part of the dilemma is teaching civilization and work skills to the students you will never see (and wouldn't keep long in any event). Half the students in the nation are below average. Everyone knows this. No one seems to understand the implications.

If there are ways to see that No Child is Left Behind, and it only takes money and better techniques to accomplish that, then the Congress ought to do it in the District of Columbia. If they are successful there, everyone will want to adopt the techniques.


We had another comment on the subject:

Dear Jerry,

I very much agree with you about the costs and other limits of contemporary higher education. And I agree that "The problems of distance learning are not primarily technical." However, the "accreditation" bodies are getting better with many working with the individual institutions to define KPIs and other quality measures - but we are still a long way from optimal. I do want to add that another major obstacle to distance ed is the faculty - many of who are unwilling to look at new Andragogy. As an example I recently offered 3 faculty in my college a $3K stipend and lots of technical and professional support to spent a few weeks over the summer to convert face-to-face courses to technology enhanced courses. They each said that an on-line format would not allow students to interact in desired ways and so turned the offer down. When I pointed out that students have lead the way in technology enhanced personal interactions they all just continued with their contentions that on-line does not offer the needed student-to-student interaction. I pointed out that I have been doing precisely this with my class group work - and they - well you get the idea.

It has been my experience that technology is rarely the bottleneck - I can imagine those old monastic students complaining about what is lost by this new-fangled machine of Guttenberg.

Be well and thanks for all you do.

Aloha

Charlie


Computing At Chaos Manor, copyrights

Jerry,

I enjoyed this month's article.

You mentioned that you had stories in both Galaxy and Analog. I subscribed to Analog for a good while, but I wasn't aware you had stories in those magazines. I would like to know how I could read your stories published in those magazines, if its possible. Also, do authors still own the copyright to those stories, so that they might be re-published elsewhere?

Google is now publishing some older magazines, which I think is a good idea. Its not likely they will get much income from old magazines, they aren't even getting much from new ones.

Regards,

Randy Lea

Most of my short stories from Analog were written in two thematic series, the John Christian Falkenberg/CoDominium stories, and the Aeneas MacKenzie series. The CoDominium stories including several novels were collected into the large work The Prince from Baen Books. Some of my Galaxy columns were collected into A Step Farther Out, which was in print for about 20 years, and which may be brought out again along with a second volume to include the other Galaxy columns; alas, those non-fiction works, which were largely about a space-faring future, are still fairly relevant and current. The Aeneas MacKenzie stories, which grew out of A Step Farther Out, have been incorporated into Exile - and Glory! from Baen books. Fires of Freedom, also from Baen, includes King David's Spaceship, which was serialized in Analog.


E-book experience enhancement

In all the discussions in mail on reading e-books I notice that no one has brought up the fact that the e-book reader misses out on the new book smell. This deficiency has been rectified by the people at DuroSport Electronics who now provide aerosols with various book smells ranging from "New Book Smell" to "Eau You Have Cats". Unfortunately my CyBook Gen 2 e-book reader is not supported so I can't report how well these oders emulate the "real" book experience.

"The smell of e-books just got better

Does your Kindle leave you feeling like there's something missing from your reading experience?

Have you been avoiding e-books because they just don't smell right?

If you've been hesitant to jump on the e-book bandwagon, you're not alone. Book lovers everywhere have resisted digital books because they still don't compare to the experience of reading a good old fashioned paper book.

But all of that is changing thanks to Smell of Books(TM), a revolutionary new aerosol e-book enhancer.

Now you can finally enjoy reading e-books without giving up the smell you love so much. With Smell of Books(TM) you can have the best of both worlds, the convenience of an e-book and the smell of your favorite paper book.

Smell of Books(TM) is compatible with a wide range of e-reading devices and e-book formats and is 100% DRM-compatible. Whether you read your e-books on a Kindle or an iPhone using Stanza, Smell of Books(TM) will bring back that real book smell you miss so much."

And:

"Warnings and Legal Disclaimers

Please use in well ventilated area. May cause dizziness and hallucinations. May cause itching and runny nose. If symptoms persist for longer than eight weeks please consult your physician. Not for use on "real" books. Do not use while riding public transportation. Discard empty container with hazardous waste. Not for use as a room deodorizer. Not for use on burning books. Do not use on a Zune."

http://smellofbooks.com/

Tim Boettcher

Uh - OK. I seem to have imported a somewhat different odeur...