Population, Programming Languages, Global Warming, Planet Defense, and other matters of gravity.

Mail 824 Sunday, May 11, 2014


You were ahead of your time, Jerry –

I recall that you once backed an idea for sending water back up into the mountains.


Of course, they’re not using recycled water, and it can’t work in both directions at the same time except incidentally. It is still subject to state approval, so they may manage to kill the idea yet.

–Gary P

It has long been clear to me that LA needs to reprocess its sewage and runoff water – it does that very well now – and pump it up into the Angeles Crest to runs down refilling the water table as it goes.


: Useless population

Hi Dr. Pournelle,

I’m not convinced that you have the right of the argument when you say that half the population will inevitably be economically useless. Perhaps there will be sufficient government incentives and disincentives to encourage indolence in half the population, but that’s not the same thing. As a thought experiment, think about the U.S. agricultural sector that you use as an example of the phenomenon.

It’s true that farm labor is a very small fraction of what it was at the beginning of the 20th century. But that reduction is a result of increased productivity through the employment of many technologies, from tractors to vacuum-packed breakfast cereal. Perhaps it would be more reasonable to add back in the employment at firms ranging from John Deere, to General Mills, to Frigidaire, to Gunderson rail cars and manufacturers of shipping containers, to financiers who lubricate the process of putting the capital investment in place. All of a sudden, there’s a lot more people employed in bringing you your Wheaties and orange juice.

Now, it’s true that those firms don’t solely service the agricultural sector—their profits are driven in large part by the rest of the industrial (and other) sectors. That is to say, the productivity of today’s agricultural sector is largely a by-product of the industrial sector. It seems to me that it’s not so much of a reach to say that productivity increases in the manufacturing sector will continue to be driven by innovations (and employment) in the technology sector, biotech sector, transportation, energy, etc. Because it takes time and capital to automate any task, and it always seems that tasks pop up faster than we can get together the time and capital, it seems to me that there will be jobs aplenty for those with a good work ethic and enough intelligence to either mop a floor or to tidy a room in preparation for the floor-mopping robot.

Think of it another way—at one time, one-third of the population was employed in feeding the other two-thirds. Later on, 1/3 was employed providing rail transportation to the other 2/3. Later on, (if memory serves) 1/3 was employed building cars, car parts, or roads for the other 2/3. Perhaps someday 1/3 will be employed capturing energy for the other 2/3.

It’s true that my optimistic take pre-supposes an educational system that equips all but the developmentally disabled with a good work ethic and the ability to do basic arithmetic and to read and comprehend instructions. But that gets back to the government incentives I mentioned. (As an aside, at one time I contracted with a foundation that employed the developmentally disabled to do part-kits for a product, so I don’t think even that is a real disqualification.)

The one fly in the ointment that I see is something that I left out of my “1/3” progression. The employment picture has gotten muddier lately, but out of the non-agriculture, non goods-producing population, one-third is employed either by government or by health and social services (which is difficult to separate from government in the official statistics). Perhaps this is an aberration, and not a harbinger, but I think it’s the source of much of our current ennui.

http://www.bls.gov/emp/ep_table_201.htm <http://www.bls.gov/emp/ep_table_201.htm>


The key to your statement is “good work ethics.” Given present trends, I’d say we were moving more to “sensitivity to entitlements”. I am not sure where good work ethics will be learned. I suspect that given incentives, freedom, and decent work ethics we would have a brilliant renaissance, but I fear that is more a hope than a prediction.

Understand that my definition of useless is an economic term, not a moral judgment. Those who produce less than they consume over their lifetimes generally have contributed little to the civilization.  Of course ‘contribute’ is subject to discussion. I sing for my supper, but I don’t produce very much in the usual sense.



Hi Jerry

You wrote that “The Department of Defense tried to get into the act with its invention of ADA but like all projects operated by committees, it grew and added features and never quite got there.”

ADA itself may not be much used but a lot of it lives on in Oracle’s PL/SQL language: http://www.dba-oracle.com/concepts/programming_pl_sql.htm

“Oracle PL/SQL was based on the ADA programming language which was developed by the Department of Defense to be used on mission critical systems. Although not a ‘sexy’ language like Java or C, ADA is still being develop and used for applications such as aircraft control systems. ADA is a highly structured, strongly typed programming language that uses natural language constructs to make it easy to understand. The PL/SQL language inherited these attributes making PL/SQL easier to read and maintain than more cryptic languages such as C.”

As a PL/SQL Software Engineer myself over the last few years I have written tens of thousands of lines of PL/SQL and I can confirm that it is widely used around the world, by most of the organizations that use Oracle databases, and is a living language that is actively updated by Oracle with new features released most years.

Having, in my time worked with Assembly language, COBOL, C, Forth, PL/SQL and other languages, in my opinion PL/SQL is one of the better development languages. Unfortunately because it’s only supplied as part of Oracle’s database offering, with very little support for using it for anything other than building and executing SQL queries it’s never going to be the language of choice for general software development.

Best wishes

Paul Dove

I was very hopeful about ADA and hoped that it would become wildly popular, but I fear the committee nature of its design, no matter how well meant, doomed it.


Compilers for number crunching

Hi Jerry.

Coincidentally, I stumbled across this article today just after the latest round of discussions on compilers got started on Chaos Manor. It focuses on FORTRAN and some up-and-coming computer languages:


From personal experience, FORTRAN is still the dominant language in both oceanography/atmospheric science and astrophysics…


Mike Casey

When I was keynote speaker to the Grand Challenges in Supercomputing Conference some decades ago, I took the opportunity to question people who used supercomputers on just what they did. I generally got the answer, “I write 120,000 lines of FORTRAN and try not to go mad.” Apparently that’s still a fair description of what some Supercomputer people do to this day.

FORTRAN can be a confusing language. It will compile nonsense including type changes and confusing data with program instructions, and allows a number of coding tricks that save lines of code and memory at the expense of understanding, but there are also programs like RATFOR (RATional FORtran) and its descendants and improvements: these are precompilers that enforce strong data typing and a degree of structuring forcing the programmer to think about the logic of the program before handing the whole mass to the compiler. I have written some FORTRAN programs including an expected value model of a nuclear exchange of ballistic missiles, and it is powerful. I’d still rather use C-BASIC for the kinds of work I tend to do with computers, but I have to confess that I’m likely to use Python if I just need something quick and dirty. But then I try to avoid programming when possible and lately I have been very successful at doing that.

I suspect that for really complex systems like climate and complex flows, FORTRAN is the weapon of choice to this day. I know that when I worked with nuclear weapon designers, most of their work was done in FORTRAN.

Subject: Cutting-edge research still universally involves Fortran; a trio of challengers wants in.



An interesting article.

Fr. N.


Coding for dollars

Dr. Pournelle,

I enjoy your occasional discussions on software coding, and agree with many of your points — but at the cost of possibly being boringly repetitive, I must stick with the position that the language is possibly the least important factor of the components of generating good software. Seriously, it is the systematic approach (or lack thereof) by the coder(s) and their organization. In my recently departed systems engineering career, I’ve witnessed lousy software written in low-level languages (or machine code) by coders with the reputation for a high degree of skill, and excellent code written in loosely typed, interpreted languages by relatively inexperienced programmers. The differences distinguishing the two extremes have always been the developer’s understanding of requirements (including security requirements), the degree of code review, integration testing, and configuration management.

Languages do make a difference in the sheer amount of good, verified code that can be generated by a given group with a given skill set. Strongly typed languages have all the advantages you list, including the ease with which the code is inspected and debugged, as well as the ease with which the project is staffed. However, if testing is given short shrift a good language is as likely IMO to produce poor results as a project run using machine code.

There is some excellent and useful software written, by a single person, much of it in assembly language, in Spinrite by Steve Gibson. I became aware of him and the software indirectly via a reference from you to TWiT. His attention to detail and commitment to thorough testing by a team set his work apart from others. He posts a lot of freeware and security utilities, along with a lot of other good information on his site at grc.com.

I do enjoy the discussion, and look forward to your comments and those of your other readers. I also am continuing to enjoy reading of your recovery and increased activity.


I don’t disagree with what you say. What Niklaus Wirth has spent a lifetime trying to accomplish is to get programmers to think a lot more about what they are trying to accomplish before they begin to write code, and to design languages that require you to do that.

Wirth’s view of ADA, incidentally, was that was based on bad principles from the beginning because it had code Exception operations (so that if you coded yourself deep into a hole you could get out by declaring and exception). “They don’t know how to program if they need those,” he told me once, I think over sandwiches at my kitchen table. I had brought him and his wife to the house when he was in town for a conference, and I recall showing him the DOS game Wing Commander: he was greatly impressed with it (as was I).


re Substandard Programming Practices and Their Effect on Our Daily Lives and the Catastrophe Waiting Just Around the Corner:

I think it is quite unfair of your correspondent to dismiss the complexity problem as nonsense. The complexity of systems has indeed become enormous.

That is no excuse for poor design and implementation, but even with good and thorough design and high-quality implementation, what we expect from a system these days is far, far more than what it used to be, and hence a huge amount of added complexity.

I know whereof I speak. I have been a software engineer for 42 years now, and have worked for Symantec and Google as well as for several not so well-known companies.

It is easy to say "poor choice of programming tools" but in reality there is little choice available – unless you want to re-educate your workforce, you use the tools they know.

There are, of course, plenty of examples of the problems Mr. Holmes cites, but there have always been inferior products in the marketplace, and a marketplace that changes as rapidly as computing has more than its fair share.

But it is not universally so, and the discipline of software engineering is advancing (albeit slowly). The most hopeful signs I see are a greatly increased use of unit tests and Test-Driven Development (write the tests *first*, then test the code as you write).

To me, the design and test aspects are more important than the implementation language, but we are not yet completely done with the language wars. The most interesting new language (for me anyway) is the "go" language. At least one of its designers was a student of Wirth’s. The language itself seems to address my biggest concern with existing languages. (Their complexity! The C++ book from its principle designer is huge, and very difficult for mere mortals to fully understand.) Hopefully go will become a major programming language that can challenge C++.

While problems such as Mr. Holmes cites are all too common (I have gone through several routers in search of a decent one for my home installation), there is hope, and there *are* companies who are doing what could reasonably be called software engineering – they design, they test (in multiple ways) and they apply the lessons learned from failures. But we do have to learn to deal with complexity in a far better way than we do now. I currently work for Panasas, a computer storage company. I can tell you that the level of complexity in a modern hard drive is truly frightening. Some days it seems like a miracle that they work at all. Then we take these hugely complex components and assemble them into systems containing hundreds or thousands of them, connected together by super-complex networking systems. So it goes. If someone does not do their work correctly, then lots of bad stuff can happen.

The best answer I know of to keep vendors of consumer equipment honest is on-line ratings and reviews. Use them to help you buy, and review the products you do buy (positive or negative – both are important).

On another topic, a quick recommendation of two webcomics:




A. Chris Barker



Subj: Languages for Reliable Programs: Don’t Forget Go!

I think you’ll find that Go channels Wirth’s spirit pretty well, if not perfectly.



Rob Pike’s "Public Static Void" talk is somewhat dated, but I know of no better concise presentation of exactly what Go’s inventors tried to do:


Pike’s more extended introduction, "Another Go at Language Design", is also somewhat dated, but, again, I know of no better comprehensive overview:


Rod Montgomery==monty@starfief.com


Operating systems


I’m still annoyed that IBM shot down OS/2, just as it was getting into orbit. Strong, stable, capable and secure, it was a victim of internal politics.

Then there were evolutionary dead ends such as Pick, which came and went.

Currently, there is really no true alternative OS to Windoze. The "IXes" — in all iterations, from UNIX to AIX to Linux to Apple — have been around since the days of DOS, and while significantly better, don’t have the muscle behind them to take on the Redmond Rangers’ marketing department (and licensing schemes which made Windows a monopoly). The only "rival" to what MS is putting out is Windows XP!

If OS/2 were still actively under development by Big Blue, MS would have been forced to fix the longtime weaknesses in the Windows platform.

Ah, well, one can wish. As the song goes, "Every OS sucks!"


I was there. IBM had no idea of what they were doing. At one COMDEX, if you came within fifty feet of the Microsoft booths and had the faintest resemblance to being a developer, A Microsoft operative would thrust a Windows Software Development Kit into your hands. Meanwhile IBM was proud to announce that you could buy the OS2 SDK for only $500. I agree that if IBM had the vision that Gates had, they’d have produced a better OS; but they did not. They never believed in a computer in every house, and in every office, and in every classroom. They believed in 10,000 big computers all running IBM software. And their dreamers thought there might be as many as a million computers by the year 2000.




Growth is the high fructose corn syrup of the financial world. The kind of growth demanded by investors is only sustainable for a new company during its expansion phase, the "exponential" portion of the growth S-curve. Such growth is unsustainable as all markets reach a saturation point. Mature companies that try to sustain growth in profits after they have saturated their market do so by pushing profit margin. A high profit margin should mean that a company is efficient — lean, spending money mostly on making product or service, not on overhead. However, there is a limit to how much overhead can be reduced relative to production. At that point, companies driven to growth will start cutting out production personnel, the high-cost (read most experienced) ones first, then moving down through the ranks till production is fully compromised and the company fails.

I have been involved with this cycle before and it is not pretty.

I do not trust companies who claim extraordinary gains in market share as they are inevitably small, unstable companies in the early portion of their life-cycle. I also do not trust mature companies who make claims of extraordinary gains in profit margin. They are inevitably destroying their means of production.

I prefer to look for companies paying a steady and healthy dividend. This is the protein and complex carbohydrate diet with a sprinkling of healthy fats that provides for a sustainable life for the company. Wall Street shuns such companies.

Kevin L. Keegan

Between the tax and the market structure of our financial system we have made it very difficult to have what used to be called stable Blue Chip companies: companies that make a good profit and pay dividends, and don’t try to buy their competitors to expand, nor do they seek to sell out for a capital gain.  For a stable Republic you need something of that sort. Schumpeter’s creative destruction needs to be combined with prevention of “too big to fail” and with tax laws that encourage “good enough” for as long as it is good enough. 

Yes. I understand that this is difficult. But we now overregulate everything, making it very hard for new companies to enter the market because they can’t afford compliance officers and lawyers, while encouraging companies to eat each other and become too big to fail.  This is a formula for disaster.  Stable companies that make a good profit  should not be forced to grow or die.  Yes, when their market vanishes they have no choice, but often that is not what forces them into unwise expansion in search of growth.  I suspect it’s too late at night for me to be writing this.



‘Despite the threat of war with Russia, the Ukrainian government is being forced by its lenders to try to militarily recapture their eastern tax base.’



Roland Dobbins

That’s a bit scary. Putin is skating as close to the edge as he dares go. If he loses control, things might get out of hand. Putin needs Russians.

The one thing we all need to remember is that Ukrainians, whether they speak Russian or not, are Slavs and related to the Russians. We also need to understand that Putin is no power mad dictator: he believes himself a patriot.


Today ends the Spring pledge drive. This is the last pitch about money you’ll hear for a while (well, there may be a similar announcement in the mailbag I’m hoping to get prepared before midnight). As we have said often, this site runs on the Public Radio model. It’s free to all, but it will not stay open unless it gets enough subscribers. I do want to thank all those who chose to subscribe this week, and particularly the new subscribers.

If you have never subscribed to this place, this would be a good time to do it. http://www.jerrypournelle.com/paying.html If you have subscribed, but it has been a while since your renewed – if you can’t remember when you renewed your subscription – this would be a great time to do that. I won’t be reminding you of it for a while, so do it now while you’re thinking about it… http://www.jerrypournelle.com/paying.html


The rising of the oceans…

Got to wondering, could the reported rising of the oceans actually be due to pumping down the fossil water in various aquifers such as the Ogallala Aquifer for agriculture and other human uses? After all, that water goes some place after its first use and many aquifers are not being replenished as fast as they are being drained. And since water vapor is considered by some to be a significant green house gas, this draining could be contributing to climate change as well.


Charles Brumbelow

Well, we know that the kilometers of ice over land areas of the Northern Hemisphere has been melting into the sea for nearly 20,000 years as we entered this interim in the Ice Ages, and we know that Scandinavia and other land areas once covered by ice have been rising. And battle hill in Hastings was once a dry road gating the way out of marshes. And, as I have said, I have seen how far from the sea the Hot Gates of Thermopylae are today. Finding a stable area to be the reference point for whether the seas are rising or falling isn’t all that easy.


Fighting Climate Change with YOUR Money

Hello Jerry,

You get right to the crux of the matter:

"We will be asked to pay lots more money to avert this new climate disaster, and the costs will be enormous because the effects of the remedies on the economy will be enormous (and the effects on the climate unmeasurable—BL), and cause famines in Africa. Now that they might get in on this industrial progress we are closing the gate in their faces, but that’s the way the climate changes.

At least there are jobs in climate change analysis. So long as you come up with the accepted results. If you don’t, well, you must work for an oil company.”

Where does the ‘Climate Fighting Money’ actually go?

No one will ever fully know, of course, but the following accounts for multiple billion of the missing dollars:


And, as you knew with the confidence of the schedule of the next sunrise, most of it can be found in the pockets of progressives in high places and their friends.

Bob Ludwick


Begley’s Best

I bought the product at a store called Good Earth, did not use it but it had Ed’s name on it and he is one of the few that actually do what they preach, so I had to buy it, would buy it again if they still sold it.



more evidence for your cocktail theory


Phil Tharp

The theory referred to is my “cocktail party” theory – i.e. a theory I would defend in a cocktail party but not publish in a scientific journal – on the importance of dogs to human evolution of intelligence. Since the same part of the brain that we use for cognition is used by dogs for olfactory sense, I propose that long ago humanity made a deal with dogs. “We’ll get smarter. You keep your sense of smell and protect out village. We’ll look after your children and you look after ours, and we’ll be friends forever, and after we get smart we’ll be better able to take care of both of us.” Human cultural evolution is by villages, and villages with dogs have a much higher chance of producing surviving descendants, and you can work the rest of it from there.


Re: Sexual Assault on Campus


A few days ago I read the article linked below that bears on the subject of your latest posting. I was unaware of many of the facts regarding investigation of rapes on college campuses and frankly, it’s quite a travesty.




The entire “battle between the sexes” has got out of hand. Girls are expected to join the hookup culture will they or nil they. People call themselves feminists shout rape at every possible opportunity, often causing authorities to become indifferent to very real cases of rape. There is little rational discussion now because attempts to talk about the situation generally degenerate into name calling and charges of gross insensitivity (and that’s the mildest charge).


The Euthanasia Coaster.



Roland Dobbins

Pope Benedict spoke of a Culture of Death.


William Harvey "Bill" Dana, RIP.



Roland Dobbins


..and one of the first astronauts…


Stephanie Osborn

Interstellar Woman of Mystery

See all my books at http://www.Stephanie-Osborn.com <http://www.stephanie-osborn.com/>


Today ends the Spring pledge drive. This is the last pitch about money you’ll hear for a while (well, there may be a similar announcement in the mailbag I’m hoping to get prepared before midnight). As we have said often, this site runs on the Public Radio model. It’s free to all, but it will not stay open unless it gets enough subscribers. I do want to thank all those who chose to subscribe this week, and particularly the new subscribers.

If you have never subscribed to this place, this would be a good time to do it. http://www.jerrypournelle.com/paying.html If you have subscribed, but it has been a while since your renewed – if you can’t remember when you renewed your subscription – this would be a great time to do that. I won’t be reminding you of it for a while, so do it now while you’re thinking about it… http://www.jerrypournelle.com/paying.html


Freedom is not free. Free men are not equal. Equal men are not free.




Bookmark the permalink.

Comments are closed.