Programming as an Individual Activity

I've decided to come back to Weinberg's classic 'The Psychology of Computer Programming' since it is at the heart of why I took up writing about programming in the first place.  At my first job oh so long ago, it was obvious that we were nowhere close to a solution to programming-in-the-large.  With applications small enough to be written by a single developer we could manage to deliver software that works, maybe it was inefficient or clunky but it worked well enough.  Sure large systems were built using the waterfall approach but they would invariably overrun both their budget and their schedule.  The industry was by no means young, even though the PC revolution was in full swing, the underpinnings had already been discussed ad nauseum. You could dismiss the waterfall method but would still be left with the same questions, 'what do they want?', 'how do we build it?', 'how long will it take?', and so forth.  The Agile or Lean movements have been able to change the discussion to ask 'what is the smallest amount of work that we need to address at one time' and go from there.

I believe that we have been avoiding the most important questions - which are uncomfortable to contemplate - like, "what is the largest single aspect that controls a software schedule?" or "How do we plan a project when the members of the team are not pluggable resources?"

This brings me back to one of the biggest unresolved question of our field, "Why can't we engineer software like we engineer buildings or spacecraft?"  I ask this not for serious comparison since the differences are legion, but rather to point at the single most important advancement in the field of engineering since the advent of the first simple tools (the lever, the wheel, etc), calculus. 

Why calculus you ask? because it gives us a way to plan for infinite variability and still have confidence in the result.  Planning a trip to the moon would be impossible without being able to calculate how much fuel is needed since how much fuel is needed depends upon the weight of the rocket which creates a recursive dependency.

Software will have no equivalent for calculus because the variability comes from the people participating in the project.  Until we accept that people are a first-order component of any software project we will continue to look to silver-bullets, the next development process fad, or the latest new-fangled technology for relief.

XSLT - The Modern Day Snake Oil

Technology seems to attract all kinds, no surprise these days when everybody and their dog are high-tech consumers. But those of us who earn their living in that fast-paced world every day can't help but get a bit jaded seeing the same mistakes repeated time and time again.  From the desperate or possibly Machiavellian managers who want to ignore common sense and decades of history and continually chase the proverbial silver bullet, to the greedy vendors selling the latest 'enterprisey' tool with promises which sound more like a late-nite comedy skit, "Not only is it a floor wax, it's a tasty dessert topping!".  In software it is, "Without writing a single line of code". Why do people fall for these pitches? You can blame ignorance, malice, or greed; in any event, it is up to each of us to keep our wits about us when evaluting claims for the latest-and-greatest doohickey the next salesdroid offers.

There is one phrase that I've seen many times over the last twenty years and whenever I hear it, I immediately think of snake oil.  Whenever you hear a consultant say anything to the effect of, "easy to use by non-programmers!", smile and quickly look for the exit.  How trite is it? Well, early in my college days I remember hearing stories from the early mainframe days when assemblers were first coming out, "It works so well you won't need programmers anymore!"  At my first job at a large company, the 4GL vendors would use the same line on the managers, distracting their mark from the large price tag with promises of savings from lower staffing.  The reality was nobody but programmers could figure out how to actually use the immense packages, assuming they could get them installed correctly in time to use on a project. Even modern tools like Requisite Pro fall, which I really liked, prey to this. They are supposed to improve productivity but the learning curve is so large that the only way to properly use it on a large project is to have someone who already knows it configure it for you.

Recently I ran accross the claim in an old book, "Professional Java XML Programming with Servlets and JSP" by Alexander Nakhimovsky and Tom Myers published in 1999 (page 526 to be exact) which attempts to summarize the ream of paper constituting the previous chapters.  In a section appropriately but belatedly (considering where it appears) titled "What's XSLT for?" the authors explain,

We are rapidly approaching the summary and conclusions to this final chapter of the book In the meantime, the main conclusion about XSLT that we hope you will draw from this chapter is this: XSLT is best described as a general-purpose mini-language, intended to make many common operations on XML documents easy to express and accessible to non-programmers.

Is the book old and out-dated? Of course. The age of the book only matters though if the authors would express a different attitude given a chance to update their book. Granted, right after the offending paragraph they attach some conditions,

Allows non-programmers to do some transformations:
- without tools, only simple things
- with tools, more complex things

When I read the above, I couldn't help thinking, "Have you guys _looked_ at what you've written lately? If it takes a book 500 pages long to explain it, you're never going to get non-programmers to use it"  Why is programming hard, I don't know and do not want to venture a guess.  What I do know is when competent programmers who are at least trained - if not adept - at imperative object-oriented languages face a significant learning curve when working with a declarative language like SQL much less a functional one like Erlang.  XSLT is like a combination of web templates with an embeded declarative language which also include functional aspects, although the authers probably disagree saying, "As of today [1999], JSP is a tool for a prgrammer while XSLT is a tool for a Competent User who is NOT a programmer."

Am I trying to call out XSLT or XML? No, they are useful tools (like automatically formatting your log4j config file) and at this point we couldn't get rid of them if we tried. The issue is quite old and better men than myself have already weighed in.  What I do want is for people who should know better, like said authors, to refrain from propagating the myth that if only we had better tools or languages, non-programmers would rush in to join the computerized fun.
...

To Buy or Not to Buy Technical Books, That is the Question

So the rookie coder is working at his new job, fresh out of college, when the team malcontent is overheard a couple of cubes over, "This place is so cheap! They should reimburse me for this technical book. It's a heckuva lot cheaper then a training class."  However true that is, it hides a simple choice: keep the book and be out the money, or give up the book and pocket the cash - assuming the boss will foot the bill.

Recently a question on Stackoverflow reminded me of that rookie coder. Being fresh out of college gives you an amazing combination of overconfidence and cluelessness.  Without any basis, it's easy to take the office attitude and run with it.  You shouldn't have to buy your tools, your company should.

I disagree.

Why? Because they are tools. It's not even a cost thing. A professional should be responsible for knowing which tools they need to be most productive. Tour an airline maintenance facility and check out the workshop.  A large area, maybe 1000 sq. ft., with nothing but large toolboxes;  each with a name and a lock. The mechanics have to buy all their own tools, think wrench sets and so forth - not large things like engine dollies.

At first it seems a bit silly and redundant but when you start to think about it, it's easy to see that not everyone is going to want to spend the extra money on the high-end tools, or they might want extras of the same set for whatever reason. Also, consider the hassles with sharing, e.g. "Where are the 1/2 inch box-end wrenches?" By not making the tools shared, the 'Tragedy of the commons' is avoided.

Another example is chefs and their knives. You don't touch another chef's knives. The restaurant provides the big things like ovens and stoves, the chefs bring their own knives.

Back in the world of computers, I've brought in extra memory before. I told my manager so he'd know that some of my personal property was in the PC in case it needed to be worked on by the internal techs. I also put a note on the box itself. I would have had a bigger issue if they discouraged bringing your own hardware. Most shops like to provide the box itself because it minimizes their support issues but they shouldn't have any problems with you bringing accessories like monitors, keyboards, mice, etc.

To summarize, change how you think about tools and strive to use whatever makes you better at your job, even if you have to pay for them out of your own pocket. It's the 'Professional" thing to do.

Martin Fowler, Alistair Cockburn, and Optimism

The effort to make Software Engineering actually live up to the 'Engineering' part has been greatly helped by Martin Fowler's book on refactoring. Martin recently posted about his decision to not participate in an effort by others in the industry, not lacking in optimism, to define a common process for software development.

Martin Fowler declined to get involved in a similar effort and for the same reason.  The key quote is,

“Why this is so was primarily crystallized for me by Alistair Cockburn who explained that since people are the central element in software development, and people are inherently non-linear and unpredictable - such an effort is fundamentally doomed.”

While my previous post about Alistair's paper covers much of the same ground I felt it was worthy of an update.  There will always be souls who crave consistency just as there will always be salesmen willing to offer the illusion of it.  Facing our own nature is one of life's hardest lessons. Some of those who ignore the human factor are optimists seeking to improve their chances of consistency, which has driven many towards the latest and greatest geegaw, a new language, a new GUI, a new design tool. Others are simply deluding themselves in the hopes that if only people would sit up and fly right (read this to mean 'do it their way') all our problems would be solved. 

Studying the process of computer programming has been around for decades.   Even so, we still have trouble internalizing human foibles - people aren't predictable and follow "standards" about as well as a random-walk.  So the most successful way to get people to be consistent is to lead by example, blaze a trail ahead and show them the way.  In this manner Extreme Programming was its own worst enemy and its own best cheerleader. By showing how a set of practices work in situ, it created a much-needed focus. The "Waterfall" model was the favorite whipping-boy, essentially dead, and the iterative approach was in its infancy.  So XP served as a lightning-rod for discussions on the craft of software development.  It provided an example of practices that could be emulated and provided an easy target for the doubters, disbelievers, and denialists; no doubt the negativity coming from the critics was channeling  their innate understanding of the darkside of human behavior.

While Extreme Programming hasn't become the standard development model, that doesn't mean it failed.  When the history of Software Development is written, XP will be given credit for re-introducing the most important factor; not tools nor process, but people.

Chapter 5 - The Programming Team

This continues my review of "The Psychology of Computer Programming" by Gerald M. Weinberg.

So far we've had individual programmers either working solo or in an environment where other programmers work but on separate tasks. As long as a single person can maintain the conception of the program, conflict resolution involves the thinking of a single individual, regardless how much advice is available or heeded. When a problem is too large for one person to hold in their mind at once, the scale of the problem goes up due to the increasing communication needs but is now one of a different kind. In this second case,

"conflicting technical demands are translated into potential interpersonal conflicts, and a social mechanism must be formed to resolve them"

This is part of the territory. Two people with solutions that are mutually exclusive. Programmers are proud of their intelligence and are more than willing to argue their side.

How A Team Forms

"As a rough rule, three programmers organized into a team can do only twice the work of a single programmer of the same ability"

I'm no longer as convinced of this as I once was. It really depends on how well defined the separation is between the modules.

Establishing And Accepting Goals

"Social psychologists have verified in other contexts that failure of one or more members to share the group goals affects the group performance -- not only through that member's share but through a reduced performance on the part of the others, for they invariably perceive the division within the group or the indifference of one of the members."

A non-programming example of this is when football teams get a new coach; the TV pundits like to use the phrase "They've bought into the new system" meaning that the players have taken the coaches goals for their own. A common programming example you might see in a development shop is where the current toolset and technology stack used is from Microsoft and there is that one guy who keeps muttering under his breath that "Microsoft sucks". His lack of cooperation will be felt by the rest of the team. The result may be that they avoid him instead of sharing the new trick they learned to solve some daily annoyance. No one wants to pair-program with him. The price being paid, at a minimum, is the reduced productivity of the frustrated Microsoft-hater.

The Team In Crisis

"The replacement of a leader--or any team member--is probably the most frequent and typical crisis in the life of a team. The effect of adding or removing a member depends quite sensitively on the group structure, and many a manager has been unpleasantly surprised at the change in performance resulting from the removal or addition of an 'insignificant' group member. "

If we need proof that programming is a social activity, we need look no further. Unless you are the sole programmer working at a company or on a project, there will always be a social aspect to what you do. I'll go even farther and say that unless you are the only person who will use the application, there will inevitably be people involved, from the person who asked you to write the app to the person who will be using the app.

"A member who is competent but who does not get along with the others can be an even more serious problem for the democratic group than an out-and-out incompetent. In an authoritarian group, such a member would not have much contact with others on a working basis anyway, so as long as he gets along adequately with the leader, he presents no particular problem. Indeed, some programmers prefer to work under a strong, centralized leader in order that they do not have to socialize with their fellow workers. But in a democratic team, an antisocial member cuts lines of communications and is a constant impediment to consensus in team meetings."

Sometimes we are our own worst enemy. If we want to stay our of our own way, we need to remember in software development, people are the biggest cause of failure.

Summary

Questions

For Managers:

- Are your hiring practices such that you get more uniformity on teams that you would like? When making up teams, do you try to see that a good "mix" of people is on each one, or do you strive in the opposite direction?

- Do you ever do things to try to inflate the appearance of your technical competence in front of the people who work for you? Describe some of these incidents, and also some incidents in which it was discovered that your technical competence was in at least one respect inferieor to one of the people who work for you. What were the consequences of that discovery, and do they justify attempting to cover up?

- In setting your own working goals, what part is set by what is passed down from above, and what part is set by what comes up from below? Are you satisfied with this arrangement, or would you like to alter it in some ways?


For Programmers:

- What part of programming work do you do best? Are you permitted to contribute that best part of your work to your team, and is it generally recognized that it is the best part of your work?

- Has your manager ever done anything to make you doubt his honesty? If so, describe the incident, what ultimately happened to your doubt, and how your work was affected.

Programming as a Social Activity

What do we imagine when we consider the individual programmer? There are many stories that liken the programmer to the cowboy who finds himself in a small remote town rife with lawlessness and misery. The cowboy grudgingly accepts the job as sheriff, to right wrongs and otherwise do deeds of much heroism; riding off into the sunset, when the crisis has been averted, to the wails of the local townfolk who think they are losing their only defense against the chaos that comes with frontier life. Is this how individual programmers actually work or is this a delusion used to supplicate the ego of socially inadequate developers?

This continues my review of "The Psychology of Computer Programming" by Gerald M. Weinberg. The previous chapter Chapter 3 - How Can We Study Programming discusses the difficulty of studying behaviors which are hard to isolate, repeat, or tranlate into an experimental setting. Chapter 2 - What Makes a Good Program evaluates the subjective and objective criteria for judging programs. The beginning of the series of reviews is Chapter 1 - Reading Programs. Techrepublic.com has a review of the same book which reinforces this book's classic status.

In Part 2 of "The Psychology of Computer Programming", Weinberg describes three different ways that programmers are organized: the group, the team, and the project. Chapter 4 is titled "The Programming Group". He starts with the observation that when programmers work together in the same environment, they inevitably seek each other out for help or advice and expresses the premise that the lone programmer is at a serious disadvantage.

Formal and Informal Organization

The first lesson is that in any assemblage of people there are formal and informal groups. Formal hierarchies like an organization chart might be nice for the manager types but do not begin to describe the various interactions between individuals in a group. One good story is from back in the batch-scheduled era when it was difficult to know whether your job had run and its output was available, which was accomplished by posted completed jobs on a board outside the computer area. Programmers in remote offices might waste half an hour coming over just to find their results were not ready. As it happened, there was secretary who sat in a location which allowed her to see the result board. When a programmer who was waiting for a job to complete happened to call her for a date. When the conversation turned back to work, she mentioned that he must have work to get back to. He told her that he was waiting for a job to complete. She volunteered that she could see the board from where she was sitting and asked him in which job he was interested. This began an informal service which quickly became known to the outlying programmers. The service became popular enough to disrupt the work of the secretary and need the attention of the administration. Wisely acknowledging that the service could not just be cut off, they decided to implement a sort of job posting hot-line, even using the secretary's original phone number, where a recording would give the latest job completion posting. Here an informal and inefficient solution was replaced with a formal but efficient one.

This contrasts with a second story where a university computing center provided a common room for working on programming problems and an adjoining room with graduate students available to answer questions ("consulting services" was the term used in the book), and at one end of the common area were vending machines. When students complained about the noise from the crowd of students around the vending machines, the administration removed the offending machines. Not long after the removal of the vending machines, another delegation came to the administration with a complaint, there wasn't enough consulting services. As it turns out, the students who were making all the noise were actually discussing their programs and helping each other with them, answering all sorts trivial questions in an informal setting. The administration, unable to believe that the removal of the vending machines was the actual problem tried to increase the number of consultants available with little success as fewer graduate students were willing to deal with the ever increasing number of mundane and trivial problems. The formal solution was not nearly as effective as the old informal one.

Physical Environment and Social Organization

Does the environment in which you work affect your productivity? Companies are all about getting the most for the least so it is no surprise that in devising the seating arrangements of programmers, minimum cost usually trumps maximum productivity. There are exceptions like early Microsoft and their "every programmer gets an office with a door" mantra, to more recent examples like Fog Creek, who also take a programmer-first approach. In Weinburg's words,

"Someone should definitely study the depressing effect that the all-too-common half partitions have on programmer productivity. They manage to cut off all useful communication while permitting all disturbing sound and movement to penetrate."

Another example was the building that replaced the old manual elevator with an automatic one. This was unfortunate for the programmers because the old elevator operator had run an informal pickup and delivery service for them between their floor -- the eighth-- and the machine room -- the basement. They saved the cost of the elevator operator but cost their programmers much time every day going back and forth to the basement of the building.

Error and Ego

"If asked, most programmers would probably say they preferred to work alone in a place where they wouldn't be disturbed by other people"

"The ideas expressed in the preceding paragraph are possibly the most formidable barrier to improved programming that we shall encounter."

Programmers tend to get invested in the programs they write, attached to them we'll say. This causes behaviors that are counter-productive. What's wrong with "owning" programs, after all artists "own" paintings; authors "own" books; architects "own" buildings. When we go to the bookstore we can get a good idea about the quality of a book if it by a familiar author. Wouldn't the same hold true for programs and their authors?

"The admiration of individual programmers cannot lead to an emulation of their work, but only to an affectation of their mannerisms. This is the same phenomenon we see in "art colonies," where everyone knows how to look like an artist, but few, if any, know how to paint like one."

Another problem with the idea of ownership happens when we start to evaluate or judge a work. Art is to a large degree subjective so the artist can dismiss the opinions of critics. Can a programmer dismiss the judgement of a computer? If we took the "ownership" association to its logical end we would reason something like this:

"This program is defective. This program is part of me, an extension of myself, even carrying my name. I am defective."

Obviously that does not happen. Programmers do not quit just because of a single bug or logic flaw but they will tend to overlook problems and strive to prove the program is correct even in the face of mounting evidence to the contrary. Proving the program is wrong turns into a minefield of self-worth where bugs become the buried mines which will injure the pride of the programmer who can't help but react in defense of his own sense of self-worth.

"Thus, if we are going to attack the problems of making good programs, and if we are going to start at the fundamental level of meeting specifications, we are going to have to do something about the perfectly normal human tendency to believe that ones "own" program is correct in the face of hard physical evidence to the contrary."

This explains the unconscious motivations behind the resistance to code-reviews, pair-programming, or any other effort to get them to loosen their hold on "their" code. If we saw this type of behavior in a toddler, we would immediately recognize it as childish, as in "James doesn't want to share his code with the others." Will James feel like he is losing something if he has to let others play with his code? It would go a long way to explain things I've seen in my programming career.

Egoless Programming

So, what do we do? The manager could insist that the coders work harder at finding the bugs in their code. He could even go around and ask them to show him the bugs in their code. Applying even the smallest amount of psychology would tell us that this will backfire. The average person will feel he is on trial and very easily take it personally. If we can change the social environment of the group, we have a chance to affect change in the values of the programmers of that group. This is possible, we know of egoless programming groups since the very early days of programming.

"John von Neumann himself was perhaps the first programmer to recognize his inadequacies with respect to examination of his own work. Those who knew him have said that he was constantly asserting what a lousy programmer he was, and that he incessantly pushed his programs on other people to read for errors and clumsiness. Yet the common image today of von Neumann is of the unparalleled computing genius--flawless in his every action. And indeed, there can be no doubt of von Neumann's Genius. His very ability to realize his human limitations put him head and shoulders above the average programmer today."

Weinburg tells another story of a programmer from the early days of space tracking systems. "Bill" was confident in a loop and gave it to "Marilyn" to informally review, as was common practice in the group. Marilyn found seventeen bugs in only thirteen statements. Bill chalked it up to a 'bad programming day' and became more and more amused as each bug was found. He took it as proof that it just wasn't his day and used the rest of the day giving everyone a laugh at his expense. Marilyn, on the other hand, did not have any false confidence in her own ability and took the code to a third person. She knew that she had looked at it so much she wasn't going to be able to see any more errors. She needed fresh eyes. With others help, she found three more errors before the day was over. The epilogue was that no more errors were found in the code despite diabolical testing and being in production in over a dozen installations even after nine years.

Why aren't there more groups like this? Weinburg asserts that such groups are not an isolated case. They regard their practices as valuable proprietary information, implying they think other would copy them if given a chance. These groups are very satisfied and stable which reduces the opportunities for programmers to have encountered these groups. Also, gypsy programmers - who never settle in one place - encourage the myth that the best programming is the product of genius, to justify ever-increasing consulting rates.

The last word of this section is on how there is an advantage to making your code as readable as possible for other programmers for that means that those who come after will have a chance to learn from such code without feeling the need to have it deciphered or translated. So, the best code isn't the stuff that is hardest to read but the easiest to read. Easier to test, maintain, understand, and learn from; who could argue with that?

Creating and Maintaining the Programming Environment

One danger programming environments face is the tendency to fixate or lock into some behavior, habit, or tool. Programming language is a common example. Even if another language is better suited for a problem, the language of choice will be preferred. More people know it, there are more tools available (to the group), more library routines, etc. In other words, it's "how we do things here" which makes choosing a different language like paddling upstream, lots of resistance and little to show for a great deal of effort.

Another fixation is the social environment which either encourages or discourages egoless programming. If the new guy is ridiculed when asking another for help or advice, he is less likely to seek assistance the next time. If, on the other hand, someone pays him the implied compliment by asking him to look over a program, he is more likely to feel secure enough to seek advice.

"To a large extent, we behave the way we see people behaving around us, so a functioning programming group will tend to socialize new members to its philosophy of programming."

On the subject of egoless programming and management, Weinburg astutely describes the culture shear between common management and uncommon programming groups. Remember that this book came out almost 40 years ago; you can draw your own conclusions about how much things have changed.

"Managers tend to select themselves from the 'aggressive' component of society and have difficulty appreciating the fact that other people do not completely share their goals of money and prestige. They are especially at a loss to understand the smooth functioning of a programming group based on mutual respect for individual talent and cooperation in the common cause. Instead, the tend to view people as working for money or under threat -- as they themselves do."

Another example given is of a group that was particularly success and producing a new system. The achievement was noticed by the company's management who decided to give them a cash award. Typically, they wanted to give the award to the groups manager but were bewildered when the manager would not accept it unless the whole group got the award. The groups manager understood that the group acted as a team and that responsibility for success could not be attributed to a single individual. Other managers thought he wanted more money, others thought he was trying to setup a "prima donna" group. The company decide to force him to take it and break up the group -- which seemed to have unhealthy ideas. For those doubters out there, I have personally witnessed this same scenario in my career. The moral of the story? Again, Weinburg says it as well as anyone,

"Had the management been more aware of what this group had to offer, and had they been more flexible, they might have worked out a solution that would have permitted this group to influence the work of other groups in a favorable way. But this is not easily done by managers, who tend to feel that when work gets done it is the direct result of the actions of some leader of outstanding ability. Even when the manager appreciates the work of the group, it is not consistent with his own philosophy to see the productivity of the group as a property of the group, not as a sum of the contributions of the individual members."

Summary

To understand how programmers perform in a group setting, you must consider formal and informal structures, the physical environment, and even the social interactions in play between members of the group. These are lessons which are still worth learning today. The hard question we must ask ourselves is why we keep searching for new solutions when we have not addressed the issues identified four decades ago. I maintain that the Agile Manifesto or the Manifesto for Software Craftsmanship, is but a few of the efforts to awaken the courage of individual programmers to affect change around them, and cause a groundswell of change which might have the momentum to disturb the inertia of the status-quo.


Sample questions:

For managers: What is your honest opinion of people who are not trying to "move up" in your organization, but who seem satisfied with the kind of work they do and the amount of money they get? To what extent is your view influenced by your own feelings for yourself?

For programmers: Do you refer to your work as "my" program? Try passing one week without using the personal possessive in reference to programs, and take notes on the effects you observe.


p.s - These chapters are large enough that I may break them down so each section is its own post.

Chapter 3 - How Can We Study Programming?

This is a continuation of the review of Gerald M. Weinberg's book, The Psychology of Computer Programming. The review begins here for Chapter 1 - Reading Programs. The post on Chapter 2 can be found here.

What chance have we to understand the human factor? How do you study programming when the important majority of it takes place between one's ears?

Perhaps programming is too complex a behavior to be studied and must remain largely a mysterious process.

INTROSPECTION

Although some modern students of human behavior tend to discredit it as nonscientific, introspection has always been the first foundation stone of their science.

The author tells a story about a student having difficulty with a particular PL/I programming problem:

ANGLES(I)=2*ATAND(SQRT((S-A(IND(I,1)))*(S-A(IND(I,2))))/(S*(S-A(IND(I,3))))));

When it was finished the author asked the student to recall the thought processes -- the difficulties -- he had experienced with this statement, here is that list.

1. He had trouble getting the parentheses matched, because there were so many of them.
2. The original ANGLES(i) was a structure, not an arry, so the compiler complained of an illegal subscript list, but he had not been able to find the problem because there were so many subscript lists.
3. When the program finally got into execution, there were these further difficulties:
1. IND was a matrix on another page of the listing, and he had trouble finding it.
2. Even though the parentheses were now matched, one pair was in the wrong place and was most difficult to find.
3. The last difficulty turned out to be a problem of precision caused by capricious declaration of different data types, by the large expression, and by the division--which would not have been necessary if the two-argument form of ATAND had been used.

The student provided a number of insights, from the proper size of statements, the choice of data structures, to the design of compiler and execution-time diagnostics.

None of these insights could be obtained without introspection of this sort, yet thousands of programmers [millions now days - ed.] each day have theses same problems and simply grumble that they are having "bugs." Without the introspection, they will simply continue to have "bugs" until, in the end, the bugs will have them.

From this single instance of introspection, we could come up with "Laws of Programming"

1. The mind cannot cope with more than five levels of parentheses.
2. Compiler diagnostics should be more explicit.
3. PL/I precision rules are too complicated to use.

Although each of these statements might turn out to be a "law" in the psychology of programming, writing it on the basis of a single introspective example hardly qualifies it as a law. To be a "law," a principle must be explored so as to set limits to its application, because all laws have limits. Indeed, the limits are Usually more important to know than the laws themselves, and it is only by investigating a multitude of cases that limits can be set. Thus, without introspections, investigation would be sterile, but without investigation, introspection will be of questionable value in application.

OBSERVATION

One way to follow up introspection is by observation of what people actually do, as opposed to what they think they do.

Like the study of parentheses and how many levels of nesting programmers use.

One of the problems with observation, however, is that it tells us what people do do, but not necessarily what they can do.

If your census showed that programmers never nested parentheses more than 5 levels deep, does that mean they are unable to?

A second problem with observation is deciding just what it is we are observing.

All you know about the programmers who limit themselves to 5 levels of parentheses is a simple fact, to know why they don't use 6 or more would of course require introspection on the programmers part. If you were designing a language, would you limit the size of arrays to three dimensions simply because you've never seen anyone use more?

A third problem with observation is the problem of interference of the observer with the subject being observed.

This is called the "Hawthorne Effect". The act of observing workers actually affected their performance, a kind of uncertainty principle.

EXPERIMENT

Here is a very interesting section where the author describes the costs of conducting experiments on programmers using typical experimental psychology methods. How do you collect data without influencing the results. To a degree it is somewhat like the problem of testing the performance of a program. The modifications necessary to observe the programs use of memory or CPU time actually uses memory and CPU time so how can you know your results haven't been skewed?

You could do experiments using programming majors in school but would the results apply to professionals? You could study professionals but that could be prohibitively expensive. The author describes doing an experiment using one-quarter time of nine experienced programmers for three months. The average salary was $14,000 [I'm assuming annually - ed.], add in overhead (workspace, machine time, etc.) came to $20,000 apiece, or $1250 apiece over the course of the study. Add to that the experimenters salaries and so forth and the bill comes to $30,000 for nine subjects. [In 1970 dollars! - ed.] Using the Inflation Calculator ($30,000 in 1970 compared to 2009) this comes out to a cost of $163,254.90. Serious money, just to study programming.

... we must deal with one other problem, which is important because of a seldom-questioned view of programming. That view is that programming is an individual activity, not a social activity.

Not that the individual level is unimportant, but we might start by asking why, if the average programmer spends two-thirds of his time working with other people rather than working alone (yes, it's true!), that 99 percent of the studies have been on individual programmers.

The answer is that it is expensive to study programmers in groups. Plus, you can't put a number of trainees together and call them a 'team'.

Putting a bunch of people to work on the same problem doesn't make them a team--as the sloppy performance in all-star games should teach us.

The neglect of social effects also casts doubt upon all individual studies, since these studies force the individual to separate himself from his normal working environment. In one of our studies, one of the programmers came to me when he had finished coding and asked who could check his work. In his home group, this was standard practice at this point in development, but it was not allowed during the study -- lest it should "invalidate" the results.

Looking back, it seems that forcing programmers to work in isolation is like trying to study the swimming behavior of frogs by cutting off their legs and putting those legs in the water to watch them swim. After all, everyone knows that the frog swims with its legs, so why complicate the experiments with all the rest of the frog?

PSYCHOLOGICAL MEASUREMENT

For the present, most of the work in the psychology of programming is going to have to be "looking for questions."

I conjecture that we will move beyond looking for questions only when we start believing and acting on the questions that we've found over the almost 40 years since this book was published. This is definitely not a programming problem. When the software managers are able to dictate practices based on a whim or the local political winds, "We don't have time to do code reviews! Why can't you just make it work? I don't understand why it's constantly breaking!", you might start looking for answers in the abnormal psychology section of the library.

USING BEHAVIORAL SCIENCE DATA

Because of the nature of programming, there is not much hope that actual results can be transferred directly from other fields to ours. Our use of results has to be for insights, not for answers.

So maybe all those comparisons with engineering and construction are invalid on the face of it. We might look at Paul Graham's book Hackers and Painters for insight.

Finally, as we pass successfully through all this, we may find ourselves in a sufficiently objective frame of mind to examine some of the myths of programming -- the articles of programming faith. Faith, as Bertrand Russell pointed out, is the belief in something for which there is no evidence; and myths, as Ambrose Bierce once defined them, are the sacred beliefs of other people. Perhaps we shall simply carry away our old beliefs unchanged, but if just one unfounded programming myth should die as a result, this book will have been worth the effort.

SUMMARY

Common pitfalls:

1. Using introspection without validation.
2. Use of observation on an overly narrow base.
3. Observing the wrong variables, or failing to observe the right ones.
4. Interfering with the observed phenomenon.
5. Generating too much data and not enough information.
6. Overly constraining experiments.
7. Depending excessively on trainees as subjects.
8. Failing to study group effects and group behaviour.
9. Measuring what is simple to measure.
10. Using unjustified precision.

1. 
   

Several of these items have had some attention paid to them over the last decade. Extreme Programming was one of the first movements focused on groups of developers to get enough traction to garner a following. There are several similar movements, SCRUM, Agile, etc. Instead of assuming the problem was in the process that humans were following, they finally addressed the human factor as a first-order success factor (thanks to Alistair Cockburn)

The question is, when will we learn that we are the problem?


This ends Part 1 - Programming as Human Performance. Next up is Part 2 Chapter 4 - Programming as a Social Activity

What Makes A Good Program?

This is a continuation of the review of "The Psychology of Programming". You can find the post on Chapter 1 here.

Chapter 2 - What Makes A Good Program?

If we plan to study programming as a human activity, we are going to have to develop some measures of programming performance. That is, we are going to have some idea of what we mean when we say that one programmer is better than another, or one program is better than another. Although we all have opinions on these questions, we shall find that the answers are not as simple as we might wish. For programming is not just human behavior; it is complex human behavior.

What do we mean by complex human behavior? This article at betterexplained.com does a good job highlighting the false dichotomy between simplicity and powerful. Another explanation my good friend Scott sent me used simple, complicated, and complex. I'll let Dave Nicolette speak for himself.

The spectrum goes from easy to understand and predict to extremely hard to understand and predict (but possible with enough effort), to impossible to understand and predict regardless of the level of effort.

In our studies of the psychology of programming, we shall be hampered by our inability to measure the goodness of programs on an absolute scale.

Plenty of effort has been wasted trying to discover that mythical absolute scale; lines-of-code, lines per day, runtime, memory usage, the list goes on and on. The only thing that is agreed it would be that there is no single metric which can be applied to all programs.

Looking honestly at the situation, we are never looking for the best program, seldom looking for a good one, but always looking for one that meets the requirements.

This might be the hidden reason for why so much lip-service is paid to software quality. Crappy code that ready right now always beats higher-quality, however defined, that is late. Unless there is a business reason which drives a requirement connected to quality (like uptime, throughput, etc.) where time-to-market is trumped by proper outcomes, managers will revert to turning a blind-eye to quality. Risk deferred is risk ignored.


Specifications

Of all the requirements that we might place on a program, first and foremost is that it be correct. In other words, it should give the correct outputs for each possible input. This is what we mean when we say that a program "works," and it is often and truly said that "any program that works is better than any program that doesn't."

Here follows a good story about a system for a car maker to handle all the options which prospective car purchasers could order. When the programmer arrives on the scene of a disaster in the making, the long-settled approach used was not only overly complicated but didn't produce the correct results. When the programmer exclaimed that the Emporer had no clothes, he was condemned for being uncooperative. While on the way home, he can't stop thinking about the problem. In a fit of insight, he realizes that a workable approach was within reach. Upon returning to the customer site, he describes his solution. Naturally the audience gave him a cool reception but they listened without questions - until the original create of the old system raised his hand and asked, "And how long does _your_ program take?" to which the response was, "That varies with the input but on average, about ten seconds per card" (an indication of how old this story is). "Aha," was the triumphant reply. "But _my_ program takes only one second per card." This seemed to settle the issue with the audience until our protagonist observes,

"But your program doesn't work. If the program doesn't have to work, I can write one that takes one millisecond per card - and that's faster than our card reader."

Here is where the sarcastic could say "We just redefine what 'done' means", or "that isn't a bug, it's an undocumented feature"

In effect, then, there is a difference between a program written for one user and a piece of "software." When there are multiple users, there are multiple specifications. When there are multiple specifications, there are multiple definitions of when the program is working. In our discussions of programming practices, we are going to have to take into account the difference between programs developed for one user and programs developed for many. They will be evaluated differently, and they should be produced by different methods.

Imagine a scene from "I Love Lucy" where Ricky Ricardo and Fred Mertz are trying to rearrange furniture at the behest of Lucy and Ethel. "Move that painting to the other wall", Lucy commands. When she then turns her attention to having the men move the couch "a teensy bit" closer to the wall, Fred moves his side an inch while Ricky moves his side a foot; after all, how many "teensies" are in a foot? Ethel, meanwhile, sneaks over to the picture while everyone else is busy with the couch and moves the painting back to its original location, "It looked better that anyway", she thinks to herself.

However comical this might be, when this happens in software it can be many, many times worse. With multiple stakeholders, multiple upstream and downstream dependencies, and a large group of developers trying to write code, it should be easy to realize how requirements (and their management) can become a nightmare.

Schedule

One of the recurring problems in programming is meeting schedules, and a program that is late is often worthless.

Most people prefer to wait a fixed ten minutes for the bus each morning than to wait one minute on four days and tewenty-six minutes once a week. Even though the average wait is six minutes in the second case, the derangement caused by one long and unexpected delay more than compensates for this. This is where a psychological study would be rewarding. Project management is not a science and only partially is controlled by mathematics. Knowing the psychological landscape in which the project functions can make a big difference in its success.

Adaptability

Few programmers of any experience would contradict the assertion that most programs are modified in their lifetime. Why, then, when we are forced to modify programs do we find it such a Herculean task that w4e often decide to throw them away and start over? Reading programs gives us some insight, for we rarely find a program that contains any evidence of having been written with an eye to subsequent modification. But this is only a symptom, not a disease. Why, we should ask, do programmers, who know full well that programs will inevitably be modified, write programs with no thought to such modifications? Why, indeed, their programs sometimes look as if they had been devilishly contrived to resist modification - protected like the Pharaohs' tomb against all intruders?

This really goes to the heart of the quality software movement. If it's shown time and again that the majority of software cost is in maintenance, why do we fail so miserably to factor in maintainability while writing software?

Adaptability is not free. Sometimes, to be sure, we get a program that happens to be adaptable as well as satisfactory in all other ways, but we generally pay for what we get - and often fail to get what we pay for.

It's as if the manager figures that maintenance doesn't come out of his budget and she's only being measured by whether the project comes in on time even if the resulting code will cost twice as much to maintain.

Fisher's Fundamental Theorem states - in terms appropriate to the present context - that the better adapted a system is to a particular environment, the less adaptable it is to new environments.

This has large effects on requirements. If one requirement is to have a database that runs on a specific system, a hidden cost is now attached where if that vendor goes out of business, the code can be next to impossible to move to another platform. You satisfied the requirement but paid a huge price. "An ounce of prevention is worth a pound of cure."

However, the same managers who scream for efficiency are the ones who tear their hair when told the cost of modifications. Conversely, the managers who ask for generalized and easily modified programs are wont to complain when they find out how slow and spacious there programs turn out to be. We must be adult about such matters: neither psychology nor magic is going to help us to achieve contradictory goals simultaneously. Asking for efficiency and adaptability in the same program is like asking for a beautiful and modest wife. Although beauty and modesty have been known to occur in the same woman, we'll probably have to settle for one or the other. At least that's better than neither.

Be carefule what you wish for, you just might get it...

Efficiency

Moreover, with the cost per unit of computation decreasing every year and cost per unit of programming increasing, we have long since passed the point where the typical installation spends more money on programming than it does on production work. [like the cost of machine time, ed.] This imbalance is even more striking when all the work improperly classified as "production" is put under the proper heading of "debugging."

Even 30 years ago, the cost of development was the largest cost in the field of programming. If you don't have a manager who understands the challenges and complexities of software development, you are facing a two-front war, 1) the challenge of writing software that works in the time required, using the resources allotted and, 2) a struggle to keep a clueless, however well-meaning, boss from making things worse through ignorance much less incompetence.

Summary

Questions to ponder:
1. Does the program meet specifications? Or, rather, how well does it met specifications?
2. Is it produced on schedule, and what is the variability in the schedule that we can expect from particular approaches?
3. Will it be possible to change the program when conditions change? How much will it cost to make the change?
4. How efficient is the program, and what do we mean by efficiency? Are we trading efficiency in one area for inefficiency in another?


I can here the pointy-haired-boss now, "Why are you asking so many (implied -stupid-) questions? I don't understand why you're making it so hard, just go write it already!"

This is a good place to point out that repeating the same behavior and expecting a different outcome is one of the definitions of insanity.

Next is Chapter 3 - How Can We Study Programming

Review - The Psychology of Computer Programming

I got a "new" book in the mail a few weeks ago and finally had time to crack it open, "The Psychology of Computer Programming" by Gerald M. Weinberg published in 1971. There is a Silver Anniversary edition that recently came out but I wanted a cheap copy in hardback so mine came from an online used book site. Why read a computer programming book that is almost as old as myself? Because it addresses the core issue of producing computer programs, the people who read, write, edit, maintain, or otherwise cuss at (other people's of course), code. People are the primary factor in all programs; before hardware, language, specifications, anything. If you don't understand how humans go about creating these abstract entities called programs, you will constantly be mystified by missed deadlines, frustrated by missing functionality, and stymied by endless scope creep. Enough of my pontificating, let's talk about the book.

Table of Contents

1. Programming as Human Performance




1. Reading Programs


2. What Makes a Good Program?


3. How Can We Study Programming?




2. Programming as a Social Activity




4. The Programming Group


5. The Programming Team


6. The Programming Project




3. Programming as an Individual Activity




7. Variations in the Programming Task


8. Personality Factors


9. Intelligence, or Problem-Solving Ability


10. 1Motivation, Training, and Experience




4. Programming Tools




11. Programming Languages


12. Some Principles for Programming Language Design


13. Other Programming Tools




5. Epilogue

"Computer programming is a human activity." A pretty bold thesis from the intro to Part I. Is there a mystique to programming? "Either you can program or you cannot. Some have it; some don't." Both quotes give you a good idea as to what is in this book, tackled expertly by Mr. Weinberg.

Chapter 1 - Reading Programs

Some years ago, when COBOL was the great white programming hope, one heard much talk of the possibility of executives being able to read programs. Weith the perspective of time, we can see that this claim was merely inteded to attract the funds of exectutives who hoped to free themselves from bondage to their programmers. Nobody can seriously have believed that executives could read programs. Why should they? Even programmers do not read programs.

I hear a similar story from when assemblers were introduced - as in, "With the development of assemblers, we won't need programmers anymore!" I believe similar statements have been propagated by hordes of 4th generation language and CASE salesmen.

"Programming is, among other things, a kind of writing."

This is not a very mainstream view in the programming world even though we work with constructs which are literally "languages".

"We shall need a method of approach for reading programs, for, unlike novels, the best way to read them is not always from beginning to end. They are not even like mysteries, where we can turn to the penultimate page for the best part -- or like sexy books, which we can let fall open to the most creased pages in order to find the warmest passages. No, the good parts of a program are not found in any necessary places -- although we will later see how we can discover crucial sections for such tasks as debugging and optimization. Instead, we might base our reading on a conceptual framework consisting of the origin of each part. In other words, as we look at each peice of code, we ask ourselves the questions, 'Why is this piece here?'"

The author begins examining a section of PL/I code showing how certain machine, language, and human limitations influence a program. Machine limitations like lack of memory to hold the entire problem set in memory at the same time causing the use of two loops instead of one. Language limitations like the lack of an end-of-file indicator which forces the operators to include a special character (or card in the 1970's punch-card centric world) and causes the programmer to account for the special character in code. Programmer limitations like not really understanding array notation of the language, in this instance PL/I. As a program is modified over time, machines change, languages are updated, and programmers come and go.

"And so, some years later, a novice programmer who is given the job of modifying this program will congratulate himself for knowing more about PL/I than the person who originally wrote this program. Since that person is probably his supervisor, an unhealthy attitude may develop -- which, incidentally, is another psychological reality of programming life which we shall have to face eventually."

Some programs have inscrutable logic like the use of special characters which are ordinarily invalid; "magic numbers" used for interim states for some long-forgotten problem.

"Once the patch was made, it worked so well that everyone forgot about it -- more psychology -- and there it sat until unearthed many years later by two archeologist programmers."

To different versions of the same PL/I program are compared, the first includes many limitations of which we've already spoken, the second much improved by removal of the previous limitations. Of the comparison,

"When we look at the difference between Figures 1-1 and 1-4, we might begin to believe that very little of the coding that is done in the world has much to do with the problems we are trying to solve"

Would we be any better off if we could use the latest code (Figure 1-4) as our spec?

"Specifications evolve together with programs and programmers. Writing a program is a process of learning -- both for the programmer and the person who commissions the program. Moreover, this learning takes place in the context of a particular machine, a particular programming language, a particular programmer or programming team in a particular working environment, and a particular set of historical events that determine not just the form of the code but also what the code does!

In a way, the most important reason for stufying the process by which programs are written by people is not to make the programs more efficient, more compact, cheaper, or more easily understood. Instead, the most important gain is the prospect of getting from our programs what we really want -- rather than just whatever we can manage to produce in our fumbling, bumbling way"

Hopefully the reason to study a 30+ year-old book is apparent; those who fail to learn from history are bound to repeat it.

Let's study historical mistakes so we can make our own new mistakes instead of repeating mistakes of programmers long-retired.

The rest of the book is just as rich in lessons we can still take to heart. By approaching the book chapter by chapter, I'm hoping to increase my chances of getting all the way through without writing a 50-page article all at once which would never get finished.

Here is the link for the next chapter: Chapter 2 - What Makes a Good Program