Backwards compatibility but not for Python

Programming languages change over time. (So do the natural, human-spoken languages. But let's stick to the programming languages.)

Most languages are designed and changes are carefully constructed to avoid breaking older programs. This is a tradition from the earliest days of programming. New versions of FORTRAN and COBOL were introduced with new features, yet the newer compilers accepted the older programs. (Probably because the customers of the expensive computers would be very mad to learn that an "upgrade" had broken their existing programs.)

Since then, almost every language has followed this tradition. BASIC, Pascal, dBase (and Clipper and XBase), Java, Perl, ... they all strove for (and still strive for) backwards compatibility.

The record is not perfect. A few exceptions do come to mind:

• In the 1990s, multiple releases of Visual Basic broke compatibility with older versions as Microsoft decided to improve the syntax.
• Early versions of Perl changed syntax. Those changes were Larry Wall deciding on improvements to the syntax.
• The C language changed syntax for the addition-assignment and related operators (from =+ to +=) which resolved an ambiguity in the syntax.
• C++ broke compatibility with a scoping change in "for" statements. That was its only such change, to my knowledge.

These exceptions are few. The vast history of programming languages shows compatibility from old to new versions.

But there is one language that is an exception.

That language is Python.

Python has seen a number of changes over time. I should say "Pythons", as there are two paths for Python development: Python 2 and Python 3. Each path has multiple versions (Python 2.4, 2.5, 2.6, and Python 3.4, 3.5, 3.6, etc.).

The Python 3 path was started as the "next generation" of Python interpreters, and it was started with the explicit statement that it would not be compatible with the Python 2 path.

Not only are the two paths different (and incompatible), versions within each path (or at least the Python 3 path) are sometimes incompatible. That is, some things in Python 3.6 are different in Python 3.7.

I should point out that the changes between versions (Python 3.6 and 3.7, or even Python 2 and 3) are small. Most of the language remains the same across versions. If you know Python 2, you will find Python 3 familiar. (The familiarity may cause frustration as you stumble across one of the compatibility-breaking changes, though.)

Should we care? What does it mean for Python? What does it mean for programming in general?

One could argue that changes to a programming language are necessary. The underlying technology changes, and programming languages must "keep up". Thus, changes will happen, either in many small changes or one big change. The latter often is a shift away from one programming language to another. (One could cite the transition from FORTRAN to BASIC as computing changed from batch to interactive, for example.)

But that argument doesn't hold up against other evidence. COBOL, for example, has been popular for transaction processing and remains so. C and C++ have been popular for operating systems, device drivers, and other low-level applications, and remain so. Their backwards-compatible growth has not appreciably diminished their roles in development.

Other languages have gained popularity and remain popular too. Java and C# have strong followings. They, too, have not been hurt by backwards-compatibility.

Python is an opportunity to observe the behavior of the market. We have been working on the assumption that backwards-compatibility is desired by the user base. This assumption may be a false one, and the Python approach may be a good start to observe the true desires of the market. If successful (and Python is successful, so far) then we may see other languages adopt the "break a few things" philosophy for changes.

Of course, there may be some demand for languages that keep compatibility across versions. It may be a subset of the market, something that isn't visible with only one language breaking compatibility, but only visible when more languages change their approach. If that is the case, we may see some languages advertising their backwards-compatibility as a feature.

Who knows? It may be that the market demand for backwards-compatibility may come from Python users. As Python gains popularity (and it is gaining popularity), more and more individuals and organizations build Python projects, they may find Python's approach unappealing.

Let's see what happens!

Variables are... variable

The nice (and sometimes frustrating) thing about different programming languages is that they handle things, well, differently.

Consider the simple concept of a "variable". It is a thing in a program that holds a value. One might think that programming languages agree on something so simple -- yet they don't.

There are four actions associated with variables: declaration, initialization, assignment, and reference (as in 'use', not a constrained of pointer).

A declaration tells the compiler or interpreter that a variable exists often specifies a type. Some languages require a declaration before a variable can be assigned a value or used in a calculation; others do not.

Initialization provides a value during declaration. This is a special form of assignment.

Assignment assign a value, and is not part of declaration. It occurs after the declaration, and may occur multiple times. (Some languages do not allow for assignment after initialization.)

A reference of a variable is the use the value, to compute some other value or provide the value to a function or subroutine.

It turns out that different languages have different ideas about these operations. Most languages follow these definitions; the differences are in the presence or absence of these actions.

C, C++, and COBOL (to pick a few languages) all require declarations, allow for initialization, and allow for assignment and referencing.

In C and C++ we can write:

int i = 17;
i = 12;
printf("%d\n", i);

This code declares and initializes the variable i as an int with value 17, then assigns the value 12, then calls the printf() function to write the value to the console. COBOL has similar abilities, although the syntax is different.

Perl, Python, and Ruby (to pick different languages) do not have declarations and initialization but do allow for assignment and reference.

In Ruby we can write:

i = 12
puts i

Which assigns the value 12 to i and then writes it to the console. Notice that there is no declaration and no type specified for the variable.

Astute readers will point out that Python and Ruby don't have "variables", they have "names". A name is a reference to an underlying object, and multiple names can point to the same object. Java and C# use a similar mechanism for non-trivial objects. The difference is not important for this post.

BASIC (not Visual Basic or VB.NET, but old-school BASIC) is a bit different. Like Perl, Python, and Ruby it does not have declarations. Unlike those languages, it lets you write a statement that prints the value of an undeclared (and therefore uninitialized and unassigned) variable:

130 PRINT A

This is a concept that would cause a C compiler to emit errors and refuse to supply an executable. In the scripting languages, this would cause a run-time error. BASIC handles this with grace, providing a default value of 0 for numeric variables and "" for text (string) variables. (The AWK language also assigns a reasonable value to uninitialized variables.)

FORTRAN has an interesting mix of capabilities. It allows for declarations but does not require them. Variables have a specific type, either integer or real. When a variable is listed in a declaration, it has the specified type; when a variable is not declared it has a type based on the first letter of its name!

Like BASIC, variables in FORTRAN can be referenced without being initialized. Unlike BASIC, it does not provide default values. Instead it blissfully uses whatever values are in memory at the location assigned for the variable. (COBOL, C, and C++ have this behavior too.)

What's interesting is the trend over time. Let's look at a summary of languages and their capabilities, and the year in which they were created:

Languages which require declaration but don't force initialization

COBOL (1950s)
Pascal (1970s)
C (1970s)
C++ (1980s)
Java (1995)
C# (2000s)
Objective-C (1990s)

Languages which require declaration and require initialization (or initialize for you)

EIFFEL (1980s)
Go (2010)
Swift (2010)
Rust (2015)

Languages which don't allow declarations and require assignment before reference

Perl (1987)
Python (1989)
Ruby (1990s)

Languages which don't require (or don't allow) declaration and allow reference before assignment

FORTRAN (1950s)
BASIC (1960s)
AWK (1970s)
PowerShell (2000s)

This list of languages is hardly comprehensive, and it ignores the functional programming languages completely. Yet it shows something interesting: there is no trend for variables. That is, languages in the 1950s required declarations (COBOL) or didn't (FORTRAN), and later languages require declaration (Go) or don't (Ruby). Early languages allow for initialization, as do later languages. Early languages allow for use-without-assignment, as do later languages.

Perhaps a more comprehensive list may show trends over time. Perhaps splitting out the different versions of languages will show convergence of variables. Or perhaps not.

It is possible that we (that is, programmers and language designers) don't really know how we want variables to behave in our languages. With more than half a century of experience we're still developing languages with different capabilities.

Or maybe we have, in some way, decided. Its possible that we have decided that we need languages with different capabilities for variables (and therefore different languages). If that is the case, then we will never see a single language become dominant.

That, I think, is a good outcome.

The IRS made me a better programmer

We US taxpayers have opinions of the IRS, the government agency tasked with the collection of taxes. Those opinions tend to be strong and tend to fall on the "not favorable" side. Yet the IRS did me a great favor and helped me become a better programmer.

The assistance I received was not through employment at the IRS, nor did they send me a memo entitled "How to be a better programmer". They did give me some information, not related to programming, yet it turned out to be the most helpful advice on programming in my career.

That advice was the simple philosophy: One operation at a time.

The IRS uses this philosophy when designing the forms for tax returns. There are a lot of forms, and some cover rather complex notions and operations, and all must be understandable by the average taxpayer. I've looked at these forms (and used a number of them over the years) and while I may dislike our tax laws, I must admit that the forms are as easy and understandable as tax law permits. (Tax law can be complex with intricate concepts, and we can consider this complexity to be "essential" -- it will be present in any tax form no matter how well you design it.)

Back to programming. How does the philosophy of "one operation at a time" change the way I write programs?

A lot, as it turns out.

The philosophy of "one operation at a time" is directly applicable to programming. Well, my programming, at least. I had, over the years, developed a style of combining operations onto a single line.

Here is a simplified example of my code, using the "multiple operations" style:

Foo harry = y.elements().iterate().select('harry')

It is concise, putting several activities on a single line. This style makes for shorter programs, but not necessarily more understandable programs. Shorter programs are better when the shortness is measured in operations, not raw lines. Packing a bunch of operations -- especially unrelated operations -- onto a single line is not simplifying a program. If anything, it is making it more complex, as we tend to assume that operations on the same line are somehow connected.

I changed my style. I shifted from multi-operation lines to single operation lines, and I was immediately pleased with the result.

Here's the example from above, but with the philosophy of one operation per line:

elements = y.elements()
Foo harry = nil
elements.each do |element|
  harry = element if element.name == 'harry'

I have found two immediate benefits from this new style.

The first benefit is a better experience when debugging. When stepping through the code with the debugger, I can examine intermediate values. Debuggers are line-oriented, and execute the single-line version all in one go. (While there are ways to force the debugger to execute each function separately, there are no variables to hold the intermediate results.)

The second benefit is that it is easier to identify duplicate code. By splitting operations onto multiple lines, I find it easier to identify duplicate sequences. Sometimes the code is not an exact duplicate, but the structure is the same. Sometimes portions of the code is the same. I can refactor the duplicated code into functions, which simplifies the code (fewer lines) and consolidates common logic in a single place (one point of truth).

Looking back, I can see that my code is somewhat longer, in terms of lines. (Refactoring common logic reduces it somewhat, but not enough to offset the expansion of multiline operations.)

Yet the longer code is easier to read, easier to explain to others, and easier to fix. And since the programs I am writing are much smaller than the computer's capabilities, there is little expense at slightly longer programs. I suspect that compilers (for languages that use them) are optimizing a lot of my "one at a time" operations and condensing them, perhaps better than I can. The executables produced are about the same size as before. Interpreters, too, seem to have little problem with multiple simple statements, and run the "one operation" version of programs just as fast as the "multiple operations" version. (This is my perception; I have not conducted formal time trials of the two versions.)

Simpler code, easier to debug, and easier to explain to others. What's not to like?

Optimizing in the wrong direction

Back in the late 200X years, I toyed with the idea of a new version control system. It wasn't git, or even git-like. In fact, it was the opposite.

At the time, version control was centralized. There was a single instance of the repository and you (the developer) had a single "snapshot" of the files. Usually, your snapshot was the "tip", the most recent version of each file.

My system, like other version control systems of the time, was a centralized system, with versions for each file stored as 'diff' packages. That was the traditional approach for version control, as storing a 'diff' was smaller than storing the entire version of the file.

Git changed the approach for version control. Instead of a single central repository, git is a distributed version control system. It replicates the entire repository in every instance and uses a sophisticated protocol to synchronize changes across instances. When you clone a repo in git, you get the entire repository.

Git can do what it does because disk space is now plentiful and cheap. Earlier version control systems worked on the assumption that disk space was expensive and limited. (Which, when SCCS was created in the 1970s, was true.)

Git is also directory-oriented, not file-oriented. Git looks at the entire directory tree, which allows it to optimize operations that move files or duplicate files in different directories. File-oriented version control systems, looking only at the contents of a single file at a time, cannot make those optimizations. That difference, while important, is not relevant to this post.

I called my system "Amnesia". My "brilliant" idea was to, over time, remove diffs from the repository and thereby use even less disk space. Deletion was automatic, and I let the use specify a set of rules for deletion, so important versions could be saved indefinitely.

My improvement was based on the assumption of disk space being expensive. Looking back, I should have known better. Disk space was not expensive, and not only was it not expensive it was not getting expensive -- it was getting cheaper.

Anyone looking at this system today would be, at best, amused. Even I can only grin at my error.

I was optimizing, but for the wrong result. The "Amnesia" approach reduced disk space, at the cost of time (it takes longer to compute diffs than it does to store the entire file), information (the removal of versions also removes information about who made the change), and development cost (for the auto-delete functions).

The lesson? Improve, but think about your assumptions. When you optimize something, do it in the right direction.

Cloud computing is repeating history

A note to readers: This post is a bit of a rant, driven by emotion. My 'code stat' project, hosted on Microsoft Azure's web app PaaS platform, has failed and I have yet to find a resolution.

Something has changed in Azure, and I can no longer deploy a new version to the production servers. My code works; I can test it locally. Something in the deployment sequence fails. This is a test project, using the free level of Azure, which means no monthly costs but also means no support -- other than the community help pages.

There are a few glorious advances in IT, advances which stand out above the others. They include the PC revolution (which saw individuals purchasing and using computers), the GUI (which saw people untrained in computer science using computers), and the smartphone (which saw lots more people using computers for lots more sophisticated tasks).

The PC revolution was a big change. Prior to personal computers (whether they were IBM PCs, Apple IIs, or Commodore 64s), computers were large, expensive, and complicated; they were especially difficult to administer. Mainframes and even minicomputers were large and expensive; an individual could afford one if they were an enormously wealthy individual and had lots of time to read manuals and try different configurations to make the thing work.

The consumer PCs changed all of that. They were expensive, but within the range of the middle class. They required little or no administration effort. (The Commodore 64 was especially easy: plug it in, attach to a television, and turn it on.)

Apple made the consumer PC easier to use with the Macintosh. The graphical user interface (lifted from Xerox PARC's Alto, and later copied by Microsoft Windows) made many operations and concepts consistent. Configuration was buried, and sometimes options were reduced to "the way Apple wants you to do it".

It strikes me that cloud computing is in a "mainframe phase". It is large and complex, and while an individual can create a an account (even a free account), the complexity and time necessary to learn and use the platform is significant.

My issue with Microsoft Azure is precisely that. Something has changed and it behaves differently than it did in the past. (It's not my code, the change is in the deployment of my app.) I don't think that I have changed something in Azure's configuration -- although I could have.

The problem is that once you go beyond the 'three easy steps to deploy a web app', Azure is a vast and intimidating beast with lots of settings, each with new terminology. I could poke at various settings, but will that fix the problem or make things worse?

From my view, cloud computing is a large, complex system that requires lots of knowledge and expertise. In other words, it is much like a mainframe. (Except, of course, you don't need a large room dedicated to the equipment.)

The "starter plans" (often free) are not the equivalent of a PC. They are merely the same, enterprise-level plans with certain features turned off.

A PC is different from a mainframe reduced to tabletop size. Both have CPUs and memory and peripheral devices and operating systems, but are two different creatures. PCs have fewer options, fewer settings, fewer things you (the user) can get wrong.

Cloud computing is still at the "mainframe level" of options and settings. It's big and complicated, and it requires a lot of expertise to keep it running.

If we repeat history, we can expect companies to offer smaller, simpler versions of cloud computing. The advantage will be an easier learning curve and less required expertise; the disadvantage will be lower functionality. (Just as minicomputers were easier and less capable than mainframes and PCs were easier and less capable than minicomputers.)

I'll go out on a limb and predict that the companies who offer simpler cloud platforms will not be the current big providers (Amazon.com, Microsoft, Google). Mainframes were challenged by minicomputers from new vendors, not the existing leaders. PCs were initially constructed by hobbyists from kits. Soon after companies such as Radio Shack, Commodore, and the newcomer Apple offered fully-assembled, ready-to-run computers. IBM offered the PC after the success of these upstarts.

The driver for simpler cloud platforms will be cost -- direct and indirect, mostly indirect. The "cloud computing is a mainframe" analogy is not perfect, as the billed costs for cloud platforms can be inexpensive. The expense is not in the hardware, but the time to make the thing work. Current cloud platforms require expertise, and expertise that is not cheap. Companies are willing to pay for that expertise... for now.

I expect that we will see competition to the big cloud platforms, and the marketing will focus on ease of use and low Total Cost of Ownership (TCO). The newcomers will offer simpler clouds, sacrificing performance for reduced administration cost.

My project is currently stuck. Deployments fail, so I cannot update my app. Support is not really available, so I must rely on the limited web pages and perhaps trial and error. I may have to create a new app in Azure and copy my existing code to it. I'm not happy with the experience.

I'm also looking for a simpler cloud platform.

After Agile

The Agile project method was developed as an alternative (one might say, a rebuttal) of Waterfall. Waterfall was first, aside from the proto-process of "do whatever we want" that was used prior to Waterfall. Waterfall had a revolutionary idea: Let's think about what we will do before we do it.

Waterfall can work with small and large projects, and small and large project teams. If offers fixed cost, fixed schedule, and fixed features. Once started, a project plan can be modified, but only with change control, a bureaucratic process to limit changes in addition to broadcasting proposed changes to the entire team.

Agile, its initial incarnation, was for small teams and projects with flexible schedules. Schedule may be fixed, or may be variable; you can deliver a working product at any time. (Although you cannot know in advance which features will be in the delivered product.)

Agile has no no change control process -- or rather, Agile is all about change control, allowing revisions to features at any time. Each iteration (or "sprint", or "cycle") starts with a conversation that involved stakeholders who decide on the next set of features. Waterfall's idea of "think, talk, and agree before we act" is part of Agile.

So we have two methods for managing development projects. But two is an unreasonable number. In the universe, there are rarely two (and only two) of things. Some things, such as electrons and stars and apples, exist in large quantities. Some things, such as the Hope Diamond and our planet's atmosphere, exist as singletons. (A few things do exist in pairs. But the vast majority of objects are either singles or multitudes.)

If software management methods exist as a multitude (for they are clearly not a singleton) then we can expect a third method after Waterfall and Agile. (And a fourth, and a fifth...)

What are the attributes of this new methods? I don't know -- yet. But I have some ideas.

We need a management process for distributed teams, where the participants cannot meet in the same room. This issue is mostly about communication, and it includes differences in time zones.

We need a management process for large systems composed of multiple applications, or "systems of systems". Agile cannot handle projects of this size; waterfall has flaws with it.

Here are some techniques that I think will be in new management methods:

• Automated testing
• Automated deployment with automated roll-back
• Automated evaluation of source code (lint, Robocop, etc.)
• Automated recording (and transcribing) of meetings and conversations

It is possible that new methods will use other terms and avoid the "Agile" term. I tend to doubt that. We humans like to name things, and we prefer familiar names. "Agile" was called "Agile" and not "Express Waterfall" because the founders wanted to emphasize the difference from the even-then reviled Waterfall method.

The Waterfall brand was tarnished -- and still is. Few folks want to admit to using Waterfall; they prefer to claim Agile methods. So I'm not expecting a "new Waterfall" method.

Agile's brand is strong; developers want to work on Agile projects and managers want to lead Agile projects. Whatever methods we devise, we will probably call them "Agile". We will use "Distributed Agile" for distributed teams, "Large Agile" for large teams, and maybe "Layered Agile" for systems of systems.

Or maybe we will use other terms. If Agile falls out of favor, then we will pick a different term, such as "Coordinated".

Regardless of the names, I'm looking forward to new project management methods.

Predictions for tech in 2018

Predictions are fun! Let's have some for the new year!

Programming Languages

Java, C, and C# will remain the most popular languages, especially in large commercial efforts. Moderately popular languages such as Python and JavaScript will remain moderately popular. (JavaScript is one of the "three legs of web pages", along with HTML and CSS, so it is very popular for web page and front-end work.)

Interest in functional programming languages (Haskell, Erlang) will remain minimal, while I expect interest in Rust (which focuses on safety, speed, and concurrency) to increase.

Cloud and Mobile

The year 2017 was the year that cloud computing become the default for new applications, especially business applications. The platforms and tools available from the big providers (Amazon.com, Microsoft, Google, and IBM) make a convincing case. Building traditional web applications on in-house data centers will still be used for some specialty applications.

The front end for applications remains split between browsers and mobile devices. Mobile devices are the platform of choice for consumer applications, including banking, sales, games, and e-mail. Browsers are the platform of choice for internal commercial applications, which require larger screens.

Browsers

Chrome will remain the dominant browser, possibly gaining market share. Microsoft will continue to support its Edge browser, and it has the resources to keep it going. Other browsers such as Firefox and Opera will be hard-pressed to maintain viability.

PaaS (Platform as a Service)

The middle version of platforms for cloud computing, PaaS sits between IaaS (Infrastructure as a Service) and SaaS (Software as a Service). It offers a platform to run applications, handling the underlying operating system, database, and messaging layers and keeping them hidden from the developer.

I expect an increase in interest in these platforms, driven by the increase in cloud-based apps. PaaS removes a lot of administrative work, for development and deployment.

AI and ML (Artificial Intelligence and Machine Learning)

Most of AI is actually ML, but the differences are technical and obscure. The term "AI" has achieved critical mass, and that's what we'll use, even when we're talking about Machine Learning.

Interest in AI will remain high, and companies with large data sets will take advantage of it. Initial applications will include credit analysis and fraud analysis (such applications are already under development). The platforms offered by Google, Microsoft, and IBM (and others) will make experimentation with AI possible for many, although one needs large data sets in addition to the AI compute platform.

Containers

Interest in containers will remain strong. Containers ease deployment; if you deploy frequently (or even infrequently) you will want to at least evaluate them.

Big Data

The term "Big Data" will all but disappear in 2018. Like its predecessor "real time", it was a vague description of computing that was beyond the reach of typical (at the time) hardware and software. Hardware and software improved to the point that performance was good enough, and the term "real time" is now limited to a few very specialized situations. I expect the same for "big data".

Related terms, like "data science" and "analytics" will remain. Their continued existence will depend on their perceived value to organizations; I think the latter has secured a place, the former is still under scrutiny.

IoT

The "Internet of Things" will see a lot of hype in 2018. I expect a lot of internet-connected devices, from drones to dolls, from cameras to cars, and from bicycles to birdcages (really!).

The technology for connected devices has gotten ahead of our understanding, much like the original microcomputers before the IBM PC.

We don't know how to use connected things -- yet. I expect that we will experiment with a lot of uses before we find the "killer app" of IoT. Once we do, I expect that we will see a standardization of protocols for IoT devices, making the early devices obsolete.

Apple

I expect Apple to have a successful and profitable 2018. They remain, in my opinion, at risk of becoming the "iPhone company", with more than 80% of the income coming from phones. The other risk is from their aversion to cloud computing -- Apple puts compute power in its devices (laptops, tablets, phones, and watches) and does not leverage or offer cloud services.

The latter omission (lack of cloud services) will be a serious problem in the future. The other providers (Microsoft, Google, IBM, etc.) provide cloud services and development platforms. Apple stands alone, keeping developers on the local device and using cloud computing for its internal use.


These are my predictions for 2018. In short, I expect a rather dull year, focused more on exploring our current technology than creating new tech. We've got a lot of relatively new tech toys to play with, and they should keep us occupied for a while.

Of course, I could be wrong!