If you haven’t heard of Alan Kay you’ll likley have heard one of his many famous quotes, the most popular, likely his 1971 gem:
The best way to predict the future is to invent it.
But for the unfamiliar, Alan has one of the most illustrious careers in computer science who’s been awarded both a Kyoto Prize and a Turing award for his work on Object Oriented Programming. He’s also one of the pioneers of Personal Computing, the Graphical User Interface, Object Oriented Programming and the inventor of Smalltalk - one of the most influential languages of all time.
Much of the writing of Alan Kay holds an insightful read into how the power of context (pdf) where he recalls the one-of-a-kind environment at Xerox PARC and ARPA which were harnessed around “visions rather than goals” and “funded people, not projects” which as a result had attracted brilliant minds together cultivating different view points essential for progress where he considers:
Point of view is worth 80 IQ points
His thoughts when reflecting back on that time:
The ARPA/PARC history shows that a combination of vision, a modest amount of funding, with a felicitous context and process can almost magically give rise to new technologies that not only amplify civilization, but also produce tremendous wealth for the society.
And that they did, which saw through the creation of the impressive array of technologies invented at PARC which were the foundations for much of personal computing and programming as we know it today, including:
- Laser Printers
- Object Oriented Programming / Smalltalk
- Personal Computers
- Ethernet / Distributed Computing
- GUI / Mouse / WYSIWYG
Whilst ARPA was responsible for the ARPANET phenomena we refer to as the Internet.
On Software Engineering
What’s interesting is that Alan still believes The Computer Revolution Hasn’t Happened Yet (pdf) where software engineering is progressing inversely to Moore’s Law which sees hardware capacity increasing each year whilst software becoming un-necessarily bloated, potentially
caused by weak and difficult-to-scale ideas and tools, laziness, lack of knowledge, etc.
This ongoing situation was eloquently captured in the funny one-liner:
What Andy giveth, Bill taketh away
Refering to Andy Grove the then CEO of Intel and Bill Gates of Microsoft.
Improving the current state of software development was the motivation behind the STEPS Toward The Reinvention of Programming (pdf) research project aimed at achieving a “Moore’s Law” Leap in Expressiveness by reducing
the amount of code needed to make systems by a factor of 100, 1000, 10,000, or more.
His enlightening 2011 talk on Programming and Scaling (video) re-iterates on this, suggests software engineering has stalled and is becoming a lost science that has not been able to advance at the same pace as hardware or other science and engineering disciplines. That large code-bases are akin to a garbage dumping ground, reaching the point where no one is able to comprehend the 100M LOC’s used to create Vista and Word, which should only be a fraction of their size. He gives the Internet and TCP/IP, LISP Interpreters, Nile (Math DSL for Vector Graphics) and OMeta (OO PEG) (pdf) as examples of elegant software with profound impact and minimal code-bases.
He refers to the Internet (TCP/IP) as being one of the few large-scale software projects that was properly engineered where its level of complication is balanced by its level of complexity which was achieved in < 20k LOC that acts behaves like a living, dynamic system able to handle billions of nodes that has never been stopped after it was turned on in September, 1969. As this was so rare many people don’t consider the Internet as normal software made by man:
The Internet was done so well that most people think of it as a natural resource like the Pacific Ocean, rather than something that was man-made. When was the last time a technology with a scale like that was so error-free? The Web, in comparison, is a joke. The Web was done by amateurs.
Nile/OMeta DSL is another example showing how the same functionality could be achieved to build a system from scratch in thousands of lines of code rather than the millions found in comparable commercial versions.
I was intrigued to find out whether Alan had any further insights on what we could do to improve the current state of software development and if there was something missing from our current languages, approaches and tools…
On Object Oriented Programming
My first point of research was to find out his thinking behind his original vision for Object Oriented Programming which were influenced by his background as a Micro Biologist:
I thought of objects being like biological cells and/or individual computers on a network, only able to communicate with messages
and a Mathematician:
My math background made me realize that each object could have several algebras associated with it, and there could be families of these, and that these would be very very useful.
Who after studying Lisp was heavily influenced by it’s extreme late-binding and powerful meta capabilities:
The second phase of this was to finally understand LISP and then using this understanding to make much nicer and smaller and more powerful and more late bound understructures.
who from that point on became a strong advocate for dynamic languages for the future of software engineering (pdf):
Until real software engineering is developed, the next best practice is to develop with a dynamic system that has extreme late binding in all aspects.
primarily as they are more easily able to be changed:
Late binding allows ideas learned late in project development to be reformulated into the project with exponentially less effort than traditional early binding systems (C, C++, Java, etc.)
with the potential for applying incremental on-the-fly changes and faster iteration times:
One key idea is to keep the system running while testing and especially while making changes. Even major changes should be incremental and take no more than a fraction of a second to effect.
that’s currently lacking in statically-typed languages:
If you’re using early-binding languages as most people do, rather than late-binding languages, then you really start getting locked in to stuff that you’ve already done. You can’t reformulate things that easily.
Surprisingly his thoughts behind OOP were limited to only this narrow scope:
OOP to me means only messaging, local retention and protection and hiding of state-process, and extreme late-binding of all things. It can be done in Smalltalk and in LISP. There are possibly other systems in which this is possible, but I’m not aware of them.
Which didn’t even require inheritance, which is not like we know it today:
I’m sorry that I long ago coined the term “objects” for this topic because it gets many people to focus on the lesser idea.
Where the big missing piece lacking in mainstream typed OO languages today is:
The big idea is “messaging”
He advocates focus should instead be on messaging and the loose-coupling and interactions of modules rather than their internal object composition:
The key in making great and growable systems is much more to design how its modules communicate rather than what their internal properties and behaviors should be.
And finds static type systems too crippling:
I’m not against types, but I don’t know of any type systems that aren’t a complete pain, so I still like dynamic typing.
Other popular languages embracing Smalltalk’s message-passing and late-binding and having implemented its message-based doesNotUnderstand construct include: forwardInvocation in Objective-C, method_missing in Ruby and more recently noSuchMethod in Google’s Dart.
Tear it down and build something better
Alan has an interesting theory on the best way to advance Computer Science:
I believe that the only kind of science computing can be is like the science of bridge building. Somebody has to build the bridges and other people have to tear them down and make better theories, and you have to keep on building bridges.
as there are benefits to starting from scratch every few months, which saw him wanting to kill his own Smalltalk and starting again:
It’s well known that I tried to kill Smalltalk in the later ’70s. There were a few years when it was the most wonderful thing in the world. It answered needs in a more compact and beautiful way than anything that had been done before. But time moves on. As we learned more and got more ambitious about what we wanted to do, we realized that there are all kinds of things in Smalltalk that don’t scale the way they should—for instance, the reflection stuff that we had in there. It was one of the first languages to really be able to see itself, but now it is known how to do all levels of reflection much better—so we should implement that.
On Messaging
As a longtime proponent of messaging for remote services, I’ve been benefiting from the inherent advantages in message-based services for some time. So I was particularly interested in Alan’s thoughts of the subject:
I could see that a more comprehensive basis could be made by going all the way to thinking of efficient whole virtual machines communicating only by messages. This would provide scaling, be a virtual version of what my research community, ARPA-IPTO was starting to do with large scale networking, and also would have some powerful “algebraic” properties (like polymorphism).
He had this to say of the Internet - which he considers is “possibly the only real object-oriented system in working order”:
To me, one of the nice things about the semantics of real objects is that they are “real computers all the way down (RCATWD)” – this always retains the full ability to represent anything. The old way quickly gets to two things that aren’t computers – data and procedures – and all of a sudden the ability to defer optimizations and particular decisions in favour of behaviours has been lost. In other words, always having real objects always retains the ability to simulate anything you want, and to send it around the planet. … And RCATWD also provides perfect protection in both directions. We can see this in the hardware model of the Internet (possibly the only real object-oriented system in working order). You get language extensibility almost for free by simply agreeing on conventions for the message forms. My thought in the 70s was that the Internet we were all working on alongside personal computing was a really good scalable design, and that we should make a virtual internet of virtual machines that could be cached by the hardware machines. It’s really too bad that this didn’t happen. If ‘real objects’ are RCATWD, then each object could be implemented using the programming language most appropriate for its intrinsic nature, which would give new meaning to the phrase ‘polyglot programming.’
Where thinking of things as objects and messages would’ve enabled an un-precedented level of interoperability, reducing the un-necessary friction and allowing seamless communication between languages that would’ve enabled a whole new polygot world in which you would be free to choose the best language to implement each domain.
On RPC, and how it distorts developers mindsets in architecting and building systems:
The people who liked objects as non-data were smaller in number, and included myself, Carl Hewitt, Dave Reed and a few others – pretty much all of this group were from the ARPA community and were involved in one way or another with the design of ARPAnet->Internet in which the basic unit of computation was a whole computer. But just to show how stubbornly an idea can hang on, all through the seventies and eighties, there were many people who tried to get by with “Remote Procedure Call” instead of thinking about objects and messages. Sic transit gloria mundi.
Carl Hewitt being the inventor of the Actor Model and Dave Reed who was involved in the early development of TCP/IP and the designer of UDP.
The last latin phrase translates to “Thus passes the glory of the world” - expressing his dismay on what might have been.
On LISP
Something that’s continually re-iterated in his papers and interviews is his high-regard of LISP which he says had given him One of the deepest insights which he considers to be the:
greatest single programming language ever designed
That all comp-sci grads should learn:
Most people who graduate with CS degrees don’t understand the significance of Lisp. Lisp is the most important idea in computer science.
Sadly we didn’t have this boat at my University - but as learning new programming paradigms and exploring different views are great ways to improve ones developer skills, I had set out to learn some LISP!
In my next post I’ll explore the history and beauty of LISP, unlock its secret magical properties Alan was mesmerized with and explain why it’s still one of the most powerful and easiest programming languages to learn today!
The Unknown side of Alan Kay
I wanted to end this post talking about a little unknown nugget of what the primary motivation behind the amazing research teams Alan was apart of, whose pioneering efforts in advancing computer-science started about 40 years ago with the goal:
to help children - and thus humanity - to learn and absorb “science in the large”.
Dating as far back as 1968, when he first met Seymour Papert the Author of Logo (a language optimized for educational use).
That is, looking at how best to use technology to empower children’s ability to learn. One way was showing them how to build and simulate their own real-world models in software and letting them experiment, tinker, measure and observe their behavior.
One of the goals was to change how children were traditionally educated where instead of dispensing them with facts, to instead encourage them to observe real-world behavior for themselves and get teachers to think of children as thinking beings in their own right
rather than as “defective adults who have to be fixed by education”
This effective method of learning was termed Constructionist learning and defined by Papert in his 1987 Constructionism publication: A New Opportunity for Elementary Science Education
Alan envisaged the Dynabook concept in 1968 and later published the concept in a 1972 paper titled A personal computer for children of all ages. This was the project that drove both the creation of the graphical user interface and the development of Smalltalk in as early as 1972.
In 1995 he helped create the Etoys computing environment a media rich authoring environment (built ontop of Squeak / Smalltalk) that was aimed at helping children learn powerful ideas by constructing them.
In 2001 he setup the Viewpoints Research Institute, a nonprofit public benefit organization to improve “powerful ideas education” for the world’s children.
In 2006-2007, the One Laptop per Child (OLPC) project had pre-installed the Squeak Etoys multimedia authoring system on all their OLPC XO-1 educational machines.
This was a fun-fact to discover that one of the most innovative era’s in computing history was a by-product of the pursuit to improve childrens education! - makes me feel a little better about the world :)
Translations
- Chinese: blog.jobbole.com/35740/