What Is Code? (2015)

Given that we probably all live in a virtual simulation of some sort, isn't everything code? A more interesting question would be, "what isn't code?"

qsort2• 4 years ago

Ex falso, quodlibet sequitur.

corsac• 4 years ago

I'm saving this comment so that next time someone uses "begging the question" incorrectly, I can show them what it really means.

TuringTest• 4 years ago

The argument that we likely live in a simulation is similar to the the cosmological argument that allegedly proves the existence of god; and it rests on the same unsound assumption, namely that the existence of entities in reality follows from the rational arguments that your mind can imagine about those entities.

Sound approaches to knowledge must operate in the opposite direction, drawing insights from observation of phenomena existing in the world.

Assuming that the world is as we deduce it to be, without then proceeding to check that the world behaves in accordance with our deduction, is a failure that delayed the advance of knowledge until the use of the scientific method became widespread.

galaxyLogic6• 4 years ago

A great all around introduction to programming.

It poses an interesting question which I think it doesn't really answer well: "What Is the Relationship Between Code and Data?"

I can easily understand what a "program" is. It is a set of instructions we give to the computer for it to do something we want it to do.

But what is "data"?

tejtm• 4 years ago

But what is "data"?

A set of instruction activators we give the computer

Oh and that easily understood set of instructions are also data

qsort2• 4 years ago

Any information (including code) that can be stored by the abstract machine under consideration.

Jtsummers2• 4 years ago

Data are instructions for the computer interpreted in the context of a program. Or no program in the case of instructions which can be interpreted directly by the computer.

That is, data are programs, programs are data. Just need to establish the environment in which to execute/process them.

galaxyLogic6• 4 years ago

Right, except what got me thinking is that "programs" can be conceptually seen as a sequence of "instructions" to the computer. Whereas it is more difficult to think of data as "instructions". Data is more like a model of the world, than a set of "instructions".

js2• 4 years ago

The Harvard architecture makes the separation clearer:

https://en.wikipedia.org/wiki/Harvard_architecture

galaxyLogic6• 4 years ago

And Turing machine? Can it modify its own tape?

Tomis022• 4 years ago

But what is "data"?

It's a sequence of bits. You take reality (or whatever you want) and represent it as bits. That sequence of bits is called "data".

atoav• 4 years ago

In the computer science world yes. But the word also exists outside of that context (and did so long before computers).

Data is the plural of datum. A datum is a piece of information. Of course not any piece information is automatically a datum, as that noun implies a certain structured nature or a normative perspective on that piece of information.

Certain information only became commonplace because it became a datum first. Before using family names as an identifier many people had no family names, they were identified by their first name and by the places they grew up in.

hurril• 4 years ago

I would say that data defines as something that might convey information, might in the sense that there are one or more interpretations of it that are meaningful to something. This sentence has plenty of open ended terms, however.

Code is data that can be executed, provided that there is a something that offers an interpretation of it. It think we must claim that this interpretation is meaningful to something.

Executability is a semantic in the context of interpretation, which ties it all back into what data is again. Something something Apply/Eval.

So more open ended terms.

akomtu• 4 years ago

If code is a relationship between the next state and the current state, then this state is data.

galaxyLogic6• 4 years ago

Well said. Code stays as it is, typically. Data changes.

Of course code can modify its own 'code". But that makes it much more difficult to reason about its behavior. It is good to have some "invariants". If we can assume that code does not change except when the human author changes it, then it becomes much easier to reason about the changes the program may cause to the data.

iammjm• 4 years ago

There is no difference between code and data. Data is code. Data is code that retains its identity.

itwrangler• 4 years ago

and there you have an unanswerable question! :-) at an atomic level in computing, as others have said, it's a stream of ones and zeroes.. after that everything else is implementation, i.e how to interpret, use, store etc. In fact it's one of the "bug bears" in computing that there are so many "data format standards" (used loosely not definitionally)..

or the age old joke: "The good thing about (data) standards is there are so many of them!"

tejuis2• 4 years ago

Data is information about things and code is information about how to transform data. Every program transforms input data to output data. Code is interpreted by the computer. Data and code are encoded to bits for the digital computer.

galaxyLogic6• 4 years ago

That sounds enlightening to me.

Code is interpreted by the computer, Data is interpreted by Code !!!

But when we say "code is interpreted by computer" we are really saying "Code is interpreted by Code" right? Meaning the interpreter or compiler.

So how come this snake doesn't eat its own tail?

The answer, the MAGIC of computers, is that at the lowest level somehow Hardware is able to interpret Code .

thundergolfer• 4 years ago

If anyone has a the print issue of this I’d like to buy it. One was on EBay a little while ago for about $60 I think

jrimbault• 4 years ago

I was also looking for one (well, two) yesterday... even sent bloomberg a mail to ask if backorder was still possible (they said no and to check amazon/ebay/etc).

klse• 4 years ago

baby don't hurt me

DeathArrow• 4 years ago

Code is a context free grammar. If we want to be more restrictive, we can call it a finite state automaton.

bmitc3• 4 years ago

This is a somewhat myopic and outdated description of what "code" or "software" is. I'm writing an article on this, but software is a much broader idea and domain than just getting computers to do something. At a minimum, it includes getting a computer to do something, communicating between humans and not humans and computers, and encoding a domain. All this adds up to software being "executable knowledge", which makes it far more powerful than thinking of it as just bits executing on a machine. Thinking of it this way also highlights all the problems with software, because the process of system design and encoding knowledge is downright hard and not something humans excel at. Getting a computer to do something is rarely the hard part aside from specific instances where pure performance is necessary.

It isn't surprising to me that the best software developers I have met have been those who either had no computer science degree or got it after having some degree or major interest in another field, typically a liberal arts or science field. Liberal arts are actually the study of systems in disguise because of the need to connect together various ideas, events, etc. that do not necessarily have clear delineations or relationships.

the_af4• 4 years ago

Are you equating "computer science" with "coding"? Because it's really not. CS studies systems, complexity, requirements, domain knowledge, etc.

I'm not arguing against people from other fields bringing valuable insight; I'm disputing that CS doesn't also involve the study of complex heterogeneous systems or that it is about "coding".

Seems like you're describing a bootcamp or coding school instead.

TheOtherHobbes• 4 years ago

CS doesn't. At least not in any useful way.

Systems Analysis and Knowledge Engineering typically assume corporate/industrial philosophies and goals.

Broader human considerations - political, social, cultural - are very peripheral in CS. They get some token attention but they're always subservient to business values.

The idea that software is primarily about human relations of all kinds is novel and rather subversive.

bmitc3• 4 years ago

No. Why?

kubanczyk• 4 years ago

I didn't really read beyond that reflection:

This is real. A Scrum Master in ninja socks has come into your office and said, “We’ve got to budget for apps.” Should it all go pear-shaped, his career will be just fine.

It's funny how organizations manage not to fall apart with so many people with so different motivations. For example, an SWE can despair in 2022 because they work on Openstack bugfixing or on mainframe JCL2 batch jobs, but here this exec worries about his position being undermined by some software project.

ChrisMarshallNY• 4 years ago

This story reminds me of the old Parable of The Toaster: http://www.solipsys.co.uk/new/TheParableOfTheToaster.html

lll-o-lll• 4 years ago

I know it when I see it.

redog• 4 years ago

TIL that I should specialize in security research.

ytwombly• 4 years ago

First, bear in mind this isn't written for a technical audience. Nevertheless, it's worth the time if you finish it.

Case in point: when this was first published, I was just starting out my career writing software, and knew very little about the industry I was entering. 7ish years later, I still think about it often.

What Is Code? (2015)

Related Stories