Inspired by a recent Twitter thread (and long-running debate anyway): are HTML and CSS programming languages?
Is it gatekeep-y to say no?
The answer is yes, it is gate-keepy, but go off
Discuss!
For further actions, you may consider blocking this person and/or reporting abuse
Oldest comments (94)
No, but it's not a value judgement, they just don't have control flow. They're powerful languages, just not programming languages. They're specialized tools designed for specific purposes.
Yes, CSS + HTML are programming languages. Not only is this scenario technically Turing complete, I think these software fit a lot of the principles of the purpose of programming languages: source code in, magic out.
You can create some of the most incredible things coding html against a browser. No, it’s not as flexible as another programming language in terms of practical possibilities, but it’s a highly specialized language for doing really interesting things.
Distinguishing between programming languages and markup languages when nobody asked you to is unnecessary gatekeeping. If you want me to talk about that distinction, than sure.
The notion of what makes computers go bleep and bloop has always been under re-construction. If you could possibly debate whether something could be a programming language, you might as well call it a programming language because.
Yes, it's "source goes in, magic out", but still constrained to a specific context. The type of magic that comes out, while very cool and very vast, is still a small subset of the magic that can come out of something more general-purpose.
I don't understand why it's gatekeeping to acknowledge this distinction. I don't agree that calling something "not a programming language" makes it "lesser than a programming language", but just helps us understand in more specific terms what the tool is and isn't. Why is using multiple categories of language when we talk about our tools dismissive?
I agree - good categorization of the languages we have is important.
I think the real problem is when people think or act as if they are superior to others because they know a certain language or category of languages and others don't.
Of course - those people don't tend to be worth your time anyway.
I agree, but I think it’s more truer that markup languages are a subset of programming language vs a different category.
And in this case I think the gatekeeping comes into play when the topic of whether or not these things are programming languages is somewhat unimportant.
So whether the answer is yes or no, pointing it out when it’s unneeded for the conversation is gatekeeping, or could be perceived that way in our imperfect online lack of tone translation.
In a sense, the true gatekeepers have ruined it for anyone coming in with good intentions.
The rest of this thread has me convinced of this as well now.
I agree, but in this case it was the conversation! You've got a point, though, the people with emotions about this sort of thing aren't usually coming to these discussions in good faith but for some sort of one-upmanship. At the end of the day, of course it doesn't actually matter - just build things. I still think it's useful to classify languages like this, in the interest of making well-informed choices about the right tool for the right job, but that sentence is so uncontroversial I don't imagine anyone disagrees.
Thanks for your response!
Is markdown a programming language?
That's another great point! IMO, markdown is not a programming language because it's more for styling text - but that's me coming from a writing background and now I'm thinking about how markdown functions and is built. It's more complicated than I thought at first 🤔
Seems to me you're rubbing up on the same issue as with HTML. It's a markup language used to program how to generate some output!
Yep, they're declarative programming languages, you can tell the browser what it should do, but you can't tell the browser how it should do it, for that you need an imperative language (whichever one chooses, I'm not going to open this can of worms now).
I think the new hot take should be that prolog isn't a programming language.
is HTML and CSS turing complete?
Odd that no one has pointed this out
Everyone is focusing on the "language" part, which without a doubt CSS and HTML are. But what about the "programming" part? Well we don't have a definition there so that bit is a lot more difficult. It becomes a matter of what we "feel" like that term means to us.
At the base level it means making a system do what you want it to do. Of course in that case, so is writing an excel spreadsheet, or drag-and-dropping, or customizing your OS with new colors, or creating a new web app account, or even just calling up IT and telling them to fix things "or else".
That feels wrong.
For some, "programming" means "general-purpose programming"- that you not only can, but realistically might, sit down and code some algorithms in it. In that case sure, HTML and CSS are not programming, and neither is Brainfuck, or Arnold, or the more basic usages of Prolog.
You can't really fault those people for being bewildered at the resistance to what to them seems like such an obvious statement. For them, "HTML and CSS are not programming languages" is not gatekeeping, it's an attempt to educate on what seems to them to be an important distinction.
So why can it be so upsetting to hear it?
I suppose there is another definition we can go with: "Things that might be a central part of the job for someone who works professionally as a programmer". In that case, not only of course are HTML and CSS programming, but hearing that phrase - for someone who is primarily engaged in HTML and CSS - sounds like a denial of their entire career and track.
Conclusion? Maybe we all need to be aware when using terms without strict definitions that they might mean different things to different people.
By this logic, SVG, PDF, and arguably even JPEG are programming languages.
No, they are not. That's why Javascript exists, a HTML page without a true programming language is as static as an image, there is not control flow or any dynamism like other programming languages.
You can have control flow in HTML and CSS by using radio buttons (and event without them). You can create a FSM/automaton with HTML and it will work as well as with JavaScript. It may not be practical, but it can be done.
I understand that you are talking about like changing the style of something based on the radio button, right? Together they can do such thing, but each one of them cannot be considered a programming language, HTML or CSS alone cannot do much as far as I know. Unless a FSM/automaton can be done exclusively with HTML, can you tell more about it? I didn't know about that.
That's a good point. A FSM/automaton could be built using only HTML with links and IDs/anchors... Again, it wouldn't be practical or pretty, but it could be done. But then someone will claim that it is really a simulation, or that they don't like it.
I help out with an after-school program where, early in the curriculum, we teach the students to make HTML "choose your own adventure" games. (I made this one). It's useful for that, at least :P
(We use separate pages rather than links to fragments on the same page.)
And they are roughly like FSMs, though I think of FSMs as having an input vocabulary that's separate from the states. For instance, with a dial-type combination lock, you can pick any number on any step. With a tree of links, you can think of each page's links as representing a
0, a1, etc., but it's not "the same"0leading out of each page. (You can link to the same destination from multiple places, but that's just going to the same state.) Hopefully I'm making sense. Maybe that lack of a separate input vocabulary just means they're more like Mealy machines than Moore machines? I'm not sure.They also feel like decision trees to me, FWIW. It could be that there's some formal correspondence between the two concepts, but it's sadly not my area :P
You may be interested in the HTML/CSS game The Mine -- the top code comment makes it sound like it might just be a big FSM made out of checkboxes. I haven't tried to confirm that though.
That’s what I was referring: each page is a state, each link is a transition, and the input is the user clicking on a link. And that can be achieved with HTML even without CSS. (Although CSS will make it prettier 😊)
Personally, I would say yes, but my definition of a programming language is more fluid. For instance, you could ask a similar question for esoteric languages like brainf*ck or data languages like yml or json. As long as the content of the text is being interpreted/ compiled by a computer, it’s a programming language.
Now, there are definitely levels to this as some languages can accomplish a lot more than others, but that’s a different discussion.
I've read from some sources that CSS3 + HTML5 is Turing Complete. Someone with the nickname elitheeli proved this with a demo of a Rule 110 cellular automaton which is Turing complete. There is not javascript involved, just a html and css which is quite surprising when you open the repo.
While very cool, isn't this an argument against? Neither language is Turing complete on their own, only in combination.
You are right and I feel the same. They aren't Turing complete on their own, only when combined. I'd have to look more into it.
Simply typed lambda calculus is not Turing complete either...
Right, because (as I understand it) you always get to normal form and terminate. It's not exactly a programming language either, though, right?
Yes. Expressions in typed lambda always terminate (you can't create Y-combinator, for example).
Why not? It's not very comfortable to use, but you can write a program in it, to calculate something
Heh, fair enough :)
I guess my eventual conclusion is that I'm wrong, it doesn't make sense to delineate HTML and CSS as something different, but also as others have noted that the distinction itself isn't all that useful. Calling something one category has no bearing on what you can or cannot do with it.
Scientific approach - given enough evidence one changes opinion
Gotta love a good ol' fashioned discuss!
HTML is a "[M]arkup [L]anguage".
CSS is a "[S]tyle [S]heet language".
SQL is a "[Q]uery [L]anguage".
JavaScript is a "Programming language".
HTML, CSS, JavaScript and SQL are "Computer languages".
Yes, they are.
There's no legitimate definition of programming language you can give that would exclude them. I consider any unambiguous language used to control a computer to be a programming language.
Even with narrow definitions, generally HTML and CSS will both fit them, if strained hard enough.
This was a great read. I think you've changed my mind, thank you for posting it.
I don't know Ruby, but using Ruby in anything doesn't defeat the purpose of being CSS only? Like, if he uses something other than CSS, it is not CSS-only anymore.
So it is like any other web application that has a programming language with a server that serves HTML and CSS?
I like to think that CSS, HTML and ETC.. are programming languages specific not to PC but to some program. Like writting a program on minecraft with redstone. You see, CSS and HTML are equivalent programming languages to BROWSER, you are actually programming how the page is going to be displayed by a browser. Different browsers will respond not the same way to your code. the insane need of some people to say that it is not a programming language have nothing to do with CSS or HTML being or not being something, but reflection of the weakness of a weak programmer. If you do something that change a behaviour, that's programming.
Some comments may only be visible to logged-in visitors. Sign in to view all comments.