Hacker News Reader

Comment by dvogel | Hacker News Reader

When I initially read the htmx documentation I was confused because it kept talking about a hypermedia client. The context clues suggested they were referring to htmx but my brain kept saying "isn't the browser the hypermedia client?" Eventually it sank in that htmx is an extension of the hypermedia client. When I first tried to use htmx I experienced a lot of discomfort regarding areas where htmx feels non-standard, such as redirects in the hx- readers on a 200 response. Once I understood that htmx is explicitly trying to move the boundary of the hypermedia client a lot of that discomfort melted away.

View on HN

Yes htmx is an augmentation of the browser, specifically through enhancing HTML by way of JS. The idea is that JS frameworks became popular due to the lack of additional hypermedia controls which are the basis for how agents (users) interact with websites through hypermedia clients (browsers).

> Once I understood that htmx is explicitly trying to move the boundary of the hypermedia client a lot of that discomfort melted away.

What do you mean by "moving the boundary of hypermedia client"?

HTMX tries to claim that hypermedia to only applies to HTMX because something something browsers and html.

Simply put, anything that talks HTTP and understands responses from a server is a hypermedia client to an extent.

You can create a client that only accepts base32-encoded cat gifs, and that will be a hypermedia client (in its infancy).

> Simply put, anything that talks HTTP and understands responses from a server is a hypermedia client to an extent.

No, anything that understands hypermedia responses is a hypermedia client.

Cat GIFs are not hypermedia so a cat GIF viewer is not a hypermedia client

Maybe I'm overly grumpy this morning but words do actually have meanings that we can look up and refer to

> words do actually have meanings that we can look up and refer to

It's odd to insist on strict word choice when transferring GIF images using the hypertext transfer protocol.

It's not though? Gifs are media, but they are not hypermedia because they don't support hypertext (can't link to to other media).

As such, they're ancillary sub-resources to hypermedia but not themselves hypermedia.

If you're going to get that picky (and please be aware I'm only doing this for the sake of the argument) media can never be hypermedia in the absence of the client. HTML opened in notepad is just text. Cat GIFs, rendered in the correct client, would absolutely be hypermedia (you could inline link data as QR codes, if you felt like being perverse).

Hypermedia starts with the client, not with the file format.

You are the one being overly picky. Of course given this gif-hypermedia-client your gif is hypermedia. But the client you mentioned above is not.

How you transfer the data is irrelevant BTW. I don't get why you include that in your argument.

I agree that a hypermedia can't act properly within the uniform interface constraint, without a hypermedia client, that is, you can't have a hypermedia system without a proper hypermedia client:

https://htmx.org/essays/hypermedia-clients/

https://hypermedia.systems/hypermedia-components/

On the other hand, there is a real difference between plain text and HTML (or HXML, don't shoot!) which is a subset of text with additional concepts layered on top of it. This is akin to how JSON (or XML) is not hypermedia, but can be used to create hypermedia such as Siren or HXML.

So I still think it makes sense to discuss if a media is or is not hypermedia without reference to the client, whereas it doesn't make sense to claim it is being used as hypermedia unless it is being consumed by a properly written hypermedia client. To make my thinking concrete, I believe Siren would continue to be hypermedia, even if it wasn't be consumed properly by a client, but then also you could not describe that pairing as a hypermedia system. (This is one reason I focus on the systemic nature of hypermedia, rather than solely on hypermedia formats)

Semantic nitpicking perhaps, but then hypermedia discussions appear to tend to invite this sort of thing.

> HTML (or HXML, don't shoot!) which is a subset of text with additional concepts layered on top of it. This is akin to how JSON (or XML) is not hypermedia

So, HTML is different from plain text because it "has concepts layered on top of text" where as JSON is not hypermedia despite "having concepts layered on top of text". And the only reason is because you said so.

> So I still think it makes sense to discuss if a media is or is not hypermedia without reference to the client

Then JSON is just as much hypermedia as HTML. Both are structured text unusable without a specific client to display them or work with them.

> Semantic nitpicking perhaps, but then hypermedia discussions appear to tend to invite this sort of thing.

They only invite them because of your insistence on calling only HTML the "natural hypermedia" etc.

> And the only reason is because you said so.

No, the reason is because HTML qua HTML has hypermedia controls and JSON qua JSON does not. Recall that, before I pointed out the widely used and accepted definition of hypermedia controls, and in particular that links and forms are hypermedia controls, you did not understand that concept, so you might spend some time quietly reflecting on that idea. It may help clarify things for you.

> Then JSON is just as much hypermedia as HTML. Both are structured text unusable without a specific client to display them or work with them.

As I have said and written previously (https://htmx.org/essays/hypermedia-clients/, https://hypermedia.systems/hypermedia-components/) I agree that a hypermedia client is necessary for a properly functioning hypermedia system that adheres to the uniform interface. However, I think that there is a good argument that Siren, for example, is hypermedia, even if it isn't being consumed correctly, just as I think HTML is hypermedia, even if someone is screen scraping it (i.e. not using a hypermedia client to consume it).

I don't think you can call those uses a hypermedia system, but I also don't think that changes the fact that the underlying formats, Siren & HTML, are hypermedia, due to the fact that they have hypermedia controls. That might be a subtle distinction, but I think it is a valid one. Again, perhaps as you reflect more on this concept, new to you, of hypermedia controls, the distinction will become easier to understand.

> They only invite them because of your insistence on calling only HTML the "natural hypermedia" etc.

I'm very sorry you that feel that way.

I would call HTML, "a natural hypermedia", rather than "the natural hypermedia". I would also call HXML & Siren natural hypermedia, due to the presence of hypermedia controls (a concept new to you) in their specifications.

Yeah. PDFs would be a better example. They can link to other media :)

But I wouldn't be surprised if there's a crazy project somewhere using GIFs as a way to render HTML pages with clickable links :D

Yeah, I went overboard with the example. The issue is that HTMX tries to take over the concept of hypermedia as if it means only HTMX and whatever HTMX is doing :)

Htmx posits that current browsers aren't "truly" hypermedia since only anchor tags and forms can initiate GET/POST requests. It is more of a tech demo showing what client with ANY tag being able to do requests would look like.

That's why whether it is library/framework is besides the point. The author posits that these features should be in the spec, and tries as closely as possible to show what something might look like if we had it in the spec

> The author posits that these features should be in the spec

Does he? The author pretends that his library is what hypertext and hypermedia are as envisioned by Time Berners-Lee and Roy Fielding, and that his approach is the only true representation of both. And that's about it. Nothing about "this should be in the spec"

> The author pretends that his library is what hypertext and hypermedia are as envisioned by Time Berners-Lee and Roy Fielding, and that his approach is the only true representation of both.

Does he? Evidence or it didn't happen.

They literally have an entire book written to contort those definitions to mean HTMX

Then it should be easy for you to find at least one passage that demonstrates that.

You are missing the spirit of the whole thing. HTMX is a polyfill for the future state of browsers.

It's not, and it's not even pretending to be

Um, yes friend, that’s exactly what it’s trying to be. Carson has said numerous times that in an ideal world, the html spec would evolve to the point the htmx becomes redundant. It’s not about htmx or any library/framework - it’s about extending html.

If that doesn’t convince you, then I’ve got nothing and suggest we both just go and enjoy some lazer horse/buffalo/pickle memes in the htmx twitter account

you keep saying this despite the fact that I explicitly include https://hyperview.org as an example of another hypermedia in https://hypermedia.systems. I am very open to other types of hypermedia and often refer people to “RESTful Web Cliebts” by mark amundsen (https://www.oreilly.com/library/view/restful-web-clients/978...) to learn how build them.

> I am very open to other types of hypermedia

Of course you're not. And I already pointed it out to you elsewhere. Your entire writing and marketing revolves around one idea, and one idea only: HTML is "natural hypermedia", and everything else is not.

At this point you are very loudly and publicly grinding your axe to the point that you’re telling someone to their face that they don’t understand their own viewpoint. Putting aside briefly the insanity and futility of that, it makes for a bad experience for literally everyone else.

OK, this is just completely unreasonable of you. HTML is a natural hypermedia in that it has native hypermedia controls. JSON & XML are not natural hypermedia because they do not, however hypermedia controls can be added on top of them, as in the case of HXML/hyperview, which, again I include in my book on hypermedia systems.

There are many other hypermedias, such as Siren, which uses JSON as a base, and I have never claimed otherwise. Mark Amundsen, perhaps the worlds expert on hypermedia, wrote the forward to my book, Hypermedia Systems, and found nothing objectionable and much worthwhile in it.

I hate to be rude but you didn't understand, or refused to acknowledge, the basic meaning and usage of the term 'hypermedia control' until I cited a W3C document using it. While I certainly understand people can dislike the conceptual basis of htmx, its admittedly idiosyncratic implementation or the way we talk about it, at this point I have tried to engage you multiple times in good faith here and have been rewarded with baseless accusations of things I haven't said and don't believe.

At this point, to be an honest person, you need to apologize for misrepresenting what I am saying multiple times to other people. It is dishonest and it makes you a liar, over something as dumb as a technical disagreement.

> I hate to be rude but you didn't understand, or refused to acknowledge, the basic meaning and usage of the term 'hypermedia control' until I cited a W3C document using it.

Just because you were correct in one small detail (citing a 2019 standard retrofitting definitions for the use in RDF etc.) doesn't make you correct in the grand scheme of things.

> with baseless accusations of things I haven't said and don't believe.

I literally quoted your own words at you.

> you need to apologize for misrepresenting what I am saying multiple times to other people.

I will not apologize for things that I even quoted from your own writing and words.

> one small detail

the presence of hypermedia controls is a defining characteristic of a hypermedia format

> I literally quoted your own words at you.

You took an essay I wrote in which I defined the term HDA specifically to contrast with the term SPA in the context of web development and spun that into an imagined philosophy where HTML is the only hypermedia in the world. You persisted in this after I pointed out that I included HXML in my book on hypermedia, and gave a clear definition of what I consider the defining characteristics of hypermedia & clarified specific examples of other formats that are hypermedia.

You have confused "X is A" with "Only X is A" and then, when large gaps in your understanding of hypermedia have been brought to your attention, you have dismissed them as small details.

> I will not apologize

I did not expect you to.

At this point I think I have taken goodwill as far as it can go. I encourage any other readers who have made it to this point in this hellthread to simply read my essays & perhaps my book, and judge them on their own merits:

https://htmx.org/essays

https://hypermedia.systems

Reply on news.ycombinator.com