Clean or pure web?

I have a feeling to be an odd bird when it comes to web development. Not using frameworks, avoiding javascript libraries, using compiled instead of interpreted language. Using vanilla of everything. And being obsessed by speed and security. And trying to be as friendly as possible to the environment.

Until I found this article that goes beyond my oddity. Interesting reading for enhancing my own bird status.

My main reasons for thinking small is speed, environment, maintenance problem with frameworks and dependencies when it comes new versions. And small often means cheaper hosting etc.

What do you think of “clean or pure web”? Why spend weeks for something you can achieve in hours with shortcuts? Or is it only money that talks?

2 Likes

I think a lot of the answer you get depends on your… position within the field, is what i’m going to call it.

“web development” is a very, VERY broad church.

Me pottering around and developing a webpage for myself, or even for my job (which is not in web development, but a tech field; so my job-based development is in the utility/presentation sort of area. Side projects to make a tool that automates this thing or another.), in which i’m developing for myself, by myself, and it doesnt matter what the code looks like, as long as it gets the job done in some degree of time, is web development.

You’ve got solo developers, who do a job, but at the end of the job it will become someone else’s job later to pick up and read their code and maintain/adjust it. And that’s a different mentality for development, because now you’ve got to care about code styling, and making it understandable, readable, concise, and operate within the delivery specs of the customer. And that’s also web development.

The person or people writing an extension for a preexisting library have to work within confines set by their frameworks, but also to have that code accepted by the original creators and/or the community that surround it, will need to meet code requirements for their deliverables. And that’s also web development.

The guy who’s creating a page or a script in a team collaboration on a major corporation’s site, who has got to encompass the ideas of multiple people into a cohesive design and work with their team, who will have varying ideas about what to use and how to use it, while keeping within the corporation’s budget for time and expense, has entirely different concerns when writing code as to the ‘problems’ of frameworks. But it’s also web development.

It’s important to recognize that to each goes their need; there won’t be “an” answer. Or “a” reason for asking the question, for that matter.

5 Likes

Thanks. Just curious how developers in general thinks.

I agree that it depends and web development is a broad area. But reading Sitepoint questions there however most questions about “ordinary” web sites. So my question are hence targeted to creating normal web sites.

At the end of the day you make a choice about environment, sustainability, maintainability, SEO (!), UI, UX, economy (short and long term), etc. And the tools you chose often have an impact of several of these areas.

And which way the direction this community is headed? More pure or more framework?

Your answer makes me feel I am heading towards and thinking in an odd direction… :slight_smile:

1 Like

I dont think “the community” is headed in one direction. Which was the point I was trying to make.

It’s not a unanimous block.

What’s your definition of “ordinary” website?

Basically a corporate or personal site with a TLD.

The question was raised mainly because of this text from the above link. And I just want to get input how developers think. If it even is under consideration?

The web and app develop­ment landscape has been heavily influenced by a bandwagon effect, leading to the dominance of React, Angular and Vue over the past decade. This has fostered a prevailing belief that using a framework is essential for building any professional-grade application.

The Pure Web Foundation was created to change the narrative that you can’t possibly create a professional site or app without a framework.

I mean, first you have to stop and ask yourself whether the article’s premise is true and pure (small pun there i suppose). It’s an article on the Pure Web Foundation’s site, saying how there is a bandwagon effect, and that React, Angular, and Vue have “dominance”. (Note: The paragraph never says dominance over… what. Over sites with frameworks? The entire web?)

If you actually look at the article’s sources for those statements, though… their “prevailing belief” comes from two sources cited: An article on FreeCodeCamp that, if you read it, makes one of its major bulletpoints “Learn Vanilla Javascript first” (so…uh… about that dominance) and the other is…themselves.

I… frankly find the authors of the article unreliable narrators with a predefined agenda. So what i think is that the question is being asked without a basis.

Yes, we can agree that I am an odd bird :slight_smile:

Hi, you’re really committed to a minimalist approach in web development, which is definitely a refreshing perspective in today’s framework-heavy landscape. The focus on speed, security, and environmental impact is commendable, especially when so many solutions today come with a hefty price in terms of performance and resource use.

The concept of a “clean or pure web” can be subjective. Many developers have adopted frameworks for the sake of efficiency, rapid prototyping, or access to a vast ecosystem of plugins and components. However, these often come with their own set of complexities and maintenance challenges, which you’re clearly aware of.

Your approach of using vanilla tools can yield a leaner and more secure final product. Plus, it can offer more control and understanding of what your code is doing, which can be incredibly valuable in the long run. While some might rush for shortcuts, the depth of knowledge you gain by building from scratch is irreplaceable.

Ultimately, it’s about what aligns with your values and goals as a developer. If clean and efficient web development resonates with you, keep at it! There’s a niche community that appreciates this philosophy, and it can lead to innovative solutions that benefit both users and the environment.

1 Like

I like the idea of a “pure web”, but that will mean different things to different people. Heck, just disable JS on that page and the site is blank. IMHO, that’s just pure ****.

3 Likes

Pure to me is to me as simple and small as possible. Shut down the engine of the car will be more pure, but also less functional.

1 Like
  • Using a framework is fine
  • Using a library is fine…

What is not fine:

  • not understanding the core “vanilla” concepts behind either of those. So you know where their approach is appropriate and where it is not.
  • letting those frameworks and libraries drive how your product is developed as it can lock things in which are not in the best interest of the clients or the products.

I have a library we use here where I work which is being used and abused because the library wasn’t understood when it was selected, nor was the version of it being used monitored and maintained by the people who selected it. Now we have this nasty library which was used as the basis of our code base which is out of date but so deeply ingrained in how our product was developed, those of us who weren’t here when it was selected are struggling to divest ourselves of it, with little support from above since it’s what our clients are dependent on, and it doesn’t benefit anyone but the developers to refactor the codebase yet, though the developers are starting to see the breaking point coming where we’re going to have to refactor it out much sooner rather than later.

3 Likes

Agree. I have used a “framework” for a few decades. And found that handing over the control and design to the vendor may lead to reduced developer times initially, but also vendor created bugs that you have no control over. At the end the customers will suffer as you cannot fix the problem asap.

Hence I am trying (I may not be totally rigid) to avoid “plugins” or shortcuts that mostly become overhead and maintenance issues.

This analogy is a bit flawed from what @ralphm pointed out. No site should be broken without JS. It should be additive, not foundational.

That site is broken without JS. That’s not a car in your analogy. It’s a lawn ornament.

This is one of the most frustrating factors of modern development for those of us that cut our teeth in days gone by. The concept that it’s easier and cheaper to throw hardware (more processors/more storage/faster networks) than to develop something elegant and clean which is easier to maintain over the long haul.

It’s why we have these massive, breaking rewrites all the time from the big players. Because they don’t think about maintainability and extensability when things are built, so it’s easier to start from scratch and re-invent the wheel.

2 Likes

I have not found any way to fetch data and manipulate the site without Javascript. So without JS it is not functional to me. But not “broken”…

Even HTMX is Javascript behind the scenes.

Sorry but this…


is broken.

Any site which renders NOTHING without javascript enabled is broken. Especially a site like that which doesn’t even need javascript for rendering anything until the filter based on role, and even that I’m pretty sure I could render with some more modern css and not need js at all.

1 Like

I agree. But using SSR (mostly), I did not think of this.

Which is why the definition of “pure web” is critical to the point, which is what ralph pointed out in post 9.

“Simple” and “small” are good things. But you have to define “pure web”.

Is pure web -
no CSS? (“Simple” and “Small” would mean removing all of that…)
no JS? (Same…)

the "simplest" and "smallest" web would be an HTML document with no styling.

But obviously, you dont mean that. You mean “the simplest and smallest… in the particular set of features that I want to use”… which as soon as you introduce “I”, is no longer a concrete definition.

shrug

Should code be slim? Ideally.

Now ask and answer the question in a fuller form:

Should you reinvent the wheel to make the code slimmer than loading a framework? Depends. How many versions of the wheel are you reinventing? How long is it going to take? Do you have the time and resources to commit to that level of reinvention?

Suddenly it doesnt quite seem so idyllic, does it?

1 Like

All code is is a means to an end. The end goal of all code is to solve a problem. If there were no problem no code would be needed to solve it.

Therefore, it always troubles me when people impose something on code that is also a goal (here: make it pure, whatever that may mean). The problem I have with it is that developers then have two goals: solving the actual problem and staying within the limits of their self imposed goal. So then what happens when those two conflict? Will they tell the business the problem can’t be solved because it doesn’t rhyme with their self imposed goal? Would that make sense from a business perspective, who are paying people to solve their problems?

There is a reason consultants are constantly saying “it depends”…

1 Like

I would just like to state that i am not a consultant for sitepoint. :stuck_out_tongue:

Also

And i’d be very much more bored.

2 Likes

There are some we can agree on.

Frameworks, plugins, orms and libraries:

  1. …can make the life easier and more profitable.
  2. …both solve problems and create problems.
  3. …loads stuff you never need and creates extra latency.
  4. …may also adds an extra “language” upon the base language.
  5. …may cause maintenance problem later as new versions released or terminated.

What we do not seems to completely agree on:

  1. This extra latency consumes extra energy and more expensive servers
  2. Often it also slows down the site or app.
  3. More impact on the environment sooner or later
  4. There is a moving tipping point where “pure” or “shortcut” is more profitable. So profitable in short term can lead to unprofitable in long term and vice versa.
  5. Asking for help (forum, AI, vendors etc) may be harder and more costly for an unknown “plugin”

Short: It depends.

Thank you for all your inputs. It was interesting reading. I have got plenty of new inputs, so I think the moderators can close this thread IMO.