>It could spur a minimalist information driven trend.
it might if anybody would actually use it, but that seems unlikely if it doesn't have CSS or JS support. Chrome and Firefox are already really damn fast if you give them a simple site without much css or js. people just need to actually make sites like that.
> people just need to actually make sites like that.
This is the problem I see with initiatives like Gopher and Gemini.
The people causing the problems with the web are not the people who will listen to these initiatives.
They are banks, advertisers, FAANG, everyone who is fine making money off the standard Chrome-IE-Edge crowd and barely even care about Firefox support.
Both Gemini support and a sane subset of HTTP / HTML require the same level of dedication that I could bring, but no commercial site will. Well, to be blunt, minimal HTTP / HTML is a lot easier. I can keep my same hyper backend, Firefox client, Nginx for TLS termination, curl and libcurl and pycurl, lots of tools that will only work on HTTP.
Plus, minimal HTTP 1.1 is not that hard to implement, so I think they're mostly attacking the wrong part of the stack while also cutting out useful performance features like QUIC or pipelining or caching.
Back in the days of Netscape 4 (I think), this is what I did. I’m not sure if it was CSS or just some custom browser applied styling defaults, but I found it difficult to read pages in all different styles so I had the browser make them all the same.
There are ‘readability’ plugins and services for browsers that attempt to provide this service for current sites. These days it’s a bit more complicated than just overriding a few styles to make a page ‘standardised’.
The plugins / services work for maybe 99% of pages I look at. Unfortunately I don’t know how to view the whole web through such a lens, without having to activate for each page.
CSS and JS are not in opposition to "information driven". Javascript is not used merely for content-free flashy pages, and it isn't going away anytime soon. If and when it does, it'll be because something better replaced it, not because a million web developers woke up one morning and said "why don't we stop using most of the capabilities of modern browsers".
If, hypothetically, browsers stopped offering powerful scripting capabilities, the app makers of the world would not suddenly say "I guess we'll make static webpages", they'll say "here's how to install our all-powerful unsandboxed application". Powerful scripting on the web means more applications running in safe sandboxes.
Browsers enable websites to be very user hostile the way it is now. Of course developers wont opt for static over dynamic, unless they suspect users might want that, which could happen in at least some niches.
Most webpages, even simple ones that do not strictly need to be designed to require js, will entirely fail to render (blank page) without js. The situation has become a lot worse in the last 2 years or so.
It’s a requirement for a general purpose, modern browser, unfortunately, even one without a bunch of bells and whistles.
This is the case even for many municipal or government sites, to say nothing of business products/services/vendors. You can’t use the web as a private citizen for normal things like banking or civic participation without js.
If it is purely for speed, then there is no need for a new browser. Simply disabling JS makes websites crazy fast. I run a documentation site for my framework. It is already pretty lean, but disabling JS makes it super fast, even though my server is cheap. Unfortunately, most modern websites display a blank page without JS, so what we really need is a change in developer attitude, not a new browser (again, assuming that you are only concerned about speed).
JS is too useful to ditch entirely. I think it would be interesting to design a language that deterministically uses compute resources, and then limit web pages to a certain amount of them.
Couple that with the idea I've been toying with for a while about a new html standard (html6?, core HTML?) that only accepts loading JS from a common repository of utility code (think useful stuff like autocomplete, partial page reload etc) and it could improve the web a lot if we got sites to use it.
In a better world, JS used to implement convenient web page features like rendering LaTeX or syntax highlighting would be implemented inside the web browser, while JS used to implement web apps would not exist and web apps would just be apps.
I like my apps to run in a sandbox, and the web is the best sandbox we have. My standard reaction to "would you like to download our app" is "no, stay in your box".
We have. They run on the web, in tabs. Or, with PWAs, they look a lot more like native applications, and still run in a safe sandbox. (There are also Android and iOS apps, which are less ideal and less portable.) Why reinvent it in a less portable, less sandboxed, historically insecure manner? People have tried, and the result never ends up as useful, functional, or secure as the existing web sandbox.
When I browse the web, I know the browser puts me in control, and keeps applications contained. The only kind of app that I know will have comparable sandboxing is a PWA. Anything with a comparable amount of control will look like a web browser, and we already have the web.
If you want the world to change, you have to offer something better.
There is great technology out there for app sandboxing. Recent Windows versions let you instantly spin up a virtual machine to run unknown apps in - I'd say that's safer than a browser and there is no reason why it can't be as convenient.
You speak of the web as an app platform - I'm not opposed to some platform like that existing (and they do exist, just look at your OS), I just think that we made a mistake when we turned the browser into one. Now we mix together hypertext and code and have so much weird legacy to maintain, not to mention the performance issues.
I think the point you are missing is that we should have 2 distinct things: applications and web pages. I should not have to run untrusted code to read a blog post, a news page or the latest PR release. I run with JS off by default and there are basic web pages that don't work with JS for example https://www.bbc.com/news not sure why without JS the layout gets messedup but the content loads.
Sure if you have a nice application that is interactive, fine use a PWA, Electron or whatever you want but for showing plain text and iamges a subset of html and css is enough
it might if anybody would actually use it, but that seems unlikely if it doesn't have CSS or JS support. Chrome and Firefox are already really damn fast if you give them a simple site without much css or js. people just need to actually make sites like that.