The end of the web browser

I thought that the point of this thread was to separate data from display. Take, for example, the weather forecast - the data may come from largely the same source, but may look very different depending on how it's delivered/received on the device. A phone browser is not the same thing as a desktop browser. Similar in functionality, but not the same thing. On a phone, you have the option to use a default weather app supplied by the phone manufacturer or a browser bookmark.

A native phone app will only request the data, receive it over HTTPS, and then use the phone's processing power to display it nicely. A browser has to also download the code to display the data, and phone browsers are generally designed to avoid that to the extent practical.
What extra "code" is this, that has to be downloaded? I have to admit that I still don't understand why a phone's web page browser cannot be called a "web browser." A laptop browser is not the same thing as a desktop browser, but we still call it a web browser. What, again, makes the phone device so special?
 
What extra "code" is this, that has to be downloaded?
HTML, images, CSS, Javascript - the web page that surrounds the raw data. The rendering instructions.

Suppose you can argue if HTML and CSS are "code".

The app will download the raw data and use local code/resources to display it.

Some (many?) apps are very thin wrappers over a web browser so won't be that much different.
 
HTML, images, CSS, Javascript - the web page that surrounds the raw data. The rendering instructions.

Suppose you can argue if HTML and CSS are "code".

The app will download the raw data and use local code/resources to display it.

Some (many?) apps are very thin wrappers over a web browser so won't be that much different.
Sure, HTML, CSS, and JS are "code" -- i.e., computer languages, interpreted by web browsers. I've been writing in these languages for years. But isn't it still true that "phone browsers" also interpret and execute the same languages? How else would they render our web pages? Forgive my thick-headedness, but I must be missing some kind of semantic fine point in this whole argument. It just doesn't make sense to me.

Edited to clarify: Again, why is it that a phone's web page browser cannot be called a "web browser?" What is it that puts it in a separate class by itself from the browsers used by larger computers?
 
Sure, HTML, CSS, and JS are "code" -- i.e., computer languages, interpreted by web browsers. I've been writing in these languages for years. But isn't it still true that "phone browsers" also interpret and execute the same languages? How else would they render our web pages? Forgive my thick-headedness, but I must be missing some kind of semantic fine point in this whole argument. It just doesn't make sense to me.

Edited to clarify: Again, why is it that a phone's web page browser cannot be called a "web browser?" What is it that puts it in a separate class by itself from the browsers used by larger computers?
I was replying specifically to the bit from astyle:

A native phone app

... but I missed this bit at the end:

and phone browsers are generally designed to avoid that to the extent practical.

So I see what you mean and I agree with you.

I don't think there's anything special about browsers on phones either (apart from Steve Jobs killing off Adobe Flash, and Apple forcing all browsers use its rendering engine, but that might be coming to an end.)
 
I always considered HTML and CSS "markup languages" because they don't compile into an executable program. They only define parameters.

Afaik, browsers on the phone are the same code as on the desktop though I've never thought about it before now.
 
I always considered HTML and CSS "markup languages" because they don't compile into an executable program. They only define parameters.

Afaik, browsers on the phone are the same code as on the desktop though I've never thought about it before now.

Valid point, particularly regarding CSS. JavaScript and PHP are not compiled either, yet, I still consider them to be computer languages, only, languages which are interpreted, rather than compiled. I started my data processing career programming in Microsoft's BASIC language, running on Tandy's Radio Shack TRS-80 microcomputers. Later, my 2nd post-college programming job was programming in BUSINESS BASIC for Nixdorf and Point 4 IRIS mini-computers. They were both interpreted rather than compiled, yet still marketed as computer languages. Machine language is not compiled, but interpreted directly and as-is, by the computer processor.

In my own headspace, owing to my background, I think of HTML as the "object code" produced by my PHP "programs" (... or should I call them "scripts?"), to be interpreted by the browser's HTML markup language interpreter. The semantic arguments regarding why JavaScript, PHP, and command shell interpreter language documents should be called "scripts" rather than "programs" still tend to elude me also. In my mind, if it's used to deliver instructions to a computer, it's a "programming language." HTML, like the old-style dumb terminal control codes, is used to deliver display formatting instructions, so it strikes me as, at least, a computer language subset.

As to the question of whether or not phone browsers should be considered to be a different or special class of web browsers, I've never thought about it either, before reading this thread.
 
HTML, images, CSS, Javascript - the web page that surrounds the raw data. The rendering instructions.
The browser downloads, for reading a newspaper I read, many megabytes for only displaying an article whose symbols make few hundreds of bytes.

Afaik, browsers on the phone are the same code as on the desktop though I've never thought about it before now.
But the design of web pages for the phone is completely different as for the PC. And this design is being imposed everywhere. Even if only the display is the difference, the effect is immense.
 
I have to admit that I still don't understand why a phone's web page browser cannot be called a "web browser."
When not in "desktop mode", phone browsers tend to render sites in non-standard ways. Often damaging the look or requiring specific babysitting in the form of media tags.

So if I open i.e an IRC client and pointed it towards a website, port 80 and it just displays a mess on the screen. Is that still a web browser?
 
The browser downloads, for reading a newspaper I read, many megabytes for only displaying an article whose symbols make few hundreds of bytes.


But the design of web pages for the phone is completely different as for the PC. And this design is being imposed everywhere. Even if only the display is the difference, the effect is immense.
In the 90s the page content that had to be delivered to Netscape's browsers was quite different than that delivered to Microsoft's Internet Explorer browser, but they were both still called "web browsers." I considered those display quirk differences to be much more technically burdensome than the formatting differences needed to display web content on a smaller sized display screen. Even in the 21st century I've had to do some browser sniffing in order to deal with different vendors' web browsers. Whereas Chrome, Firefox, Opera, and Safari have been fairly standards-compliant, Microsoft's browsers continue to be problematic, and in some ways the Edge browser is even more of a pain than Internet Explorer.
 
Whereas Chrome, Firefox, Opera, and Safari have been fairly standards-compliant, Microsoft's browsers continue to be problematic,
My argument is that the "development" is bringing the web browser and WWW to an end, you are concentrating too much on technical issues, so that you ignore the big picture.

And yes, it is a question of standards. That is what www and gopher made, a standard, before them, we had to login with telnet to different servers designed by each information giver. Now we are using different apps in the phone for things that were done with www.

There are for example jet "new banks " that offer online banking only on their phone apps, not anymore in the browser.
 
My argument is that the "development" is bringing the web browser and WWW to an end, you are concentrating too much on technical issues, so that you ignore the big picture.

And yes, it is a question of standards. That is what www and gopher made, a standard, before them, we had to login with telnet to different servers designed by each information giver. Now we are using different apps in the phone for things that were done with www.

There are for example jet "new banks " that offer online banking only on their phone apps, not anymore in the browser.

I wouldn't open an account at such a bank. I have a friend whose eyes make it difficult to use a small display and those tiny touch screen keyboards. It takes a certain amount of technical expertise to use assistive technologies like voice recognition. I don't think streaming services like Youtube, Vimeo, Disney+, HBO Max, and Netflix will be in any hurry to abandon the devices that allow them to display widescreen videos on laptop and desktop monitors. I suppose one could connect an external monitor, keyboard, and conventional mouse to a phone, but how practical is that? In an office environment? I don't foresee such things happening on any large scale, at least, not in my lifetime.
 
Vull : For purposes of this conversation, desktop/laptop are the same thing... yeah, a laptop's screen is smaller, but there are 17-inch laptops on the market...

Also, a phone's browser is pretty different from a desktop/laptop - due to size and processing power...
So if I open i.e an IRC client and pointed it towards a website, port 80 and it just displays a mess on the screen. Is that still a web browser?
There's a proliferation of rendering engines - there's Gecko by Mozilla, chromium by Google (and forked by Microsoft for Edge)... oh, and WebKit for Apple's Safari... and they all kind of need to take advantage of the underlying hardware. Standards help, but we still gotta keep up with the ever-evolving hardware and security issues.
Very soon most web pages will have 5 nested Linux kernels in VMs.
yeah, and the browsers will be tasked with downloading and running that truckload every single friggin' time... because everybody wants to do it their way, rather than let someone else (IETF, W3C, etc) call the shots :p
 
I don't foresee such things happening on any large scale, at least, not in my lifetime.
But the deteriorating quality of Web-Sites is in large scale. The proliferation of Apps that take the work of the browser and WWW also. Perhaps we get soon such apps for Windows and Mac, eventually some apps also for linux, but for no other OS. And with the apps no web sites more for such services, like the new banks.
 
But the deteriorating quality of Web-Sites is in large scale. The proliferation of Apps that take the work of the browser and WWW also. Perhaps we get soon such apps for Windows and Mac, eventually some apps also for linux, but for no other OS. And with the apps no web sites more for such services, like the new banks.
Such apps are already here: Maple/Mathematica/Matlab, Quicken, enterprise accounting apps, JMP, misc/orange3, devel/RStudio, and more... heck, even /bin/sh making use of /bin/fetch :p
 
Have you got any examples of what you mean by this?
Did some research, and got surprising results:
1676062066483.png

1676062122108.png

The desktop installer is about a 40-50 MB app that downloads close to 400 MB... and www/firefox tarball is close to 500 MB... that is in line with my experience... but over 1 GB for Android was a surprise... My guess is that the Android app contains rendering code that gets downloaded on demand by the desktop version...
 
It is a wonderful nettop, small, light, with an integrated modem that works without problem with OpenBSD and supports many frequencies. The only thing that makes problems are now web browsers. Why should a web browser be inflated to be like a virtual machine with own OS?
It is a potato computer and performs as such...
 
I never used it but there is "hugo"
Hugo was president of Venezuela until his death around 2013.

... because everybody wants to do it their way, rather than let someone else (IETF, W3C, etc) call the shots :p
I do it the W3C way...

The desktop installer is about a 40-50 MB app that downloads close to 400 MB... and www/firefox tarball is close to 500 MB... that is in line with my experience... but over 1 GB for Android was a surprise... My guess is that the Android app contains rendering code that gets downloaded on demand by the desktop version...
Curious as to what the download size is for Mac OS. My file: /var/cache/pkg/firefox-107.0_2,2.pkg is only about 59 MB. I'm thinking Microsoft's Firefox download is small because they are only downloading a wrapper that goes around their own internal web browsing software. At least, that's the way it was with Internet Explorer before Edge came along.
 
Did some research, and got surprising results: [...]
The desktop installer is about a 40-50 MB app that downloads close to 400 MB... and www/firefox tarball is close to 500 MB... that is in line with my experience... but over 1 GB for Android was a surprise...

Don't believe anything any one google result tells you.

I'm running the older FF for android (68.11.0), before its clearly deliberate destruction by the new google $upplied team, so I can't auto-update, but see the latest offering:

Version: 110.0
Size: 70.64MB
Required OS: Android 5.0 up.

My guess is that the Android app contains rendering code that gets downloaded on demand by the desktop version...

I've seen no evidence of anything like that. I suspect they changed languages.

My guess is that that the post about FF on android refers to the amount of memory in use by a running FF, without however specifying how many tabs are a) available and b) in use then or recently.

I have about 110 tabs in FF 68.11, perhaps 20 recently used, memory use 778MB.

When I briefly ran FFnew last year, when you couldn't even reorder or arrange tabs so I had maybe a dozen, memory use was well in excess of 1GB, and that after they'd ditched saving to PDF, viewing page source and numerous other major features that many users complained about, in vain.

On a 3GB RAM phone, ~2.7GB available for apps, 1GB is a serious hit.

Fortunately I had the older version on an older, synced phone - there was otherwise NO WAY to revert - so saving a couple of years' work; others were less lucky.

So once bitten, I began using the installed Samsung Internet, which is actually the best browser I've used on any platform: fast, clean rendering, excellent favourite and tabs management, and which has never crashed or lost tabs, currently with 68 tabs using 631MB.
 
One thing I like about the phone browsers is that they display tabs in a 2 column thumbnail gallery. I also keep scores of tabs open at once, and this makes finding the correct tab much easier. However, I'm unable to figure out how to reorder the tabs on the phone like I can do with the large monitor browsers. I'd like to see someone program a browser with the best of both worlds, something like a multiple column thumbnail gallery with the ability to reorder tabs using drag and drop.

Using FreeBSD or Linux on a large monitor system with as little as 2 GB of RAM, I always seem to have plenty enough memory to manage hundreds of open and sleeping browser tabs at a go. Surprised to learn that a 32 GB RAM phone only allocates a max of about 2.7 GB RAM per app. Another point in favor of non-phone systems in my opinion.
 
I tried to find more info for WebAssembly (again). I am not sure but it seems it does not have its GUI. I expected to find libraries for things like window maganer, easy way to create menus and other GUI dialog options.
 
Back
Top