Vivaldi opens up an exciting new front in the browser wars, seeks to get around blocking with cunning code

Browser maker Vivaldi celebrated its last release of 2019 with a handbags-at-dawn move that will see it don a Google Chrome disguise. The move comes as the Oslo-based outfit reached the end of its tether with web sites rejecting its Chromium-based browser, while waving the similarly Chromium-based Google Chrome through with a …

[thosrtanner]

I'm not really surprised. the ability to spoof user agent strings has been around for a long time now, because of similar problems, mainly because the designers (I use the word loosely) of web sites seem to be somewhat reluctant to test for feature existence. Or to fix their own sites because they rely on the undocumented and non standard behaviour of browser x.

[tiggity]

@thosrtanner

.. Just what I was going to say

"but in Microsoft's defence it is very clear on what browsers are supported."

Not a defence, it's not difficult to make things work cross browser.

If you insist on using some bleeding edge browser functionality (hint, chances are you really do not need to as usually simpler options will exist unless its weird areas such as DRM laden media playback stuff), you do not test on browser "name", you check if that functionality is supported (or not)

[sabroni]

re: Not a defence, it's not difficult to make things work cross browser.

It's obvious that you weren't working in mobile ten years ago. Why do you think jQuery was a thing?

[Anonymous Coward]

Re: re: Not a defence, it's not difficult to make things work cross browser.

"It's obvious that you weren't working in...."

Wait what? I think present times are implied here and even when jquery was still a thing to use, you still had non-standard... standard methods that even old jquery wound up including. TODAY, I can't help to agree that if you need to sniff the agent, you've lost. But do you feel like a winner (that's the trick)?

Spoofing your users agent is a sign of somebody breaking things somewhere.

That said, today browser extensions are the new "Quirks Mode". There's more fucked up shit going on with "frameworks" and their interactions with browsers than people care to talk about, and every single one of those are proprietary regardless of origin.

But to solve all this, everyone could stop willing giving their users identity to Google, Facebook, Amazon, etc. and do proper design.

[sabroni]

Re: I think present times are implied here

Right. So by cross browser support you mean "works on chrome"?

[Michael Wojcik]

Re: re: Not a defence, it's not difficult to make things work cross browser.

Why do you think jQuery was a thing?

Because web developers are lazy and would rather use a poorly-designed library written by people who can't even be bothered to read the language specification and do things correctly, than write their own code.

[david 12]

Re: re: Not a defence, it's not difficult to make things work cross browser.

Personally, when I use an unsupported browser, I don't care what the site looks like. If I render tables, side panels and transparencies wrong, I know the blame is on me. On the other hand, if they just refuse to provide a page at all, that's a problem, and it sucks.

jQuery started out offering transparent support for multiple browsers, then, as you no doubt know, later choose to interpret 'does not support' as 'will deliberately break support for'. So that when using an older browser you would get a blank screen on sites using current versions of jQuery.

jQuery pioneered "break site for unsupported browsers".

[Michael Wojcik]

Not a defence

Came here to say the same thing. It's not a defense; it's an admission that the development team is incompetent.

Write to the standards, and in the case of (stupid) "living standards" promulgated by the likes of WHAT-WG, don't require bleeding-edge features. If your web application can't degrade gracefully, get out of the business.

[teknopaul]

And if you want to use bleeding edge features do feature sniffing with a fallback.

[Bronek Kozicki]

Filtering by browser is a throwback

... to 2001 when IE6 was released and so many websites wanted to make a point that they no longer want to talk with IE4 or IE5. Now we have HTML5 universally supported by almost all browsers, and I frankly am fed up with web developers making a stance "No, we are special because something something something". No you bloody are not, you are just ignorant git!

Particularly good (bad?) example is Barclays banking - connect from Linux Opera or Firefox, current stable version, and it will suggest that my web browser is obsolete and I should use (possibly older, but hey, running on Windows!) versions of Chrome or ... Firefox. At the same time when trying to stuff 15 tracking cookies from 3rd parties on my computer - because it's only banking! Duh!

[jaywin]

Re: Filtering by browser is a throwback

Now we have HTML5 universally supported by almost all browsers

If only there was one common agreed definition of what HTML5 consisted of, and how it should all work.

[BenDwire]

Barclays

That's not my experience - running Debian 10 with FF 71 and Barclays (.co.uk) works fine for me. As far as I recall, I've never had an issue with any version of FF under any recent (8+) version of Debian. I've recently switched to Brave, and that works fine too.

Of course YMMV, but that's what I'm seeing here.

[Bronek Kozicki]

Re: Barclays

FWIW the website is www.smartinvestor.barclays.co.uk, my Linux is Ubuntu Bionic with XFCE and browsers are FF 71 (from [e-mail]) and Opera 65 (from [e-mail], stable stream), both 64bit.

[BenDwire]

Re: Barclays

I just tried that website link and it works fine in both my browsers. I have cookies enabled if that makes a difference? (also Gnome3 with Wayland, but we won't go there ..!)

[Martin an gof]

Re: Barclays

Probably not connected, but I've come across a couple of websites recently that don't work in Firefox for "no apparent reason". It often turns out to be some hidden (iframe) content which requires script to run that NoScript is blocking, but which doesn't show up on the menu bar NoScript menu. Of course, I don't have NoScript on other browsers. It's only when you dig around and find the correct site to "allow" that the whole thing springs into life.

I first met this kind of symptom on shopping sites, which use Arcot or Sagepay or Worldpay or similar in a "hidden" window to verify your card details. Unless you already have those sites allowed the transaction will fail.

M.

[Doctor Syntax]

Re: Barclays

"I don't have NoScript on other browsers"

Presumably you have the same set-up as I do. One browser has little filtering installed but just deletes all its history when it closes. Fired up as needed ans closed down immediately afterwards.

[Bronek Kozicki]

Re: Barclays

It only bothers me about my choice of browser after I'm logged in. So, unless you hold an investment account there, you won't see anything out of ordinary.

You might be onto something with the cookies, though. I block quite a few (might be fewer than 15, but still)

[jelabarre59]

Re: Filtering by browser is a throwback

... to 2001 when IE6 was released and so many websites wanted to make a point that they no longer want to talk with IE4 or IE5. Now we have HTML5 universally supported by almost all browsers,

But we *also* have MSIE6-revisited (AKA GoogleChrome) mucking (or some other rhyming word) the whole ecosystem up. And anything based off of it is just going to make things worse.

[W.S.Gosset]

Re: Filtering by browser is a throwback

Embrace

Extend

Excrete

[ma1010]

Re: Filtering by browser is a throwback

Totally agree!

I use the current version of Waterfox on Linux, and every time I go to Bank of America, it gives me bullshit error messages about using an unsupported browser. I ignore those messages, and the web site works fine.

[dvd]

Re: Filtering by browser is a throwback

I use Firefox on Linux Mint and the Barclays site works fine...

[Neil Barnes]

Is there any rational reason why a browser needs to announce its pedigree in this day and age? Has no-one heard of standards? Surely, all browsers should behave the same way, no?

[Charlie Clark]

The two questions are distinct: standards are a moving target so it's unlikely that all browsers will provide the same level of support at any one team, even such close cousins as WebKit and Chromium. But, it should be a matter for the browser to resolve and the user agent should be dropped.

[Doctor Syntax]

"standards are a moving target"

That's a problem in itself. At least it's a problem when they don't move in concert.

[jake]

"At least it's a problem when they don't move in concert."

Sometimes a split in the standard is logical, and not a problem. See email vs. USENET headers, for a simplistic example.

[jelabarre59]

Is there any rational reason why a browser needs to announce its pedigree in this day and age?

The fact is, there needs to be MORE browsers based on Gecko/Mozilla code, and less on Chromium. Or someday we'll see a core vulnerability in Chromium that will take the whole web down.

[Dave K]

Unsurprising

I feel for Vivaldi, and for other smaller browser companies as well. I've often had similar problems with Pale Moon on sites - yet if you spoof the latest Firefox UA, hey-presto the site works and renders fine! Pale Moon currently has a pre-configured "list" of naughty sites for which it'll spoof a Firefox UA by default to work around blocks, which isn't ideal but does mean that most things work correctly whilst reporting the Pale Moon UA to everything else.

The problem here is with over-zealous web developers who have an infuriating habit of testing their site with a few major browsers (which is fair enough), but then blocking everything else (which is stupid). The result is smaller browsers that technically have no problem rendering a site - but are prevented from getting the chance to do so because they aren't one of the "big-few".

Of course the best approach would be for the site to check for required browser features and only pop up a warning message if a feature used by the site is reported as not supported by the user's browser. But of course this approach takes effort - far easier for the lazy developer to just block everything that isn't Google Chrome and then tell everyone to jump on the Google bandwagon...

[werdsmith]

Re: Unsurprising

They probably make a cold hard decision based on effort vs return.

At a minimum a web site should support all browsers that JQuery supports.

[Dave K]

Re: Unsurprising

But why do they block other browsers arbitrarily? I'm fine with a website not going out of their way to explicitly support Pale Moon or Vivaldi (etc), but why do they just block everything that they haven't explicitly tested? How would it have worked for Chrome back when it first started if everywhere just blocked it due to it not being Firefox or IE?

[Cavehomme_]

Re: Unsurprising

Because in these day of litigation, they want to be clear as which browsers are supported. Especially when it comes to banking.

[Doctor Syntax]

Re: Unsurprising

In that case just stick to basics and don't try to use browser-specific tricks. It shouldn't be a function of banks to act as marketing departments for Microsoft, Apple, Google or whoever.

[Charlie Clark]

Re: Unsurprising

I've been advising on this for years and feature detection is the easy and safe way to go. This can includea a cut-off for non-standards compliant browser, basically IE < 11, but the rest should be handled by feature detection.

A bigger problem is keeping a codebase that allows you to jettison framework code as native support becomes available: jQuery is no longer required for many things it's been used for for years.

[vtcodger]

Re: Unsurprising

"The problem here is with over-zealous web developers ..."

Are you suggesting that there might be intelligent lifeforms involved in website development? That's an interesting theory, but I can't imagine how one would go about testing it.

[Doctor Syntax]

Re: Unsurprising

"Of course the best approach would be for the site to check for required browser features and only pop up a warning message if a feature used by the site is reported as not supported by the user's browser."

A better approach still would be a plain vanilla test browser approved by someone such as the W3C. Devs could use it for their own testing but if it passes W3C or whoever could also hand out an approved badge based on independent testing. Banks and other financial institutions would be required to have it.

Lack of the badge would become an indication of cowboys and smartarses at work.

[Charles 9]

Re: Unsurprising

Or a firm unwilling to pay through the roof for yet another badge of approval. Besides, how can you be sure the standard will have teeth and not get shoved aside for (or even Borged into) the Next Big Thing?

[amanfromMars 1]

A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

Web browsers are the new operating system for operating systems/SCADA Operations.

A true fact rather than fake fiction, and that has profoundly deep and dark consequences for future failing leaderships, methinks, for all are next to impossible to control absolutely to one's will.

[jake]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

You've said that before, amfM. You are still incorrect. A browser is an application, not an OS, no matter how hard you squint at it.

Unless your browser controls your hardware, of course ... in which case, I submit that your actual OS is very, very b0rken.

[amanfromMars 1]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

We'll have to agree to disagree again then, jake, whenever one of us, or even both of us for that matter, are still incorrect.

However, share a sensitive operating secret available for simple viewing in a browser and just watch how compromised hardware tries to react and close vulnerable operating systems down.

That has escaped info leading intel communities responsible for the staging of a right royal merry dance, surely ‽ .

[JakeMS]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

Count me curious...

What is it that makes you feel the browser is the OS?

Bearing in mind I very much think the browser is merely an application. (I offen don't use it* when working on my computer).

* I don't need a browser for running commands or editing code

[ArrZarr]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

I think it's more of a reference to how many tasks that used to be done in specific programs launched from the OS have migrated to web pages, web clients etc.

For most people, the OS is primarily a means of getting to the things you want the computer to actually do. Browsers are now doing very much the same thing for a hell of a lot of use-cases.

Unsurprisingly, there is an XKCD for this - https://xkcd.com/934/

I'm sure you & Jake can suggest actions that a browser cannot do, but speaking generally, I believe I'm correct.

[amanfromMars 1]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

That pretty much lays out all the Base ICQ Basics, ArrZarr. And that is just the beginning of what IT has discovered its simple services in applications can do ....... and what it really means for the human race and humans racing ahead in/of current presentations and/or Future ProgramMING Projects..... Mined IntelAIgent Networking Games, where the simple virgin browser provides Sensitive Secret Instructive Services to Attending Attentive Masses and Persons with Unusually Specific Interests.

[nagyeger]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

Isn't debating with amanfrommars almost as much a sign of madness as trying to extract meaning from it's high-faluting algorithm?

[jake]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

"Isn't debating with amanfrommars almost as much a sign of madness as trying to extract meaning from it's high-faluting algorithm?"

No more so than debating with you is as much a sign of madness as trying to extract meaning from your high-faluting(sic) algorithm, nagyeger.

[ROC]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

What about Chrome OS? Seems as though it has been approximating that goal.

[jake]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

"Seems as though it has been approximating that goal."

So has EMACS, these last forty-odd years. And systemd is attempting a latter-day coup. But that still doesn't make either of them operating systems. And they won't be, either, until you can boot from them as a stand-alone kernel (monitor, executive, whatever).

[jake]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

As I said, "in which case, I submit that your actual OS is very, very b0rken.".

[phuzz]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

(Always fun watching unsuspecting people argue with bots aliens, thinking that they're people)

[Anonymous Coward]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

Agreed, but this exchange was disturbingly coherent. Don't know if AMFM is adapting or if's my brain that's changing.

[jake]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

amfM isn't (entirely) a bot. There is organic intelligence in there, if you look for it.

[Doctor Syntax]

Re: A Ubiquitous Weapon for Mass Distraction and Destruction and Disruptive Creation?

The concept of what an OS is has been rather fluid ever since virtualisation raised its head on mainframes. If you run Windows as a virtual OS on BSD or Linux on Windows - or these days just running WSL as part of Windows - what's the OS as far as the application is concerned. We're dealing with layers and all the application can do is take whatever provides its API is the OS. For an application running in a browser that's the browser. If you want to look as far as the hardware* then the OS is the entire stack up to and including the browser but all the application sees is the browser.

* Given that the processor's instruction set is apt to be an artefact provided by microcode what actually is the hardware as afar as the OS is concerned?