Apple's Safari browser runs the risk of becoming the new Internet Explorer – holding the web back for everyone The legacy of Internet Explorer 6 haunts web developer nightmares to this day. Microsoft's browser of yore made their lives miserable and it's only slightly hyperbolic to say it very nearly destroyed the entire internet. It really was that bad, kids. It made us walk to school in the snow. Uphill. Both ways. You wouldn't …
I think we have different memories of Internet Explorer, but IE's inability to implement standards wasn't the reason it was bad. It was the audacity IE had to think it's implementation of various things should be the standard. Which meant when it came time to coding a website, you would often have it working perfectly in Firefox, with cascading versions of the CSS to suit IE8, IE 7, and the total car crash of IE 6.
The issue really isn't Apple's reluctance to implement various API's in to Safari, it's Chrome's insistence that API's it includes should be an industry standard. Like what's been said, a web developer really shouldn't have or need access to various parts of the computer viewing the site. So why is there an insistence that this should be the case?
Web browsers are more than 25 years old. From an end user's perspective they don't need new features. Perhaps better ways to stop Evil Corp (aka Google, Facebook) from slurping their data. But so long as we can watch video, browse websites, see images we diont need anything else.
All of this is just a Google whine, Apple should allow us to just block all ads coming from a the Googleplex and be done with it. SHut the fuckers out of my phone once and for all.
While I agree in general, I have encountered specific instances of Safari not implementing certain standards correctly, specifically to do with bubbling of events within certain contexts. Every other browser handled the issue I was having without problems, but Safari just plain refused to even attach the listeners.
I could go into great detail about the time I spent trying to debug a click listener that failed to capture an event on a child element, but this isn't really the place to get into a lecture about event bubbling and how it should work vs how it did work in Safari.
Pointing out that Safari does have a specific issue, while pointing to the general area in which it has that issue in a discussion about Safari's potential to screw over standards compliance, is plenty meaningful to anyone who actually has to deal with this sort of thing on a day to day basis.
This. One thousand million trillion zillion times this.
The world needs web content and web browsers that are primarily *display* tools. Want more than that? Lots more than that? Then it probably doesn't belong on the web, and it's not the job of the web browser to provide an all singing all dancing "portable" runtime environment (same nuisances incompatibilities and vulnerabilities everywhere? [1])
The web is dying anyway. The walled garden approach now universally found in "modern" mobile applications will see to that. Sometimes simpler really was better.
[1] "Same idiocies everywhere linked into everything" is systemd's job.
In tomorrows news - how Google/Chrome are breaking the internet all over again by rushing out a feature that nobody other than the advertisers actually want.
I would rather have properly-thought out stuff than Google's slapdash approach any day.
I think we need to take a step back from shoving everything on the browser, and work out what a browser needs to do, particularly the privacy, and reliability implications.
Everything you do on a browser will require server access at some point. That access can be (and probably is) logged and profiled. You also don't know what access others have to the data you store on the cloud, which has massive privacy implications. Even if you have it written in a contract that Microsoft et al don't have access to the data you store on their servers, all that says is you have some legal recourse if they do access it, which is essentially shutting the barn door after the horse has bolted.
I mention this, because Web Apps tend to store their data in the cloud, and even where they don't, they probably have to send locally stored data to a server somewhere so they can process it. It also means they won't work if you lose your connection. As you might on a train journey for instance.
I think we need to stop pretending it's good to bung everything on the web. We don't need our browsers to become almost replacements for our operating systems (as they seem to be doing), and start to move back to using locally stored applications. If the hardware we have is too slow, then optimise the code. Remove unnecessary bloat, such as features that aren't necessary for the core function of the suite or application. E.G. I need Outlook to access my email and manage my work calendar. I don't need Outlook to analyse my work patterns and suggest when I should break for lunch, and I don't need it to analyse my grammar. Yet it is using extra resources in doing so. And before you say Libre Office, I know there are alternatives, but I have to use Exchange 365, and our IT department have disabled access from all non-Microsoft apps, so I can't even use the built in clients on my iPhone any more :(.
@IGotOut
Absolutely agree. but what's worse is web devs immediately latching onto the new "features", making older browser unable to render their productions. That in turn forces the churn that sustains the update rate - a vicious positive feedback loop that serves nobody except the technocrats.
Many flat content web pages now refuse to render except in the very latest generation of browser. When you examine the source it's typically a bloat of obfuscated client side code that can neither be checked for security or run in the older browser. For example, until a couple of months ago, the RS components web site (which we've been buying from for years and years) worked perfectly in a slightly older version of Firefox with javascript disabled. One simply searched for a component and the item came up on the screen. Then suddenly one day any attempt to search for a component yielded nothing except a permanent spinner, apparently waiting fro something that never occurred. So no more component searches and RS lose a customer.
As another poster asked, "cui bono?"
I’m Safari user on macOS and iOS, by choice, don’t work for Apple and am heavily involved in some parts of the Web Platform Tests.
While I recognise some of what you’re saying here, each of the three browser vendors is behind the curve - as it must be. The spec moves forward and the browsers play catch up. Where they focus is inevitably a decision that rests with them, but each of them has coverage issues, some of which are very long standing.
Personally I don’t see that Safari is noticeably further back, and in some areas it leads (I believe CSS Color 4 is fully implemented in Safari, including mobile. Chrome’s work is ongoing, and Firefox haven’t started as far as I know). I can’t speak for all areas, and I know in the areas I work in it’s probably the weakest of the three. But not by much.
It could be that the author has a point on certain API’s, I wouldn’t know as I don’t use them. But we’re a long way from IE6.
Okay, you found one thing Apple leads on. Unfortunately, that really doesn't impact my day to day usage.
But you know that if anybody's not implemented a feature yet, it will be Apple. Take Object.hasOwn(). Core. Useful. Okay, trivially polyfillable. But I'm carrying around a pollyfill because Apple hasn't got round to it yet. And will have to continue to carry around that dead weight if I want to support older iOS devices.
There's Safari's half-hearted supports for custom elements which is holding back modularisation of code.
And then there's the buggy implementations, too. Of stuff like CSS grid.
why exactly do you need all that again? What does it REALLY get you?
My suspicion is you're trying to do too much, your web site is a monolithic high-bandwidth scripty nightmare, and end-users have performance issues with simple UI actions, on ANY browser, because of it all. Or maybe not, but it kinda *smells* that way to me.
A lightweight web does NOT need a heavyweight browser. And yeah, I hand-code my HTML and use limited CSS where needed, usually in a "<STYLE>" element block in the head section and NOT some ginormous all-in-one style sheet that's necessarily delivered by a CDN.
Still wanting to do a webkit browser that looks more like OLD Firefox with the 3D Skeuomorphic UI.
(I just do not want to do it in Python)
Wrong comparison - IE6 (and IEs in general) added all sorts of MS-specific shit that was not thought through or actually desirable to most folks.
Now it is Google Chrome that is adding all sorts of shit that is not desirable because of privacy or security. Google don't care because the business model is about whoring you from advertiser to advertiser, and web developers have drunk their kool-aid.
Pfft. If it's not on the CGA color palette I'm not interested.
You can keep yer Farrow and Ball or Little Greene. I'm painting my house 13/Light Magenta.
at least there's no ActiveX any more - that was the DUMBEST thing IE supported. It may have sounded 'cool' in 1998 and I actually wrote a couple of these, but there is NO security in ActiveX and it's a HUGE vector for infecting computer systems (and S. Korea had ActiveX required for anyone doing online banking within that country for NEARLY 2 DECADES)
This post has been deleted by its author
Why do you feel the need for constant spec updates ?
Other presentation systems don't do this. If some feature isn't provided, you implement it from the existing features. If those aren't flexible enough you look at the core features and redesign on a sensible schedule
Adding to the basic spec and expecting it to be available on every one of billions of browsers worldwide is just insanity.
stability has its virtues
bleeding edge is overrated
"moving targets" for development (especially in the planning stage) is NEARLY ALWAYS a BAD thing.
it's why I will not target any new, shiny from Micros~1 (like UWP and previously, Silverlight). So why tolerate it from GOOGLE?
"I should probably admit now that I hate the modern web. I find the experience of modern websites so unreliable, slow, and overall user-hostile that I prefer to do literally anything else. I dislike the experiences enabled by JavaScript-driven web APIs so much I've taken to browsing with JavaScript disabled."
Some websites try to load scripts from 15+ different sites. Many times they try to load multiple analytic websites; why is one not enough? There are multiple unscrupulous, abhorrent privacy-hating advertisers; internet ads once worked quite well without tracking, which means they can work today. And worst of all, some websites will not load unless you enable javascript. All this is bad and wrong.
It seems to me that modern programmers have myopia. They use Google Chrome, they test on Google Chrome, therefore it works. They don't test with Firefox or Safari. If it doesn't work, their solution is to use what they use. They create websites on powerful computers with ultra-fast internet. If a website is too slow, their solution is to buy a stronger computer and get a faster internet. Not everyone can afford a new computer, not everyone can get or afford fast internet. Instead of people conforming to you, you need to conform to your users. And that means assuming they use a different browser, assuming they have a weak computer, and assuming their internet is slow.
It's not only that modern programmers are myopic, they are worse: arrogant. They make a decision, they believe that their decision can only be the "right" choice, and stick to it through thick and thin.
MS Ribbon. GNOME. Systemd. Google (just about anything, actually). Facebook.
The examples are endless. User preferences be damned, better ideas be damned, [I / we] are going to stick with our idea because it must be the "right" one. What is right for me by definition will be "right" for you, because I'm the programmer, I know more about tech than you, so that makes me fundamentally correct from the very outset.
It is absolutely killing our tech experiences.
young kids these days. they need to program with punch cards for a while on an old big-iron machine that only has FORTRAN 66, a card reader, and a line printer.
yeah they lack the experience to know when their ideas are stupid and meaningless, but they *FEEL* good about them (and are arrogantly willing to foist this upon the rest of us, with NO concern about customer wants or needs).
and there is also the "it's MY turn now" syndrome. Fresh out of school, want THEIR CHANCE. Everything ELSE done before THEM is all CRAP and needs CHANGE. That sort of thing.
I'm not sure its (just) arrogance. It's ignorance (though they seem to be opposite sides of the same worthless coin these days). Rather than try and understand what they are trying to do for the user they seem to just grab libraries that go some way to what they think they want and then find other libraries that do another bit and we end up downloading massively complicated bucket loads of javascript that spin your cpu fans and delay proper rendering due to time-outs when something infinitely simpler can be achieved with simple understanding of what the fuck they are trying to do in a simple client server setup.
Though as with all coding ignoring the marketing departments demands until you have a working bit of code can help.
I have often made similar appeals to logic. Back in the Win '95 beta one of the engineers agreed with me on some of the performance problems with the 'exchange' mail client it shipped with, it's RAM hungriness, and other serious performance problems. He even tested things with a slower 386 machine. Yet, the product shipped 'that way' anyway. So the problem may be way higher up the chain when "the boss" has a more capable machine than the engineers or customers... (and therefore can not see the problem).
Hear, hear. A long time ago, when I developed desktop applications I wrote them on a PC that was available to my customer base perhaps 4 years before, because *That Is What Most Punters HAD*. I learnt that lesson when running a business for someone else. I oversaw a project where we wanted a system to produce an application report for farmers based on chemical analysis of their soil and growing crops. It looked really impressive on the developer's Pentium. Unfortunately much of the intended customer base had a 386SX, usually without a co-processor that they used to "do the books". Something that ran in a few 10s of seconds on the Pentium took >15 mins on a SX, and usually finished by truncating the expensive report without the totals at the bottom (the bit that was actually useful). Much to my surprise, when I took took some of the design - VB4 with Jet and Crystal Reports (which was usually dreadful, and often applied badly) and bunged it across to MS Access V2 (not known for being speedy compared to the largely machine-coded Access 1.1), and using the Access band report generator produced reports in a couple of seconds, but did need a whopping 4-6MB of memory (we only warranted it to run on 4MB). When I asked the developer why he was using the latest PC, he said his time was valuable and that we weren't paying him to sit around while stuff compiled, and it wasn't his job to keep "obsolete" kit around! We didn't use him again, even though he was the acknowledged expert in that field... As I recall 2MB of RAM to add to the 2-4 that PCs shipped with back then was <$200 whereas the developer machine was >>$2,000.
THIS! A trillion times this!
*RiverDances on the UpVote button in hopes the backend will register a few billion clicks*
I'm using Win7Pro64, Jaws 2021, & IE11+ with everything as up-to-date as their makers provide. I run with JS (and all scripting, 99.9% of the "enhancements" disabled, etc) so the sites I visit will hopefully load faster, cleaner, & safer. If I block JS then those ad slinger JS exploits can't even get in the front door. If those "enhancements" are disabled then the vulnerabilities they create are also stopped cold at the door. If the site refuses to load then I do a DDG search for a cached version of the site, read the plain text copy, & consume the content in safety. If no plain text cache is available then I go elsewhere.
My bank doesn't require JS, ElReg doesn't require JS, my homepage doesn't require JS, and none of the news sites I (RSS) subscribe to need it, so any site that *requires* me to turn JS on in order to deliver content is given TheFinger as I close the tab.
"Before you use this car wash we require you to lower the top on your convertible, roll down all the windows on your hardtop, & stick your head out to smile into the oncoming stream of hot wax so our spinning brushes can scrub your teeth." Since I don't consider using a car wash to power wash the *inside* of my car to be a very good idea, and there's no way in hell I'd agree to let them brush my teeth, I'll go find another car wash that doesn't require such stupidity.
Change "car wash" to "web site", "windows" to "firewall", & "brush your teeth" to "bend over & take it up the arse". =-/
Please understand that the web moves at breakneck speck partly because the majority of its users (read: NOT THE LIKES OF US HERE) demand it, and partly because the pace drives demand from the aforementioned, creating a vicious cycle and leaves us either hanging on or giving up and getting off and getting left behind. Sorry, but that's just the way it is. If you want to change things for the better (like for me, a split between a passive web and an active graphical terminal interface), you'll need to convince Joe Stupid and all the Facebook Friends first.
"Please understand that the web moves at breakneck speck partly because the majority of its users (read: NOT THE LIKES OF US HERE) demand it"
Nonsense. The vast majority of users don't have the slightest clue about anything regarding computers or the internet. They just take whatever they're given. If they're given browser updates every half hour to allow sites full of tracking, infinite scroll, flashing adverts, dogs and cats living together, and so on, then that's just what they'll use and will quickly consider normal. Give them a stable browser with a consistent interface that rolls out relatively minor updates once a year, then that's what they'll use and will consider normal. Not a single part of the internet was created because it was demanded by ordinary users before they even knew it might be possible.
You've missed the point. Nobody cares how often their browser updates, or whether it supports some crappy API or not, or whether it transmits every twitch of the mouse to Google or the Kremlin. Nobody, for statistical values of nobody, even notices any of this.
What they care about is being able to direct the dancing cats that someone told them about. If they can't do that in one browser, they'll do it in another. And then they'll conclude that the latter is better.
"because the majority of its users ... demand it"
Citation needed.
When was the last time anyone heard a typical web user exclaim. "What I really want is for my phone/PC to run slower ... and more erratically. And I **LOVE** ads. Please send me more of them. And I don't think I'm being spied upon anywhere near enough."?
I cited firsthand experience. I deal with Dave's on a daily basis. These are people who think Facebook is the Internet and can't be convinced otherwise, when told to turn left turn right and when told to turn around spin 360 degrees. They believe their way is the right way and FU! AND they have the money.
This is why I keep saying you MUST fix stupid...before they take the rest of us with them.
no, the users do NOT demand a rapidly changing web. At least, none of the users I have known.
It's the MARKETEERS and AD SLINGERS who demand it. "New Shiny" - right?
Most people get tired of the CONSTANT UPDATING very quickly, especially if they get throttled for excess cell phone bandwidth usage or have their windows workstations taken over by all-too-frequent updates for several hours (when they really needed to use their computers) a few times too many.
All that moving target updating and "New Shiny" is HIGHLY overrated. And I think nearly everybody KNOWS this.
I think it’s great that they don’t give in. A browser should be as simple as an email client. Websites should be cleaner, simpler, faster. And follow standards. For all other advanced features there is the OS to support them.
Chrome will soon make french fries (a saying in my native language which means it will start doing stuff a browser should not do).
Personally, I hate that Google and others refuse to make native apps and insist on making all apps web-based. But they’re only doing it to promote their own browser and also ChromeOS.
The cost of me to have to build apps for Android, iPhone and the web so that all users can access my systems anywhere is prohibitive. Browser based access simplifies design and testing.
Having web standards to allow access to Bluetooth or RFID or photos to scan a barcode allows customers to configure systems with a simple click of a button from a web app that will run everywhere that the support is available. It means that the UI is consistent and testing and bug fixes are readily available.
I don't want to have to buy an iPhone for the 2 users that I have that use them to view data on them. I don't want to pay for a service to test them on. I want to run on a selection of browsers that run on my developer machine and verify that it works. I can automate testing easily and quickly.
I want to push out a bug fix in days for customers because we can test it readily on all browsers. I don't want to test on 30 different versions of iOS, windows, android, etc to support native apps across multiple OS versions.
I love that the one thought-out response from a developer explaining why this model actually works for them is immediately downvoted.
The modern browsers that you all seem to hate create real value for developers who can create a single web app (using those APIs you see no value in) that provides a consistent user experience on any platform, all using web standards. Developing native apps for iOS and Android is time consuming and complex and involves writing two entirely different implementations of the same app. Want to target Windows users too? Another implementation. Linux users? Another one. It's impossible for a one-man operation to do. But a single modern web app can do it all with a single implementation that's actually less likely to have bugs and can introduce new features more quickly.
I actually signed up to make this post to support developers like Michael. You guys all want a good, consistent experience? This is what these developers are trying to do. They're not arrogant and they don't think they know better than you. If you think this, chances are you don't know a professional web developer.
No, it's not.
Are you aware that you have to grant access to a web app for it to access Bluetooth or your battery status? The web browser won't just allow websites and web apps to access that data without asking your first. Just because your browser has an API, does not mean it's accessible to any site you browse to. As a developer, you have to request access and the user must explicitly provide it, the default behavior is to block access.
But let's take an example. You buy a smart light switch and need to set it up, so you head to the app store and download the app and - read carefully - give it access to your Bluetooth stack so it can detect the light switch.
That developer had to create an Android app for you, and an iOS app for Apple users plus a Windows app, etc. But they could just develop a single app *with the same permissions* that targets all devices.
You're giving the same developer the same permissions and getting the same experience - possibly a better one. Modern web APIs allow all this, as well as that web app to be look and behave like a native application.
"download the app and - read carefully - give it access to your Bluetooth stack so it can detect the light switch."
OK. So let's assume (incorrectly, I guarantee you) that the Bluetooth specification is secure and watertight (which it isn't), and that every implementation of that specification is flawless (which it isn't).
But anyway, assume perfection. Which parts of the flawless specification and flawless implementation assure the user/manager that any given app will access only those Bluetooth facilities as are relevant to configuring a particular smart light switch? And nothing else at all will be touched? E.g. (outside Bluetooth) that the app in question won't quietly be snaffling every entry in the Contacts list, whether it needs to or not.
References welcome.
"You're giving the same developer the same permissions and getting the same experience - possibly a better one. Modern web APIs allow all this, as well as that web app to be look and behave like a native application."
That's not a feature. That's a privacy/security catastrophe in waiting.
You are aware that you can adjust the UI to suit different screen resolutions in a web app? It's a relatively minor overhead written in the same language/framework which ensures that things work consistently. I can have one set of code that runs everywhere, the UI may look slightly different on your phone or pc. Hell, I can disable or remove features that won't work or be used on a mobile and make them appear on the desktop version.
Want to enter the barcode on the desktop you can type or use a barcode scanner. On your phone the camera. Provided the API supports it.
Alas people use responsive HTML without actually testing it on the devices, they just seem to assume that a form with 30 entries on it will work on a phone because it works on their PC.
I use one set of code that works on any device - but that's on the server and it check what you device you are using so it can server you something ergonomically useful for the device. Phones work better if you feed them small pages of info rather than the big pages that work fine on a PC. Responsive HTML is never going to solve that problem but too many people assume it does.
@gilphilbert I have not voted up or down on either yours or Michael’s posts. I understand the value to a developer of writing a single application that will run everywhere that the support is available. But browser/web apps are not the way this should be done.
Browsers and web standard were originally designed to display data from the internet in the form of a web page. There was no thought in that original design for running apps and security that having that functionality would require. We should not be trying to shoehorn in and bolt on functionality to create apps. This just creates security risks, by adding APIs that bad actors on the web can use to get out of the sand box to access other parts of the computer.
For write once run everywhere we need something that was designed and built for that purpose. Now I am not a web or Java dev but maybe Java (JVM) or something like it is the way to go. Now Java may not be up to the job, but that is the approach we should take. Not trying to co--opt a system to do something it was never designed to do.
Now before you say it as I think this maybe the case Java (JVM) does not run-on Android or IOS only on desktop OSs like Windows, Mac, Linux, etc. But can you really have a UI that works for both touch screen Phones and WIMP?
@Michael “I want to run on a selection of browsers that run on my developer machine and verify that it works. I can automate testing easily and quickly.”
Oh, how I wish testing was only done on the developer’s machine. No longer will I and my fellow developers utter the cry to testers ‘’well it works on my machine’. ;)
But seriously you can’t just test on your machine, you can’t just assume that it works on other platforms even if it is a browser/web app.
Let’s assume your dev machine is windows using the latest release of Firefox as an example. Just because your app passed all your tests on your machine for Firefox. Does not mean it will for Firefox on Android, Mac or Linux.
Maybe Firefox’s release point is different maybe in front or behind your dev machine. Even if they are the same it is still a different build of the browser. There is no guarantee that just because it works on yours, it will pass all the tests on those other platforms.
I understand your don't like Google, that's your right, but don't just make stuff up. The web goes through standards bodies of very hard working people from many companies, all trying extremely hard to make a safe, secure web experience.
The fact that you feel native apps are actual safer prices my point, they are not and not getting upset with Apple about this again proves you don't understand what's actually happening.
Don't like Google AMP? Either do I! That's something you should be upset about, I agree with you! But that's not part of the web standards process.
Safe and secure? Like the Battery status API?
From the spec of that API
The Battery Status API specification defines a means for web developers to programmatically determine the battery status of the hosting device. Without knowing the battery status of a device, a web developer must design the web application with an assumption of sufficient battery level for the task at hand.
A web developer should NOT have access to the battery status of my device under any circumstances. It can tell when you are on battery or power which is all part of tracking the user and device.
The reasons quoted for needing is info is IMHO total spurious. Who would design a web page that would suck the life out of a battery? Oh yes, crypto mining.
The Battery Status API specification defines a means for web developers to programmatically determine the battery status of the hosting device. Without knowing the battery status of a device, a web developer must design the web application with an assumption of sufficient battery level for the task at hand.
Determine is a funny word.
Determine, verb 1. cause (something) to occur in a particular way... 2. ascertain or establish exactly
So they can 1. use all my battery charge and 2. ascertain my battery level so they can... what? Display a huge pop-up with a Dennis Nedry video telling me I am about to run out of juice?
The web goes through standards bodies of very hard working people from many companies, all trying extremely hard to make a safe, secure web experience.
you have minutes from these collaborative meetings and/or mailing list archives available someplace? just curious...
(yes I'd like to see the actual discussion/debate that reportedly did NOT happen)
Most likely the "hard working people" are just a bunch of moderately sycophantic 3rd party marketeers brown-nosing Google to get what THEY want (and damn the end users, we're just being monetized after all)
"The web goes through standards bodies of very hard working people from many companies,"
Many of whom come from Google and do whatever Google says. That's why we have, in addition to the open standards of HTML which don't change very often, Google's proprietary DRM system as a W3C standard (well, technically, a standard that Google implements, but it goes the other way). This despite several things:
1. It's not a standard. Google wrote it and doesn't give out the mechanism.
2. Google decides who gets to use it, and if you don't have their permission, it's illegal.
3. It's not even very good, but nobody can improve it, because Google doesn't allow modification.
They do the same thing with a lot of other ideas they come up with. They shove API suggestions through the W3C all the time, and then they usher those through after they've already created them in Chrome. This means they can say that other browsers aren't adhering to standards rather than that other browsers don't immediately adopt all of Google's code.
The standards bodies try to work on solutions, but they don't have many resources and nearly everybody there is there because their employer wants them to change the standard in some way for that company's benefit. Google doesn't control it, but they influence its direction and the others there are not there because they want to protect the standard, so Google doesn't have to go to pains to get them to agree.
> developers who want to use "cutting-edge" web APIs
...should get publicly flogged! Novelty for novelty's sake, having to constantly add some new "feature" to justify your salary is stupid, counterproductive and utterly dangerous. It leads to today's horribly bloated web, where the least meaningless web page loads tons of 3rd party JavaScripts for no other reason except justifying somebody's salary.
Here is a crazy idea: What if (let's imagine) they started building web sites just for the purpose of informing/selling/whatever? Streamlined, perfectly tuned for the purpose they were built for? I know it would be utterly unprofessional and not command a huge price, but I think the web would be much better for it.
In the same spirit, fixing the browsers' development cycle to 12 months would be great. Not only would we be spared all those constant meaningless updates ("added a line break in the About menu")), but one could even hope they would have time to squash more bugs before release (yes, I know that's unprofessional too, testing is supposed to happen after release).
Here is another crazy idea. Rather than wait for "them" to build the Web just the way you want, why not have a go at building it yourself. Then you can sell what you have to others.
That would surely prove that you know better than "them" about what "they" want.
Tip: I wouldn't recommend you bet your mortgage on this.
I actually know enough about website building to be able to see through the hype. All those useless bells & whistles are pleasing to the marketing goons paying your bill, but simple, clean code is way better for the end user, and I happen to be an end user, the poor sod who has to use your monstrosity afterwards, even depend on it for important goods and services.
"Build it yourself"
Because, when Average Joe user knows what they dislike...they automatically know how to build what they DO like??
Great idea! So when they don't like something about their car they'll just crack open their 100-ton bodywork press, robot welders, casting machines and CAD/CAM systems and make their own! Why didn't the world think of this elegant solution before?!
It wasn't actually that IE6 was a long way behind as such. It was more that MS seemed determined to make their browser totally proprietary in every way possible. Their plan was obviously to try to leverage their dominance to such an extent that websites would only work on their browser, so that people would have to at least keep a copy of IE6 for those websites. But of course what this meant in reality is that everybody coding a website had to put in loads of browser specific code*
So the difference between Safari and IE? Microsoft were malicious, Apple are merely hopeless.
*This does not include those sad individuals who did their best to code their minority interest websites so that they would only work on one browser (usually firefox) in order to try to force the uptake of their favourite browser.
I’ve had some fun browsing Gemini sites of late. Since Gemini is text-only (you can link to an image file for download, but no embedded/inline media), most of the browsers are terminal applications.
I’m (just) too young to remember the internet pre-web, or even the very-early web when terminal browsers were more common. But if it was anything like Geminispace then it must have been lovely.
$ lynx
-bash: lynx: command not found
$ brew install lynx
$ lynx
LYNX - The Text Web-Browser (p1 of 2)
__________________________________________________________________
http://lynx.invisible-island.net/
Copyright 1997-2017,2018 by Thomas E. Dickey
__________________________________________________________________
$ lynx
-bash: lynx: command not found
$ brew install lynx
$ lynx
LYNX - The Text Web-Browser (p1 of 2)
__________________________________________________________________
http://lynx.invisible-island.net/
Copyright 1997-2017,2018 by Thomas E. Dickey
__________________________________________________________________
Web standards have grown too complex (& sold out re DRM, I sympathise with EFF feeling they had no choice but leave W3C).
I hope Apple hold out against some of those Google APIs, which are privacy invading (& a security issue too).
Apple may be small but have enough market share to make a difference, e.g. Apple unilaterally stopping Flash support was the thing that really killed off Flash.
Disclosure: I am not an Apple user (partner is), hate the whole idea of web apps, a browser should not be a "mini OS", give me proper applications, don't try and do everything in a browser.
It is pretty well known that Flash died because of Apple, or specifically Steve Jobs. Maybe it would have died on its own eventually, but Jobs categorically refusing to implement Flash on iOS devices and the growing importance of iOS devices ~10 years ago led web sites that cared about mobile access to phase out their use of Flash.
Don't you remember the early Android vs iPhone wars, back when Android users claimed superiority based on 1) removable batteries 2) SD card expansion 3) Flash support.
I remember that time. I remember the ability to switch in a new battery when traveling. It was great. I remember adding storage via and card. I still do.
I don't remember using a single site that used or needed flash at that time. I certainly don't remember anyone saying flash support was a benefit. I do remember people saying why does android not drop support too.
I think what finally killed it wasn't Apple but increased regulations that mandated that web pages be more accessible to sensory-disabled users. Flash inevitably didn't take the blind and others into consideration and was getting clunky on top of it and too much of an exploit vector. Not so much one big decapitation but several grievous wounds.
I don't need autoplay videos, I don't need my browser to know my location, I don't need any of that needlessly complex, info sponging, buggy spyware enhancing 'features'. If my desktop can't do it...
Why can't I render a page on an old Pentium? Crap features that I don't need or want that's why. I don't buy off those ads anyway... Just heat up my phone, kill the battery, and blow up memory usage. And I have yet to see the killer feature any of the junk/bloat provides.
A perfect example of this is the official government of Canada weather web page (weather.gc.ca). For years it worked just fine. Had a button for radar maps, which came up and didn't change other than when they had to wait for radar data to fill in the space.
A year or two ago they started messing with it. They have continued to make it worse.
Now, it bounces around two times before stabilizing (I suspect async stuff, and the actual data arriving after the basic web page is shown).
I once tried it on my old Android phone. The weather radar stuff was slow to load. I scrolled around to try to find actual radar imagery, and it let me scroll to places 1000 miles away. Eventually I realized that its another async monstrosity - I should have simply waited after the page first came up, hoping to see local radar data. Only when it was done should I have tried scrolling the page to center the radar imagery. Ick.
If Apple allowed Safari to actually compete, it would be better for web developers, businesses, consumers, and for the health of the web.
I still have the occasional nightmare about using Safari on Windows, so on that basis alone I couldn't possibly countenance Bruce Lawson's argument.
I really don't think Apple cares about the web other than as a check box item feature: yes, iOS has a browser. And ancient style email support.
But their ecosystem is the AppStore as their sole source of presentation layers for their devices.
I don't think they're *trying* to hold back the web; I just think they consider it largely irrelevant to their vision.
Funny thing is that Internet Explorer 6 shouldn't exist after 2007, when Microsoft released Internet Explorer 7 as an update for WIndows XP. But back then, connections were slow in most of the world and people avoided upgrading, and Microsoft had the stupid idea of bundling WGA with Windows updates, so that's how we ended with Internet Explorer 6 being what most people recognized as Windows XP's browser.
Well, guess what Chrome does, it pops up a request and the user presses yes to make it go away, then the website spams the user forever more.
I am not the unsuspecting user, I don't even have Chrome installed on my phone, but there are plenty of users who are, and a couple have asked me how to make it spamming them. It's not a useful feature and the web and phones aren't better for it.
We've already seen with the Extension API how Google is abusing its powers with Chrome. They're now altering the API to make it impossible for ad-blockers to function properly. This goes to show that it was a bad idea for other browsers to relinquish control of their Extension API to Google.
I don't buy the argument that Apple is good for the web. At best it's indifferent towards web standards and may actually prefer to hinder them to increase profits (hiring developers to implement those web standards costs money, you know).
Our best bet would be to support Firefox, but it too has been hijacked by dollar-eyed executives who're dreaming they're running a billion dollar company instead of a non-profit foundation, and demanding commensurate compensation.
Apple are slow in implementing new image and video formats (they’re just rolling VP9)… ….but most websites use HTML frameworks to mask the differences between browsers.
There isn’t really a big group of people grappling with low level CSS features. Or if you are: you’re probably doing it wrong.
Sounds like a bit of Apple fanboiism to me.
The 71 percent compliance is not Apple not implementing the web as determined by Google. It's a standards complaince test, not a "What Google does" test. Not implementing something like USB access is sensible, but most of this is simply Apple not keeping up with standards.
I don't know if modern features on web sites are responsible for this, but all the browsers I have tried on Linux use a massive amount of configuration data, and slurp RAM like a drunkard. I am running Vivaldi at present. When I installed it, I had a look in ~/.config/vivaldi. Without actually opening any websites, there was 57M of config data, in 725 files. Firefox and Chromium are similar. I have no idea what most of these files do. Many of them are SQLiite mini-databases, which are not amenable to poking around with grep and vim.
I have a number of quite complex applications on Linux, and none of them have a configuration system as complex and opaque as any web browser. I am talking about applications such as PCB and mechanical CAD, and electronic circuit simulation. Probably the most complex non-browser application on my system is TeX. However, I would not call TeX opaque. I don't mess about with my own style sheets. My boss wrote a house style sheet, and it is fairly clear what it does.
On the RAM usage, my laptop has 8G of RAM. Not many years ago, that would have been an impossible dream. And yet I have run out of RAM after a few days uptime, to the extent that the system eats into swap. If I am not careful, the machine will freeze when the swap is finally used up. I am pretty sure the RAM slurpage is the web browser, because top() tells me most of the RAM is free if I close the browser, while leaving everything else running. My SPICE circuit simulator gets nowhere near slurping all the RAM, even on a big circuit and very fine time resolution. Vivaldi might possibly be better than Firefox, but you have to look carefully for the difference.
> I am pretty sure the RAM slurpage is the web browser
My old laptop (32-bit Linux) had only 4 MB of RAM, and I never experienced RAM shortage. So it might be a little more complicated than just accusing nasty browsers of gobbling up all memory. For instance, I never have more than a dozen tabs open. Seriously, I can't possibly work on more than one at a time, and I'm not a collector, so there is little point in having a bunch of tabs open I won't need in the next hours/days/weeks (suggestion: You could just bookmark them, and load them when you actually need them). Don't say that's your problem, but something must explain why 4 MB is enough and 8 MB isn't.
You probably need a better tool than top to measure the amount of memory a web browser is using. Chrome grabs all of the memory you have free to cache webpages... but it is just a cache - the OS just reclaims it if it is actually needed.
Having 6Gb of 8Gb RAM doing absolutely nothing in your computer would be a waste, and would make your web browsing much slower...
Note I have disabled the predictive webpage loading (in Firefox). I don't see the point in Firefox trying to pre-load all and every webpage linked to from a given page, just so if I click on a link it arrives instantly. It might make sense on dial-up, but I'm on fiber, loading times haven't bothered me in ages.
Erm, in my experience this issue is that everyone is writing to work on Chrome in a way that isn’t supported in other browsers like Firefox (I’m looking at you HSBC BIB site) and isn’t necessarily supported in the next version of Chrome (which may have been pushed out overnight without so much as a by your leave).
Safari isn’t holding the web back, it’s people coding for one rendering and JavaScript engine (again), and Google shifting goalposts overnight.
The web isn’t Chrome
Mozilla with all the ad blockers, script blockers, and delete all website data and cookies afterwards.
oh, wait.... nothing works any more.
It's funny that my iphone xr is almost not usable anymore, because everything was "upgraded" for bigger screens that I can't even fit in my pocket. I thought that was the purpose of a mobile device? Maybe I have it wrong. Maybe the purpose was to sell me more shit. Maybe the purpose is to make a select few very, very rich on my data? Maybe I'm just being paranoid again. Wait.... am I paranoid?
So, let me get this right. When IE6 came out everybody complained because it introduced all sorts of non-standard features which had to be supported because it dominated the market, and this meant that other, standards-compliant browsers often wouldn't work.
And now everybody is complaining because Safari sticks to standards and doesn't keep inventing new features like Chrom*?
I really, really, hate seeing my CPU cycles being stolen to animate a F**ing advert for something I will never, ever, have an interest in.
Bring back text only and being able to navigate from the home keys.
OK, Yes, I have a number of browsers installed here. I can take my pick of them, but every time I turn to one of them, it feels the need to 'update' and deliver some shiny feature that adds complication to my life.
Simple is efficient, and should be the default. If you want to run some sort of sparkly this week's 'thing' then turn it on and pay the sponsor - in power, in adverts, in plain old subscriptions - but stay out of my interface.
Microsoft was a lost cause years ago, and Android started with the obvious intent of being rabidly commercial. Linux is the way to go, but still mostly needs user input to refine and tune it. Apple is as simple as the power supply. You plug in and it works, and works reliably and consistently across releases. Almost as if they test stuff before release. But you pay. My time and involvement versus the money. My time nowadays is used for stuff that doesn't involve IT (mostly) or only uses it like a real user - just doing what I want to do without all the config and debugging.
So I'll stay with Safari (mostly)
....Safari Technology Preview release 133 (the current release), a.k.a Safari 15.4, clocks in at 85 on the same compatibility test page, which better than halves the distance to Chrome. Do we know if the remaining 11% is evil Web "standards" that seek control of your devices, or if there's useful functionality in that difference?
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
