One important thing not mentioned here: Apple should make sure to keep up Safari Development. It's by far the worst browser available and the web platform is only growing.
On a Mac, Safari is the best browser (as defined by being fast, battery efficient and native app) by a far margin compared to other mainstream choices, only matched by the Orion browser (but I may be biased).
Since Internet Explorer is not supported anymore by most websites Safari has become the new worst browser in regards to having really bad CSS and Web API Support.
Yeah maybe you as a developer want some fancy CSS and bleeding fresh Web API but I as a user don't.
The web is too bloated as it is, I feel that Apple is on my side putting a damper on that.
Safari is powerful enough as it is. You want voip, notifications, codecs and so on - do the work and learn to write a native app. I'd rather pay for that than paying with my attention to ads in your "free" web app.
I was a web developer in a previous life. You can write basic standard code and Firefox/Chrome/even old Edge works fine, but Safari has a million little things that are buggy or don't work, or break at random in new versions. I wasted so much time hacking random bits of code to support iPhone. In 2021 now that IE is finally dead, I'm sure Apple is singlehandedly keeping BrowserStack in business.
The first one that comes to mind is the <input> select method doesn't work[1], although MDN claims it does. It works _sometimes_, for reasons I couldn't discern. I'm not sure about you, but I wouldn't call selecting text a fancy or bleeding edge feature.
Why? I want it to run on all platforms. That's one of the best things about the web. I'd rather just not support Safari than write and deploy separate MacOS, Windows, and Mobile applications.
+ I don't want my users to download to have download a native binary that has unrestricted access to filesystem and I need a Apple Device to develop even if it's the smallest thing in the world.
Web browsers are the best of both worlds, secure enclaved sandboxes, easily reachable by just a link, and powerful enough to do many basic tasks, and cross compatible.
Web browsers are also huge, complexity-ridden, resource-hungry monsters long past their original mandate and currently well on their way to become a gigantic mess that include everything and the kitchen sink.
As an user, I do not mind downloading and installing slim and fast native application which take advantage of the hardware my devices provide and integrate nicely with my OS.
I personally do not mind the state of web browsers anymore, to be a web browser today is quite akin to the operating itself now (Let's not talk about Chromebooks), It's a very complex, resource heavy monster that abstracts away hardware differences from the software, and I love it. Web today is a mess, What we thought a browser will do, read HTML and render it on the screen, has been well offloaded to well, "web apps", and web browser just is a tool that runs them, quite similar to what the OS does.
What I like about web is how "safer" it is compared to well running binaries on the OS, me visiting a site would not give it access to all my private photos, and potentially allow it to add itself to startup everytime I open the OS without my permission.
Well I as a developer just want a browser which doesn't break existing features: localstorage, indexeddb were completely broken for months this year.
Like CSS containment is feature that was shipped in Chrome in 2016, Firefox 2019 so how many years have to pass for Safari to implement it so it isn't just 'some new shiny' feature anymore?
Yeah, this is the kind of entitled mentality that is going to sink the MacOS ship. I'm not writing a native app just for basic web browser functionality. If the choice comes down to re-writing everything for MacOS or dropping support altogether, I have no qualms cutting out less than 20% of the desktop market share.
As an end user, I felt like Safari should have done better. And extension for Safari are gated behind app store only (no sideloading the extension).
Also I came across a few sites that Safari couldn't render it correctly whereas Firefox and Vivaldi show the site just fine. I only use Safari as a last resort and only used it for Zoom/Doxy/Whereby/etc.
It's interesting that your list defining the "best browser" doesn't actually include properly rendering web pages and the general issues that safari has as a web browser outside what you mentioned. "Fast" is interesting, as if something doesn't work, is it really fast? I routinely see how Safari fakes speed. Being able to read a web page before someone else doesn't make it faster if you can't also interact with that web page. And while battery efficiency is critical, rarely do I know someone who is only using safari. And while being "native" is nice, it also means its use is limited to just the Apple ecosystem.
Best means different thing to different people. I have opted to define exactly what best means to me, so that we can get comments like yours to further the discussion.
Yes - Chrome, Brave, Edge, Firefox... all use non-native UI components in their macOS apps (even iOS), which makes them look and feel out of place on a Mac.
Same experience. The amount of css "quirks" I have to work around trying to get a site "mobile friendly" is absurd. The navbar issue, and the scrolling still not being disabled with overflow:hidden makes me bang my head against the wall.
For iPad oh boy they sure added new niceties such as:
- Not being able to click a tab unless you drag it to the center of the tab bar
- Make a "drag a link/anchor into a new window" feature; except it triggers way too easily from ordinary scrolling, forcing you to spend the next 15-30 seconds figuring out how the FUCK do you close it with an non-obvious swipe gesture. Apparently buttons with "Close" or a big X are too complicated.
Oh my god the inadvertent link drag to new window (or just overwrite current window) has been killing me, glad I’m not the only one, means there’s a chance it gets addressed.
It is always behind on web technologies and hasn't implemented may features that have been around for a while.
It only gets updated once a year, so any updates they do make are slow to hit users. This wouldn't be so bad if they were more up to date with Firefox and Chrome.
Not having web push is a feature. They're the 3rd party toolbar or BonziBuddy of the modern web—often enabled, but rarely wanted—however nice an idea they were in theory.
It's a pure loss for non-geek users. It should be disabled—no dialog allowed, even—by default, if the feature's gonna exist. That hitting "allow" on a pane that pops up on a site—something users are conditioned to do without thinking by all kinds of shitty but extremely common web patterns—affects something outside the current site's browser window is unexpected and unwanted by normal users. Cases where it's easy for sites to affect the behavior of a machine, or of the browser, outside the current session on that site, have historically been a big problem for non-geeks, and this is no exception.
That sounds pretty reasonable to me. But that still doesn't excuse Apple not implementing it. Apple doesn't implement features like this because they want to maintain app store sales. They do this for them, not their customers.
I think if Apple really really wanted to implement it if not for the risk to app sales, they'd have it on desktop Safari, at least. As it is, I just think they don't want to implement it, App Store or no.
From observing non-geeks use the "feature", I think that's the right call regardless. It needs to go back to the drawing board, or be scrapped. In general, anything that expands the reach & capabilities of websites should be treated with a ton of suspicion, as a likely vulnerability vector (leaking tracking info; social engineering/phishing vector; straight-up exploitable bugs) or simply a net-negative annoyance for most users.
You must feel that way about push notifications from apps too then? You seem strongly against push. If you only dislike push on websites, why?
If Apple thinks push notifications are annoying, why does iOS have them? Again, Apple is protecting their walled garden. That's all this is. There really is no other good faith argument for this.
People don't switch between ten just-installed apps in under a minute. They do that all the time on the Web. There's more focus on what you're doing when you open an app for the first time, and installing the app in the first place is a strong signal that you're open to entertaining requests to expand permissions, like to receive push notifications. Browsing to a website is something people do on a whim while barely paying attention, which they don't consider any kind of commitment at all, and they're used to sites going away entirely when they close the tab, not sticking around because they clicked "allow" on a dialog that they don't even remember because the action didn't rise to the level of conscious thought, since the web bombards them with "allow/deny" boxes all day long, where "allow" is just the thing that consistently makes the dialog go away the fastest so they can get on with looking at the site.
Firefox lets me simply deny a web page the ability to send a notification when it attempts to do so, allowing me to only let the apps I want do so, while not bothering me about the majority of web sites that don't use it.
I'd love some data on what percentage of web-push notifications going out are to users who want them, versus to users who didn't realize browsers had added a feature to let websites shit up their desktop, and, like most folks, just click "allow" on everything that pops up on a site (as there's rarely any reason, for normal people, not to) and don't know why or how some site they visited once is sending them stuff.
My experience from my immediate friends and family support circle is "0%" and "100%", respectively, but I'd be open to data showing that the first number is slightly higher than 0%. The "UX" for this feature for normal users is, as far as I can tell, exactly like the bad old days when people'd accidentally install 3rd party "search bars", then not know how to get rid of them.
Historically, browsers have worked to keep web sites from doing things that affect the machine outside that site's browser pane, that normal users don't want or expect them to be able to do.
You cannot use anything outside of Safari on iOS [0]. FireFox/Chrome aren't allowed to have extensions on iOS for that reason (Regular Safari can because antitrust is MS-only).
WebP is a terrible practice and and it has no real value vs jpg https://pagepipe.com/dont-use-webp-image-format/ - just because endorsed by google doesn't mean it's good (i would even argue, the contrary is true)
> Cloudinary says we can reduce that image’s page weight from a 4.6k PNG to a 1.5k webP image. That saves 3.1k. In big letters, they tell us that is a 32-percent savings.
This article seems outdated. Webp supported on Safari. Also this article itself confirms that webp saves image size, just author for some reason thinks that it does not matter. Well, this page is not so fast to open for me, definitely not 0.5 second, more like 5 seconds, so I wouldn't be the one listening to his advices.
I'm pretty sure the parent poster is talking about the page that says you shouldn't use Webp, not about low-tech machazine. The former claims to load in 0.5 seconds, the latter doesn't.
their argument is that filesize doesnt matter because images download in parallel... thats just idiotic. 10x 100MB images downloading in parallel is still going to take longer than 10x 1kB
AVIF is still in process of adoption [1], while WebP is virtually everywhere and you can even ignore fallbacks if you don't support MSIE [2].
Also while AVIF supports lossless images, its lossless compression is known to be weak and can be even inferior to PNG, while the WebP lossless mode is a dedicated algorithm and almost guaranteed to be better than PNG. This is a reason why JPEG XL can be better than AVIF in the long term, where AVIF is just another intra-frame format extracted from video codecs while JPEG XL is explicitly built to support both use cases.
The company I currently work for still actively uses classic ASP in production and they still actively develop that product. Currently trying to get away asap, because I don't wanna touch spaghetti vbscript from '99 with a ten foot pole.
I have originally seen this back on the OG post a couple of months ago.
This has really evolved quite a bit in that time and I really enjoy working with primo.
If we ever plan to make a manned mission to that place, we better pack enough water filters and a nuclear reactor, else we solely rely on a mad man in a blue box.
"The typical use case for this high speed Node.js module is to convert large images in common formats to smaller, web-friendly JPEG, PNG and WebP images of varying dimensions.
Resizing an image is typically 4x-5x faster than using the quickest ImageMagick and GraphicsMagick settings due to its use of libvips."
"libvips is a demand-driven, horizontally threaded image processing library. Compared to similar libraries, libvips runs quickly and uses little memory. libvips is licensed under the LGPL 2.1+."
