I'm also interested in new browsers. I mainly use Firefox, and have been using it since its Navigator days (although for a long period I used a derived version for OSX, Camino), and I use Chromium and Safari when I have too. A new browser is welcome for multiple reasons. But the article is very light on details, such as: what features does/doesn't Orion's engine support? It's fast, but the author doesn't even mention if Orion uses Google Safe Browsing or an alternative (which could be why Safari briefly lingers).
As it turns out, Orion is webkit based. That would explain how there can be an iOS version in the works. That seriously reduces my interest. I've signed up, because you never know, but Orion is not an undertaking comparable to Chrome, Safari or Firefox, that will add a much needed breath of fresh air/diversity to the browser landscape.
Orion uses the latest webkit release so it's usually ahead of Safari. The team is working to implement the web extensions API. It's not perfect yet but I use a number of extensions with no problems (Bitwarden, Dark Reader, Request Control, Sponsor Block) and a couple with minor issues (Reddit Enhancement Suite, Mate Translate).
Many extensions are not functional at the moment, owing to the incomplete extension API implementation, but the progress so far is very encouraging.
I also use Dark Reader on the iOS app which is pretty cool (haven't tried any other extensions on iOS).
I can see the appeal in particular on iOS: having real extensions is attractive. It will influence speed, though: Firefox is much faster when I start it without them. So the perception of Orion as a snappy browser might conflict with its ability to load extensions.
I think the time taken by extensions loading is order of magnitudes lower than the time saved by skipping the hundreds of MBs of ads every page has to offer.
> the author doesn't even mention if Orion uses Google Safe Browsing or an alternative (which could be why Safari briefly lingers).
Aside: do you keep Safe Browsing on? I always turn it off, as I'm pretty wise to phishing / download buttons etc. I don't know how much of a performance increase there is, but I'm sure I've saved quite a few seconds from not having it on.
I have it on, but I can only remember one time it warned me, and that was when I knowingly opened a potentially malicious page (after a colleague had clicked a suspicious email). I'm always doubting about turning it off: Google is not your friend. But I'm not even sure how it works. I did read that Apple proxies it on iOS, so I suppose Firefox does expose part of the URL and your IP address.
As a product manager I just want to say how impressed I am about Orion’s decision to not do any form of in-app tracking and instead to actively solicit and act on user feedback.
This kind of decision takes courage in this day and age of detailed usage analytics. I might try it out in my own projects.
This is not true. The only way they will be earning money from Orion is through an optional "Orion Pro" subscription of $5 dollars a month, or $50 dollars a year. This gives you direct communication to the developers, allows you to influence the development of the browser, and have access to a nightly builds of Orion. Check it out here: https://browser.kagi.com/faq.html#business
I think what the comment you're replying to meant was that the product team isn't capturing and reviewing user activity to figure out what the user's are doing or wanting.
As GP, to clarify, that was my takeaway - that they intend to forego in-app telemetry completely. Apologies if I was mistaken in my assumption but that was my read. If true, it is a bold move and a strategy I would love to employ if I could convince others on my teem of the merits.
See my sibling comment to yours for clarification on my mistake, but to respond I don't think Vlad intends on selling the browser since he needs it to promote his search engine, Kagi Search. Doesn't completely mitigate the risk but I think enough since its now tablestakes in the search engine space to offer your own browser, Brave is the same way.
> Orion supports both Chrome and Firefox extensions
Wow, this sounds amazing!
from a reddit thread: [1]
> Overall really snappy, but almost all of my Chrome extensions didn’t work properly, except for UBlock Origin which was the headlining feature on their website.
That's not good. I guess I'll download and see if my preferred extensions work with it. Other redditors didn't have issues with their extensions (though noted it was "hella buggy").
UPDATE: It does seem snappy (not much faster than Brave, if at all), but the first two extensions I tried didn't work great. Tree Style Tabs (Firefox addon) only appears on the right side of the window, and I can't find a way to move it to the left. BeeLine Reader (Chrome extension, I am the creator) runs fine, but the settings panel shows all the tabs at once, creating a very wide mess of settings.
I second the bugginess - I gave it a spin (despite my paranoia...) and a lot of my favourite ff extensions don't work well, if at all. Chrome extensions fare better. I'm really tempted by the idea of a way more battery efficient browser that can still use ff extensions, but this isn't it (yet).
Extension support in Orion is still experimental and product is in beta. The way to make it better is keep giving it a spin and contributing feedback to orionfeedback.org (a big ask I know!)
A significant chunk of the web extension API is not yet implemented so some extensions do not work. But having my bug reports for extension issues addressed is very encouraging, as are my interactions with the team creating Orion. I think they will make it to the finish line.
I’ve been using the beta for both iOS and macOS and tbh it’s just “meh”.
It’s still in beta and as such has quite a few bugs. I’ve drifted back to FF.
Although I gotta say I love Kagi search. I hope there’s a way to use the service TRULY anonymously though, because a paid service requires an auth token. If they provide a hidden service, a useful api, a cli tool, and a way to pay/use anonymously I’ll gladly pay a hefty annual/monthly subscription.
It’s meh compared to existing mainstream browsers but being in beta still I don’t fault the project.
The support for standard web extensions is incomplete and even ublock origin doesn’t work properly. Hopefully they can iron out the kinks of translating the standard web apis to WebKit.
For me a browser needs to be available on all my devices. Android, iOS, Mac, windows and Linux. And it needs to sync bookmarks, history and passwords across all of those. Anything less and I go back to chrome or Firefox. Extra features just does not seem worth losing this when they are both excellent browsers already.
I partly agree with you. To me, bookmark sync is essential (I didn't look into options), history sync isn't as important, and password sync is essential (Firefox is a second-class password manager on Chrome or Windows/Linux or Android, and BitWarden is a second-class password manager on the Web and not perfect on Android either).
Sure I agree. Password is a strong first, bookmarks is a second but still very essential and history a fairly distant third. But Firefox and Chrome both have these so I don't feel like downgrading to not having them.
BTW, I would like to have live sessions where I can have the same tab session open on a second computer/phone, or connect to different tab sessions from different computers. I don't think anyone currently offers this.
I have that issue as well and end up using xBrowserSync to keep bookmarks synced between Chrome (Android) and Firefox (MacOS and Windows). Maybe worth a shot. I'm on the waitlist for Orion so I don't know if it works on Orion yet, fingers crossed!
I wish Firefox would natively let us silently take full page screenshots from a keyboard shortcut. It is (currently) a three click operation, but they seem to change the steps and graphics for it every year or so.
I just got a new MacBook after using Windows, and am trying to do things the "Mac" way, before I attempt to customize it too much. I kinda like Safari, although it has weird bugs (e.g.: if you use tab groups, it will randomly navigate back and forth). So I like the idea of a better version of Safari. The killer feature is the support of WebExtensions (ad blocker!)
One thing that is really important to me is bookmark sync. It would be really great if you could implement Firefox sync (the source is available, I'm not sure if there are license problems in using it from a third party browser. In any case, if your extension model is powerful enough, somebody could write an open source extension to do it). This is the one feature for which I could imagine buying a pro version.
Another thing that I think would be cool is a tab filter. Press a hotkey, get a list of all tabs. Then press "hacker" and it highlights all tabs with hacker(news), and ⌘-w to close them. Or you could move all tabs with "office" to a new window. Or maybe you find a clever UI to select and close all old, not recently used tabs. This would be great to clean up all the old tabs that accumulate.
> I just got a new MacBook after using Windows, and am trying to do things the "Mac" way
> Press a hotkey, get a list of all tabs.
I recently tried a Mac for six months, coming from KDE on Debian-based systems. I also tried to do things the Mac way.
I hate to break it to you, and am really trying to not be condescending, but despite Mac's famed usability a hotkey and then filter is too "hacker" to be "the Mac way". The Mac way is one click to get to things, even if that click generates dozens of possible things to get to. Filtering, completion, etc feel like a Linux accent on Mac (I don't know about Windows, haven't used it in two decades). You probably already know that e.g. Tridactyl offers such features in Firefox but I get trying to use a Mac browser.
If you're jumping into the Mac Way, go all the way. Maybe it is better, I don't know, but sticking to your old ideas of what a workflow should be will just make the Mac frustrating.
There is the way the Mac operates for a novice. That is entirely left clicks on icons.
There is the way the Mac operates for someone like me, who has been using them since the days of Classic. That is full of hotkeys and advanced features discovered over the years. A hotkey to invoke a filter would be entirely in the scope of this Way of the Mac. (And the filter would also probably be implemented as a “search” field somehow connected to the tab bar, for the novices to discover it.)
The Mac Way includes things like Automator (or whatever they're calling the current incarnation of their attempts to make generating simple scripts a point-and-click task). The Mac Way includes the Keyboard panel in the prefs, which lets you change the menu hotkeys of any app on your machine.
> I hate to break it to you, and am really trying to not be condescending, but despite Mac's famed usability a hotkey and then filter is too "hacker" to be "the Mac way".
This has been built in to Safari for quite some time now. View -> Show Tab Overview (default hotkey Cmd + Shift + \). And then you type to search.
A great example of Cunningham's Law, thanks :-). You can also make a Pinch gesture. And with a German keyboard, it is Cmd + Shift + #.
This is great for finding a tab. The thing I was proposing would go a step further, and allow you to close all the filtered tabs or to move them to a separate window, to clean up your browser.
> despite Mac's famed usability a hotkey and then filter is too "hacker" to be "the Mac way". The Mac way is one click to get to things, even if that click generates dozens of possible things to get to. Filtering, completion, etc feel like a Linux accent on Mac
The ubiquity of spotlight (for example, the global CMD-space hotkey) throughout Mac OS would seem to belie that view
Yes, spotlight is the single, centralized such facility on a Mac. That kind of proves my point - other Mac applications (as opposed to cross-platform apps such as VS Code) do not reimplement this interface.
That said, I'm happy to be shown wrong by someone with a username "swiftcoder"! I do not pretend to be a Mac expert, just someone who got frustrated trying to do things the Mac way. I've got a list of frustrations, if you'd like to dispell some more.
You are correct, but spotlight is a lot more ubiquitous than folks tend to give it credit for.
It powers the search in Finder's view of the filesystem, as well as "smart folder" filesystem filters. It powers search boxes in the Open/Save dialogs across every application. Every native app can choose to integrate its internal data model into Spotlight (via Core Spotlight), and use it to either provide in-app search or plug it's own data into the system-wide spotlight search...
Unfortunately, a lot of popular cross-platform apps (VS Code et al being obvious offenders) don't choose to provide spotlight integrations, so a lot of folks have never really been exposed to the full suite of capabilities
To be clear: I think that feature is something that would be useful on any desktop OS. I actually started implementing it as a firefox extension on Windows/Linux, but never had time to finish it. A feature like this would be orthogonal to the "look & feel" of the platform.
With "Mac way" I mean forcing myself to learn all the shortcuts and gestures, instead of redefining them to be like on Windows, and using built-in or native apps when available.
Nevertheless, if I look at my colleagues computers, the "Mac way" seems to include hundreds of open tabs just like the "Windows way", and I was suggesting a new tool to deal with that issue.
> Another thing that I think would be cool is a tab filter. Press a hotkey, get a list of all tabs. Then press "hacker" and it highlights all tabs with hacker(news), and ⌘-w to close them. Or you could move all tabs with "office" to a new window. Or maybe you find a clever UI to select and close all old, not recently used tabs. This would be great to clean up all the old tabs that accumulate.
Idk if this is a good thing or a bad thing, but a browser is one of those things where I simply am not comfortable trying out options that aren't big or well known. At this point, the browser is essentially a window into my entire life, and I do things like check my bank balance on it. While Google isn't a trustworthy company by any means, I can be reasonably certain they aren't looking into what I'm doing. So trying out a new, closed source browser - even if it is as great as everyone makes it seem - is a bit too much for me.
If you don't think there's a big difference between "indirect means" and literally examining bank records then please try this new browser I just invented.
Are there banks that don't? I haven’t come across a bank (in South Asia at least) that didn't have the option (and encourage you to use it) to receive statements as password protected pdfs.
Here in Australia, I get an email when I have a new statement ready. I have to log in to their website to download the PDF. Its pretty annoying, but I suppose it keeps me safe from things like this.
Google has a pretty good idea of your household income, among a huge load of other things. Go check the "Ad personalisation" page in your google account settings.
I agree -- also, in my (limited) experience, keeping a modern browser up to date security-wise is a major undertaking. They talk about forking WebKit -- how well are they doing at keeping that fork up to date, when 0-days appear worryingly often in modern browsers?
Orion is usually ahead of Safari as it has a much faster release cycle and uses the latest Webkit version. I understand there was a bug in a recent webkit release so Orion rolled back to the previous webkit release - but besides situations like this, Orion uses the current webkit.
Thought experiment: What if there was one browser to use for the "serious" interactions with the web and another, or several, to use for non-serious, recreational uses of the web. The later might not even be a browser, it might, e.g., be a small utility to download content that can be consumed offline. Arguably the title hints at this idea with its inclusion of the adjective "main".
This is not a thought experiment for me. I have actually approached the web this way for many years now. Using a single program for 100% of web use has become unthinkable.
I am in 100% agreement. The best tools have enough power and flexibility built-in that they end up serving use cases that the authors never dreamed of. Meanwhile, these days when developers attempt to write a piece of software that's everything to everyone, they only end up satisfying the most basic possible use cases because anything more would cause distraction and/or intimidate the users away from their product. Those who want to actually harness the power of what's possible with software are left out.
If Vim and Emacs were released as new projects today, most of HN would roll their eyes and complain that neither is "opinionated" or "minimal" enough to rely on for serious work, and let's not even get started on the lack of built-in GPU accelerated text rendering...
"("using the API of local services" seems to be an ongoing theme, one of my favourite things is to use secret undocumented APIs where you need to copy your cookies out of the browser to get access to them)"
The aboove is from another submission's comment thread currently on HN front page.^1 Not everything can be done satisfactorily within the browser. Other (smaller) programs are useful. In the later case, the programs are not supported by online advertising.
I do something similar. I have a "hardened" main browser for my serious stuff and everyday browsing needs, and I have a Chromium+uBlock install, which I mainly use in Incognito mode, for sites that break in the "hardened" configuration.
A perfectly reasonable/sensible point. If you're like me, though, and want to test-drive a browser for the hell of it, you might do your banking, shopping, flight reservations etc using your usual browser, and just use Orion for general info/entertainment surfing. A bit like driving an ordinary car for the daily urban grind, but riding a motorbike out in the countryside at the weekend.
Edit to add: yes, I realise that Orion wouldn't be your "main" browser, with just the casual usage I envisage, but it could be interesting - and maybe, if it eventually takes off and earns a rep for privacy/security, you (and/or I) might feel sufficiently informed to make the switch. I guess I feel that the browser space lacks choice/variety atm.
Orion solves the question of trust by being a zero telemetry browser by default.
If a single byte of unwanted information ever leaves Orion (and this is easily detectable with any network proxy), that would spell the end of Orion's credibility.
A web browser is more than just for webapps. You don't have to use the same web browser for everything.
Surely there is a case to be made for a web browser that caters to a need for focused reading, where a quick context switch to logged-in sessions wouldn't be something you would appreciate having.
At that point, you wouldn't need much of the fancy javascript served to you, and can trivially cut down on attack surface with minor sacrifice.
> While Google isn't a trustworthy company by any means, I can be reasonably certain they aren't looking into what I'm doing.
In what way are Google not a trustworthy company in your view?
Expecting that they'll be looking into what I'm doing is exactly what leads me to consider them untrustworthy so I'm just curious what else you're thinking of.
>In what way are Google not a trustworthy company in your view?
Google "snowden nsa revelations" see all the companies sucking up all your data, the NSA and google and all the tech commpanies have removed everyones privacy. It's called "total information awareness".
If your definition of being snooped on is whether confidential information is leaked or nothing is happening, that is short sighted.
Google analytics is ubiquitous and you're tracked to your bones especially if you're logged into your Google account and while they don't read your Gmail email content, they put you in the social graph according to any email address or URL in your email and they have better idea about your past more than you do.
I still use Google search as there aren't better alternatives but I never use Gmail nor Chrome as I have plenty of alternatives for those.
I'm not happy about tracking (I personally use Firefox + uBlock) but with the state of the web as is, I'll settle for the browser not stealing confidential info, thanks.
Depends on what you mean by "read". Google (and MS, and likely others) certainly parse the content. For example pulling out times for upcoming flights you've booked without you needing to click into the email.
Obviously I have no way to know if they actually do anything else internally with the result of parsing. I have to assume that they do though.
Chrome's entire reason for existing is that Google wants control over the client, so they can control the capabilities of the web. If they are snooping on the data of webpages or what you type, they will lose that. Also the fact that millions of eyeballs are on it, including security researchers who would be more than happy to report it.
There are dozens of "phone home" features in Chrome that by default leak everything from what you type to every single url you visit. Check out https://github.com/Eloston/ungoogled-chromium to see the extent of patching they need to do to shut some of those pipes off.
This is a bit off topic I guess, but I don't think the goal of fixing a slow Web can be achieved by building a faster Web browser. Developers will build and test (or not test) things on their powerful Macbooks, and usually ship things that aren't noticeably slow for them, then a user with a $300 Windows laptop from Walmart will struggle to click on things. This is the nature of web dev. Increasing the speed of the browser just means web devs will accept that code is fast enough when they use it, regardless of whether or not its fast enough when a user on a less powerful machine uses it. Essentially, web pages will always expand to fill the available resources of the dev's machine and if you use their code on a less powerful computer you're out of luck.
If we want a faster web we need to get web devs to test on slower browsers and more realistic target devices.
(I can actually think of an alternative solution that browser vendors could do ... Browsers could just display a blank screen whenever a web site falls below 60fps. Websites would get a lot faster pretty quickly if Chrome did that.)
EDIT: Actually... This is possible. Chrome should display a blank screen when a website falls below 60fps when devtools is open. That would have no impact on users, but would push web developers to fix their slow pages. I'm genuinely tempted to find the time to make a PR...
I'll make the PR if you don't have the time. It's a great feature and aligns perfectly with Chrome's (original, stated) reason to exist: to make the web better.
Another trick to deal with heavy web requests is artificially limiting bandwidth. By simulating a slow 3G connection using devtools (yes, its a feature!), developers should see how much time does it take to load. Personally, I would love to see the slow 3G speed the default
I think it would be great if devtools had a very prominent and easy to enable "test mode" that switched on a few settings like network and CPU throttling. It should be a top level tab like Elements and Recorder.
If browsers were to artificially limit network speeds, it would force web page authors to slim down their pages, which would in turn reduce the aggregate costs to run servers, backbone links, and last-mile connections in developing countries and the rural US. Maybe as a nice bonus, we'd get fewer auto-play videos on websites that the browser authors somehow still haven't figured out how to prevent.
Anyway, it would be far from the first time the web browsers have thrown their weight around to effect changes across the whole Internet...
If browsers were to artificially limit network speeds, it would force web page authors to slim down their pages...
I wish that was true. The truth is that a lot of developers don't know how to do that work, or that don't care, or (very weirdly in my opinion) are actually proud of the fact they build a huge and bloated app.
Moreover though, sometimes those large apps are actually relevant. Having a background video on a brochure site isn't automatically a bad thing - those features can and do drive sales. You might disagree with that about some feature you don't like, but chances are there's some other feature you do like that other people would happily limit because they don't need it. We should be wary of imposing limits on everything because we would lose some things that are actually good for some users. To use a lovely British idiom, let's not throw the baby out with the bath water.
I appreciate that people really need cross-platform functionality, but given that their own website says that the browser is currently in Beta, I think it’s fair to try and restrict criticism to the actual product as it currently is.
Personally, I’ve been using Kagi Search (the same company’s search engine, also in beta) for about a week, and have had a great experience.
I'm very happy with kagi search and gave switched over too. It does rely in google (and others) but seems to retain the "good" aspects of google search without all the crap (ads) and nagging. And by virtue of it being smaller, I don't care as much about them presumably harvesting as much data as possible from me. It will be interesting to see how they scale.
Well, I understand that on some level, a search engine only gets better by utilizing the data in its searches.
Sharing data, or using it for purposes other than improving search, sure, that’s a problem. But I think it’s reasonable to say that they gotta learn something from their users.
The author says that this new browser is very similar to Safari, as if it is a good thing. Maybe it's just me, but to me Safari is a horrible, terrible, very user hostile browser that is only used once to dowload Firefox. And on mobile it is even worse. So having another browser that feels like Safari is a big instant NO in my book.
Stable WebKit (whats on iOS 15 and Monterrey) is breaking <canvas> with its recent GPU Acceleration flag toggled on, look at Issues or boards for any browser game engine based on <canvas> or the Apple Developer forum. I can't believe they didn't test that flag sufficiently like this.
are there any web browsers out there which are created from scratch? ie not based on chromium, or webkit but every single component used in the browser was created for it?
Bonus points, if that browser is actually usable and not just a toy/academic exercise.
Firefox? Konqueror seems to still support the KHTML renderer - not sure how well that works though. Opera moved to webkit. I used to use lynx/links on the terminal.
But these are already existing engines. I believe the engineering effort to build a rendering engine from scratch makes it very unlikely we will see another engine.
More intelligent search would be a huge bonus. I frequently search for text and it turns out that it's hidden on the page, but there's no indication of that in the search results, so it manifests as just a weird 'phantom' hit I have to skip past.
Makes claims about being faster and more leightweight than Chrome or Firefox. Looks interesting, but seems to be another MacOS and iOS exclusive app, so I won't be able to use it. It's a shame that the developer chose native over cross-platform and the FAQ doesn't mention any plans to support other platforms.
> It's a shame that the developer chose native over cross-platform
As I'm sure the developers will read these comments, not it isn't. Thank you for choosing to go native. There are dozens of other non-native, cross-platform browsers. Why would we need another one?
Precisely. I seek out native Mac apps and actively avoid cross-platform junk, since it is slower and more bloated, and doesn’t match the native UI (though Apple itself breaks its own conventions these days).
The founder has expressed interest in developing a Windows and Linux version once they're out of beta and "once they can afford to hire 20 more engineers". Linux would probably come first: https://news.ycombinator.com/item?id=28803789
I think focusing on macOS first is a good move. macOS users are currently stuck between a lagging-behind-standards Safari and a memory-consuming Chrome (not sure how Firefox performs), and there is more of a market for paid software (which is Kagi's business model).
Author describes a lot of features with words, but would have really been great if they added some images of what they are describing for better visualization of it.
I've been trying Orion these days. It's pretty awesome although there are some rough edges. The most annoying quirk is using my Airpods Pro with Google Meet, for some reason the mic input does not always get recognized (. This behaviour also happens from time to time in Safari so this may be not an Orion issue.
Looking at the orionfeedback.org, I'd say that their most important challenge now is deciding what not to do.
One thing that I've noticed is that when running Orion in the macOS Control Center appears a new widget called "Mic Mode", which I've never seen while running Safari.
But yes I am with you, I feel that this issue is not Orion-specific. But surely this thwarts my adoption efforts ;)
I've tried it. The few bugs are expected for being so new but memory usage is the big issue for me.
If you ask me what an ideal browser in 2022 is, it would essentially be a VM hypervisor with UI made for browsing.
I would have paid for a browser that resurrected Servo or something but as much as I like Kagi from the same company, I just didn't see enough differentiator for Orion compared to say Brave.
The problem now is the web, and browsers can no longer fix it without VERY wide adoption of the fixes, and that isn't happening. So any attempts to resist result in being unable to render the web as it exists out there.
One can say browsers created the problem by letting the web get this way, but it is very hard to turn back now that we are here. It will only get worse.
I just requested access earlier this week; for work, I have to use a browser that uses the host's certificate store, and am not enthused to use Chrome or Edge. Safari is OK (not a fan of the UI), but no ad blocking is the worst.
Well, I've been trying Orion these last weeks. I've seen some issues related to Google Meet (but they are probably due to Webkit). Apart from that, Orion browser seems stable enough to be a main browser.
As it turns out, Orion is webkit based. That would explain how there can be an iOS version in the works. That seriously reduces my interest. I've signed up, because you never know, but Orion is not an undertaking comparable to Chrome, Safari or Firefox, that will add a much needed breath of fresh air/diversity to the browser landscape.