Also, any thoughts on having official communication channels on some open, freedom-respecting platforms, rather than Discord only?
And we're perfectly happy using proprietary services like GitHub and Discord as long as they make our work easier and more enjoyable. We recently evaluated a number of alternatives, and found that they all introduced more friction than we were comfortable with.
Although the task of building a browser is itself challenging, we're a pragmatic project :)
Hey, just a reality check: in the event that you actually do become wildly successful, this means that others (Google, Microsoft, etc.) will be able to fork the browser and then develop it faster than you - thus leaving you behind and taking away your users! Would highly recommend leaving yourself some mechanism to prevent that, unless you're really okay with the project defeating itself through its own success.
Neither is Safari. Safari is actually part of the solution. Safari has saved Firefox and other browsers by being the only option on iOS for a long time and a better choice for many (because of battery usage) on Mac OS. Without Safari I am afraid we would all be locked into Chrome now.
The problem is that Google, like Microsoft before them,
1. used their dominant position in one market to force their way into dominating another market,
2. used various underhanded tactics to make users think Chrome were better while in reality it was just given better treatment by their backend servers and also the Googles frontend devs[1]
3. and that unlike Microsoft they still haven't got a multi billion fine for it and haven't been forced to advertise alternative browsers for months.
[1]: see various bugs[2] in everything from the core of the Angular framework to Google Calendar to YouTube
[2]: yes, I am generous enough to consider them bugs. I am fairly certain though that bugs that doesn't affect Chrome aren't exactly considered top priority.
Amazing. https://en.m.wikipedia.org/wiki/1984_(advertisement)
But more seriously: it is actually the truth.
Kind of in in the same way that people are thankful for Churchill: not because he was a fantastic man in every way (he wasn't) but because he saved us from something even worse.
If you have to convince people that you are seriously telling the truth, you are probably making an unproven assertion that relies on many benefits of the doubt.
I even thought it wasn't necessary to test them separately but I recently heard from someone with more and more recent experience that some differences exist, particularly around prefixed css attributes. Can't say for sure though, but that was why I wrote my comment above somewhat defensively.
> Google, like microsoft, <1-3>
If you're going to complain about 1-3 for google and ms, I don't think you can praise safari in the same breath.
Apple's abused their position with the iPhone to make safari relevant, and unlike Chrome and IE, users can't just install another browser.
Apple's behavior is the only reason I can't run my own addons I've written for firefox on iOS (they run _fine_ on android of course), why I can't run uBlock origin on iOS, and so on.
Apple's behavior on iOS is far more egregious than anything microsoft or google has ever done.
I never once had to run IE or Chrome unwillingly since I could always install netscape, or mosaic, or firefox.
I'm forced to run Safari, unable to decently block ads, unable to use the adons I've written, unable to fork and patch my browser to fix bugs, and I've generally had my software freedoms infringed... and if I don't run safari, then I can't talk to my family group chat (no androids allowed, sms breaks the imessage group features too much) or talk to my grandma who only knows how to use facetime.
I wish so much I could use a phone with firefox, but I can't justify having a spare iPhone just to talk to my family, so I'm kinda forced to suffer through safari, held hostage by apple's monopolistic iMessage behavior.
The only thing that comes close to Apple's behavior is Google's campaign to force Chromebooks upon children in classrooms, requiring them to use Chrome, but at least Google isn't holding their grandmother's hostage... and managed work/school devices already are kinda expected to have substantially less freedom than personal devices, so it feels much less egregious.
If people don't support Safari, it's because the free market has spoken and overwhelmingly chooses alternative options: https://gs.statcounter.com/browser-market-share/desktop/worl...
The story would be different, if Apple wasn't miserly with their native APIs and App distribution. But this is indeed a harmful and competition-restricting decision, even in Mozilla's opinion: https://mozilla.github.io/platform-tilt/
So I think we can safely assume that Apple's policy harms browser diversity by forcing their users to support a single minority option. If their users preferred a more feature-filled browser, we would never know; they aren't sincerely presented an alternative choice. If Apple wants their users to defend Safari, maybe they should invest in it until their browser (or Operating System, for that matter) competes with Chrome. Until then, they're promoting a megalomaniac solution and being a sore loser about it at the same time.
You mean the company dominating the internet heavily promoted and pushed users towards its own browser.
> If their users preferred a more feature-filled browser
Where by "feature-filled" you mean "all the Chrome-only non-standards because free market or something"
If the company dominating their hardware did any better, maybe the majority of them wouldn't leave Safari. If Apple doesn't want to build a competitive browser, then they need some (non-anticompetitive) strategy to retain their users. Otherwise we're doing the Microsoft Shuffle again.
> Where by "feature-filled" you mean "all the Chrome-only non-standards because free market or something"
No, at this point I really do just mean "feature-filled". iOS has notoriously restrictive APIs and it makes full sense that those users would want a browser do do the things Apple prevents their iPhone from doing natively. At the rate Apple's heading, I wouldn't be surprised if next-gen iPhone apps were just PWAs that hook into WebGPU. Big-business has no reason to keep living under Apple's thumb, and market regulators can't justify it in Europe, Japan or even the United States.
Apple doesn't dominate all of hardware. Google, however, dominates major access points to the internet, and used it to aggressively promote its browser.
> No, at this point I really do just mean "feature-filled".
I doubt it
> iOS has notoriously restrictive APIs and it makes full sense that those users would want a browser do do the things Apple prevents their iPhone from doing natively.
Ah. So you are talking about Google-only non-standards
> I wouldn't be surprised if next-gen iPhone apps were just PWAs that hook into WebGPU
Android has been the dominant OS for over a decade now. It has no real or perceived limitations of iOS. We've yet to see a single amazing PWA future we hear so much about.
Then maybe it's time you gave Android another try. Chrome runs on mobile just as well as it does on desktop, so any of the web apps you use on your computer work fine on phone too. It makes modern Safari look like a tofu browser substitute by comparison.
So?
> so any of the web apps you use on your computer work fine on phone too.
So where's the amazing PWA future we hear so much about. All the "amazing web apps" we hear about are shitty slow bad monstrosities that can barely display a few lines of text without jank.
The very few actual great apps which are made ad great engineering effort and expense (like Figma) don't run in full mode on mobile for obvious reasons.
So, my question remains and you haven't answered it.
Edit: There are some web apps here and there which are surprisingly good. E.g. I'm quite impressed by Foodora's app. And it runs well on iOS, too. However, 99.9999999% of the "great PWA future" is just garbage despite the "Chrome runs just as well on Android".
Although Orion also has built in a (simpler) implementation the most important Firefox for me and I assume many others, tree style tabs. Orions built in version doesn't have the full customizability from TST but it works and presents tabs nested by what tab the descend from which is the most important feature.
How is MIT any worse at preventing this, when GPL didn't?
B: "Ah I see that means we should just give up all measures, instead of you know advocating for stronger measures or anything else obvious and logical like that."
This only means we must start any projects today with stronger GPL or similar variants such as AGPL.
You had a security breach, despite using a better set of technologies and techniques.
During the postmortem, you suggest this means we should give up all security or just use the weaker solution, since its all the same, the server would have gotten breached in either case.
Instead of advocating for building a stronger security.
If someone forks our code and does a better job with it than we do, fair game. :)
But ultimately this is all developers' decisions and I respect that. If anything, if a major company decided to take off and invest, they could do it in any case, publishing their modified source code would not make that much of difference essentially. It is really refreshing to see at last a browser that does not absolutely depend on google's resources in any way.
The browser: Chrome. could be kept closed because the LGPL allow the integration of libre libraries in closed products as long as the library itself remains "libre". In this case the library is the renering engine: Chromium.
As a counter example MacOS was built on top of decades of work on the BSD operating system and Apple is under no obligation to give the code back to the BSD project... and it doesn't.
So the most valuable company in the planet took from the community and it doesn't bother to give back.
For some of us that is unacceptabme.
Although... this still doesn't explain why the other parts of the browser besides the rendering engine are open source? i.e. if the license was the reason, then presumably Google would've made the rest of the browser closed source, but that wasn't the case for most parts.
Let me see if I have this right:
>For some of us that is unacceptabme.
1. So, Apple, creator of macOS, "the most valuable company on the planet", followed the rules of the BSD licence,and that is unacceptable?
But, Google, a company that is also highly valuable, and the creator of Chrome, also followed the rules of the LGPL licence, but that is acceptable?
With BSD companies are under no obligation to release their changes, and like any self interested party, most don't.
ANd give very little back. Like Apple on top of BSD or AWS on top Riak.
For people like me that find that unacceptable exist the GPL licenses.
No you couldn't. What would happen is your project would be a complete non-starter for many companies; either in use of, or developing for.
"Ah that means we must remove all security protections, instead of you know further strengthening security."
If older GPL failed, this means we needed a stronger license...such as AGPL, or in future something even better, instead of giving up and saying we should have just given them shit on a platter.
But then Badcorp takes the code and builds their own varient with extensions. Badcorp is big and has lots of market share. Lots of people use Badcorps's browser, and because lots of people are using it, lots of web developers code for it, including coding for its extensions.
Soon, lots of websites -- including Badcorp's own websites, and they have lots of popular ones -- use the extensions in the Badcorp browser.
Then people still using Ladybird can't use it for most websites. They have lost something.
What have the original BSD users lost? Absolutely nothing. BSD still exists, it’s still maintained, and people can still use it. They can also use fruit BSD if they want.
With an OS core, interoperability isn't really important. Existing BSD users presumably weren't too interested in buying shiny new Macs to run their BSD OS on, so Apple using BSD as the core of their OS really didn't affect them. Moreover, existing BSD users didn't need to interoperate with the new MacOS users. An OS isn't some kind of network protocol. BSD users could work with MacOS users just like users of any other OS, using existing network protocols and other standards.
The poster child for the BSD/GPL argument on the GPL side is usually Microsoft's "embracing and extending" of Kerberos. It's a network authentication protocol, licensed with a BSD-like permissive license, and Microsoft infamously forked it, creating their own proprietary extensions. This resulted in only non-MS users not being able to fully interoperate with MS users.
We do already see cases now where web developers write websites targeting Chrome-only browser extensions instead of sticking with standards. In theory, if this happened with Ladybird, it should be possible for the original devs to simply add their own versions of these extensions, but how feasible that it I'm not sure. Currently, there's Chrome-only extensions which apparently haven't been implemented by Firefox for some reason, so maybe it's not as easy as it sounds.
1. All the BSDs have been out there for decades without anyone running with it.
2. Google and Microsoft - while being a shadow of their former selves technically - are probably still very capable of reimplementing whatever they want.
3. If Ladybird gets so wildly popular, lets celebrate wildly!
I am aware that it builds on BSD.
Yet BSD is very alive and nobody who wants BSD is lost to Mac.
At least I personally have never heard anyone deliberating over a free BSD vs Mac.
Edit: and of course upvote. Apple ran with it. But they didn't run away with it. We still have it. Actually we have some patches thanks to them. As I mentioned in my other reply: Open source is not a zero sum game.
In a relative sense, I would argue that Apple has pilfered an order of magnitude higher value from the community than they have given back. The only example of Apple's net-positive contributions seem to be CUPS and LLVM, both of which were cross-platform before Apple took control. Compared with how much networking and userland code they've taken it feels like a trillion-dollar pittance. Even Microsoft chips in more.
They do? With what? (Besides Linux kernel drivers that are only useful for running Linux on their own VM solution.) I guess VScode, for people that use it?
CUPS is fantastic, though you're right, it was cross-platform before Apple took control.
I take objection to the use of "pilfering" to describe usage of software according to the terms specified by its authors.
Or would you somehow argue that some features disappeared from BSD thanks to Apple copying their code as they were expressly allowed to in the license?
Furthermore, even if it wasn't free software but rather MS Windows or a "pirated" movie many people here would argue it wasn't theft but just unauthorized use.
As their primary workstation?
(Yes, PS3 ran Linux in the beginning.)
Have you caught anyone deciding to go with Cisco instead of BSDs on their servers or their laptop?
I'm serious here: Open source isn't a zero sum game.
Partially thanks to the permissive license of BSD we now have both Mac OS and JunOS (edited: it said Cisco first), which is a good thing, not a bad thing.
The problem with Chrome isn't that it exist but that it has been forced upon us and the fact that we know they have used questionable methods to establish it as the dominant browser.
Cisco IOS is absolutely not based on BSD - it is a proprietary kernel, and such that it even has a “userland”, a proprietary userland.
IOS XE is based on Linux.
Most of the voice stuff is Linux.
Perhaps you are thinking of Juniper’s JunOS, which is based on FreeBSD?
It is used in Cisco's email and web security appliance, which is also their hosted offering. This appliance was previously known as IronPort, before being acquired by Cisco.
Slow computing it's sometimes called [0]
I sometimes experience some friction (really acceptable though) on Firefox, it has never lured me to Edge of Chrome. Some people have standards you know ;)
It's rather condescending of you to assume that the developers of Ladybird aren't fully aware of the consequences that their choice of license entails.
Also, I don’t think intellectual property is real and so I don’t think I can make demands on the users of my code: it seems to me that there’s an implicit contradiction in the GPL between the FSF’s anti-IP stances and their attempt to control how their software is used using IP constructs.
Secondly, my game theoretic argument. Lets say a powerful corpo takes your code and makes something that turns out very useful and popular. If it was BSD, they have no obligation to share anything back and your original project is left to dry and rot. If it was (A)GPL they are obliged to return the changes, and then you can absorb their changes and beat them to the punch. Its more competitive and creates a stronger capitalist environment. BSD's end state is feudalistic, GPL is capitalistic.
Also, I trust you'd be happy to have your code MINIXed, to have your code end up in closed source bootloaders locking down your new ARM laptop, and so on. At least with GPLs you get some code dumps. At least with Androids we end up with a begrudgingly shat out Linux source dump, that helps a little at least. With iPhone you got, absolutely fucking nothing, nil, nada. I am just suggesting please do not complain if BSD works ever end up creating such a world.
Yes, this is implied by my view that IP is a fiction
Whereas if the company had forked a BSD project, there is no such legal recourse for me if the company chooses not to share the sources, at best you can hope to talk to them/pressurize them but of course that is most cases futile. As a user its much more inconvenient for me, I need to use advanced debuggers, disassemblers, etc to debug or modify. Sometimes even that does not work.
As a user, GPL/AGPL provides me far more convenience by default than "permissive" licenses do. It gives me an assurance I can just as easily see and modify the sources of any forks of the project, and if in any case such fails, its only because it wasn't strong enough, for example using GPL software in SaaS, due to which stronger licenses like AGPL were invented.
So, that fetish for infinite growth... can we get rid of it?
Firefox keeps trying to grow in various directions and look where it's taking them.
License "permissiveness" is a relative concept. From the point of view of the users of your software, the GPL is more permissive than MIT, since they have permission to see the source code. If you release software under MIT or BSD licenses, you allow middlemen to strip this right to users of your software.
I know that’s not what you meant, but it is what you said.
Permissive licenses doesn't grants you less freedom than GPL, infact it grants you more because the user also has the freedom to modify source code without being enforced to make it public.
Companies copying the codebase to their propietary ones won't automatically strip right of users, licenses don't work like that, the original codebase will still be fine. Whether said companies will contribute back is irrelevant.
I license all my stuff with permissive licenses because (in my opinion) they are more free than the GPL and such licenses. I don't have any massive for-profit company pushing me to do so. Mr. Kling is also not a massive for-profit company, he's just a guy making the software he wants to make. Your argument is in very bad faith.
That's not true.
Somebody can take the source code and build something closed on top of it, but the original code will be already free, and you will always have the right to see it.
For example, PlayStation OS is based on FreeBSD (AFAIK). They took it, adapted it and added a lot of stuff. Did you lose the right to see the source code of FreeBSD ? No. Can you see the source code of PlayStation OS ? No, but you never had that right, so you have not been stripped of anything.
Of course it doesn't change the original project. But when people take the codebase and build a new product on it, what GP says is absolutely the case. The devs can withhold all code and rights to it from the next user. This is most commonly an issue when it comes to libraries rather than end products, but not always.
It doesn't also have to mean that the original project dies or disappears, it can just rob from their growth potential. Examples are quite easy to find. There's been a big hullaballoo over cloud providers taking open source projects and competing with them by offering managed versions of the service that are well-integrated into their ecosystems. Economically this is also a problem because the cloud provider can then undercut the price of the managed service compared to the official one since they aren't bearing the burden of building/maintaining the codebase.
I'm by no means against "permissive" licensing (MIT, etc), I think they have their time and place just like GPL, etc, but I am against dismissing valid concerns with shallow replies.
The first freedom that GPL-lovers have is whether or not to use the project.
At least not now or the foreseeable future. I also don't think community support would work towards that.
I'd favor the more permission mit/isc as long as reasonable myself.
No you don't. You're being extremely disingenuous with this phrasing. No matter how many other parties take the source code and make a closed source product out of it, the users of your software will always have the same rights you granted them to begin with. No freedom has been lost.
And before you say "but your users won't have the same rights to the derivative works", that isn't a loss of freedom. They never had those rights to begin with, therefore they cannot lose them. Not gaining something is not the same as losing it.
https://gavinhoward.com/2023/12/is-source-available-really-t...
Github I can understand to some extent, it's a convenient temporary staying place until they can afford, community-wise, to move to something truly open, but Discord? In this day and age?
What’s your recommended alternative?
So, IRC, Matrix (these can even be interconnected) for synchronous, mailing lists or forums for asynchronous.
And of course issue tracker, where some topical communication can happen as well, but that could completely be covered by mailing lists.
There's no reason to ever have anything non-open in your FOSS project's infrastructure.
Except, that it is easier to set up? What you describe sounds reasonable. But someone needs to set all that up and host it and needs to be trusted etc.
So I guess if many people get involved and do set this up, they would convince the rest of the team to join them. But right now it is just a demand for more work for them.
Because that is a project in itself and as much as I despise discord, I understand why they just went with it.
And as for me being there for years to come, who can say that for sure? I would absolutely try to. For that matter, will Discord still be here next year?
Probably way more likely, than a random volunteer. Kling said they used discord because it was easy. No energy spend there. Migrating the communication structure is work and energy. Friction.
But so far I also did not followed developement closely, because I did not set up discord yet and do not really want to. But I realize I am not in the majority here and I do not think I am in a position to demand anything here. Your words sound a bit like demanding something.
As for random volunteers disappearing, the protection against that is easy - increase the bus factor, by multiple people having necessary access, and by keeping your infrastructure documented. This should be table stakes for any project that becomes even marginally popular.
Then, if a random volunteer disappears, things might slow down due to lack of manpower, sure, but not come to a screeching halt.
You can't increase bus factor with Discord, since it's out of your control.
That is not easy. It works till the first drama and then some angry person deletes it all, or changes and takes the keys with them. So you have to really take care who to trust.
"If the project leadership does not see a value in using open infrastructure"
And I do not think he ever said that. He just said it was easier this way for the moment. Not that discord is meant to be the solution till eternity.
Simply put, this approach is proven to work. We can always "butwhatif?" ourselves into a corner and get paralyzed because of that, but it's senseless.
So I totally agree, that there are tons of examples where it worked and in the long term I also think they should switch to something better, but for the moment I see why they don't, as it is working for them as it is.
(However, GitHub is accessible by git in case you only want to download the repository, regardless of what else they do; however, having multiple mirrors on other services as well can be helpful)
Discord IS the platform of this day and age, what the hell are you talking about? You might not like Discord for whatever reasons, but trying to make it sound outdated or legacy is very weird sounding.
It is a non-starter.
Not public. Got to register an account and accept Discord ToS first.
There was a hype for discord about 5 years ago for EVERYTHING - discord servers were made for every little thing and there was little to no objection about it.
But in recent times, I have seen many people complain about the lack of searchability, discussion-thread management, and other stuff in Discord and moving away to forums, especially for software projects. There is definitely a lot more disgruntlement with Discord today, so their statement makes sense.
So he knows that corporations can take open source browsers and make it proprietary.