Darling – macOS Translation Layer for Linux • Hacker News Board

Re: Darling – macOS Translation Layer for Linux

#31

Post by viraptor » Mon, Apr 29, 2019, 12:41 AM UTC

saagarjha wrote: post #18

viraptor wrote: post #16

Earlier quoted context omitted.

You can't fully test things this way. Not enough to make a serious release anyway. You can easily pass all tests while not being able to even start the app. You need at least one person with the real hardware to do testing.

This really depends on the tests you're running.

True. But in context of: "tests in a CI environment which costs you less than a MacBook", I don't believe anyone would go to the level that avoids those issues. It would take a lot of work and resources to replicate a "real run in clean, real environment".

Re: Darling – macOS Translation Layer for Linux

#32

Post by bgorman » Mon, Apr 29, 2019, 12:41 AM UTC

veridies wrote: post #21

lostgame wrote: post #15

I wish I had the time to contribute to this. Getting Logic running on linux would let me ditch the Apple hardware. It's usefulness is entirely limited to the implementation of the core library frameworks such as Core Image, Audio, MIDI, Animation, Data, etc...which will be very, very difficult, I think, while maintaining FOSS status.

I wish Ardour had some of the features Logic does (most significantly flex time), as well as a nicer UI. Literally Logic is the only thing keeping me on macOS.

Bitwig has a Linux version now.

Re: Darling – macOS Translation Layer for Linux

#33

Post by lostgame » Mon, Apr 29, 2019, 12:43 AM UTC

veridies wrote: post #21

lostgame wrote: post #15

I wish I had the time to contribute to this. Getting Logic running on linux would let me ditch the Apple hardware. It's usefulness is entirely limited to the implementation of the core library frameworks such as Core Image, Audio, MIDI, Animation, Data, etc...which will be very, very difficult, I think, while maintaining FOSS status.

I wish Ardour had some of the features Logic does (most significantly flex time), as well as a nicer UI. Literally Logic is the only thing keeping me on macOS.

That, Final Cut, and the occasional need to use Xcode.

Re: Darling – macOS Translation Layer for Linux

#34

Post by AntiRush » Mon, Apr 29, 2019, 12:44 AM UTC

Maloader[0] is a similar (smaller, easier to understand) project that runs in userland. It's a nice code base to read and there's this nice presentation about it[1].

[0] https://github.com/shinh/maloader [1] http://shinh.skr.jp/slide/ldmac/000.html

Re: Darling – macOS Translation Layer for Linux

#35

Post by sjwright » Mon, Apr 29, 2019, 12:51 AM UTC

linguae wrote: post #17

What would be amazing is if Darling could join forces with the long-running GNUstep project in order to provide access to the Cocoa APIs. This would provide a means to run Cocoa apps on Linux, which will make Linux a more attractive operating system for prospective Mac converts and may also increase the popularity of GNUstep. Also, with many Mac users expressing anxieties about the future of the Mac given Apple's bus…

The problem with replacing MacOS is that its key value lies in aspects that the open source community have yet to succeed in, or are structurally unfit to succeed in.[0][1]

— System-wide UI consistency, especially in the fine details

— Singular UI/UX vision eliminates points of confusion and everything-is-a-compromise choices for third party developers

— Navigable by non-experts, even when things turn to shit

— Nominally "perfect" hardware support

— Robust colorimetry

— Millisecond audio latency

I use Linux extensively, but I'm not aching to leave macOS as my primary platform. Yes, they did start to go a bit loopy and off-track around Mac OS X 10.7–10.9, but from my vantage point it has been all uphill again since Mavericks.

The idea of running Mac apps in a Wine-like compatibility layer sounds worse than anything Apple could ever do to macOS. Yes, I like free. And yes, I like personal choice. But I value my time and I'm sorry to proponents of other platforms, but macOS just values my time more than any other platform.

________

[0] The achilles heel of the open source community stems from the lack of a unifying vision and a top-down approach. This has major advantages as well as major disadvantages. But nonetheless it affects the result. And depending on your priorities, the advantages might outweigh the disadvantages or vice versa.

[1] And to be fair it's equally true in the opposite direction. Open source succeeds in many aspects like robustness, hardware compatibility, longevity and transparency/deep trust. Hence why it has utterly destroyed the server market. These areas of success are not accidental or arbitrary.

Re: Darling – macOS Translation Layer for Linux

#36

Post by netheril96 » Mon, Apr 29, 2019, 12:57 AM UTC

sjwright wrote: post #35

linguae wrote: post #17

What would be amazing is if Darling could join forces with the long-running GNUstep project in order to provide access to the Cocoa APIs. This would provide a means to run Cocoa apps on Linux, which will make Linux a more attractive operating system for prospective Mac converts and may also increase the popularity of GNUstep. Also, with many Mac users expressing anxieties about the future of the Mac given Apple's bus…

The problem with replacing MacOS is that its key value lies in aspects that the open source community have yet to succeed in, or are structurally unfit to succeed in.[0][1] — System-wide UI consistency, especially in the fine details — Singular UI/UX vision eliminates points of confusion and everything-is-a-compromise choices for third party developers — Navigable by non-experts, even when things turn to shit — Nomin…

— Internationalization

Re: Darling – macOS Translation Layer for Linux

#37

Post by apetresc » Mon, Apr 29, 2019, 1:00 AM UTC

beardedwizard wrote: post #23

I'm puzzled, wont most console apps already cross compile without much modification? Isn't this basis of brew/etc? The GUI piece seems like the key, but without it I'm failing to understand why this is significant.

Well, this runs _binaries_ that were compiled for macOS. Yes, practically speaking, the only binaries worth running on macOS tend to be open-source utilities that can be compiled for Linux anyway. But it is a fundamental difference in implementation that makes this project truly exciting.

Re: Darling – macOS Translation Layer for Linux

#38

Post by zapzupnz » Mon, Apr 29, 2019, 1:07 AM UTC

solarkraft wrote: post #26

scroot wrote: post #22

Earlier quoted context omitted.

Every year or so I take a look at GNUStep just to see what's going on. Crazy it hasn't taken off.

Unfortunately - call me superficial - I can't really get past its crusty look.

I'm with you on that, and it seems to be symptomatic of the project's refusal to step away from the legacy of mimicking OPENSTEP. They've made 1995 look like a goal towards which to aspire.

They've added theming support to GNUSTEP, but that's really lipstick on a pig. Then again, I can't blame them too much, the project is mostly quiet and doesn't attract as much attention as the other popular Linux toolkits.

Re: Darling – macOS Translation Layer for Linux

#39

Post by sjwright » Mon, Apr 29, 2019, 1:09 AM UTC

netheril96 wrote: post #36

sjwright wrote: post #35

Earlier quoted context omitted.

The problem with replacing MacOS is that its key value lies in aspects that the open source community have yet to succeed in, or are structurally unfit to succeed in.[0][1] — System-wide UI consistency, especially in the fine details — Singular UI/UX vision eliminates points of confusion and everything-is-a-compromise choices for third party developers — Navigable by non-experts, even when things turn to shit — Nomin…

— Internationalization

— Accessibility (deaf, blind etc)

Re: Darling – macOS Translation Layer for Linux

#40

Post by Mediterraneo10 » Mon, Apr 29, 2019, 1:10 AM UTC

linguae wrote: post #17

What would be amazing is if Darling could join forces with the long-running GNUstep project in order to provide access to the Cocoa APIs. This would provide a means to run Cocoa apps on Linux, which will make Linux a more attractive operating system for prospective Mac converts and may also increase the popularity of GNUstep. Also, with many Mac users expressing anxieties about the future of the Mac given Apple's bus…

> At the risk of going off topic, I still don't understand why GNUstep never seemed to have reached critical mass despite all these years of development.

As someone who was part of the GNUstep scene and contributed a tiny bit of code, I think the reason is obvious: GNOME was getting the big corporate investment, and KDE also had a stable foundation somehow. As you say, the "Linux desktop" very quickly came down to a GNOME/KDE duoply. Plus, GNUstep was written in Objective-C which was outside of a lot of devs' comfort zone at that time. So, GNUstep only managed to attract a few passionate hobbyists but no more.