Back to the Fuchsia, part IV: Google's in-development OS now open to community contributions Google has opened its forthcoming operating system, Fuchsia, to community contributions, but has not addressed the question hanging over it: how will it be used? The OS's Developer Advocate, Wayne Piekarski, has written about a new stage in opening up the project to external input and scrutiny. Put into place were mailing …
Well there are plenty of problems with Android, and Linux, whilst good, isn't the best starting point for all applications - especially those with very limited hardware. Is Fuschia the solution? I don't know. And maybe not do Google. But there are extant problems in need of fixing.
There has been work on other microkernels with a view of making security a formal process rather than a game of bug bounties and whack a mole. That seems a worthwhile endeavour.
> just how limited is your hardware that Linux isn't suitable?
Embedded systems, sensors that process data before transmission to save bandwidth, Pi-like devices that can be powered with smaller solar panels or more novel forms of harvesting energy from the environment, mesh networks of smart dust...
Who knows? But why limit possibilities by starting with something that had a very different design intent?
"What sort of applications are you going to run on meshed dust?"
Localizers were a crucial part in the sf novel A Deepness in the Sky.
Very much speculative fiction as it is.
Is Fuschia the solution? ...... Dave 126
Would Google deny or confirm Fuschia [sic] provides anything more than just another novel gateway for problem solvers?
Ye Olde Worlde divide and conquer methodology that attempts to prevent an overarching and overwhelming singularity of purpose and vendor/virtual agent with dedicated direct streaming of shared thought to unite a mass audience?
While there is little appetite for anther competitor for Apple and Windows (evidenced by ChromeOS), a microkernel operating system is an ideal platform for for Hypervisors.
Fuchsia / Zircon however is more likely to drive the open-source of the Microsoft Executive as a firmware kernel operating system to replace/augment UEFI
And I believe there are L4 microkernel derivatives on enclaves in SoCs by Apple and Qualcomm amongst others.
An episode of the Voices from DARPA podcast looks a project to put tiny microchips on microchips destined for mission critical applications, to ensure that they are the genuine article and haven't been interfered with by adversaries.
Linux has shown that the world really wants an open-source operating system, no matter how geeky and badly architectured. MINIX could've been that operating system were it now for Tanenbaum's restrictive licensing. And the world would've probably been much better off that way, but Linux filled the niche here that MINIX left open.
Fuchsia is an attempt to "do it right" the first time, although I'm skeptical they'll succeed, even with the resources of Google behind them. Microkernels are theoretically much safer and easier to understand but suffer from poor performance and the ones that don't are much more complex.
The problem is that it'll take an enormous amount of effort to replicate the features in Linux and / or Windows in Fuchsia. It will take many years for it to become marginally usable and even longer to get alongside either of them.
and while you can argue all the benefits of the microkernels, I do like the fact that once hardware support was added to Linux once, it stays there for a very, very long time, there is nobody to pull a Creative and decide that no, kernel version 5.0 won't support your soundcard because we said so
"no matter how geeky and badly architectured. MINIX could've been that operating system"
Linux isn't badly architectured. Just because some ivory tower university academics have an issue with monolithic kernels because they're still living in the 70s doesn't mean they're a bad idea. Similarly micro kernels and message passing have all sorts of problems, not least speed - or lack thereof - due to constant message processing and unnecessary context switching. Given that no mainstream OS currently uses a micro kernel whether OSS or paid for I think tells you all you need to know about how professionals on the coal face rather than ones in ivory towers playing with rainbows and moonbeams feel about them.
I personally find microkernels much easier to understand and a more elegant solution than monolithic kernels. There are already quite a few microkernels out there, but the most abundant one seems to be....surprise MINIX which is used Intel's processors. Save for a few very poorly programmed insecure services running on it it seems to be secure.
In addition the newer ones like RedoxOS and TokOS which are starting to gain traction.
The biggest problem seems to be the sheer amount of code which needs to be ported or written for them to catch up with Linux and Windows.
Both Windows NT and Darwin/MacOS use microkernel architecture, but include additional core services within the kernel to trade safety for performance. The Hybrid kernel uses message passing as the API protocol, but skips on permission checking on calls within the kernel.
When you consider NUMA computers, message-passing becomes more efficient because it avoids slow access across the NUMA backplane. Not for a very long time has message-passing involved actual copying because the OS can use copy-on-write pages to transfer pointers between user and supervisor code.
Linux is not a monolith kernel like Unix used to be because i uses modules, but user-mode TCP has highlighted that loading everything into the kernel is not always desirable.
Microkernels might yet have their day for containers hosts and hypervisors
This is the meat of the issue. Is this going to be another Google playtoy to be abandoned when it grows bored? Their track record certainly discourages me from contributing or buying their products.
I only have an Android because Apple is worse and Microsoft is gone.
>> Their track record certainly discourages me from contributing or buying their products.
Indeed, this is exactly what came to my mind as well. The only things that have a chance to endure are what they acquire, everything else they do inhouse is retired, with terrible transition plans to whatever some Google exec power play has got.
Play music, Talk/Wave/Hangouts/Duo/Allo/Meet, Reader, Inbox, Trips, Wear OS
This idea of using the public as alpha testers has damaged their brand - it is difficult to invest in new Google services or ecosystems. The release date seems to be the countdown to closure. Change for change's sake, or sometimes for worse.
I suppose if this is just an OS, and not an ecosystem something might happen.
This is the meat of the issue. Is this going to be another Google playtoy to be abandoned when it grows bored? Their track record certainly discourages me from contributing or buying their products.
I only have an Android because Apple is worse and Microsoft is gone.
I like that you think Microsoft is better having already done what you're worried about Google doing is.
We will adopt this until.
A) we get bored then abandon it.
B) we will get you to do all the hard work, but slowly but surely put in features that are only in our interests and screw everyone else. Sure you can fork, but well have such a dominant presence your forks will be minor niggles to our plans, but in the mean time, hey open source etc.
Interesting that the article's author utterly fails to take note of the most fundamental characteristic of this new OS, that, while the OS is "Open Source", it is *not* "Free Software". Just as Google's Chromebooks ship with "broken" Coreboot, advertised as a form of "security", there is only "security" for the business model, the "walled garden", the "golden handcuffs", of "eyeballs" and "user data". Google's new OS appears to be little more than a path to a future without "Freedom" - all about politics, not technology.
Google has been at it for more than 4 years and the system isn't even halfway there, it seems. There are only short-term technical goals and the overarching goals are vague and too general.
That Fuchsia is developed in several different languages (C, C++, Rust, Flutter) doesn't really bode well for the future and complicates things enormously.
Also, I'd like to see Google hire some external companies to do some pen-testing on the OS so we at least have some indication of its security. Now all we have are promises. A micro-kernel is theoretically safer, but there's very little anecdotal proof in this regard.
Linux adoption has been slow for several reasons. One of the major roadblocks for many is what software titles are available on Linux even if there is an option available that may be completely suitable for the user. This is a major problem for any new OS; users want to be able the box to do something. Bloatware sucks but it has the most titles available many of which are extremely familiar to users. Fuchsia may be best OS ever developed but without a robust title list it will be ignored by most users. True Chocolate Factory is try to develop an OS to use in place of Android and ChromeOS but Fuchsia must have titles available on release. Commercial vendors might be very slow to release Fuchsia versions. Here Linux is a good example, many common titles do not have a Linux version available.
I can continue with other issues about Linux I use Linux as my daily driver.
Linux has been slow to be adopted for the desktop but this is really only a relatively small slice of the overall computing market. Other environments - Windows, most notably -- have a lot of presence in the desktop simply because of institutional intertia, it does the job for the user base and the cost of changing is too high to justify changing.
The key here is to understand what exactly is an 'operating system'. You'll notice that I don't refer to something like Windows as an operating system because it isn't, its a user environment. The OS lives underneath this and like just about every other OS it has some baked in problems caused by initial design oversights. The 'ix' OSes are generally much better in design concept but even they have issues although many are subtle enough to escape notice by the majority of users (unlike Windows........). I'm tempted to go and have a rummage around Fuscia, see what's radically different (or not, as the case may be).
As that wikipedia page tells us, 2+2=5, and it surely should have been included somewhere in the El Reg tale of Google's Fuchsia .......
Some of Apple's personnel and design concepts from Pink and from Purple (the original iPhone codename)[55][56] would resurface in the late 2010s and blend into the Google Fuchsia operating system. Intended to envelop and succeed Android, its open-source software repository was launched in 2016 with the phrase "Pink + Purple == Fuchsia".
.... for that's a pretty impressive creative construction base.
I am not an advocate of opensource and FOSS. I have never contributed or taken from opensource. Outside of occasionally developing programs for myself, I find developing software for no remuneration quite alien. I am not anti opensource and FOSS. Just when it comes to software development, I am quite mercenary to me it is a job.
But despite my viewpoint or maybe because of it I find what Google do through opensource and FOSS to be a perversion (or is it corruption) of the original ideals and values of opensource and FOSS.
Let me explain, Google’s first business is Ads more precisely Ad delivery and the medium of delivery is the internet. The more control they have of the medium the more, they can control what internet users see, more importantly must see. Control to give their Ad network unfettered access to place ads and collect info for targeted ads.
They give away Chrome to get control by getting a say in the setting of standards and setting defacto standards through features they add in their browser. Android is just a vehicle for their ad network. Sure, it is opensource but they control what gets added so there are proprietary bits. Those proprietary bits lock you in because if you fork, your App ecosystem will shrink drastically. Not only that the missing apps will be the ones that people want, the ones that do something, but they tend to need something from a proprietary bit.
So, regarding Fuchsia and maybe I just can’t see from my mercenary view of software development why would any developer want to contribute to it? Going on Google’s history Fuchsia is just to maintain dominance of and extend the reach of their ad network making them money.
For the same reason why would a company adopt Fuchsia for some new piece of hardware? I mean “Hey Google I think I have found a new market with this hardware I really think Fuchsia is the OS. I would love to hand control of the market to you so you can make even more money while I establish Fuchsia as an OS while locking myself to Fuchsia”.
Finally, Annatar just a nerd reference to Tolkien’s middle earth Google it. ;)
Media in all of its popular past and present forms, both print and electronic and word of mouth confidential, have been doing a Google since long before world wide webs internetworking were even a twinkle in someone's eye, Annatar.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
