[ / / / / / / / / / / / / / ] [ dir / animu / asmr / rel / strek / sw / travis2k / vore / zoo ][Options][ watchlist ]

/tech/ - Technology

You can now write text to your AI-generated image at https://aiproto.com It is currently free to use for Proto members.
Name
Email
Subject
Comment *
File
Select/drop/paste files here
Password (Randomized for file and post deletion; you may also set your own.)
* = required field[▶ Show post options & limits]
Confused? See the FAQ.
Expand all images

[–]

 No.895633>>895707 >>895720 >>895753 >>896022 >>896117 >>896265 >>896633 >>897262 [Watch Thread][Show All Posts]

https://fuchsia.googlesource.com/

Microkernel means I'm interested, beats Linux and could even beat Darwin.

Not sure about the platform, I'm in as long as it's open but it probably won't remain so under MIT. Perhaps many of the pitfalls Android falls into in regards to manufactures will be avoided.

I hate Android as a platform and I'm look forward to a fresh start on a new open platform. Android's aged API with 5 years of backwards comparability all for the sake of selling App Play store apps to 3rd worlders still running 4.4 because their devices don't get updates... Not to mention Java and the horrible multiplatform support for C and others. Even NeXT did better in' 95.

Interesting to see the Android dev team compete with the Fuschia team for relevance. Android has introduced Treble, Android Auto, a new official IDE, API "cleanups". Seems they wish to entrench Android into a permanent legacy position in order to sandbag Fuschia to an early grave before it's even out.

Speaking as an expert from Cupertino, I suspect their efforts will keep Google's existing mobile offerings strong just long enough for Fuschia to come out and replace it...

 No.895634>>895891 >>896317

File (hide): 6db6c04e4a13445⋯.jpg (142.88 KB, 600x877, 600:877, 1523083059536-pol.jpg) (h) (u)

>As an Apple insider

Has Tim "the faggot" Cook realized that it's not 2008 any more and that he should stop relying on marketers like Steve did at that time?

The new iMac Pro and the ARM Macs down the pipe is the first time I've seen Apple shine so bright in nearly a decade.

I doubt Fuschia will do well, it seems like a webOS type thing with TypeScript (Microsoft JavaScript) and HTML and languages so hipster that Rust and Go look as legacy as FORTRAN by comparison.

I doubt you'll see C/Obj-C, or C++, or even Python or Java on it, at least not from Jewgle.

Expect Electron apps, expect Atom to be the IDE.

We need a Phone that's LISP, maybe include Ruby and Obj-C for newbies but make it primarily LISP.


 No.895635>>895666

File (hide): 73b5430bb00e072⋯.jpg (237.58 KB, 900x900, 1:1, 1522922277304-pol.jpg) (h) (u)

I want to see GNU/Linux phones instead. And then I want to see a good microkernel being offered as an alternative to Linux and BSD Unix.

GNU is a great POSIX Operating System but Linux is only okay and BSD already serves as a monolithic kernel POSIX.

Linux is redundant.


 No.895636>>895640

Oh the microkernel is awesome. Its just everything by jewgle is botnet. So avoid like the plauge.


 No.895640>>896268

>>895636

GNU/Linux Phones are finally here. Just in time in case Android does to Fuschia (which depends entirely on license).

If Fuschia remains MIT, expect everyday be to fork it like Tizen, WebOS, etc.

If Google wants to avoid the Treble situation and rule with a clenched fist, then manufactures will keep on using Android and Google will happily keep milking that cow.


 No.895662

Microkernels are hard. I hope it works because that would be cool, but then again, it's Google, so even if it fails, it might get pushed by Intel Agencies.

Linux is here to stay, Android, not so much.


 No.895666

>>895635

Linux is just a kernel. Literally the only thing it has to do is be posix-compliant and have a lot of drivers. Why do you think that all the gnu/kFreeBSD projects fall through? One Unix kernel is really no different than another, except maybe more or less monolithic.


 No.895707>>895716 >>895721

>>895633 (OP)

>Android's aged API with 5 years of backwards comparability all for the sake of selling App Play store apps to 3rd worlders still running 4.4 because their devices don't get updates

The fuck you talking about? That's still 100x times better than having to throw away your perfectly working phone every 1-2 years because all your shit dropped compatibility.

You should blame jew carriers and driver makers for not releasing drivers and not allowing you to install updates instead.


 No.895716

>>895707

iOS does not have this issue. Just because you're using a 5 year old phone with 8 year old software doesn't mean developers should shit their apps up with some weird Frankenstein amalgamation of new and old APIs including some that are "deprecated" but still used and recommended by Jewgle.

Complete mess.

Blame your manufacture for not updating the software, blame Jewgle for marrying he drivers and APIs to the OS. What's not good is being forced to not improve the platform and support 5 year old cludges because Shaniqua's CheapMobile-exclusive LG Envy H9 doesn't get updates anymore.


 No.895720>>895729

>>895633 (OP)

>microkernel

Enjoy losing all your performance by having to context switch 5 times to do trivial operations.


 No.895721

>>895707

>5 years is 100x better than 2 years


 No.895729>>895731 >>895733

>>895720

>implying switching contexts isn't 100x faster than the gay privileged memory table hacks across rings that monolithic kernels do

mach was called mach for a reason.


 No.895731>>895765

>>895729

Its called mach because its faster than the other slow as shit micro kernels. Context switching absolutely is slower


 No.895733>>896028

>>895729

>privileged memory table hacks

<standard permission bits are a hack


 No.895753

>>895633 (OP)

Listen you faggot when google aquired android they could still can changed it into somthing actually good because it was still early in development. But instead android continued to turn into shit it's not gonna be better this time, it's gonna be the same, google developers are just incompotent.


 No.895765>>895769 >>895838

>>895731

>mach because its faster than the other slow as shit micro kernels

Mach was the first-ever microkernel. It had to have been faster than existing (monolithic) kernels.


 No.895769>>895838

>>895765

Mach was not a micro kernel until version 3 in the mid 90s. It was BSD with message passing.


 No.895838>>895840 >>895885

>>895765

>>895769

Microkernels are associated with being slow by laymen because the first ones were slow.

http://wiki.c2.com/?MicroKernel


 No.895840>>895847

>>895838

At this point, even if microkernels are slower than monolithic, the security benefits would be more than worth it.


 No.895847

>>895840

More than security, they are reliable.

https://youtu.be/vlOsy0PZZyc?t=126


 No.895885>>895893

>>895838

> the first ones were slow.

And they are still slow. They never solved the fundamental hardware problem of context switches being expensive.


 No.895891

>>895634

>We need a Phone that's LISP

So you want the lisp machine back ?

> maybe include Ruby and Obj-C for newbies

>maybe include Ruby and Obj-C

>Ruby and Obj-C

>for newbies

Scheme is a better options for young ones.


 No.895893>>895896

>>895885

Read the link, dumbfuck


 No.895896

>>895893

I don't think you read link you posted. Mind quoting the spot you think contradicts what I said?


 No.896022>>896058

>>895633 (OP)

>microkernel

>beats Linux

pick one


 No.896028>>896060

>>895733

enjoy your Meltdown, faggot


 No.896058

>>896022

Linux is kinda garbo, even BSD beats it.

>muh userspace compatibility

>muh cruft and bad implementations

Fuck off with that shit. Not my problem you use proprietary shit.


 No.896060>>896071 >>896104

>>896028

Atleast I will enjoy A SYSTEM. Not like all that microkernel vaporware bullshit. Best things out there are proprietary QNX and shitty Minix with some BSD userspace.


 No.896071>>896077

File (hide): e990cd4256e1bca⋯.png (110.35 KB, 625x773, 625:773, Brainlet9.png) (h) (u)

>>896060

>microkernel vaporware bullshit


 No.896077>>896080

>>896071

Where is my microkernel desktop OS?


 No.896080>>896098 >>896100

>>896077

NeXTSTEP.

System 7

System 8

System 9

Early OSX


 No.896098>>896102

>>896080

First off thats shit is all 20 years old. Not modern functional systems. Second it being a derivative of mach does not mean that their code uses a microkernel architecture (which it doesn't).


 No.896100>>896102

>>896080

>where is my desktop OS

>NeXTStep

Thanks for the giggles mate.


 No.896102>>896104

>>896098

>First off thats shit is all 20 years old.

Hardly vaporware if we've had it for 20 fucking years, is it retard?

>Not modern

Nice Goalpost move

>Second it being a derivative of mach does not mean that their code uses a microkernel architecture

What are you talking about retard? Their code? The only thing that's micro is the kernel. What the fuck do you mean derivative? Pulling things out of your ass?

Dumbfuck, those have 100% Microkernels.

>>896100

Kek


 No.896104>>896113

>>896102

>A derivative of something keeps the same architecture and would never extra things into the kernel

AYEEEEE

>Nice Goalpost move

Maybe you should read the context of >>896060


 No.896113>>896118

File (hide): 2f3a2d00b9496d1⋯.jpg (32.45 KB, 505x567, 505:567, peabrain.jpg) (h) (u)

>>896104

>S Y S T E M

>vaporware

You have your systems, they've been out for decades. Hardly vaporware.

>>A derivative of something keeps the same architecture and would never extra things into the kernel

What the fuck are you talking about? If only you were white and not autistic then you'd actually know what you were talking about.

All of those kernels are 100% grade A certified Microkernels.


 No.896117>>896123

>>895633 (OP)

>many pitfalls Android falls into in regards to manufactures will be avoided.

>MIT

Yeah, no. It can only be worse than Android.

>Android's aged API with 5 years of backwards comparability all for the sake of selling App Play store apps to 3rd worlders still running 4.4 because their devices don't get updates.

What are you talking about? Android 6 is currently the oldest officially supported version. Support is always only available to the last 2 big versions prior to the current one. So they're literally supporting only versions up to 3 years old.

Only less than 15% of users are still on KitKat or older, and those are usually smartwatches or cheap tablets. Or as you've said, very poor people.

Some developers support older versions simply because it's either profitable or the app doesn't implement features which require newer versions.


 No.896118

>>896113

And they are all irrelevant bullshit from the 90s.


 No.896123>>896125 >>896129

>>896117

>Android 6 is currently the oldest officially supported version

The vast majority of apps out right now target Android 4.X though.


 No.896125>>896131

>>896123

the vast majority of apps are trivial shit that will never be updated again


 No.896129

>>896123

The vast majority of apps haven't been updated since 2016


 No.896131>>896135

File (hide): 654c9ecde3b4775⋯.jpg (63.62 KB, 591x636, 197:212, Android Pajeet Dev Crew.jpg) (h) (u)

>>896125

...Okay.

Why is it that I can still use Android 2 widgets in Android 9 though? Why is it that half of Google's current code samples use 'methods' so deprecated that they are crossed out in the IDE?

It's a complete mess and it bloats system images, it bloats the source, it prevents the developers from "doing the needed" as they say over at Android HQ because they have to prioritize comparability over making their API good.


 No.896135

>>896131

If you actually knew what you're talking about, you'd know that you're wrong and overdramatic.


 No.896265

>>895633 (OP)

> beats Linux

Don't kid yourself


 No.896268

>>895640

>GNU/Linux Phones are finally here.

Post some.

Reminder that all phones have a proprietary baseband processor running proprietary code and it manages all GSM radio operations, and that it's actually illegal to run your own software on that processor.


 No.896297>>896305 >>896309 >>896318 >>896407

The idea of a microkernel is good (if your hardware sucks, i.e. RISC) because it adds to reliability and turns kernel drivers and "components" into normal programs that can be written in any language, but these microkernels all suck because they're written in C and designed for UNIX. Like JavaScript is better than C, these C/C++ microkernels are better than UNIX, but they still suck. If there was a microkernel that's not written in C or C++ and not designed for UNIX/POSIX, it probably wouldn't suck.

Subject: I Have Seen The Future Of The Future And It Sucks Too.


So I figured I was better than your average unix-haters
reader since, hey, I hack Mach kernels; I'm not just
bitching about how bad things are, I'm actually doing
something about it, damn it!

Well, it all started with when my research version of Mach
went down in flames and took bits and pieces of the file
system with it. In hindsight, I suppose I should have just
lived with things like not being able to print to certain
printers. Well anyway, I had the Facilities people "add 10
megs or so to the root partition, oh and while you're at it,
why don't you rebuild the disk since the filesystem's been
kind of flaky since that, uh, big kernel crash last week..."

So anyway, I came in today and found a brand spankin new
filesystem. All I had to do was replace the v2.5 production
kernel with a v3.0 research kernel. Yep, just replace a
couple files on the root partition and reboot, no problem.

Well, it's a little more complicated than that. For some
unknown reason, mach machines have a "root" directory ("/")
and a "superroot" directory ("/../.."). Once someone tried
to explain why but I didn't quite get it.

Anyway, you have to have hard links to your kernel file in
both root directories. Then since mach is not a kernel but
a microkernel (the micro kernel binary with symbol tables,
debugger, etc is a svelte 950k on a MIPS box, unix emulation
is only another 1500k; I actually *needed* the extra 10 megs
on the root partition) you have to create a mach_servers
directory and populate it with various programs that mach
uses to emulate unix.

Now the tricky part is that since you have two root
directories you can put the mach_servers directory in the
root directory and then put a symlink in the superroot or
you can put the mach_servers directory in the superroot and
put a symlink in the root directory.

The documentation says you can do it either way.
Unfortunately, the documentation doesn't say that it only
works when you put mach_servers in the superroot with a
symlink in root. This also happened to be the second case
given in the documentation and I had the misfortune to use
the first case given in the documentation. Silly me.

So I proceeded to boot my machine 10 or 12 times only to
crash into the kernel debugger every time. I stared at
stack traces and read code and read code and stared at stack
traces. The error messages were no help, they said "init
died". Thanks a lot.

So I finally figured that one out and redid the mach_servers
directory and rebooted once more. It crashed again at about
the same place complaining about "error 6000". Now the
annoying thing about mach is that since your operating
system is spread out over 69 different address spaces, you
never really know for sure who is generating an error code.


You just have to memorize a mapping from error code ranges
to subsystems. Eg errors in the 2500's are from the
device-independent device routines (go figure), -300's are
arg checking errors in an IPC stub routine, you get the
idea. Well anyway, I never ran across 6000 before. So I
hit the source code again... Oh. Error 6000 means some
file is not executable.

One of those damn programs in mach_servers was trying to
exec another of those damn programs in mach_servers which
unfortunately didn't have it's execute bit set so naturally
the exec function, in it's penultimate moment of existence,
exclaimed "6000!" before it crashed the whole show.

I guess the moral is that Mach looks like unix on the
outside but does totally different stuff internally.
However, when you strip away the tasks, threads, ports,
messages, and memory objects, when you get down to the very
essence of the system, the thing is still unix. There is no
hope.


 No.896305

>>896297

>asshurt lisp nigger is back

Posted from my ARM smartphone :^)


 No.896309

>>896297

you already know it's ya boi when you see the wall of text


 No.896317

>>895634

What Tim Cook doesn't know could fill a book. Google is playing with a handicap too, but they have the educational market wrapped up. As it turns out, nobody wanted to use iPads for schoolwork, just like nobody wanted a Trashcan Pro or iMac with 18 cores for a workstation. Not that any work is being done on Macs because the pro AV software has been flushed down the toilet. The remaining x86 Macs will sit in a warehouse until investors cook Cook, because fire sales are so gauche. Google will take Apple's Macintosh customers because Tim simply doesn't want them anymore.


 No.896318>>896407

>>896297

Your posts suck because they're written in English and designed for HTML.

Although not limited to a prison setting, the
implications of this study are perhaps most directly
relevant in the type of social setting in which
it was conducted. In such female "people process-
ing" organizations, women interact with other
women and a small proportion of these women
are incumbents of social roles to which formal
decision-making authority is attached. Both the
interaction and decision-making processes have
implications for the effectiveness of the organiza-
tion in terms of the degree to which it approxi-
mates the realization of its intra-prison (social
order) and societal/community (rehabilitation, se-
curity) goals. How can knowledge of the findings
'2 For an excellent discussion of the distinction be-
tween operant and respondent classes of behavior,
and of the processes by which respondents may become
operants, see S. Bijou & D. BAER, Cam DEVELOP-
3IENT (1961). Among well-known theorists with pri-
mary interests in aggressive behavior, Berkowitz, The
Concept of Aggressive Drive: Some Additional Considera-
Lions, 2 AnvANcEs im
EXPERIMENTAL SOCIAL
PsyvcoL-
oGY 301-29 (L. Berkowitz ed. 1965), focuses exclusively
on '"non-instrumental" aggressive behavior, as do
almost all frustration-aggression theorists, while A.
Buss, Psycaorory or AOGRESSION (1961) is much
more concerned with instrumental aggression.19711
MENSTRUAL BEHA VIOR
of this help improve social interactional and de-
cisional outcomes within prisons?
First, to the extent that the outcome of inter-
action between a prison officer and an inmate has
been adverse to both, and to the extent that
"intentional meanness" is used by both as an ex-
planation of the behavior of the other, the results
of this study might be relevant in that it may help
each interactant understand that some biological
processes over which neither has control may be
partly responsible for their behavior. Such explana-
tions may help dissipate the angry feelings of
revenge or spite which may otherwise accumulate
and assume an even greater explosive potential
over time. Where adverse direction of the inter-
action is one way, as in punishments meted out
by disciplinary committees, an essential pre-deci-
sion item of information should be the menstrual
condition of the prison officer who reported the
girl, the menstrual condition of the girl herself,
and that of the female members of the committee.
Great care, however, should be taken to avoid the
creation of a situation in which either inmates or
officers come to believe that during certain phases
of the cycle they can "get away with anything."
In this discussion, we have assumed that the effect
of adverse interaction on the prison effectiveness is
mediated by the explanation provided by both
interactants for the behavior of each.
Intra-prison effectiveness is also a function of
the numbers of aggressive interactions. The results
of this study suggest that one way of reducing the
frequency of aggressive inmate/inmate or in-
mate/officer interactions is to put in temporary
seclusion those girls whose past behavior justifies
their requests to be alone just prior to or during
menstruation. Alternatively, decisions may be
made in such a way that girls who become mark-
edly more aggressive during these phases of the
cycle are placed in prison jobs and accommoda-
tions requiring little interpersonal contact with
others. In the context of jobs and accommoda-
tions, similar considerations should apply in the
case of prison officers. This would require the
completion of a daily menstrual log by inmates and
prison officers, and should be made a part of
administrative routine.


 No.896350>>896500

This thread started out OK, but has quickly turned to shit.


 No.896407


 No.896500

>>896350

due to, in no small part, your post.


 No.896609

"Above link"

>Pink + purple = new os

I bet the creator of this must take LSD for this kind of shit


 No.896633

>>895633 (OP)

>it's literally called "Fucks ya"

they are completely shameless at this point


 No.897262

>>895633 (OP)

>beats Linux

>not sure about the platform

>I'm in

What the fuck is this OP anyway, reads like a Pajeet wrote it.




[Return][Go to top][Catalog][Screencap][Nerve Center][Cancer][Update] ( Scroll to new posts) ( Auto) 5
55 replies | 5 images | Page ?
[Post a Reply]
[ / / / / / / / / / / / / / ] [ dir / animu / asmr / rel / strek / sw / travis2k / vore / zoo ][ watchlist ]