Linux is pozzed

Linux is pozzed
Firefox is pozzed
OpenBSD has clinton donor contributors
MPV has security flaw
TAD is dead
Intel is jewish
AMD is shit

When did you realize technology is dead?

Attached: kingdavis.png (424x424, 325K)

Other urls found in this thread:

youtu.be/BF0sHQxgnLk
softwareengineering.stackexchange.com/questions/279334/why-was-the-itanium-processor-difficult-to-write-a-compiler-for
twitter.com/SFWRedditGifs

everything is corrupt

>When did you realize technology is dead?
when Microsoft started asking you to pay for _Free_Cell

What is wrong with OpenBSD? So what if companies who donated to Clinton also donate to OpenBSD. OpenBSD is explicitly non political and pro meritocratic; the antithesis of Linux in its present state. Moreover, that donate to the OpenBSD project do so because they use its products, perhaps not the OS itself but the firewall, networking, etc. I know that Amazon does.

>AMD is shit
That's intel propaganda.

Obama gave away the keys to the Internet

technology was never good

>when a black man btfo'd republishits so badly for 8 years they can't stop seething even after he's gone

Clinton foundation employs death squads, and social justice is the greatest threat facing humanity.
Do you even Jow Forums?

after microsoft updated my os to windows 10 for me without asking

None of those are real problems

The real issue is that Linux is shit, all web browsers are shit, programmers are getting dumber, and we use x86, the shittiest architecture, on our main computing devices.

>MPV has security flaw
Elaborate

>and we use x86, the shittiest architecture
I don't agree with that, agree with the rest.
x86 wins on the performance front because it turned out that small instruction size is more important than easy decoding.
If you asked cpu designers 20 years ago they would say risc is perfect because it saves on decoder complexity and we would all use something like powerpc now.
Today decoding complexity is irrelevant, but keeping more code in cache as opposed to ram gives great performance and power gains.

>caring more about politics-of-the-month than actual product quality (the former has yet to impact the latter btw)
>see above
>personal politics don't automatically bleed into other unrelated interests in the real world, see above
>what is a patch
>what are ME debloater scripts
>not 2012

seems fine to me

Attached: 1537283809104.jpg (540x405, 28K)

It wins on the performance front because it has been worked on for so long. If we used a different architecture originally, we'd be at higher speeds.

no we wouldn't, did you read what I wrote?
the core of the cpu is risc, modern decoders translate instructions into that. There's zero speed penalty.

I don't mean RISC VS CISC, I mean architecture in general, possibly another CISC.

what's wrong with x86 then?

It's over-complicated, and not just in the CISC-vs-RISC way.

that doesn't tell me much. what's overcomplicated?

First off, no immediate division/multiplication (aside from imul), and instead you have them weirdly split across multiple registers instead of using a single 32 bit one.
Then, the boot process is unnecessarily complex, and there are obsolete structures, modes, and systems that are still used for backwards compatibility with a previously bad designed system.
Also, some instructions are redundant, or try to do too much (like mov, see movfusticator for an extreme example), resulting in a bloated architecture.
None of these really effect program speed significantly, but they do create for a hotter, larger, and more complex CPU.
Even Intel engineers will admit that x86 is nowhere near being 'elegant', and

he was right the entire time...
youtu.be/BF0sHQxgnLk

RISC-V is just getting started
It's gonna be a wild ride

>First off, no immediate division/multiplication (aside from imul), and instead you have them weirdly split across multiple registers instead of using a single 32 bit one.
irrelevant, all these instructions are translated into internal ops anyway with thousands of registers.
>Then, the boot process is unnecessarily complex, and there are obsolete structures, modes, and systems that are still used for backwards compatibility with a previously bad designed system.
Again irrelevant, this doesn't impact speed or power in any way.
>but they do create for a hotter
no, it's all implemented in microcode on top of existing instructions.
>Even Intel engineers will admit that x86 is nowhere near being 'elegant'
yes it isn't, that doesn't mean it's slower or more power hungry because of that.
The actual complexity, and source of all recent security problems, is in speculative execution, but that's independent from the architecture.

The only segment in which modern x86 has problem is ultra-low power, as there the additional complexity from the decoder is handicap. That means 100mhz microcontroller, in smartphones etc. smaller code size pays for itself.

>When did you realize technology is dead?
When IT became "technology"

you will have spying shit in the binaries

Structures and modes definitely create more to implement, as I'm fairly sure not all of that is microcode. But yes, you are correct in that the speed impact may not be significant or may actually be an increase, however, we could've have a much better designed architecture, which was my original point, x86 is a shitty architecture by design. It works, but it is definitely not pretty doing so.

>When did you realize technology is dead?
When people started to use JS in backends and desktop applications

>follow-the-leader politics makes you miserable
gee who knew

>we could've have a much better designed architecture
We almost certainly wouldn't, 'perfect' instruction sets in the past would be a disaster today.
softwareengineering.stackexchange.com/questions/279334/why-was-the-itanium-processor-difficult-to-write-a-compiler-for

Not only the 'perfect' architecture would be slower, but it would have to break its elegance anyway to add unforeseen new features. In 20+ years you would get an evolutionary design like x86 too, but instead of a minimally functional core design (8086) you would have the base designed for paged 64 bit computing in ~2000.
x86 is so nice to cache because when 8086 was designed 1kB of ram was a normal amount.
Itanium was designed when ram speeds didn't yet diverge much from cpu speeds + decoder was still relatively big as % of all transistors.

In 20 years the then-perfect architecture is going to be fundamentally different from what today would be 'perfect'. It appears the future platform is going to use millions of cores in a 3d arrangement.
A perfect architecture designed today would expect 64 cores at most.

Don't worry though, x86 isn't going to last forever, I think the future cpu core is going to be more like fpga with lots of specialized cores, with decoder implemented on fpga too. This would allow eg. jvm to run 'directly' on hardware.

fairly sure that openbsd is completely free of binaries actually
theo doesnt like blobs

>OpenBSD has clinton donor contributors
Sauce?