Slashdot Mirror

There's plenty of blame to go around on both sides here.

The motherboard manufacturers – pretty much all of them – are to blame for developing and shipping really crappy firmware. (Unfortunately, this is pretty much par for the course – 95% of all firmware is crap, no matter what it's for. Modern hardware companies, with a few obvious exceptions like Apple, just don't do software very well at all.)

The Linux kernel devs are to blame for being stubborn about "standards-compliance" versus the real world. From what I can tell in clicking through a few links, the ACPM feature was working in the past, but the kernel devs then deliberately broke it by changing it to only work if the BIOS advertises it properly. Yes, the standard says that's what is supposed to happen. But we know from experience that manufacturers often don't follow standards. Linux needs to deal with the world as it is, not as the devs wish it would be.

Best not to say "Try it in Linux" on Slashdot, you're a lot more likely to run in to someone who already has. My laptop and server are exclusively Linux and my desktop dual-boots. Ubuntu LTS all around, 14.04.1 on the desktop/laptop and I haven't gotten around to upgrading the server from 12.04.5 yet. AMD even lost performance per clock compared to themselves with their recent chips.

My home server previously ran a Phenom II X4 945, a 3 GHz quad core released in mid-late '09. That motherboard blew up after a power event so I switched to an A10-7850K, a 3.7-4.0 (turbo) GHz quad core released in January of this year. It's both faster clocked and a full four years newer, plus I threw double the RAM at it since I had to get new sticks anyways for DDR3 vs the old DDR2, yet somehow it's slower in the real world. My usenet downloads parcheck/extract slower, my Minecraft server bogs down more often, and it can't even manage to proxy Steam traffic at the full 100mbit/sec my internet connection allows.

As for Phoronix, how's this one? The very processor I'm running, the top model of the latest core AMD has released.

In looking at the results the AMD A10-7850K is supposed to be in line with the Intel Core i5 4670K according to AMD's expectations. However, with Ubuntu Linux on this hardware the Core i5 4670 (non-K) was generally running noticeably faster than the Kaveri APU. This is a big deal since the Kaveri APU sells for $190 USD where as the i5-4670 is not much more at a price of about $218.

It barely holds with the Core i3s on average.

I have historically defaulted to AMD. My last Intel outside of laptops was a 300 MHz Pentium II. I regret going with AMD for the server and unless they pull something huge out of their ass in the near future I'll be changing my desktop over as soon as the prices drop on the now last-gen Intels.

Maybe it is just me but when I see these things, I sometimes get crazy ideas. And I think:

Might as well translate into LLVM bitcode and recompile the code:
http://www.phoronix.com/scan.p...

Hell, maybe it's even faster if you compile the LLVM bitcode with emscripten and use asm.js to run into the browser. :-)

1. Agreed, that sounds stupid.

2. AMD has a somewhat tarnished reputation for the performance of their FX CPU line. So far, NOT over lack of reliability. I hope they won't acquire that now...

[digression]From Nvidia, the only really bad thing I remember is that their mobile Geforce 8xxx had a reputation for dieing early. The 8600 GT in particular.
They are known for not caring about Open Source, and that is why I would currently prefer an AMD GPU (even if the GeForce 750 Ti looks really nice in terms of performance/watt).
But I'm probably in the minority there, and Nvidia's binary drivers have a good reputation and fairly long support time frames, longer than binary AMD drivers anyway.

For a real mess, consider this: http://www.phoronix.com/scan.php?page=news_item&px=MTMyODA (status of Intel Poulsbo on Linux ;-)

Take the absolute lowest Intel and AMD quads, the Atom and Jaguar respectively, and put it against the most expensive top 'o the line ARM quad and what happens? the ARM gets a curbstomping

Wrong, you can see the Atom chips getting smashed by the Exynos 5 chip.

And here in monte carlo and FFT benchmarks.

And here in h.264 encoding, zip compression and PHP compilation benchmarks.

Also here it's more of a mixed bag but the Atom gets thoroughly beaten and the Tegra4 and Jaguar trade the lead.

I understand running a business that depends on PCs is where your obvious bias comes from but the facts don't lie, this isn't to say that ARM is better than x86 but in some cases it is and it most certainly isn't the "curbstomping" you claim it to be.

But your analysis is only correct if you live in the past. With baytrail Intel is again stomping ARM and this is only getting worse now that Intel has set ist sights on this market segment.

Take the absolute lowest Intel and AMD quads, the Atom and Jaguar respectively, and put it against the most expensive top 'o the line ARM quad and what happens? the ARM gets a curbstomping

Wrong, you can see the Atom chips getting smashed by the Exynos 5 chip.

And here in monte carlo and FFT benchmarks.

And here in h.264 encoding, zip compression and PHP compilation benchmarks.

Also here it's more of a mixed bag but the Atom gets thoroughly beaten and the Tegra4 and Jaguar trade the lead.

I understand running a business that depends on PCs is where your obvious bias comes from but the facts don't lie, this isn't to say that ARM is better than x86 but in some cases it is and it most certainly isn't the "curbstomping" you claim it to be.

But your analysis is only correct if you live in the past. With baytrail Intel is again stomping ARM and this is only getting worse now that Intel has set ist sights on this market segment.

Take the absolute lowest Intel and AMD quads, the Atom and Jaguar respectively, and put it against the most expensive top 'o the line ARM quad and what happens? the ARM gets a curbstomping

Wrong, you can see the Atom chips getting smashed by the Exynos 5 chip.

And here in monte carlo and FFT benchmarks.

And here in h.264 encoding, zip compression and PHP compilation benchmarks.

Also here it's more of a mixed bag but the Atom gets thoroughly beaten and the Tegra4 and Jaguar trade the lead.

I understand running a business that depends on PCs is where your obvious bias comes from but the facts don't lie, this isn't to say that ARM is better than x86 but in some cases it is and it most certainly isn't the "curbstomping" you claim it to be.

But your analysis is only correct if you live in the past. With baytrail Intel is again stomping ARM and this is only getting worse now that Intel has set ist sights on this market segment.

Take the absolute lowest Intel and AMD quads, the Atom and Jaguar respectively, and put it against the most expensive top 'o the line ARM quad and what happens? the ARM gets a curbstomping

Wrong, you can see the Atom chips getting smashed by the Exynos 5 chip.

And here in monte carlo and FFT benchmarks.

And here in h.264 encoding, zip compression and PHP compilation benchmarks.

Also here it's more of a mixed bag but the Atom gets thoroughly beaten and the Tegra4 and Jaguar trade the lead.

I understand running a business that depends on PCs is where your obvious bias comes from but the facts don't lie, this isn't to say that ARM is better than x86 but in some cases it is and it most certainly isn't the "curbstomping" you claim it to be.

Take the absolute lowest Intel and AMD quads, the Atom and Jaguar respectively, and put it against the most expensive top 'o the line ARM quad and what happens? the ARM gets a curbstomping

Wrong, you can see the Atom chips getting smashed by the Exynos 5 chip.

And here in monte carlo and FFT benchmarks.

And here in h.264 encoding, zip compression and PHP compilation benchmarks.

Also here it's more of a mixed bag but the Atom gets thoroughly beaten and the Tegra4 and Jaguar trade the lead.

I understand running a business that depends on PCs is where your obvious bias comes from but the facts don't lie, this isn't to say that ARM is better than x86 but in some cases it is and it most certainly isn't the "curbstomping" you claim it to be.

Take the absolute lowest Intel and AMD quads, the Atom and Jaguar respectively, and put it against the most expensive top 'o the line ARM quad and what happens? the ARM gets a curbstomping

Wrong, you can see the Atom chips getting smashed by the Exynos 5 chip.

And here in monte carlo and FFT benchmarks.

And here in h.264 encoding, zip compression and PHP compilation benchmarks.

Also here it's more of a mixed bag but the Atom gets thoroughly beaten and the Tegra4 and Jaguar trade the lead.

I understand running a business that depends on PCs is where your obvious bias comes from but the facts don't lie, this isn't to say that ARM is better than x86 but in some cases it is and it most certainly isn't the "curbstomping" you claim it to be.

In summary: I don't believe that Facebook will be implementing this technique in their servers. If they really wanted encrypted DRAM, they would pay Intel or AMD to build a semi-custom processor with encryption techniques built into the DRAM controllers. They bought this company because they want to hire these guys who have a lot of kernel and hypervisor knowledge.

More proof to your hypothesis:Facebook is currently hiring kernel hackers. With a humorous "we gotta beat FreeBSD!" target, but still. BSD-jokes aside, it's another proof that they are interested in increasing kernel performance and thus people with very good low-level knowledge would be welcome, no matter these people's current product has very few practical application.

when gender issues are poaching away resources from real work.

Your gender issues seem to be poaching away resources from real thinking. How is it related to web browsers what other people may or may not do with their nether appendages?

Btw I'm queer and I'm sad about how some marriage advocates made Brendan Eich quit. But I'm confident there's still "real work" going on at Mozilla.

Are you familiar with the debacle where the Gnome Foundation went broke because they blew all the money on their Outreach Program for Women?

Here's coverage if you're unfamiliar, although if you're a queer slashdot reader you probably aren't:

http://www.phoronix.com/scan.p...

The Eich issue showed the world that Mozilla is chock full of the same sentiment. And Mozilla's lost so much market share that they're only a bit player now. When push comes to shove, their "Real Work" is not cutting the mustard.

I've worked on technology projects with people who didn't agree with my views on the issues, and done volunteer work on community projects with people who didn't agree with my views, but everything worked because the projects were focused enough that they became something we could both agree on.

The reason gender issues are screwing up technology projects is because technology projects are extending their mission statements in political directions and it's removing the focus that made it possible for people who disagree on political issues to work together.

Expressed simply, if you stand for one thing, you get half the people agreeing with what you stand for and half of them not agreeing, and 50% of the people give you their support.

If you stand for two things, half the people who were supporting you will no longer feel comfortable supporting you, and they will leave. You shrink your support from 50% to 25%.

It's not that we disagree. It's that I can't actively support organizations that vocally espouse things that I think are nihilistic and therefore immoral, and logic dictates that if I was the only one, there would be no controversy, so therefore, I'm not the only one.

The ability to agree to disagree has been removed, and it's not going to do anything but harm.

Phoronix

Symbolic links are used, just not often-- theres rarely a need. I have myself used them, however.
File locks are supported by many filesystems, and generally its not Windows doing the locking, its an application.
Read speed-- even on Linux with ntfs-3g-- is apparently better than both ext3 and HFS+. Generally, as a journalled filesystem, NTFS isnt going to be quite as fast as unjournalled systems like ext2 and FAT32, but AFAIK its actually one of the faster filesystems out there:
http://www.phoronix.com/scan.p...
NTFS @ 127MB/s vs EXT3 @ 75 and EXT4 @ 130

It really sounds like you dont know what you're talking about (and Im not sure what you mean by "stat"). There are some things NTFS does well, some it does less well, but all around its a pretty decent filesystem, and if you think its horrendously slow you're doing something very wrong.

Did this dude get so excited over the release that he peed his pants? http://www.phoronix.com/image-...

There are varying degrees of stability and I felt that after 7 years of development, official inclusion into the Linux kernel, Facebook deployment and the default fs on OpenSUSE that it's good enough for my laptop, workstation and a few other systems. Having that said, I've not migrated by backup drives yet, they're still on XFS. It may be a while until I migrate those. http://www.phoronix.com/scan.p...

The GeForce 9600GSO was running into font rendering issues.

Maybe stick with the PDP-10 ;)

*golf clap* How it must feel to be worse than even Apple at supporting OpenGL 4.x versions.

To be fair, blissfully ignorant, as I am incredibly happy with Intel graphics on linux. I would care if I was interested in the BSD License, with Apple failing to follow the spirit of the license...but then here is the real truth.

Here is back last year when Linux outperformed Apple http://www.phoronix.com/scan.p...

Here it is this year still performing faster on Apple hardware http://www.phoronix.com/scan.p... ,,,in the case of one 10year old engine 20 times faster.

Your troll comment only serves to exposes weak software...and its "short arms...deep pockets" open source mentality on its overpriced hardware...it is pretty though.

*golf clap* How it must feel to be worse than even Apple at supporting OpenGL 4.x versions.

To be fair, blissfully ignorant, as I am incredibly happy with Intel graphics on linux. I would care if I was interested in the BSD License, with Apple failing to follow the spirit of the license...but then here is the real truth.

Here is back last year when Linux outperformed Apple http://www.phoronix.com/scan.p...

Here it is this year still performing faster on Apple hardware http://www.phoronix.com/scan.p... ,,,in the case of one 10year old engine 20 times faster.

Your troll comment only serves to exposes weak software...and its "short arms...deep pockets" open source mentality on its overpriced hardware...it is pretty though.

It's not about addressable memory space.

64-bit usually yields better performance due to more registers and the fact that i386 was a register starved architecture.

I thought there was a tradeoff between register starvation and data cache starvation. There are fewer registers on x86, but the pointers are half the size (without the so-called "x32" ABI that didn't catch on).

I don't run Linux, and I've never had to deal with X, so this isn't an emotive issue for me. But your post was clearly not a serious of facts, more of a backlash against perceived slights from Wayland developers and/or "fanboys", the latter being a useful way to smear people who disagree.

Your post might make more sense if the existence of Wayland meant that X no longer existed, but as that's clearly not true it's hard to take claims of "throwing out a perfectly good system in favour of an ideological rewrite" seriously. And ideology? There have been plenty of technical arguments since the beginning, here's one set that was posted on /. a while back, and it's just the first one I found.

As far as I'm aware X currently does act like VNC in most cases, except without any compression at all and a synchronous API - so nobody uses it directly because of the performance issues, instead using ssh as a tunnel. Even having it act like per-window VNC with H.264 compression would be an advantage. But anyway, that's all part of the compositor, which now has RDP as part of the core, and I've yet to see any explanation of why or how X forwarding is different or better than rootless RDP.

The assumption that forwarding is a critical feature is based on the idea that your personal requirements are the only important ones. If a piece of software doesn't do what you want, don't use it. As it turns out, they are supporting it (as I'd read from pretty much day one), it's just taken time to get to that point... as you'd expect from alpha software.

Incidentally, I just re-read my post from this morning to you. I've got to start being nicer :/

Same. I've been acting like a bit of a dick on slashdot recently. I've been grumpy AFK, but no need to bring it online.

Anyway, compared to other OSs. There's not all that many benchmarks, but some crop up now and again. OpenBSD is good but not all that fast (not a design goal). FreeBSD is probably the closest competitor. Other OSs are either too obscure to get useful figures out of (e.g. VM) or specialised for things (e.g. realtime) which generally have a negative effect on general performance.

For some reason for many benchmarks, people take something CPU bound and run it on various OSs. FreeBSD hasn't helped by moving on over to LLVM, since that still loses out to GCC in terms of performance so far.

In practive, Linux has a lot of people beating on it, especially in terms of very high performance hardware.

Anyway here's some more not very good benchmarke :)

http://www.phoronix.com/scan.p...

It would seem that the gap is closing, in terms of both the performance (of the compiler and the executables it produces) and the error messages:

http://gcc.gnu.org/wiki/ClangD...
http://www.phoronix.com/scan.p...

As charming as your characterisation of /.s membership is, I'm more interested in whether or not there is any truth to the assertion that Gnome's funding was eaten up by outreach programmes. I managed to track down this article, so there does seem to be a certain amount of legitimacy to the claim.

You can actaully find more or less the same thing from GNOME themselves: https://wiki.gnome.org/FoundationBoard/CurrentBudgetFAQ. It states:

What is the problem? The Foundation does not have any cash reserves right now.

Why has this happened? The Outreach Program for Women (OPW) has proven to be extremely popular and has grown quite rapidly.... GNOME, as the lead organization, has been responsible for managing the finances for the entire effort. However, as the program grew, the processes did not keep up.

That being said, the original poster's sexism and cisgenderism is obviously out of line in any case, but it does appear the growth of this program (which undoubtedly is largely cis women) was a large factor in creating the current financial situation. They also except to have it resolved within a month or so and don't seem to be too concerned about it.

I'm not. Sadly, this is precisely what happens when non technicians do technical decisions on a tech Foundation.

Gnome Desktop 2 was one of the main reason I jumped ship from Windows and spend 2 excelent years developing on a Linux box. Almost everything just works, and the few that didn't, I managed to tweak it into production with little effort - I'm a tech guy, after all.

And then came Gnome Desktop 3. And I decided that the migration efforts would be better spent on MacOS X - that I'm using since that days. No regrets.

I think the time for a MATE Foundation has come. :-)

This is a screaming message to every Open Source Foundation around (yes, Mozilla, I'm talking to you): do what your users *NEED* you to do, not what your non techies "advisors" *want* you to do.

There's no space on a tech industry for "politically correct" tech solutions that doesn't cut it!

http://www.phoronix.com/scan.p...

As charming as your characterisation of /.s membership is, I'm more interested in whether or not there is any truth to the assertion that Gnome's funding was eaten up by outreach programmes. I managed to track down this article, so there does seem to be a certain amount of legitimacy to the claim.

I fail to see how an individual's sexuality is a prerequisite for FOSS involvement. How does being "women (cis and trans) and genderqueer" matter in this context?

As charming as your characterisation of /.s membership is, I'm more interested in whether or not there is any truth to the assertion that Gnome's funding was eaten up by outreach programmes. I managed to track down this article, so there does seem to be a certain amount of legitimacy to the claim.

The correct answer would probably be that there is already competition in this market. By changing to a Linux os perating system you can maintain your 15 year old computer fully supported. Unfortunately, in many cases that's not true. Device manufacturers only provide full documentation and support to Microsoft and the Linux drivers cannot be guaranteed. This means that while your computer will work and your operating system will be supported, your actual whole system may not be.

Three years ago, the developers of Mesa dropped support for some old graphics cards:
http://www.phoronix.com/scan.php?page=news_item&px=OTg0Mg
Now those cards were badly obsolete and rare even in 2011, but this shows that even the Open Source community will at some point lose interest in supporting old stuff.
Today, you can maybe cobble together your own distribution that still contains those old drivers, or pay someone to do it. But for most people this won't be an attractive solution.

It *IS* easily done, right?!?

What are your thoughts on this one?

While Keith is not developing Wayland as far as I know (he does more management these days), he fully supports the idea of X on top of Wayland, and gave presentations about that.
X.org and X11 will be around for some time, if only to support legacy software. But the new development seems to focus on Wayland, with or without X running on top.

He called LLVM a 'terrible setback' and BSD a 'pushover licence'

http://www.phoronix.com/scan.p...

Mantle seems to have woken up the OpenGL folk too.

The slides are here. While I respect the idea that video is an inefficient means to convey information but, since this is an issue that you seem to care about, you may want to take the time to educate yourself. I believe it is possible to speed up playback of youtube videos. This article conveys some similar points, but not much depth. Here is an architectural overview.

Why don't you go over some of this information, and we can take this conversation over again from the top. You should find most of your concerns answered. I'm not particularly interested in spoon-feeding it to you, however -- please accept my apologies on that score.

The benchmarks on Phoronix did temperature, and commented on (though didn't measure) noise. Was actually a fairly comprehensive, well done benchmark, the only thing missing was frame latency measurements.

I don't really care, yeah, it's an early release, people who want a stable desktop maybe shouldn't use it.

It just recently went into feature freeze, so I guess it's "done" as far as deciding what will go in at least:
http://www.phoronix.com/scan.p...

Here it says mystery release 2k14:
http://www.phoronix.com/scan.p...

But on the previous link it says:
"there is currently no set date or estimated date for E19. My only goal with regard to time is to release before July 2023, though I may be forced to delay until September 2026 depending on celestial alignments." .. but in reality it seem to be expected in this year, e17 and e18 was released in December the last two years, so no 10+ years in-between those releases.

I don't really care, yeah, it's an early release, people who want a stable desktop maybe shouldn't use it.

It just recently went into feature freeze, so I guess it's "done" as far as deciding what will go in at least:
http://www.phoronix.com/scan.p...

Here it says mystery release 2k14:
http://www.phoronix.com/scan.p...

But on the previous link it says:
"there is currently no set date or estimated date for E19. My only goal with regard to time is to release before July 2023, though I may be forced to delay until September 2026 depending on celestial alignments." .. but in reality it seem to be expected in this year, e17 and e18 was released in December the last two years, so no 10+ years in-between those releases.

Time For Feedback On A New Phoronix Design

Well if AMD or someone can figure ways to make popular open source stuff run more than 20% faster on AMD CPUs than they do currently, let's see it. Otherwise all that is pointless talk- most stuff will be slower than Intel if you need the speed (and thus are willing to pay more).

See for examples: http://www.phoronix.com/scan.p...

X doesn't need to run as root any more. There are other ways due to people like this guy.

Is was that in the past, but is that true today?
According to the latest benchmarks in Phoronix http://www.phoronix.com/scan.php?page=article&item=llvm34_gcc49_compilers&num=5:

Clang 3.4 offered faster performance of compiled C/C++ code in several areas but GCC 4.9 also brings some performance improvements of its own over the current stable release. Clang still certainly outperforms GCC when it comes to compile times, but aside from that the compiler performance competition is rather mixed depending upon the particular code-base, workload, and processor.

For being a much younger project than GCC, LLVM/Clang is certainly running nicely and now building with almost all C/C++ code-bases tossed its way, and with the 3.4 release it's one step closer to having performance parity (or superiority) to the GNU Compiler Collection on modern x86 CPUs.

http://www.phoronix.com/scan.php?page=news_item&px=MTQ5NzQ

The Acer C720 Chromebook is Intel Haswell-based, and perfectly compatible with most Linux distros.

Phoronix did an awesome review of linux on the C720 several weeks ago, and in short: it's awesome. It runs everything you'd need - movies, internet, USB 3.0, streaming, 7-8 hours of battery life. There is some issue to work out with the touchpad, but it's possible to run most distros out of the box with an external mouse, or by applying a kernel patch. This is temporary though - I'd expect the touchpad to be incorporated in due time.

For $199, there's no better laptop on the market for Linux.

I replaced my old go-to laptop with a Haswell-based Acer C720 Chromebook that I put Linux on. I regularly get 7-8 hours of battery life, it has a decent matte screen (1366 x 768), USB 3.0, x86 dual-core Haswell chip...plenty of stuff under the hood for $199. Yeah it's got 2GBytes RAM (the 4GB RAM version is out of stock...) but for $199 I'm not expecting a gaming monster.
 
  The C720 is one of the few x86 Chromebooks on the market and the best damn value I think for a portable Linux laptop.

Before the using via network (remote X) is available for wayland, and includes root windowless mode, most of non-hobbyist users (in heterogenic corporate environments and such) will never switch over. There's nothing to get excited of if you aren't at even half-way feature parity.

There is now support for remote use. There is also a full screen mode, if that's what you mean by root windowless.

The main issue at the moment is that it is new - I doubt if it is as stable as X - and the limited supporting applications.

Funny thing is, everyone always complained about latency and if you look at this graph you'll see Windows actually generally has higher latency than Linux these days: http://openbenchmarking.org/embed.php?i=1307193-SO-WINDOWSIV40&sha=869d65c&p=2

From this article: http://www.phoronix.com/scan.php?page=article&item=intel_ivbmesa92_win7&num=2

It's good enough for valve to base its console on (and not wayland), it's also good enough for me (FWIW) in that it works and at this point wayland does not as it is nowhere near feature parity with X11.
And when comparing X11 vs wayland for a simple desktop: wayland loses every single benchmark.

why? you want nvidia only piece of shit hardcoded by former nvidia employee?

SteamOS also works on AMD and Intel GPUs.

why? you want nvidia only piece of shit hardcoded by former nvidia employee?

SteamOS also works on AMD and Intel GPUs.

Ironically while Netflix's infrastructure runs on Linux

And even Phoronix managed to get that one right...

I guess you would agree Phoronix is qualified to install the first beta of a Linux based OS? How about they being unable to boot anything but Windows 8 (.1) on hardware with UEFI after lot of trying?

I would have thought a Phoronix writer would be smart enough to realize the real problem was with that bootloader not allowing 64bit EFI images which is not something most (if not all other) UEFI systems do.