* Posts by Lennart Sorensen

265 posts • joined 1 Jan 2008


Trust the public cloud Big Three to make non-volatile storage volatile

Lennart Sorensen

Re: Will we soon be reading about...

I believe a single TRIM command to the drive should wipe it, so I doubt they haven't implemented that properly. Issue wipe command before attaching to new instance.

Really not any different than the all other storage for the cloud.

Time for a cracker joke: What's got one ball and buttons in the wrong place?

Lennart Sorensen

Re: You can also get...

That sounds like an eSATAp connector. Not that weird, other than not being an official standard. Many systems have had them.

What's big, blue, and short on Intel? The supercomputer world's podium: USA tops Top500 with IBM Power9

Lennart Sorensen

The core counts for the Power9 boxes look to be totally wrong. I suspect they were counting threads instead of cores given they mention 4600 nodes with two 22 core Power9 chips each, which does not equal over 2 million cores. At 8 threads per core though (if that is the version of the power9 they are using) it gets close. The numbers for the second place machine don't even divide by 22 evenly, so that's just more wrong. Someone somewhere goofed.

Do not adjust your set, er, browser: This is our new page-one design

Lennart Sorensen

Well it has been quite clear that every time the register changes the layout, they ask for feedback, get lots of it, and ignore 99.9% of it. The opinion of the readers simply doesn't matter here.

Nokia reinstates 'hide the Notch' a day after 'Google required' feature kill

Lennart Sorensen

The notch is an indentation in the screen where the speaker (usually) is. So the screen is not actually a complete rectangle. Hiding the notch is done by not using the part of the screen next to the indentation so that you only use the part of the screen that is a complete rectangle.

Oldest swinger in town, Slackware, notches up a quarter of a century

Lennart Sorensen

Re: Cheers, Slack

Yeah I went:

SLS 1.03 (Wow something better than DOS)

Slackware 3.0 (Wow shared libraries but same dreadful package system as SLS)

Redhat 2.0 through 6.0 (Wow useful package management)

Debian 2.0 through today (Wow package dependencies that actually work, upgrades in place, and packages that aren't full of bugs)

I haven't seen anything improve on Debian yet, so no need to switch.

You wanna be an alpha... tester of The Register's redesign? Step this way

Lennart Sorensen

As if it matters what we think.

Maybe you could just undo the previous change and go back to a proper page width? I don't want the page to look the same on my phone and my desktop. They have vastly different screen sizes and shapes. I still want my screen width used properly.

And I am not convinced by the idea of highlighting anything. I go daily to read the days stories in order and don't care to have them shuffled around since that just makes catching up a pain.

But well what we thought didn't matter last time I have little hope it will this time either.

Oh and having the area outside the articles on the home page be a clickable link to some add or other sponsored content should be punishable by death. I have only ever clicked on it by accident when changing focus of my windows. If it doesn't look like a button it should not be a button.

Well at least so far the new layout doesn't seem obviously worse than the current one, unlike the last update. Of course it also doesn't fix any of what the previous update screwed up.

Intel confirms it’ll release GPUs in 2020

Lennart Sorensen

The term GPU is much older than nvidia making it popular (in 1999). The i740 was 1998. So GPU was a thing, and the i740 probably qualified (as a bad one) even if people tended to call them graphics accelerators at the time instead.

Sysadmin's PC-scrub script gave machines a virus, not a wash

Lennart Sorensen

Re: a Mac SE FDHD installed as a gate guardian

The amiga got 880k and 1.76MB, so there. Of course while it could read floppies from other machines, no one else could read the amiga ones because they had no sector gaps.

Industry whispers: Qualcomm mulls Arm server processor exit

Lennart Sorensen

If the chip makers wanted there to be an arm server ecosystem, the first thing they should do is actually sell the stupid things to anyone that wants to buy one. So far it has been just talk with pretty pictures and lots of specs, but just about nothing anyone could actually buy if they wanted it.

If you think you can stick with only talking to the large cloud providers and people doing custom designs, well then you are not going to get any ecosystem going because the people that would actually play with it and get it working with lots of software can't get your system.

Mum? Dad? Can I have a 3D XPoint disk for my birthday?

Lennart Sorensen

Re: What is the market for these

I thought the 960 EVO was a conventional SSD, and hence the claim of 640MB/s is nonsense given clearly conventional SSD outperforms the intel.

Bright idea: Make H when the Sun shines, and H when it doesn't

Lennart Sorensen

Re: carbon monoxide as byproduct???

So yet more stuff released we are trying to make less of.

This sounds like a far from clean method for making clean energy.

So far I have no hope at all for hydrogen as a fuel source because there simply is no good way to make it, store it, or transport it.

AI smarts: IBM pushes out 'faster than X86' POWER9 servers

Lennart Sorensen

Re: Price-performance

No IBM makes PowerLinux 7rX machines that only run linux and not aix or i. I have never heard of Lenovo having anything to do with those.

There are other non IBM systems out there though. The Talos workstation for example.

Intel Management Engine pwned by buffer overflow

Lennart Sorensen

Re: WTF?

Certainly none of what they listed is required and many people would rather not have it.

Sure Microsoft wants you to use secureboot, and it does have some good features. So why is it in the same environment as remote management which clearly is not required or desired in most cases? Do not combine useful local stuff with optional risky remotely accessible stuff.

Lennart Sorensen

Re: Who is behind all this bashing against Intel ME , uh?

Actually they don't. Some AMDs have something similar. IBM power uses a different chip outside the CPU for management and it has its own network interface that you are not required to connect if you don't want to. No idea what sparc has (are they still doing anything?). ARM supports running trustzone code, although not all of them use it, so it is certainly possible to buy arm systems that don't have it enabled at all. \

Intel's big mistake is putting optional stuff and essential stuff together in the same place. The essential system startup stuff has no reason to have network access at all, and the optional stuff that does have reason for network access should be able to be turned off so it should have been an independent device from the essential startup stuff. Secureboot and remote system management have no reason to share the same CPU and OS.

Want a new HDMI cable? No? Bad luck. You'll need one for HDMI 2.1

Lennart Sorensen

Article is wrong

Well the article is wrong. You will need a new cable to use HDMI 2.1's new higher resolution and refresh rate. Your existing cables are fine for VRR, eARC, dynamic HDR and the other features that are not using the new higher resolutions. Only the higher resolutions require the new 48Gbps capability, which is the only thing that requires new cables. Since every feature in HDMI 2.1 is optional, a device that implements just one of them can call itself HDMI 2.1 and it doesn't have to be the 48Gbps feature.

So for the features that might actually be relevant anytime soon for most of us, existing 2.0 compatible cables are fine.

A cable meant for 1.4 might handle 2.0 in some cases, and it might also fail at the edge cases (when you go pushing the full 18Gbps, not just 12 or 15) in the case of 10 bit HDR at 60Hz 4k resolution. Fortunately premium certified cables can be bought for about $5 and work great for HDMI 2.0 stuff.

Now, now, Qualcomm... Don't play hard to get, grins Broadcom

Lennart Sorensen

Re: Next, hostile

I suspect Qualcomm is very correct about regulators. Broadcom and Qualcomm together own the wifi AP chipset market. Everyone else combined makes up almost nothing in that market. I am sure there are other markets where they would be totally dominant.

So at least some parts of the business would need to be split out, and many of Qualcomms advantages is that they provide all the bits needed so splitting wifi from cellular or processors just doesn't make sense.

Card shark Intel bets with discrete graphics chips, shuffles AMD's GPU boss into the deck

Lennart Sorensen

Re: Great for competition

Oh yes the intel i740. Let's offload all video memory to system memory over the AGP bus. Sure it makes geometry move faster, but sure didn't help the texture performance.

$10,000-a-dram whisky 'wasn't even a malt'

Lennart Sorensen

Re: Carbon dating ?

Things have been radioactive forever. It's part of reality.

Some things are just more radioactive than other things.

For example: https://en.wikipedia.org/wiki/Natural_nuclear_fission_reactor

Tesla share crash amid Republican bid to kill off electric car tax break

Lennart Sorensen

Maybe someone is assuming Tesla will actually make 300000 model 3s in 2018 and sell them all in the US. Or they expect the Bolt to sell a lot of cars.

Lennart Sorensen

I believe the requirement is that the electric range has to be larger than the non electric range (this affects the BMW i3 with range extender for example which has an artificial fuel tank limit in the US). Regular hybrids (like a prius and such) does not apply since the electric range is much shorter than the non electric range.

US voting server in election security probe is mysteriously wiped

Lennart Sorensen

Re: "Low level format"

Actually there is, and you haven't been able to do one for about 25 years now. It is not something done on modern drives, but used to be required on old drives a long time ago.

Magic hash maths: Dedupe does not have to mean high compute. Wait, what?

Lennart Sorensen

Re: unique mathematical hash

Yes hashes are clearly not unique (and hence the article is just plain wrong about that).

Any dedupe system that assumes they are unique and doesn't verify the data is the same after getting a hash match is insane and should not be used. It doesn't matter how low the probability is, I do not want to risk my data getting destroyed because it just happens to have the same hash as some other data.

And yes there are programmers out there dumb enough to assume the hashes are effectively unique because the probability is so insanely low of them having a collision and then treat them as if they were actually unique. Some day their customer is going to pay for that mistake and they might not know it for a long time after it happens.

Bill Gates says he'd do CTRL-ALT-DEL with one key if given the chance to go back through time

Lennart Sorensen

No he meant the keys to hit to get the login prompt in Windows NT. Not to reboot the machine. It happened to do that when running DOS because that is what the BIOS did with that interrupt. Windows NT and newer does something else with the interrupt.

Lennart Sorensen

Re: O la la!

And amazingly control-shift-esc did the same thing 20 years ago too.

Core-blimey! Intel's Core i9 18-core monster – the numbers

Lennart Sorensen

Re: Nice L3 cache you've got there

No on UEFI systems they start in 16bit mode then the firmware rather quickly switches to 64 bit mode and that's the mode it starts the OS in, unless you enable legacy boot mode, in which case it switches back to 16 bit for booting.

Done and done: BlackBerry ties up $940m settlement with Qualcomm

Lennart Sorensen

Re: Who's their Lawyer?

The article was wrong. It has been fixed after it was pointed out. It is Blackberry that is getting a nice pile of cash.

Netgear says sorry four weeks after losing customer backups

Lennart Sorensen

Re: "the outage lost all of his photographs"

It sounds like they had a brilliant design where the local ReadyNAS you can hold in your hand would wipe itself if the ReadyCLOUD account was marked as closed. So if the cloud server makes a mistake and decides the account is closed and wipes the data and then the local NAS gets told that the account is closed, then the data is wiped everywhere. If it really does work that way, then it is a mindbogglingly stupid design. Essentially you are not in control of your local NAS at that point and it can only be considered a backup device, not primary storage, which it seems a lot of people thought it was with cloud backup.

Crafty Fokker: Norfolk surgeon builds Red Baron triplane replica

Lennart Sorensen

Re: I wish him well

The rotary has a much larger mass rotating than the radial, which is not a good thing in an airplane. Tends to make things want to turn when you don't want them too, especially when changing power settings. Some of the high power WWII planes had enough trouble with the mass of the prop when increasing power quickly. Doing that with essentially the entire engine spinning is a bad idea. But that is what they did early on.

It does seem terribly wrong to put a 4 cylinder in a Fokker DR I though.

The DR I replica that my dad flies sometimes in Brampton, Canada at least has the right type of engine in it.

Inside Intel's Optanical garden

Lennart Sorensen

Re: NAND SSD are crap! Stop with the lies against 3D XPoint !

Based on the endurance numbers intel is providing, XPoint isn't looking a whole lot better than NAND, which is rather contrary to the original claims about the technology.

So either intel's specs are wrong, or they are having early production problems causing endurance problems.

Certainly 30 complete writes per day is not a lot more than the 17 complete writes per day that the intel NAND SSD is rated for, especially if it is for 3 years versus 5 years. Certainly based on intel's hype I was expecting 1000 complete writes per day instead.

The latency is rather impressive though.

Inside OpenSSL's battle to change its license: Coders' rights, tech giants, patents and more

Lennart Sorensen

The openssl advertising clause is so obnoxious that if you even say:

Our product has secure connections.

You actually have to do something like:

Our product has secure connections (using OpenSSL copyright x, y, and z, blah blah blah).

Every single time you talk about any feature that relies on what OpenSSL provides. Does everyone do that? Well no, but the license does appear to say exactly that. It is very hard to comply correctly with that license.

It is not just the documentation that has to list the copyrights, or the about info for the application. It's documentation, advertising, discussions of product features, etc.

Microsoft loves Linux so much, its OneDrive web app runs like a dog on Windows OS rivals

Lennart Sorensen

Re: Fixed now

Except as the last few comments there say, it isn't actually fixed for many linux users.

Devs bashing out crappy code is making banks insecure – report

Lennart Sorensen

Re: Agile Payoff

Wait, you mean there is a step to the process after fast and cheap?

Lennart Sorensen

So 40% had bad security and 40% were written in Java-EE. Just a conincidence, right?

Sir Tim Berners-Lee refuses to be King Canute, approves DRM as Web standard

Lennart Sorensen

Re: All the whining in the world...

DRM does not stop piracy.

What would help piracy is to make buying it legitimately actually be more convenient than pirating.

If you could buy it easily, in a format that would be yours forever (so no server shutdowns to worry about), that you could play on whatever device you wanted when you wanted, could re-encode into a format needed for your device easily, then that would be what one would do. As long as a legitimately bought version is less functional than the pirated version, people will be willing to go through the hassle of finding a pirated version.

One IP address, multiple SSL sites? Beating the great IPv4 squeeze

Lennart Sorensen


Better email a correction request.

Germany, France lobby hard for terror-busting encryption backdoors – Europe seems to agree

Lennart Sorensen

Strong encryption already exists. So no matter what new encryption you invent with a backdoor (ignoring for the moment that you can't do that while also making it secure enough to be worth using), there is nothing stopping the criminals from just continuing to use the strong encryption, leaving the new backdoored garbage for the rest of us. So no help for law enforcement, just harm for everyone else.

Softbank gros fromage: ARM will knock out a trillion IoT chips by 2040

Lennart Sorensen

Re: "a trillion chips"

Security is primarily a software problem.

A lot of arm chips are designed to support signed code execution all the way from the initial boot process, so they do provide a lot of features that allow software to be done securely. Of course you still have to make bug free secure software on top of that, but not much arm can do about that problem.

Lennart Sorensen

I did both. :)

Lennart Sorensen

Which kind of billion are we talking? I thought they were pretty much doing that already each year, so unless we are talking the 10^12 type of billion, this doesn't seem like news. Certainly for the 10^9 type of billion, I believe they were at 12 billion in 2014 alone, and probably higher now, so a couple of decades at that rate with a bit of growth might very well hit the 10^12 type of billion (although that meaning seems uncommon in English these days from what I can tell).

Reading the Reuters article in fact says Trillion, not Billion, which seems to clarify things.

Intel reveals Optane will need a 7th-gen core and a PC-centric launch

Lennart Sorensen

Re: Oh just what I need

The DIMM ones are very non standard and I haven't actually seen OS support for them (Diablo seems to be totally failing to provide sources for any drivers as far as I can tell, so no idea how they are supposedly working with linux at this point). At least Optane appears like it might be NVMe compliant, although the fact they claim you have to have the NVMe mapped through the PCH is a potentially bad sign for actually being standard. I am suspecting that they are saying that so it can be used as a cache for another drive, which is one of the RST features intel offers for windows users and if the Optane drives are only 16 or 32GB, then caching might in fact be their best (only?) use, rather than primary storage.

In fact based on what I have read elsewhere, an Optane drive will in fact work in an NVMe slot on a Z170 board, but it won't work with RST as a cache drive, which is the new feature of the Z270. So if that is actually true, then the Optane drive looks like any other standard NVMe drive, and you can use it as such, although since it is rather small, you probably don't want to just use it as a stand alone drive. Linux users can probably use it as a cache drive for another disk using bcache or lvmcache.

Intel Atom chips have been dying for at least 18 months – only now is truth coming to light

Lennart Sorensen

Re: First gen 22nm FinFET / 3D transistor / whatever?

I don't think so. From what I have seen it is simply something got left out of the design that had to be there, and is why there is a workaround involving resistors possible.

Lennart Sorensen

Re: sanmigueelbeer

The pentium was socketed and easy to replace. This one is soldered on the board.

Lennart Sorensen

These are chips for embedded systems with long term supply promises. This is very much not a chip that is end of life yet. It was supposed to be available for at least 5 years I suspect, maybe more.

Intel's Atom C2000 chips are bricking products – and it's not just Cisco hit

Lennart Sorensen

Not rebooting is not good enough. The clock signal is used for quite a few things inside the CPU.

Having the system sleep when not working will reduce wear on the clock and make it last longer. Makes sense that things that are off last longer than things that are on. :)

Lennart Sorensen

Re: This:

They are not saying they will all fail, they are saying that the rate of failure starts to go up more than normal for intel's chips, due to a design mistake on the LPC signals.

So you might have a system that fails in 18 months, or you might have one that fails in 36 months, or one that never fails. Intel almost certainly has statistics of how much the expected increase in failures is after a given amount of time, but they aren't likely to share that. Could be the failure rate is 50% higher than normal, or 5000% higher (I have no idea what the normal failure rate for intel chips is, although based on the ones I have dealt with over the years, I have no seen very many fail). If the normal failure rate was 0.1% and it is now 1% or 5%, well that's certainly a problem, although it might still mean that most systems will be OK. Unfortunately intel isn't likely to share that level of details although I am sure they have done the calculations and hence determined it was bad enough that they had to admit to it.

Lennart Sorensen

Re: So basically...

I suspect their customers might in some cases want to continue to sell products even if they have to replace them and swap the CPU on the one they sell now. It is their choice to take the risk after intel tells them about the problem (And intel will probably insist on them signing something to continue to receive the chips with the known problem to reduce intel's risk at that point). So I would think it is still shipping although probably not in the same quantities as before.

New chip revisions take time to make and validate, so certainly not fixed yet.

Lennart Sorensen

The only fix so far is to change your own board to add the workaround. New chips don't exist yet so no one is getting those until they exist. So everyone is at their own mercy about how long it takes to change the board design and get new boards made, or they can wait for the new chips and hope for the best in the mean time. Doesn't matter if you are Cisco or some tiny company. Of course I suspect Cisco might very well be able to get a new board revision design made a lot faster than the little guys.

Lennart Sorensen

The work around means adding some resistors to the design, which on most boards is not something you can just do, since this is a clock line. You can't just add wires and resistors since that would mess with the clock signal. So it is either change the board design to add the resistors, or wait for the next version of the chip (which will probably take months to happen). Of course since the chips are soldered to the board (not a socket), they are not easy to replace either.

Chinese bloke cycles 500km to get home... in the wrong direction

Lennart Sorensen

So one month into his trip and he made 500 km? At that rate he would need 4 months to get home. Was he only biking 1 hour a day and very slowly?


Biting the hand that feeds IT © 1998–2019