754 posts • joined 21 Sep 2010
Re: large POWER8 servers are out:
"Bye, bye POWER and AIX. HP has also left the high end server segment. Left is Oracle who is betting
more money than Sun ever did on large Unix servers. "
Hmm, article about IBM going fabless (like SUN, Oracle, MIPS and ARM) and you post the same old FUD about the death of POWER... Are you trying to tell us that fabless = death of an architecture ?
A cynic could be forgiven for concluding that Larry has forgotten to give you new instructions over the past 12 months. I wonder if you will get a new shilling to go with the "Oracle loves Clouds" message ?
Re: Microsoft loves linux
One of my favorite cartoon moments, thanks for the memory launcap. :)
That said, I think Nadella reminds me more of Skeletor.
Re: Such hatred
"Why do I get the feeling that a lot of the people that want to keep sysvinit have never tried to read a service script, let alone write one."
Speaking for myself I have read a fair few of them and written a couple and I didn't find it very hard the first time around when I was a PFY either, they're just shell scripts after all. The nice thing is with init.d/* is that you could swap the shell scripts for something else if you really hate them, binaries, makefiles, or even roll your own special init.d shell if that is too easy.
Sigh, DOS as a security feature...
I can see how Vodaphone came to make the schoolboy error of locking accounts after N failed attempts. Microsoft made DOS by login failure a feature on nearly every workplace desktop and people have copied it ever since...
It would be nice if some big names made some moves to fix this particular misfeature.
Re: Embrace. Extend. Extinguish. from 20 years ago. Do get with the times.
"No need - current versions will still open files even as far back as Word 6!"
Try opening a Word 6 document written on a Mac way back when see how far you get. :)
Re: @AC - Windows 10 vs. Linux
"Or at least PCs not infected with SecureBoot, I would still be happy with that."
That doesn't seem like an unreasonable request, yet we have some down voters who are against other people having a choice of what they run on the hardware they paid for.
"Well, one of the WinNT features from the early days is the POSIX and OS2 subsystems (the latter being removed finally). The mechanism is already there. MS could add a full on Linux subsystem to run existing Dockers."
1) The POSIX support wasn't actually POSIX compliant.
2) Containers are totally different in their scope, design and application.
Re: Embrace. Extend. Extinguish. from 20 years ago. Do get with the times.
"I'm sure we could find some VM running Office '95, then open/save the file through each version until we get to now."
Naw, you'd need to start with MS-DOS 5, and find a licensed copy of Word on media that hasn't turned to compost.
"The biggest surprise is how you still seem to comment on something you know nothing about..."
That barb can't be taken seriously when posted by someone called "Anonymous Coward"...
Re: Contempt for UK law
"These people clearly have something to hide. They are ducking and weaving around the law put in place to protect us."
Sounds like the perfect assignment for PC Simon Harwood. He has a cracking track record of protecting the public from protesters too, so I am sure that he will help them make their way home safely if they have nothing to hide.
Finally, a decent replacement for SliTeX & WTF is with the down vores for Matty B? (was Re: Terry 6)
"Hmmmm, I think the intended market is the real "99%", as in the 99% of smartphone, tablet and PC users that aren't power-users, that don't want or know how to use PowerPoint or Photoshop, but want to quickly be able to put a 'slideset' together for a limited audience"
I know this is a Matt Bryant post, but it makes sense, it isn't insulting anyone and I can even agree with it. I really can't see any point in down voting it which makes me wonder WTF the down voting clowns are up to.
Kinda weird seeing MS produce a SliTeX for the smartphone generation. :)
"In my old age you just shake your head & remember what on-line games were like back in the 90's. Oh the same! People trolling, just not as much of it due to the number of players online back then?"
Trolling has always been an integral part of games (and sport) both online and offline, Aussie cricketers like to call it sledging.
Re: Trolling for suckers
My personal favourite was posted to alt.flame, it went something like this:
"Hi, my name is Mel, can I be your friend ?"
It generated a *lot* of traffic. :)
Re: Best case
"It's been happening for the past 2 decades with the OS, too."
Fair point, and probably a better (if inflammatory) example. Have an upvote. :)
Re: Best case
"The other option is so poisonous that the entire software industry will fail, leaving at best, a weak freeware / open source model since there isn't anyone worth a lawsuit. I really doubt that will be allowed to happen."
Doubtful. I suspect that a few big players & trolls will survive in the US/European markets, everyone else will be sued to oblivion, customers will pay a lot more for a lot less and the industry will stagnate. Meanwhile in places where the choke hold of IP lawyers is less pronounced (e.g. Russia,India,China) the software biz will thrive while the IP titans of the West will continue to produce overpriced crap that fails to sell in the global marketplace, WTO@USofA notwithstanding.
There would be second order effects too, Western businesses being forced to use uncompetitive crapware will not be as productive relative to the places that allow competition in the market place.
We've seen that happen with the hardware business already...
Re: Oracle are most definitely evil
"Google have been making money from the innovations of others for far too long wihout paying adequate compensation. I hope The Borg gets badly burnt!"
You must find it a delicious irony that the plaintiff in this case ships a clone of RedHat Linux, and actually makes it a selling point that it's a total rip off with the stated intention of driving the original developer out of business. They make Google look like cuddly kittens by comparison.
Re: God, keep me from harm and working on FPUs
"The problem lies in guys like me having to make transistors give you an answer in a small amount of time and guys like him expecting results regardless of how long it would take."
No, the problem here is the quality of Intel's documentation.
In the 80s at least one chip vendor managed to come up with formal specs for floating point instructions and then prove that the hardware meets them so no one gets any unpleasant surprises. By contrast Intel's published documentation has a habit of specifying complex behavior using ambiguous waffle, so it's hardly a surprise that the hardware fails to match expectations of software developers.
"“Better products will do better because buyers will quickly be able to determine that they’re better,” he said."
Kinda weird seeing such a naive comment emerge from Bruce. He must have been asleep in the 80s, 90s and 00s.
"For all your snarky sniping and your dedicated vitriolic fanboy bullshit, Mr Anonymous Shill, you never do address these issues. And at the end of the day, I'm just tired of fighting the battle. Against Microsoft, against you, against the legion of paid "evangelists" that make truly obscene money to "control the message"."
That's the point really. It's wasted energy, which is better directed towards using gear that works for less headache and expenditure. The shills had already lost this debate in 1992, the current crop are too dumb to know it or too scared to admit it.
Re: Calm The F**k Down!
""Isn't that how we've been treating Windows for years?"
Until we recently realised that Open Source is probably even worse...."
Naw... At it's worst Open Source is as bad as payware, because the payware makes liberal use of Open Source code anyway... The TCP/IP standard was initially defined by a 'libre' style code base for starters...
Payware has done good stuff too. :)
Re: How long before....
"To fix that, one had to set the file to 'read only.' Not sure if that still works. Anyway, that was back in the days of dial-up."
Even in those gnarly old days OpenBSD firewall FTW... Mainly because it was the smallest download, but in practice it turned out to rock solid and mostly excellent in day to day usage too... :)
Re: All the more reason...
"In essence their addressable market consists of Windows fanbois who don't like Microsoft..."
Hmm, down vote. Is that because they don't like the truth or is it because they disagree ? They haven't posted a rebuttal, which suggests that they don't like the truth.
Personally, I'm quite happy to see ReactOS thrive and do well. More OSes, more choice is a good thing. Maybe the effort will yield something new and useful - I just think it's a very very very long shot with ReactOS.... :)
Re: All the more reason...
"How come no one is giving ReactOS any love at all?"
Err, because it's pitched as a clone of Windows and you can get the real thing without the threat of MS cutting you off at the knees for *£0 when you buy a new computer. * = the MS Tax.
Even if you ignore the competing with the superficially free real deal, their blurb about how much UNIX sucks suggests that they are doomed to repeating the same mistakes that Microsoft made 20 years ago, which is also a big turn off if you're looking for something similar but better.
In essence their addressable market consists of Windows fanbois who don't like Microsoft...
“Full encryption of communication and storage online will make life very easy for the criminals and terrorists and very difficult for law enforcement and law abiding citizens."
Making the failure to hand over passwords/keys on request a criminal offence covers that one (lucky us in the UK).
Surprised that he stopped at private comms & data, he will be asking us to leave the keys in the ignition of our unlocked cars on the grounds that it'll make policing car crime a bit easier next.
"Technically, at least according to the Department of Defence and the Executive Branch, the FBI could be an accomplice to acts of war against foreign nations. "
Starting wars is BAU for the US government agencies, and as most of the death and destruction happens elsewhere and no one is prosecuted for their criminal behavior there is no incentive for them to cease being criminals.
No, it's the other way around. Sabu was committing those offences at the behest of the FBI in return for more lenient treatment. Would be nice to see his handlers pay for the damage they have willfully caused out of their own pocket. IIRC in McKinnon's case the US Navy assigned a value of ~$250K per login with default password, and inflation has happened since then, so I think the FBI handlers could be on the hook for a few $m a piece, perhaps more if Sabu had to apply a bit more effort than using the default password to crack those machines & accounts.
Re: Get back into the HPC market
"Synchronization messages over a cluster are slow, to nodes far away in the cluster. A tight SMP server is much much faster...""
For a small subset of definitions of "cluster" "tight" and "SMP" that is true (specifically the non-standard hand-wavey definitions you use). Meanwhile in the real world where people pay real money for stuff to go faster, I have spent over 20 years migrating workloads from monolithic DB centric designs to federations of SMP boxes. People pay me (and lots of other people) real money to help them do these migrations because there isn't a "tight SMP" box fast enough to handle their growing workloads.
By and large any application that uses a DB as a central point of synchronization simply can't play in the big leagues anymore. Sure, there are a bunch of folks out there with more modest requirements, and maybe a SPARC box that doesn't set the SPEC world alight will suit them fine, but that is a pretty tiny and shrinking market - as you well know, most *new* customers will go the x86 route in this day and age.
Re: Lot of ignorance from the IBM crowd here
"I answered and wondered why you compare a 2-socket POWER7 server to a 1-socket Oracle SPARC server in terms of SPEC benchmarks."
Those two bits of kit were the closest I could find in terms of specification and shipping date. My intention was to illustrate that POWER boxes have quite a big margin over SPARC T5s per core when running highly parallel compute intensive workloads.
"Is that fair you think? Let me compare a 32-socket SPARC M6 server to a 1-socket POWER6 server - would that be fair you think?"
If they were shipping at the same time cost a similar amount and take up a similar amount of power and space, I figure it's a fair comparison... I found that Oracle are quite cagey with their SPEC results, and that is ultimately what drove me to choosing those particular machines for comparison.
"In short, it is you that FUD about me, trying to imply that I avoid SPEC posts."
Naw, there's no FUD, all I do is repeatedly point out that you compared single core results for POWER boxes to multi-core results for SPARC boxes.
I accept your quibble about 2 sockets being better than 1, the extra pin count gives those POWER boxes more memory bandwidth and more I/O bandwidth - and that is a USP for many of the folks who buy POWER boxes.
"This is supposedly "proving" that SPARC is much slower on SPEC, because I avoid those posts. "
What actually happened is that I spent a bit of time to find SPEC submissions of comparable age and hardware config which happened to show that the POWER7 system with a similar number of cores beat the SPARC box by a big margin in a set of compute throughput benchmarks. In case you hadn't noticed the POWER box was using SLES + GCC, while the SPARC box was running Solaris + Oracle's amazing SPEC pwning compiler. If I was a mean bastard I would have posted the AIX figures.
"Why stop there, why not compare four POWER8 to one SPARC instead, while you are at it?"
That's an easy one: Because it would be as pointless and ridiculous as your comparison of SPEC to SPEC_rate results.
Re: Get back into the HPC market
"HPC is fast growing out of their niche market. Oracle should try and get back into the HPC market...look beyond just running their suite of software on SPARC."
Exactly... Fujitsu have done it, with SPARC, repeatedly, so it's clearly possible. But then you look at the state of the art and you are left wondering whether Snoracle have the stones to leap frog stuff that is currently shipping, sadly I think the answer is no. I wish it were otherwise, more competition would be good in the marketplace, but the last time we saw a quantum leap in performance was the DEC Alpha, I suspect that kind of leap in scalar performance isn't repeatable (and in fairness Intel got pretty damn close to it fairly soon after with the Pentium Pro). :)
Re: Lot of ignorance from the IBM crowd here
"trying to trash talk and spread negative rumours of the SPARC M7 cpu. For instance, say that the M7 memory protection new functionality is nothing more than an ordinary MMU"
In fairness the quote from Larry didn't exactly spell out what was new or different from anything that had gone before. If you understand the difference you could spell it out for the folks who want a bit more substance than Larry's content free blather.
"let us see IBM POWER8 do 120GB/sec querys."
Nah, let's wait and see if M7 boxes actually show up and do this first in independently conducted repeatable and verifiable tests, after all Larry has incentive to lie and exaggerate his product's performance, and we can't trust anything you say because you keep comparing SPEC to SPEC rate figures... Come to think of it you keep ignoring any SPEC rate figures that beat the SPARCs as well.
"Well designed specialized hardware accelerators are typically >10x faster or so."
You really shouldn't have bothered making that comment, it's content, context and fact free.
"For instance, compare GPU vs CPU."
GPU = executes thousands of short lived threads with very few branches/conditionals, extremely regular memory access patterns, low-medium single thread clock rate...
CPU = executes code with a branch/conditional every 3-6 instructions, highly irregular memory access patterns, very high single thread clock rate.
"And the SPARC M7 is 4-5x faster than the SPARC M6. Let us see the benchmarks, until then you can continue spread the FUD and negative made up rumours."
While we're waiting for some independently verifiable benchmarks, it would be great if you could pop over to spec.org and learn how their benchmarks work so you don't make the mistake of comparing the rate figures to the single core figures again. It shows a lack of professionalism that reflects poorly on you and the products you push so vehemently.
Re: " instantaneous response to changing conditions is required, such as lighting management"
"I could rub two boy scouts together" - have you consider Catholicism?"
Nearly gave an upvote for appropriate use of A/C. :)
Re: What else?
"I can see that after openssl and now bash, lots of open source tools whose code has not been looked at for decades because they have, up to now, just worked, will come under intense scrutiny."
That's no bad thing IMO, but this process isn't a new event in the Open Source world either.
I get that some people honestly can't tell the difference...
I can tell the difference with my battered Rega speakers & amp, with most recordings... Stuff like Peepshow from Siouxsie and the Banshees sounds completely different in a compressed format. Like another commentard said, storage is cheap, FLAC shrinks most stuff to 60% it's original size or better so it's not a big deal to rip CDs to FLAC. It's a great format, simple and effective. I guess a lot of folks are desensitized to compressed music, I don't resent that, it's what suits them, just let me carry on finding new music that sounds good to me. :)
I think it's more of a case that companies figure the cost of doing it properly outweighs the cost of mitigation. I would be surprised if many of them have even considered the customers losses in that equation.
And what 'frank ly' said too.
"Was he presenting his images/thoughts to fellow passengers like some kind of infomercial?"
The linked article states that another passenger was reading his notebook over his shoulder, and made a complaint to the flight attendants. Presumably the accuser felt that the victim was being disruptive because he was talking up valuable leg/elbow room.
"I meant was not a file system, per se, but of just taking the data out of the database itself a slapping that crap right onto the disk itself"
People have already done that with RDBs as standard operational practice... Still not seeing what's new or better here.
"a huge number of SQL-specific cores thrown in there too, similar to modern supercomputers and using GPU to do the heavy lifting"
That particular wheel of reincarnation has made several revolutions already... I really don't see how your approach would yield anything that hasn't already been tried or yield a better result than previous efforts. Today general purpose processors such as ARMs, Xeons and POWER8s come with app specific acceleration HW.
"Yes, Linux and Solaris have 45+ years of OS expertise, but almost all of that is on things that aren't databases."
Yeah right, all that boring stuff that isn't database specific, but also happens to be absolutely essential for modern DBs such as resource management, I/O, authentication, access control, diagnostics, networking, device drivers, etc, etc, etc...
"I mostly mentioned ditching it since the GPL would cause a huge legal clusterfuck"
Err, they could pick up a BSD licensed OS instead if that's really a problem.
"Network stacks aren't all that hard to write (I've built a few myself to run on controllers) especially when you know what is going to be sitting above it."
Sure, but with a modern DB that just isn't the case, these things plug into networks that are host to all kinds of surprises, subtle networking bugs and belligerence, all of which is outside of your box's control.
I suspect at some level we share a preference for simple modular systems, although my perspective is colored by studying 50+ years worth of whacky systems, some of which failed, some of which flourished, and many features of which have now ended up borged into mainstream systems. :)
The surviving architectures are the product of decades of innovation, optimization and winnowing, I'm fairly confident that they are not wildly off the mark in terms of trade-offs.
"*Build a new File System specifically for databases, possibly to the point where partitions/directories/files no longer exist, only databases/tables/lines"
Err, like Pick in the 80s/90s, and like what Microsoft has spent 15+ years trying to do but has repeatedly shit-canned. Oracle are *weaker* than Microsoft are building that kind of software at present (IMO)...
"*Build a new processor centered around SQL commands, allow database locations in place of addresses, etc."
It's pretty clear that 40 years of microprocessor design has passed you by, there are some very good reasons for ISAs being the way they are right now. There is quite a lot of material available on odd-ball designs and the strengths/weakness of various approaches. A few years back Teradata came up with DB specific hardware - worth a look at some white papers if you are genuinely interested.
"*Forgo Linux and Solaris in favor of building out their software to run directly on bare metal"
I really can't see how ditching ~45 years of OS expertise is going to move them forward... They'll have to write an awful lot of drivers again for a start off - which is a time consuming and painful process at best... Presumably they'd be writing their own network stack as well... I really can't see that one working for a R&D dept run by Larry, Hurd & Safra.
Re: Re : Asperger syndrome...
"Unless you mean our political idiots here in the UK will make a law making it legal for foriegn agents to hack computers willy-nilly which I cant see them slipping past the electorate with much ease (well I hope not anyways)"
Err, laws, infrastructure, and a huge whacky shaped building already exists for exactly that purpose, surely you must have noticed all that heat and noise about Snowden, GCHQ & the NSA...
Opening the stable door so the horse can bolt...
"there's evidence that the FBI is already involved in overseas cyber-ops of one form or another"
So they are making it all legal so no one has to be investigated and the horse can frolic happily in the ragwort infested meadows. Well played.
Re: Sympathies to Oracle
"Engineering will still report to Larry.
So, the death by a thousand Hurds of any R&D in Oracle is postponed for the duration."
I don't think that would make any material difference to the customer, Oracle's R&D strategy over the past decade appears to consist of two plays "buy next big thing" and "gouge the customers".
What a total bunch of Assanges.
"FinFisher continues to operate brazenly from Germany selling weaponised surveillance malware to some of the most abusive regimes in the world."
Looks like Wikileaks has gone into the end of the weapons business, didn't see that one coming...
The most likely net result of publishing "weaponised surveillance malware" at zero cost is going to be a few more innocent bystanders pwned. At a push it may put a dent in FinFisher's sales for a quarter - until their customers realize they need an update to keep it useful.
Re: Political more than commercial (wrong end of the stick there Led)
"Abandoning WP is a daft idea from Huawei if on purely commercial grounds, because it gives them less of a stick to wave at Google,"
The problem for Huawei is they were holding the shitty end of the stick, the one with low sales revenues and no reason to believe that the turd coating will spontaneously transform the entire stick into solid gold.
That's a shame...
Another fine product doomed to be bled ugly by a corporation. :(
I would be *very* happy if Microsoft proved my cynicism to be unfounded.
Re: RISC, not IRONIC
"Yes, I deliberately didn't mention those because I'm not sure they actually help with code density."
I got the impression (years ago, in the days when people were speculating about EPIC) predicated instructions were about improving performance rather than coding density.
Re: 384 cores in one memory pool?
"is it coherent? If so, how the $%^&*( do they manage to do it?"
Here's my value-free unresearched cynical speculation : Magic Smoke.
Re: Low Hanging Fruit
"Once again the British Government, and the media barons, show a complete lack of understanding of how the internet works, and concentrate on the search engine indexes."
It pays to judge people by their actions rather than what they say...
In this case you will see that the folks making these huge blunders in understanding are trying to change the internet into something completely different that suits them better (typically: from unrestricted distributed control to an authenticated & centralized control model). It is easy to make fun of these plonkers, but some of them carry enough to weight and influence to make life very miserable for everyone.
This blather is corrosive, some people have questioned why I choose to run my own mail server in preference to using a more "legit" service like gmail... Go figure where that kind of thought process is going to lead...
Re: Count on it.
"Bottom line: state-of-the-art microprocessors run 24 x 7 are going to have a limited life. Credible speculation is that this could be as short as a few years. And nobody appears to be seriously thinking about the cost of end-of-life replacement."
Precisely the premise of the early BlueGene machines. They used tried & trusted embedded cores at larger feature size & lower clock (better FLOP/W *and* higher MTBF). Superficially it looks as though BlueGene/Q is following the same path. Someone might take ARM in a similar direction, it has already been done with MIPS64 (SiCortex).
Ballmer 0, Stroustrup 3 (HT)
I have to say that Bjarne's choice of face furniture is awful. He carries off the 70s badman speed wasted biker look far better without the inverse Dama Edna effect spectacles.
Argh !!! My fscking eyes !
I just happened to be browsing the Reg this evening when my eyes were violently assaulted by the For Dummies themed advertorial backdrop. I can't tell you what it was advertising because I had to vanquish said abomination before it had a chance to progress further into my precious grey matter. The only thing I remember was the outline of a severe looking owl.
Note to el Reg: Think of it as a back-handed flame.
Anyone else been struck senseless by a piece of advertising @ El Reg ?
Re: There's no such thing as a secure platform...
"Apple might have screwed the security pooch."
Nice to see some upvotes shielding the bearer of bad tidings. Thank you. :)
- Vid Google opens Inbox – email for people too thick to handle email
- RUMPY PUMPY: Bone says humans BONED Neanderthals 50,000 years B.C.
- Pic Forget the $2499 5K iMac – today we reveal Apple's most expensive computer to date
- Is your home or office internet gateway one of '1.2 MILLION' wide open to hijacking?
- Review Vulture trails claw across Lenovo's touchy N20p Chromebook