Posts by Lysenko 986 publicly visible posts • joined 23 Jan 2015
I find it interesting no-one ever considers that there is potentially a genetic engineering answer to the problem of getting humans to other stars. Immortality. We've had immortal human cell lines (e.g. HELA cells) for decades. Scaling that to an entire organism doesn't require rewriting the laws of physics just ethics and that's a much more malleable area.
“In the old days, ... optimised for reducing Mean Time Between Failure [MBTF]. This was appropriate in a world with a high cost to repair, ... In the new DevOps world, you optimise for Mean Time To Remediate (MTTR), because you make the cost of repair very low by allowing very rapid deployment and limiting exposure,”
...is just our old favourite: "Beta test in Production". No surprise it comes from an industrial scale patch factory like MSFT. Poster children for this approach are IE, Java and Flash. Don't worry too much about shipping security holes because you've got high performance patch servers that allow you to fight a continual rearguard action against your own continuous delivery of agile defects.
The problem is 'cost of repair' isn't a simple metric. The cost to Adobe of shipping continuously insecure Flash is destruction of customer confidence in the brand, which is not something a DevOps toolchain can track and churn out pie charts about.
Cost of remediation is already way too low and has been encouraging bad practice for years. Lowering it further is solving the wrong problem and it only takes one serious economy updating it's Trading Standards and Consumer Protection legislation to invalidate the entire premise.
[*] And lets not even think about IoT with millions of units with embedded firmware that is next to impossible to update efficiently if at all.
WTF is a "resource action" and what is a member of an employee organisation doing babbling like a retarded Catbert?
If they are terminating the employment of staff en masse then that is "Redundancy". It has a legal definition (in the UK at least) and you can't weasel out of it by using nouveau terminology dreamed up by a "differently abled" Personnel Manager (pronouncing HR just sounds like phlegm).
@Graham
I covered that in the second paragraph. You can trade profit for goodwill (or other investment) so long as the final analysis works out as enhanced shareholder value. What you can't do is act to consciously diminish shareholder value in the pursuit of some other objective. Not without explicit shareholder approval anyway.
To voluntarily forego an available tax avoidance strategy the Board would have to be able to show that the goodwill accrued to the company is more valuable than the cash expended or that it is approved by the company's articles (e.g. investment squeamishness by the Church Commissioners).
>>Because Google is such a small company that they *need* to use those >>rules
No, its because Google is a publicly traded company and it *needs* to maximise profits (and therefore minimise taxes) in order to discharge its legally binding fiduciary duty to its shareholders.
You can trade off a higher tax payment against intangible goodwill but in the final analysis the deal has to pan out in the accounts as more value for Google shareholders.
No it didn't, it was/is called tax avoidance and it has always been legitimate. The difference is primarily one of intent. If you honestly and reasonably (i.e. not culpably irresponsibly) believe you've found a technical ruse to avoid paying tax then that's legal. If you're wrong then the worst you can generally expect is a retrospective bill (with interest). Tax evasion is dishonestly misrepresenting your affairs to reduce or eliminate your tax bill and can land you in jail. Mistakes are not fraud and being wrong doesn't necessarily make you guilty.
The Government is not allowed to win with the argument: "Yeah, yeah whatever! It doesn't matter what Parliament SAID, what they MEANT was...". Damn good thing too given the abuses it would lead to.
No-one actually knows. Until you take it to a tax tribunal, appeals and then potentially Judicial Review (all at vast cost/profit to m'learned friends) all you've got are opinions of varying levels of mendacity.
Unless you're prepared to threaten retrospective legislation there is the real possibility of going after Google for a lot more money, losing and getting nailed for both your own legal costs and theirs.
This sort of law is essentially Poker and HMRC vs. Google is a no limits game between the highest of high rollers. Such players generally avoid each other for the very good reason that it is too risky and rarely profitable. The "correct" way to play is to get into games with opponents (SMEs) who can't match your resources - then you can buy the game with a big bet and a bluff.
Tax law has damn all to do with finding the truth, being fair or punishing cheats (that's what Criminal Law is for). It is about efficiently hoovering up cash for HMG. That means targeting weak opponents and staking as little Government cash as possible on each hand.
...that this has nothing (NOTHING) to do with using shoddily weakly typed languages like Pointy Haired Programming and would have been inevitable even if the whole thing were written in something with a type safety and didn't allow EXECing of strings.
Glad that's out of the way.
They have a point in some cases though. If you're in an ER you might need to pull someone's medications, allergies and blood type *NOW* and playing pin the tail on the password might kill someone (and no, you can't take off the gloves for a fingerprint scan).
They need something more like an NFC wristband so the terminals unlock and login automatically when the user is within six inches of the keyboard. (Some) Doctors may be arrogant but they're not stupid. If IT security impedes patient care they'll find some way to subvert it.
...might be to charge Corporates and (especially) Politicians membership/usage fees. They've got a critical mass of customers/voters and they're giving away all this free advertising and customer support. The US Election cycle alone (hell, Donald Trump alone) could probably kick them into the black.
If losing one exec could be characterised as unfortunate inadequate, and two as carelessness promising, we’re not quite sure how to describe losing four – hang on – five, over one weekend.
Four: Definitely Encouraging
Five: Clearly Serious
Whatever you think Twitter and its Twits, what is not in question is that it is a gaping abyss where money goes to die. The 8% of staff who were axed previously were not responsible for that. Either the problem is the senior management or else the whole business model is flawed. My money is on the latter, but either way a "twilight of the suits" was overdue.
I worked on a project like this 15 years ago ...up until Training & Enterprise Councils were abolished. The systems could never be completed because of perpetual feature creep which was in turn a direct function of "Budget Announcements".
Every year there was some pointless tweakery implemented with impossible timescales and no forewarning ...because budgets are a secret!! The result was naturally the an Agile methodology with continual delivery of half finished rubbish. Unsurprisingly, nothing much seems to have changed.
That hasn't crossed my path personally, but I don't doubt it has happened. Most CoLo users I deal with have redundancy so if a BigData Inc. DC goes down for any reason there's a HugeData Inc. hot standby which has to be at least 100 km away.
The point I'm making is this isn't really about cloud vs. CoLo vs. in house. It's about redundancy. People without DR plans get burned no matter where or how they implement their systems.
Not even space? I'm not being sarcastic - I'm thinking CoLo.
People have been outsourcing the physical hosting of their servers along with associated power, cooling and network backbone services in third party data centers for decades. If one of them suddenly blinked out of existence then re-locating all that computing hardware would arguably be a bigger headache than just porting data somewhere else.
Doesn't happen often with CoLos of course because there is a tangible asset involved (the DC) and that gets sold on as a going concern. "Cloud" can be fragile because there are too many entirely virtual outfits with no tangible assets backing them up. That's fundamentally from the AWS and Azure cloud propositions.
That's my point. If you do that then you're not an "in-memory" database. All C/S RDBMS engines have read ahead buffering, LRU page caches and RAM based lock managers. Some also have MVCC (e.g. FireBird) and cache complex structures like views and OLAP cubes. No-one calls them "in memory" databases just as no-one (sane) calls MongoDB ACID.
I'm familiar with some of Stonebraker's papers. I've been using his stuff since Ingres on DEC Ultrix. As far as I know none of his work refutes any of this. Neither does Jim Starkey's stuff, and he disagrees with Stonebraker on many issues.
>>Splice Machine RDBMS, a hybrid in-memory technology that incorporates >>Hadoop, ANSI SQL, ACID transaction
Unless they're counting an SSD as "in-memory" (stupid) or redefining "in-memory" to mean "has read caching" (mendacious) the only way to make DRAM based data engine ACID is to disallow writes.
D=Durability. A session cannot be allowed to StartTransaction until the previous Commit has a confirmed clean write to non-volatile storage. Doesn't matter if you replicate the transaction log to other nodes (e.g. MongoDB) to reduce single point failure risk, it doesn't qualify as ACID. Proper ACID systems go to great lengths to disable or force flush OS and storage controller write buffering for exactly this reason. That doesn't mean all (or even most) systems need ACID of course.
Not bad at all.
Mentioning Coho Data and Andy Warfield (who he?) puts it in quasi-Advertorial territory since they're irrelevant to the subject matter, but there was no gratuitous plugging of "DataStream".
Moderately informative content.
Conference plug at the end: innocuous and easily ignored.
Tricky. Depends on the recipe.
Using rabbits as a proxy: the carcass is about 60% edible and gross weight (cat) about 4kg ...so 2.4kg of usable cat. Assuming high quality catsup with at least 50% emulsified cat and other ingredients liquid, that's nearly 6 liters of "brown" catsup per cat.
However, "red" catsup is a different matter. You only get around 50ml of blood per kilo of cat so that's about 2.5 cats per liter.
[*] I keep pet rabbits. If Gastropubs think it is OK to serve rabbit without general outrage, I reserve the right to devise cat recipes ;)
I was thinking of the USAF/Boeing X-37.
Getting side tracked into the precise flight characteristics of a "space plane" seemed a tangent too far, particularly since most of its capabilities are classified. What is known is that it is aircraft (specifically, a glider) capable of going into orbit (like a satellite) and staying there for at least a couple of years before re-entering the atmosphere and landing (like an airplane).
@RIB
* I'm arguing that high speed strike aircraft vs. SAMs 40 years ago has little predictive value regarding huge, slow loitering drones vs. SAMs in the future.
* Powers. Yes, they would care. Grabbing a body was a bonus. The wreckage was enough to prove the US Government were lying through their teeth - which was the primary issue.
* Swarms of drones. No. Not loitering around anyway. You don't need long range detection if the target is going to obligingly stay put in a predictable place ergo you don't need to activate ground radar. If the "drone" you are thinking of is just a parachute retarded anti-radiation missile then that's existing technology. Useless against IR, beam riding or active homing missiles.
* Attacking SAM sites right next to a strategic asset is indistinguishable from an attack on the asset itself. You don't know what warhead the inbound missile has until it explodes. If you're not prepared to risk MAD then you can't shoot at all. That means at the moment the ICBM launches you're either engaged in evasive action or already in the process of exploding.
I don't expect to see laser weapons on anything smaller than a ship in my lifetime. The Standard Missile (RIM-161) is much more credible technologically and possibly also the good old "lead laser" (CIWS/Vulcan etc). Given that a laser drone would be the size of a Hercules, fitting a rotary cannon is no more impractical.
* None of those conflicts involved drones.
* In Kosovo a F117 "Stealth Fighter" was taken out by a '60's era SA-3 SAM system ...and lets not forget Gary Powers.
* None of the aircraft involved were slow, lumbering, AWACS sized and in a holding pattern over a defended target.
You can't loiter over defended airspace in any sort of aircraft (bar satellites) and if you're going to take out the defences with conventional munitions you might as well take out the missile silos while you're at it. The argument that ballistic missile silos are harder targets than SAM sites is bogus because once the opposition realise that's your logic they'll put the SAMs in silos too.
This is before you even consider the fragility and vast power (i.e. weight) requirements of a laser weapon. The last attempt needed a 747 ...a platform that would be a sitting for any SAM in the last hslf century and any interceptor since the Gloster Meteor.
Assuming we discount the Russians and Chinese (because they actually have fighter aircraft and drones would just be target practice) you're left with Iran and the Norks. Both probably have S-300 class SAMs and there is no way the cost equation between a SAM and a giant laser toting UAV favours the latter.
Drones are having an easy ride right now because they're deployed against adversaries with pre-WW1 aerial combat capability.
...I remember quite a bit of angst about a new ice age and the chaos it would cause for humanity. Turns out in an unprecedented feat of planetary engineering a coalition of coal miners and oil companies may have averted the disaster! Not a word of thanks either ...such ingratitude!! ;)
@Hans1
>>Human rights (as defined by the European Commission)
They aren't. The ECHR is a Council of Europe convention; the European Commission is the Executive branch of the European Union. The two are not related. They share some of the same members, but then so does the European Song Contest.
>>European laws were drafted with not only Britons...
It has nothing to do with European Law because it has nothing to do with the EU. It is an intergovernmental treaty which includes countries like: Russia, Georgia, Turkey, Armenia, Albania, Azerbaijan etc. You've probably noticed those aren't EU members.
You're possibly confusing the European Court of Justice with the European Court of Human Rights ...or else it was just an excuse for some anti-Anglo ranting ;)
@Adam
A Minister wouldn't dare just pull such an interpretation out of his ass. He'll have documented advice from Government law officers and probably "independent" QCs as well. The fact that this Court disagreed with the rationale of the lower courts (albeit reaching the same verdict) clearly illustrates that there are solid alternative arguments and "reasonable doubt" is all it takes.
Further: the Court only ruled that there is an incompatibility. Both the HRA and the Terrorism Act remain valid law therefore so long as the Minister is following either one of them he's in the clear.
Finally: Crown Immunity. You can't prosecute a Minister of the Crown for actions taken as part of his official duties. Bliar would be in serious trouble otherwise.
>>the company is not short of resources to deal with any problems >>discovered
This sort of simplistic thinking always winds me up. Nine hours of heart surgery doesn't become 30 minutes because you throw 18 surgeons at the problem. Cash, staff numbers and a capital equipment budget won't materially change the timeframe needed to fix a live site ... in fact "resources" can be a liability because PHBs don't get the first point. Adding people to a problem frequently just adds to the problem.
Preventing these problems arising is a different matter. A fully resourced security and QA section is invaluable. I'm only discussing dealing with issues after discovery (i.e. after QA and "DevOps" already failed).
...are people still using Oracle? Serious question. I mostly do embedded stuff and the higher level analytics don't remotely tax PostgreSQL. I implemented Oracle as an alternate ORM target, but that was just to add a bullet on a data sheet. We've never actually used it.
MSSQL I can see has attractions if you're developing in Visual Studio, but Oracle? What is its USP these days? Does it trounce Postgres with truly colossal datasets? Or is it mostly market inertia?
@Big John
I didn't "imagine" being less than 500m from terrorist bombs here in the UK on three different occasions. It happened. Once it was Jihadis (7/7) the other two were IRA.
I'm still a damned sight more concerned with incompetent idiots in motor vehicles and if I were still in the USA I'd be far more worried about mouth breathing retards with a quasi-sexual firearms fetish. Terrorists. Not so much.
We're supposed to be IT professionals. We apply maths and statistics to decision trees, not FUD and rhetoric (that's Consultant territory).
@Code
Even if Asus & Acer were 100% x86 (they aren't) their combined sales would still be a rounding error in phones. Apple going x86 would only give 14%.
Anecdotally, in tablets, I have three Atom units (Lenovo and Linx) running Win8.1 but I'm the only person I know who does.
In embedded, where I know the hardware better, iMX6Q boards eat Galileo's lunch (FreeScale ARM7 vs. Intel's offering) in both electrical and computing power and Intel has nothing at all in the STM32 or CC2650 space (to run IoT from a coin cell).
I even use ARM for half my desktop stuff because switching back to the main x86 machine and cross compiling is slower overall than compiling in situ (SECO A62).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
