Forgot your password?

Comment: Have a look at Teradata (Score 1) 147

by golodh (#48339571) Attached to: Ask Slashdot: Choosing a Data Warehouse Server System?
I've recently had good experiences with running SQL queries on fairly large (# records: 200 mln. plus) databases on a Teradata machine in a corporate environment. I wasn't involved in any sysadmin work, just the statistical modeling / analysis side of things.

The company I consulted for uses SAS (on the mainframe, AIX boxes, and PC's) for almost all of its dataprocessing needs, including ETL work. Now they're looking at "Big Data" and discovered they need parallel processing to make it cost-effective (outperforms the mainframe, no per CPU-second charges, ability to let analysts work on AIX boxes or PC's etc.).

I was able to show significant cost and performance savings in SQL queries over the mainframe (and AIX boxes). Interestingly substantial (50%-100%) speedups were also possible by accessing the Teradata machine in its native SQL (bypassing the SAS "in-database" Teradata support).

The interesting thing about Teradata is that they offer genuine parallel processing (like Hadoop), but offer it as an end-user ready SQL interface to a database engine (you still need sysadmins though). Contrast this to Hadoop where the Hadoop layer is basically the start of the road and you usually have to worry about hardware issues and software architecture issues (such as which database engine to choose) as well. Sometimes you have to take the custom-made route (e.g. Wall-street firms doing automated trading) but sometimes it's an outright liability in a DIY-hostile environment (e.g. in large corporations).

The teradata machine I worked with supports SQL, SAS, and R (which competes with SAS of course, and usually out-competes it when it comes to advanced statistics if you know what you're doing but we had to use SAS exclusively, by order) and could easily handle terabytes of data.

So my suggestion is to take a look at it.

It's not Open Source (although it does support R), and it's less fun for tinkerers, and it's harder to custom-parallise your own algorithms on (I hear, I never tried). On the other hand it does provide a ready-to-run parallelised SQL database and lots of storage. It's not cheap though, but in a corporate environment that's usually not the first consideration.

Comment: Re:Nothing. (Score 2) 209

by golodh (#48315691) Attached to: What People Want From Smart Homes

So do I, but the mainstream seems to be moving towards something very different.

As in: the majority of consumers seems to want maximum "comfort" (read: "ease of use and no hassle", a.k.a. "I'm lazy and dumb so I need smart appliances"), and that's what industry will provide (on pain of being marginalised and ultimately disappearing).

And guess what? Ease of use and "no hassle" means offloading lots of detailed control decisions to the manufacturer. And that means that said manufacturer has got to distinguish themselves by offering comfort and taking away decisions and cares from home-owners.

It is understood that home-owners are willing to pay for that and that manufacturers incur no penalties by offering dumb gear and putting the "intelligence" on their servers. Those decisions (blinds closed or open, heating higher or lower, anticipating the home-owner's homecoming, level of lighting, when to switch on the air conditioning, burglar alarms, suppressing false alarms cause e.g. by pets etc. etc.), still have to be taken of course. Just not by the home-owner.

Taken together this means a big fat premium on supplying dumb, (but sensor-rich) proprietary hardware, collecting as much data as possible on the habits and preferences of the home-owner, his/her family, children, pets, neighbours etc.etc., storing and analysing all that on the company's servers, and selling the resulting control information to the home-owner as a service. Look for upcoming legislation that not only allows but also compels "domestic service" companies to "share" their information with everyone from law-enforcement, insurance companies (think fire insurance, burglary insurance, health insurance (!)), medical care providing companies (think monitoring of elderly people), market research companies, advertising companies and any other interested party you can think of.

I'm pessimistic about being able to opt out, let alone to stop this kind of thing. For one thing, mass-production will drive down the price of the "mainstream" systems (whatever form they will take), thus marginalising any non-mainstream hardware. Of course manufacturers have zero interest in supplying hardware that will work without their (or another company's) service package so stand-alone or "user-controlled" hardware will come at a premium. In addition you may find that your insurance premiums are higher than without "smart home" automation.

All in all, the stable market situation will probably be a load black-box hardware that needs daily updates and tuning by proprietary off-site control software that eats your privacy for breakfast (on an ongoing daily basis).

Comment: Oh boy ... (Score 5, Insightful) 424

by golodh (#48233881) Attached to: Law Lets IRS Seize Accounts On Suspicion, No Crime Required
First we pass a law that is an open invitation to unintended use (like this seizure law) because it conveniently neglects to mention where it is to be applied and where it isn't.

Then we come over all indignant when that law (which is "on the books") is used outside its originally intended area of application.

Am I the only one who thinks that Congress is to blame here (for passing sloppy legislation), not the IRS or The Government?

Might it not be a good idea to work harder to phrase legislation in such a way that it's difficult to abuse? Or would that cramp the style of "tough-on-crime" politicians?

Comment: your thoughts ... (Score 4, Insightful) 372

by golodh (#48218807) Attached to: NY Doctor Recently Back From West Africa Tests Positive For Ebola

IMHO your "opinion" is very very humble indeed and belongs in the category of "uneducated careless speculation with a sensationalist bent".

It may have escaped your notice, but doctors who help out in West-African hospitals come into close contact with a constant stream of very ill people who are in the stadium where they really are contagious, every day for months at a stretch.

Their protective clothing prevents transmission in the vast majority (say 99,9%) of cases (something you can tell by the fact that we still have doctors left treating Ebola patients). The real danger comes when you take off your protective suit. That has to be done carefully so as not to touch the splatters of blood, muckus, tears, sweat etcetera that very ill patients secrete and if possible it has to be decontaminated first.

Now I'm sure your "humble" and uneducated opinion never has been schooled in elementary probability so you wouldn't understand things like P(contagion_after_100_days) = 1 - [P(no_contagion_after_1_day)]^100, but try it this way.

Playing the lottery every day makes it unlikely that you won't win a single prize.

And so it is with medical personnel who treat Ebola patient for months. They run a risk.

So it's no conspiracy (I can feel your incredulity and disappointment) and no case of "fsking idiots" (a term which I'd like to reserve for you personally).

It's easy to shout your (thoroughly humble) head off about stuff you don't understand, but it's not helping anybody and it stands in the way of a rational attitude towards Ebola.

P.S. there is absolutely nothing "insightful" about your post. On the other hand it's revealing. Revealing of a mindset that couples a penchant for conspiracy theories with a complete lack of understanding of risk and a disdain for plain ordinary everyday scientific commonsense that seems to have whizzed over your (so very humble) head.

Comment: What surprises me here ... (Score 1) 79

by golodh (#48204865) Attached to: DHS Investigates 24 Potentially Lethal IoT Medical Devices
is that the Government is actually doing something sensible.

Like airing the vulnerability, launching an investigation, and giving off a signal that the *manufacturers* should pay attention to security and at least make a reasonable effort to make their kit tamper-resistant

It would be in total accordance with a certain political outlook to suppress the news, pose as being "tough on crime" by imposing ridiculous penalties on offences that could be construed as breaking into medical equipment, and criminalising research into and publications of weaknesses.

Perhaps I'm being optimistic ... perhaps this will still happen. That "certain political outlook" I mentioned could be a bit behind the tech news on this issue. We can still hope though.

Comment: Re:The essence of enterprise (Score 2) 148

by golodh (#48175369) Attached to: Cisco Exec: Turnover In Engineering No Problem

You argue that "capital" and "labor" are essentially equal to the identity of an enterprise.

I really didn't: when I said "bundle" I left the relative proportions unspecified. But I agree with you in that the relative importance of people's identity varies sharply with the scarcity of people's skills and that depends on the setting.

We agree that in a environment where people do routine work, so many people share the required skill that identity of who provides this skill no longer matters. And that's where vast majority of the working population is employed.

Of course there are settings where individual skills matter to a greater degree. One can think of e.g. professional sportspeople, scientists, lawyers, entrepreneurs, politicians, artists, inventors.

But their numbers are small compared to "ordinary" workers, so that by and large I think the proposition holds. Yes, there are exceptions, but 99% of the working population is rather un-exceptional.

Comment: The essence of enterprise (Score 4, Insightful) 148

by golodh (#48174963) Attached to: Cisco Exec: Turnover In Engineering No Problem
I fear that the negative reactions here indicate (once again) that Slashdot readership consists mainly of techies. And such people often have difficulties understanding understanding how society works (even if they tend to have vocal opinions on any subject that comes along). Let me try to bring some perspective into the discussion.

Lest somebody misunderstand, the very essence of an enterprise (any enterprise) is that it is a bundle of labour and capital whose essential structure and identity is independent of and more persistent than the labour it employs. The identity behind its labour component is no more important than the identity of its capital component.

It is for this reason that any contemporary HR policy is aimed at (and this is important) divorcing the work from specific individuals.

What this means is that all and any employees must (and this is essential) be plug-replaceable as a matter of policy. Those that aren't should either be unique individuals like Steve Jobs, Bill Gates, the actual owners of the company, or leaving.

That is one of the drivers (not the only one of course) behind the desire for standardisation of work procedures and documentation of ideas and knowledge.

The result of careful execution of such policy is a situation in which personnel really is replaceable. Even when it concerns 10%-30% of the employees. Which is what we are now seeing illustrated at Cisco.

So there's no need to be surprised. And no need to be disgruntled. It's simply the consequence of a certain feature of our society we collectively decided we want and actively maintain. And it has truly served us well for the past century and a half and its end-result is the envy of our neigbours.

Unfortunately the current economic tide makes the downsides (for such they are) of this state of affairs more visible: i.e. employees are just another commodity and any successful enterprise will treat them as such. . As a result, employees can get a rough deal (if they get any deal at all, i.e. if they are employed). Let's be clear about this: I don't know how to make those downsides go away without wrecking the competitiveness of enterprises. But I suspect it will involve a realignment in the balance of power between labour and capital.

One way of achieving this is through the use of force. Also known as "legislation". Fortunately we have a mechanism in place for effecting change. It's called Politics. But what actual policy should be enacted through Politics? If knew (and could prove it) I'd tell you, but I don't.

One of the problems is the constraints imposed on all of us through competition. I.e. if the policy we adopt is too disadvantagous for enterprises, they will simply take their capital, set up shop elsewhere, and drive the disadvantaged enterprises off the market.

So it's up for debate really, and this isn't a new debate. It's a debate about a basic balance in our society that needs to be realigned from time to time.

Comment: Things to know before you ask ... (Score 5, Informative) 187

by golodh (#48132671) Attached to: Interviews: Ask Florian Mueller About Software Patents and Copyrights
There are a number of relevant things to know about Florian Mueller before you start asking him questions.

Things that people with short memories will have forgotten by now since they happened all of three years ago. Detailed summaries of his doing can still be found on Groklaw though.

You see, mr. Mueller is not just *any* publicist. He's a publicist who is, basically, for hire by large companies to provide a congenial account of their doings and their position. In short: he is a lobbyist. His (former) clients seem to include SCO (the company who tried to claim crippling copyrights on Linux and engaged in an intense campaign of legal blackmail aimed at companies using Linux) and one of his current clients seems to be Oracle (the company that reied to shut down Android by claiming copyright on Java library API's).

As summarised by the following posts:

My only question to him would be: who is on your current client list?

Comment: Re:Only 100 you say? (Score 1) 104

by golodh (#48121055) Attached to: Only 100 Cybercrime Brains Worldwide, Says Europol Boss
Err, sorry, but how would *you* know anything about that?

Did you do any kind of analysis tracing existing malware to point-sources? Or did you see any data on that and did you identify and count those point-sources?

No? Then what is your opinion worth?

You seem to be confusing *operators* (i.e. the ones that actually push the button and run botnets, burglarise computers, and/or spread malware) with *researchers*, *designers* and *programmers* who never hack, but who write (and sell) the tools the operators use.

If you had actually read the article, you would have noticed that it's talking about those tool-makers, not operators. I could very well believe that those toolmakers number only about 100 world-wide.

Comment: Re:Makes Sense (Score 1) 225

by golodh (#48053927) Attached to: Google Threatened With $100M Lawsuit Over Nude Celebrity Photos
You're probably speaking in jest, but unfortunately it's true.

If Google focuses on filtering content rather than providing it then it can certainly comply quicker and more completely with all such take-down orders.

The question of whether Google can " control and censor every last thing" is totally irrelevant, as the suit is addressed to Google on basis of what you can find using Google ... as opposed to what you can find "on the Internet".

It's simply a matter of where you put your priorities. Which in term depends on how reasonable you think the demands to censor search results are.

As noted in earlier posts, techies don't appreciate the extent to which society can suppress behaviour it doesn't want.

Lawsuits like this may well lead to a shift in Google's priorities and a substantial increase in the extent to which it filters search results.

Comment: Shooting the messenger ... (Score 1) 195

by golodh (#48025675) Attached to: CEO of Spyware Maker Arrested For Enabling Stalkers
It's quite OK to mass-produce cellphones that can be tapped and controlled in this way.

But apparently it's not OK to sell software to allow people to use their perfectly ordinary cellphone to pick up other conversations from its vicinity.

How about securing the transmissions of cellphones instead of prosecuting someone for doing the obvious?

Genius is ten percent inspiration and fifty percent capital gains.