Forgot your password?
typodupeerror

Comment: Re:Not just Google (Score 1) 543

by daniel_newby (#32659250) Attached to: At Google, You're Old and Gray At 40

However, if you have no honors and no record of having worked full time - then you slacked off. Statistically, if you slacked off and did just well enough to get the degree in college, ...

A EE degree averaging 17 credit hours/semester is slacking off? Who knew?

No, we bring in H1bs because there isn't enough native talent.

At a price you are willing to pay. Start coughing up $500k/year and you'll find a lot of native talent magically appears -- and the finance people will hate you. H1bs are all about the benjamins.

Comment: Re:C++? (Score 1) 546

by daniel_newby (#32418512) Attached to: GCC Moving To Use C++ Instead of C

I never understood the animosity against C++.

Many of us were around for the long trial by fire that was needed to figure out which features of C++ to use and which to avoid. "Modern C++" gets good results from a careful set of language features and libraries, but many folks still get flashbacks from the language itself.

Comment: Re:Take some time and think (Score 1) 537

by daniel_newby (#32037536) Attached to: Juror Explains Guilty Vote In Terry Childs Case

I don't see this as one of our own being unjustly persecuted.

I do. His moral crime was failing to turn over the password, but that is legal! If someone simply refuses their assigned employment duties, the only remedy is to fire them and write better procedures so it doesn't happen again. Incompetence at your job is not a crime. What he was convicted of was the act of setting a password that all other authorized users did not know in advance, which he did do, but that is a terrible law that should be changed.

It sounds like he knowingly lied in the course and scope of his employment, which can probably be cooked into a crime (fraud? sabotage?) by a decent lawyer. When somebody thinks they have acted immorally, they often run an illegal cover-up that can be prosecuted. Let the guilty incriminate themselves. (The lesson to the "innocent" being to never lie, profess confusion, and never talk to cops without a lawyer present.)

Comment: Re:Missing something (Score 3, Insightful) 441

by daniel_newby (#31653466) Attached to: Best Way To Land Entry-Level Job?

He's graduating now, so that means at the end of his second year he couldn't figure out why a string named string was a problem, ...

If you reject based on this, you will have carefully selected, at great expense to your company, a workforce consisting entirely of people who hide all their sophomore homework assignments.

A recruiting manager who inflicted this on me would not get a favorable performance review. In fact, I would consider it career limiting. You might as well measure how well they glued macaroni to construction paper at age 8.

I agree with others who state that they only hire the best people they can find.

In which case you will select people based on reliable, major positive measures of skills (loops, pointers, recursion), and ignore unreliable, minor negative measures of problems.

The challenge is to find the 0.5% of applicants who can solve the FizzBuzz problem at all. That means 200 applicants to consider per position on average (and a pool of 600 resumes if you need a guaranteed high-quality hire). If you weed out the half of applicants that don't have a squeaky clean Google image, you'll have weeded out about half of the most skilled, which means an extra 200 applicants (1200 resumes total for guaranteed hire). Either you pay horrifically inflated costs, or you are forced to compromise on quality -- probably by self-delusion.

Comment: Re:Due to RBI regulations (Score 1) 509

by daniel_newby (#31058358) Attached to: Paypal Reverses Payments Made To Indians

That seems to be the real complaint, and the nutters can't understand the difference between $1 that's promised to 9 people and $10 printed by a bank for every $1 deposited.

This is your misconception. Money is a social construct that assigns numbers to mental states. Humans are optimistic and trusting, so they consider a promise to be given an apple tomorrow to be nearly as good as stumbling across an apple today. On their mental balance sheet, a promise from a trustworthy partner is the same as the thing promised. When one person finds a windfall and puts it into this network of social trust, it can therefore create secondary trust and productive activity at a multiple of its immediate value.**

Money is just a way of assigning numbers to the activity in the social trust network, whether we account for it in oxen-equivalents or dollars. The money supply is just the sum of everyone's mental stuff-I-more-or-less-own numbers. Increasing the average trust grows more money supply from nothing. Lowering the bank reserve fraction is just a way to way to increase the trust level, which as far as human psychology is concerned is about the same as the gods sending manna from heaven. In the true context of the system, printing/destroying money really is an accurate metaphor for tinkering with the reserve ratio.

Note that it doesn't have to be this way, it simply is, as a matter of convention. The accounting conventions reflect the underlying psychology.

**Which answers your objection of why anybody would broker promises at a loss. They don't. A successful society is not a zero-sum game. They can charge interest and the borrowers can end up with more than they started.

P.S. One's ideas are often incomplete. Using those ideas to insult people makes one look childish when, as so often happens, the ideas turn out to need improving.

Comment: Re:And this is news why? (Score 1) 285

by daniel_newby (#30732776) Attached to: CES Vendors Kicked Out of Hotels For Showcasing Wares in Room
The RICO Act is civil law and provides for harsh punitive damages (up to three times actual losses). It also has a rather expansive definition of "conspirator", so CES exhibitors who requested or knew of any racketeering could be attached as defendants. On the other hand, RICO is a Federal law, so prosecution is expensive and time consuming.

Comment: Re:Why (Score 1) 926

by daniel_newby (#30665918) Attached to: Slovak Police Planted Explosives On Air Travelers

Why didn't they use blocks of plastic COATED with a thin layer of real explosives?

Coatings are hard to do reliably. What they should have used was something like a NESTT simulant, a granular material mixed with explosives at a low enough concentration that a detonation cannot be sustained. That's what we used at an explosive detection company where I used to work.

Comment: Re:why would an adult talk to another child? (Score 1) 596

by daniel_newby (#30372606) Attached to: Canada Supreme Court Broadens Internet "Luring" Offense

How long before the US enacts the "me too" version of this law, potentially exposing us to criminal/civil liability just for letting this kid into our lives?

Never. "Congress shall make no law ... abridging the freedom of speech, or of the press". This was tested when the Federal courts struck down the U.S. Child Online Protection Act (COPA).

Comment: Re:Overkill... (Score 1) 524

by daniel_newby (#27670579) Attached to: Should Network Cables Be Replaced?

But grounding the shield at both ends creates ground loops. You might not notice them right away, but you sure will the first time the MOVs in the surge suppressors at one end or the other shunt a spike to ground, and some of that current decides that its preferred path to ground is over your STP Cat5.

Eventually, after you blow up enough switch ports, you'll stop doing it that way.

Sort of. Shielding must be contiguous to be effective. A shield with a gap is often as bad as no shield at all, and can be worse by focusing the interference on the single most sensitive point in the system, the panel. (Says the electrical design engineer who has had to design equipment that does this right.)

The issue is that much networking equipment is designed for unshielded cable only. You know, like the Ethernet standard specifies. Such equipment often lacks the provisions needed to handle shields properly.

It's generally pretty bad form to ground both ends of any shielded wire that traverses any real length.

Nope. Either the shield must be grounded at both ends (and maybe have a thick drain wire in parallel), transformers must be used to isolate the endpoints, or unshielded cable must be substituted. You know, like cable TV systems, which go to a lot of trouble to do end-to-end shielding properly.

The lesson is that Ethernet's built-in transformer isolation is a gift from the Gods. Don't spurn the Gods with your puny "shielding" unless you have made the necessary sacrifices. Which may very well involve burnt offerings.

Comment: Re:Either trivial or bullshit (Score 1) 305

by daniel_newby (#27467123) Attached to: Coders, Your Days Are Numbered

Write simple code. Build complex behaviour from simple start points. Unless you're doing some serious mathematics (in which case it'll have an elegance of its own to those that comprehend the maths) then there's no real need for complicated code that is hard to understand.

You seem to not understand that some people can (1) bang out a big pile of work that is complete but scattered, then (2) see through it to the relationships underneath, then (3) perform a symbolic reduction to reach a state of simplicity, much like simplifying a complicated algebraic equation.

As a matter of course. On everything that falls into their hands. Often doing step #1 mentally.

They will find it painful and frustrating if the organization forces them to plod along, laboriously making every step of the work explicit to the back seat driver. The work will also tend to be of lower quality because their mind is constantly interrupted. Therefore they will tend to flee to another organization.

Which means that they will not be there to promote to system architects and other leaders. The organization will have to make do with people who have a proven lower ability for seeing the big picture and doing synthetic reasoning. Or the org will have to recruit leaders from outsiders who (1) don't know the business, and (2) who want to divorce themselves from hands-on work (because the chain gang approach would drive them crazy).

Now there may indeed be people who are more productive with pair programming. Or with whatever other scheme tickles your fancy. Their existence does not imply that the technique is generally, or even frequently, productive. There is so much variation in human cognition, not to mention personality, that any business operation scheme claiming generality is automatically suspect. There Are No Silver Bullets.

You have a tendency to feel you are superior to most computers.

Working...