Please create an account to participate in the Slashdot moderation system


Forgot your password?

Comment And half a dozen people .... (Score 1) 465

We usually still end up needing to sift through 40-50 resumes and end up interviewing 4-6 people (over the phone or on-site) ...

Read this to my wife, who responded: "Thank you for wasting my time."

You may have hundreds or thousands of people who wasted a minute reading the ad. You have 40-50 (possibly out-of-work) people who wasted maybe 10 minutes and a buck or two sending you a resume, or an hour composing a cover letter. You have a half-dozen who spent hours, and maybe auto expenses, coming to and attending your interview, and experienced hours of stress responding to your people's questions. You wasted maybe a day of their time in the midst of a job hunt. All when you had a candidate you were happy with.

Did you EVER hire anybody except the one you'd started with? Or did you just get a warm fuzzy feeling from having found some possible alternates in case he drives into an overpass pillar during some night's commute?

But thank you for your honesty. Job hunters really need to understand the viewpoint from the other side of the desk. It will help them accept the repeated rejections that result from such "fair hiring" practices.

Comment Re:To hire specific people (Score 2) 465

Slight problems with your theory:

It wasn't the poster's theory. It was what the professor told the class.

It's also been well documented. Recruiting firms have seminars for their customers on how to do it, and videos have been smuggled out by irate HR people and posted to the internet, sent to congresscritters, ...

One, it happens in countries that don't have H1-Bs.

Even countries that don't have H1Bs - or companies that operate there - often have requirements that job openings must be posted openly and the most qualified candidate(s) be offered the job. The same procedure works when an administrator has a particular candidate (or relative) he wants to hire, and HR procedures require the posting be open, for whatever reason.

Two, unless these H1-Bs are coming from Gallifrey there's no way they have 279 years experience in Java 147 on RHEL DCXLIV.

That sounds like another hiring pathology, which has caused much hilarity in the past, the result of clueless hiring managers and HR departments.

Back when Unix was first becoming broadly used in private companies, want ads were filled with job offers for low-level unix sysadmin positions - all requiring experience with Unix and its tools that could only be met - if at all - by Ken Thompson, Dennis Ritchie, M. D. McIlroy, and J. F. Ossanna. B-)

On the other hand, H1-B candidates often have their qualifications "boosted" on the resumes presented by the recruiting agencies representing them. (Often the candidates themselves have no idea this is being done - and are horrified if they discover it.) They may be enhanced beyond real-word possibility, either because those doing the enhancement are as clueless as in the Unix ads case, or because it's a way to insure that no read candidate can meet them (without also lying on the resume).

Comment When you INSTALL it? (Score 1) 346

You have the right idea, big time. But IMHO you didn't go quite far enough.

Ubuntu should follow the openSUSE way: when you install it, it asks you which desktop you want.

The screen server and all its friends get restarted when your session is launched at log in. There's no reason the system can't run the desktop of each user's current preference, and give him the opportunity to change that preference - for the session or persistently - every time he logs in.

IMHO It should be a drop-down menu opton on the login, with a well-documented and easy way both for the user to edit his settings and for an admin (or even the user, in his own account) to add more desktop software.

The install - actually, the new-account create - would just ask about the acconts' initial preference.

Heck: It REALLY should ALSO be an item on the background pulldown menu, or equivalent, of every windowng system to switch mid-session. But there are a lot more worms in that can, starting with convincing the developers of all the windowing systems to provide a hook for users to drop their baby and adopt children of their rivals.)

Comment That doesn't follow at all. (Score 1) 255

Note from the article "Removed reference to Google stating the app was not in violation of TOS â" this was a mischaracterization of Googleâ(TM)s statement."

Well presumably if they removed that statement because it was a mischaracterization of what Google said, then Google said it was in violation of the ToS.

That doesn't follow at all. You're engaging in "The Falacy of the Excluded Middle."

There's an enormous difference between Google saying that installing the app isn't a TOS violation (thus committing themselves to supporting the phone) and claiming it is.

For instance, they could be reserving the right to make the claim later, on a case-by-case basis depending on what was installed and/or how it was done. Or they could just be avoiding potential legal entanglements without extensive (and expensive) study and/or risk of litigation if their lawyers guessed wrong about how juries, judges, and regulatory agencies around the world might act.

Comment Also, they're posting "return to factory" images. (Score 2) 255

Actually, that rule changed a few weeks ago: ...

From the article:
  - They are also retroactively reinstating the warranties of people who already asked for an unlock code and had their warranty voided as a result.
  - They are posting "return to factory images". (Nice pun, that. They let you flash your phone back to the factory image, which you'll want to do before returning it to the factory for service.)

I guess losing a touch more than a third of a billion dollars ($342 million) in one year CAN sometimes get executives to look at customer complaints and try to address them. B-)

Comment Re:I hope the laptop shell's monitor is LED. (Score 1) 98

I've had a number of laptops where the CPU was doing just fine but the monitor died due to fluorescent lamp end-of-life.

So, um. Why didn't you just, uh, you know, fix them?

Because they belonged to my, um, employer, who by then had handed me a later model for my primary machine.

I've only had one of my own die so far, and it wasn't the screen backlight that failed.

I intend to pop that one open and see if it is something simple, like a loose connection. But it was one of two identical ones I bought used for like $50 or less, and had already run for me for two years. So I swapped the disk into the other one and I'm not in a big hurry.

One nice thing about Linux: It runs just fine on older hardware, as long as you don't need the extra crunch of a new model for some application. (Better, even, since the driver guys have had time to figure out the peripherals.)

Comment Maybe the charge jcreeping deeper into the crypt? (Score 1) 424

TFA says exactly where the power goes: the car's electronics don't sleep when the car is off.

So the next question is: Is the fix only partial, leaving a lot of stuff still awake? Or is there something else? It looks to me like his measurement of the leakage and vampire load has a methodology problem.

I note that the test performed by the author would find another "drain": Charge diffusion in the batteries.

When a battery is first charged the surfaces of the electrodes become fully charged, but the bulk of the plates are still not quite full. As the charge carriers diffuse deeper into the material, the cell still has the same charge (except for leakage, which is a separate issue). But the surfaces become less charged as the core comes up to match them. (This is why the last stage of charging is slow - to get MOST of this done before charging cuts off.)

This is a BIG DEAL on lead-acid batteries. I'm not sure how much of this effect is present in lithiums, which is what Tesla uses, but I bet it's lower but not zero. (I note that my cellphone does both bulk and "topping" charging of its lithium batteries - ramping up quickly to the 80% level and then trickling its way up to "100%".)

So if he's running his car in the day, charging it until it hits 80% in the evening, unplugging it overnight, and plugging it back in, I bet that, even in the absence of ANY leakage, the charge controller will see the lower voltage as the cells from surface charge diffusion as the battery not being quite at the 80% setpoint. So it will "top them off" to the setpoint level again. If the computation for the display doesn't take this into account it will look like a lost of charge.

He should:
  - Leave it on the charger for several days, not driving it (while the whole bulk of the cell material reaches equilibrium), then
  - Unplug it overnight, and
  - Plug it in in the morning.
Then he'll be measuring just the leakage - from the cells and any vampire loads.

Comment Re:Vampire? Huh?! (Score 1) 424

Again, I love it when non-engineers talk out of their ass.

It's fun when engineers do it too. B-)

They all need current clamps which are just one side of a transformer coil to step that current down into a usable range.

Transformer-type current clamps won't measure DC.

You need either a (powered) hall-effect sensor type clamp or one where the magnetic field directly moves a mechanical needle. Those are a lot harder to find.

Comment Re:Now if only it could TRANSMIT. B-) (Score 1) 75

The dongle receivers are typically I/Q receivers.

Yes, I understand that. I guess I phrased it ambiguously.

What I meant is "convert data from the USB to I/Q OUTPUT, i.e. do the TRANSMIT side of a transceiver, too, not convert the receive side to I/Q from something else.

Then we need a local oscillator and mixer to boost it back UP to the desired frequency band (which might be done with the companion block downconverter chip if the appropriate signals are accessible or if it is actually also a transciever chip). Add a "power" amplifier (for suitably small values of "power"), a diplexer (if you really need to use a single antenna for both directions) and you're done.

Comment (Supposedly) Broken for only some buses (Score 1) 75

One of the cited article talks about the system having two cases:
  - The buses with the tracking hardware are displayed based on the tracking.
  - the buses without the tracking hardware are displayed based on the schedule.

Now maybe the line you're on has buses without tracking. (Or maybe the tracking system doesn't work and it's all a crock.) But the anecdote that your particular line is just showing an automated schedule doesn't show that all others are doing the same.

Comment Now if only it could TRANSMIT. B-) (Score 1) 75

Its the RTL-SDR project. A Linux developer discovered that a digital TV receiver chip made by Realtek (used in $15 dongles) had the ability to receive the raw sampled RF data. The bandwidth is nearly 3Mhz so that means you can view a HUGE chunk of the RF spectrum at once and decode the signals via software.

Now if only it could transmit.

Or if it could also convert digital signals into I/Q and we could feed that into the Rx mixer of the block downconverter, run backward. Then two $11 - $15 dongles, one of them hacked slightly and with a small power amplifier added, would be a two-way software defined radio for very cheap.

Comment Re:The public Internet is NOT a government project (Score 1) 1030

... the Telco and Cable companies bringing you this commercial Internet, this bastion of free enterprise you are talking about, is about as heavily subsidized as an industry can get?

The public Internet wasn't developed by the Telco and Cable companies. It was developed by garage shops that started as small ISPs or equipment companies. Telcos fought it, while cable companies watched from the sidelines.

The "Mom and POPs" built the public net at first. Some of them were literally in people's bedrooms. (At least one I know used rack-mounted equipment but built its own 19" rack panels out of two-by-fours.)

Many of the equipment companies, too, started in garages. Cisco, for instance.

Once things were up and running the Telcos decided they were missing out on a good thing and tried to enter the marketplace. But at first they did it by trying to sell their own overpriced ATM-based services. Others continued to compete rings around them - though often leasing their copper wires for the last mile and various digital carriers for long-haul - or leasing those from the more competitive long-distance carriers.

DSL and cable modems were both developed, not by the Telco and Cable companies, but by private equipment manufacturers (including one spun out of Bell by the antitrust decision), trying to sell boxes at a profit. Some cable companies used this new stuff to leverage their installed base and get into the ISP game. Other ISPs, such as Covad, used DSL to push fat bandwidth through legacy Telco copper leased at regulated wholesale rates.

What finally happened is the FCC relaxed the access requirements on the legacy telcos - deciding two competitors was "competition" (when it takes three to four, minimum, to destabilize defacto price fixing and drive the price down towards cost). The tellcos immediately started squeezing their competition in the ISP market (for instance, Covad), eventually doing in or crippling pretty much everybody but the Cable companies (who also had legacy subsidized copper in place) and some rural little guys.

Telcos and cable companies, with their government subsidized infrastructure and rights-of-way, are the bulk of the ISPs NOW. But they AREN'T what "build the Internet". They're the big fish that ATE it.

Comment Stop thinking of the GOP as a monolith. (Score 1) 1030

... but it's abundantly clear that the GOP is not seriously opposed to government intervention in energy markets.

You make a big mistake when you think of the GOP as a uniform monolith. It's composed of about five major factions, and much of its recent behavior comes from the Neocons' iron grip on the party machinery (and the others' attempts to dislodge it).

Of particular interest is the Liberty Movement faction - with a primarily libertarian and/or constitutionalist ideology, but far better tactical savvy than the Libertarian Party's people. They're gaining power rapidly. On this issue they're bovernment-hands-off: Don't subsidize: It actually retards development and deployment. Don't interfere for entrenched interests: Ditto. They also want the people energy-independent, and thus better able to resist external control (both foreign and institutional).

Comment The public Internet is NOT a government project. (Score 3, Insightful) 1030

It's ironic that you're posting this on the Internet which was invented by government funding.

This isn't about invention of the fundamental underpinnings. Plowsharing is a grand tradition.

This is about development and deployment in the public sector. Bringing the Internet to the masses wasn't government funded. It occurred when the government got out of the way and let commercial interests play with the new toy. (THAT's what Gore rightly claims substantial credit for.) Scaling it up and the burst of innovation in using it was done with private money in a largely free marketplace, not government subsidies.

In fact, government subsidies HURT this development-deployment phase. The picked winners have no incentive to innovate - they're paid to work on what is already there. The non-picked have no incentive to innovate, or even enter the market - they start at a big competitive disadvantage, and if the did succeed they can expect the government's cronies to get still more subsidies (unless, like Solyndra, they collapse so fast the pumping is ineffictive).

Solyndra failed because they spent the government money like water, ending up with a product that was slightly MORE expensive than the non-subsidized competition - when moving potential customers to a new variant of an existing technology requires a substantial improvement in price-performance - and about a factor of ten to obsolete the previous mainstream approach.

What's driving the current burst of innovation and deployment is the loss of government subsidies around the world. Now the playing field is closer to level. More companies are playing with private investment. The products must compete with existing grid systems, so innovation is occurring and price/performance is improving to where they ARE competitive in progressively more situations.

Indeed, panels are now available at less than a dollar per watt, which is about the point where solar starts beating grid costs in most places where there's enough sun, rather than just remote places or small loads where it's cheaper than running miles of new lines.

Slashdot Top Deals

The clearest way into the Universe is through a forest wilderness. -- John Muir