Forgot your password?
typodupeerror

Comment: Re:A DC resident replies (Score 1) 167

by kriston (#47957247) Attached to: Washington DC To Return To Automatic Metro Trains

Washington DC has a very strong and proud tradition of African-American railway workers. From the old railways to the RF&P to the Metro and also operating the premier Acela Express maintenance facility in Ivy City, Washington DC has a very successful and proud heritage of railway workers.

Comment: Re:What's wrong with American drivers? (Score 1) 167

by kriston (#47957221) Attached to: Washington DC To Return To Automatic Metro Trains

No. The old system was perfectly reliable except in certain slippery weather conditions. WMATA was in the process of replacing the old system with a new system from a different vendor. In the process they mismatched sensors with lineside equipment. The result was death and injury.

Please read the depositions. It reveals much more than most news reports say, with the possible exception of certain Washington Post articles.

Comment: Re:What's wrong with American drivers? (Score 1) 167

by kriston (#47957211) Attached to: Washington DC To Return To Automatic Metro Trains

Washington DC Metro and San Francisco BART are very similar systems and were designed from the outset for fully automatic operation. The platform position errors are learned by the system and corrected over time, and this calculation also measures the weight of each car.

The real challenge is that under crush loads the system tries to get a good idea how much braking force to apply with the extra weight of the passengers, but often gets the calculation wrong because even though the system "knows" how much each car weighs, people sway back and forth as the vehicles slow down and we undershoot the platform. So, what? The train pulls ahead (automatically). The operator just opens the doors.

I've never experienced a train overshoot a platform but I've had many trains undershoot. It's not a big problem, more of a curiosity.

Comment: The automatic system was fine, not the upgrade (Score 3, Informative) 167

by kriston (#47957183) Attached to: Washington DC To Return To Automatic Metro Trains

The automatic system was fine for nearly 40 years. The upgrade process killed nine people and injured 80 and caused a safe system that ran reliably for nearly 40 years to run in "manual" mode for five years because of a maintenance error.

Several depositions from railroad workers who were tasked with upgrading Metro's nearly 40-year-old system mentioned a real problem. The signals were "bobbing." This happened because the older signal system was being replaced with a different vendor's technology in two phases with catastrophic results.

In phase one, the lineside signal cabinet equipment was replaced but the original track sensors were left. In other cases, it was reversed: the track sensors were replaced but the lineside signal cabinet equipment was original. In both cases the vendor was different and not totally compatible.

Naturally, as we would expect, the two different vendors' equipment was not interfacing perfectly. This caused signal "bobbing," where track occupancy would "bob" from red to green repeatedly. Trains would vanish from the system. Phantom trains would appear in the system.

This massive oversight reported in the depositions wasn't really taken seriously in the press nor by the accident investigation. From this point of view, the system failed due to incompatible equipment made to interface in order to save money and service disruptions. They didn't interface properly, and people died as a result. Nobody seems to care about what appears to be the real problem: incompatible vendors made to interface to save time and money.

But we now have faster trains with shorter headways that sometimes fail to stop at the correct spot in stations, so we have that going for us. At least the lineside cabinet equipment and track sensors are now from the same vendor, eliminating the problem that killed those people and put hundreds of thousands of others at risk for a couple of years until that deadly day in 2009.

Comment: It's mostly a USA problem (Score 3, Insightful) 200

by kriston (#47822709) Attached to: Hitachi Developing Reactor That Burns Nuclear Waste

It's mostly a United States problem that waste isn't reprocessed. This is now and has been done on an industrial scale in Europe and the U.K. for several decades. For some reason the United States, under the guise of non-proliferation, will not permit reprocessing of spent commercial nuclear reactor fuel.

The story in this article isn't news. Everyone knows how to reprocess spent fuel since before the 1960s. What would be actual "news" is the time at which the United States allows the well-proven, industrial-scale reprocessing to be applied to its own reactors.

Even Canada does it. The United States' nuclear energy policy is laughably stupid. It's a shame, really.

Comment: Thirty minutes is ridiculous. Swap out the packs. (Score 2, Interesting) 190

by kriston (#47726523) Attached to: How Does Tesla Build a Supercharger Charging Site?

Thirty minutes is ridiculous. That is not "rapid" ANYTHING.

The only real solution is to streamline the process of swapping out battery packs, or, ideally, hydrogen fuel cells.

This is where hydrogen fuel cells really make sense. They are the ultimate battery pack. They are interchangeable modules. You stop at a filling station and replace your depleted fuel cell with a full one in fewer than five minutes.

I know Tesla has a battery pack replacement service, but it really needs to be affordable and streamlined and not require expensive robotics.

NOBODY wants to wait thirty minutes for "rapid recharge." The money spent on this infrastructure should, instead, be spent on optimizing the use of hydrogen fuel cells. They are the ultimate battery and they don't wear out.

Comment: If it's not written policy, it isn't done (Score 1) 116

by kriston (#47625635) Attached to: Ask Slashdot: Datacenter HDD Wipe Policy?

The rule of thumb here is:
If the process you are expecting is not written into your agreement or documented as a matter of company policy, then the process is not done.

Likely you're not using a data center certified under HIPAA, PCI, SOX, SSAE/SAS-70, otherwise it would be documented and you'd already know.

Comment: Re:It was pretty cool in its day (Score 1) 192

by kriston (#47518261) Attached to: The Almost Forgotten Story of the Amiga 2000

Wow I remember Degrader, too.
There was a compatibility problem I had with my original Amiga 2500. It was supposed to be sold with 1 MB chip memory but for some reason it had 512 KB instead, and this caused so many problems until I discovered why.

We figured it was probably from a dealer swap or manufacturing line error, even though the Wikipedia article discusses that the A2000 could have 1 MB or 512 KB, the A2500 line was always supposed to have 1 MB of chip memory. Mine came with a real Amiga 2500HD sticker. *shrug*

Comment: Re:It was pretty cool in its day (Score 1) 192

by kriston (#47503387) Attached to: The Almost Forgotten Story of the Amiga 2000

We experienced timing-related problems in early demos when running them on a more modern Amiga.

However, the worst was running software written for PAL mode. The chief complaint about these PAL games was that the audio played at the wrong speed. Major offenders were games that play lots and lots of sound clips, like Llamatron. PAL mode always messed up the length of the sound clips so they get cut off or played at the wrong pitch.

I had an Amiga 2500 which was the one that came with a 68020+MMU accelerator. I later upgraded it with the 68030 accelerator and Univ. of Lowell A2410 workstation graphics. I also still own an Amiga 3000T with a QIC tape drive with Amiga UNIX and an Amiga 3000 with a floptical drive. Playing old demos is a pain in the neck on these new machines.

A bug in the code is worth two in the documentation.

Working...