User: tiohero

Application vs Task Centric OS

On: Dvorak: Linux too much like Windows

I guess what we disagree upon is whether the OS should be application(program) centric or task centric.

Application vs Task Centric OS. (Points #3,#5,#5) (rebuttal)
In my opinion, the OS should be task centric (based around the user's current project) but currently it is application centric. (for the most part). The individual applications are not aware of what you are working on NOW. Yes, they do sometimes store cached histories of work done in the past, however, there's no authoritarian central command to tell the individual apps that I (the user) am working on my thesis now, put all that information at my fingertips. One extension is to use multiple desktops, but that still ideally would require each desktop to be assigned a project by the OS. That really hasn't been offered in a major OS, AFAIK.

(Point #4,) You can't just open a project where you left off. There is a whole ritual of closing and opening windows to move from one project to the next and back. You should be able to click one icon: World Domination and have all your rantings and calculations/CAD drawings for your doomsday device open all together. Do a Google search on the history of the OpenDOC project to get an idea of the concept I'm talking about.

(Point #5) I'm saying that the user shouldn't see ANY application binary/system files or directories in the file explorer. They should be inaccessible under normal means. (unix does this moderately well) Separate application code from user data and make it hard to mix the two.

Windowing is inefficient. Points (#1,#2) rebuttal
The other point I was trying to make (points #1,#2) is that the whole concept of the windowing OS using resizable/overlapping windows may be flawed. If a window is hard to see by being covered up by other things, maybe it shouldn't be a window at all? Seriously, an overlapped window is worthless other than to serve as a reminder that it is there this takes up unnecessary screen area. That's why I suggested the idea of a smart frame based system. I'm arguing for a more efficient use of screen space.

The mouse should be used for cutting and pasting, not for having to deal with arranging the softwares' interface dialogs around the screen. Most computer use involves a data source(web-browser/e-book/calculator) and a data output (word processor). Most people would like to see both at the same time. That's what tiling and multiple monitors are for, but why? Wouldn't it be nice if when working between two documents, the OS should display content from both documents simultaneously without a bunch of unnecessary manipulation? I want the window manager to manage my windows for me! (Yes, I'm a lazy person!)

Re:A Perpetual Second Place.

On: Dvorak: Linux too much like Windows

I agree with Dvorak that there should be alternatives to Windows other than the command line. Windowing was a good idea in 83 with the Mac and its handful of applications, however as computing complexity progressed the windowing concept has not kept up. Bottom line is that computers today are a Pain in the A## to use. There are many small chores that need to be done 100s of times per day in order to get to your work done. My personal preference would be some kind of frame-based GUI.

Here are some of the problems with current generation windowing OS architectures.

#1 Why do I need to resize windows?
This is one chore I could do without. The user of a windowing GUI is always fighting a battle of resizing windows to try to dig through and format the information they need. Right now new, old, useful, and less useful windows are treated equally on the desktop. In my opinion the most utilized windows should be right at ones fingertips and the least utilized/oldest windows should be least accessible from the users standpoint. The OS should automatically adjust window sizes/layouts for easiest access to the most utilized information.

#2 Why is it so hard to cut/paste information?
A windows in this case is a blind, covering up the information you need. If Im using one application(browser) and then I switch to another (word processor), why cant the previous window resize and sit to the side so I can cut/paste the data easily. IBMs TOPVIEW basically used 2 window panes. This is a functional idea since the whole concept of windowing GUIs comes from the need to copy data from one App to another. Ive seen one shareware utility that actually punches a hole in the top window to see the data underneath.

#4 Wheres my data?
Working on a project? How many times do you have to open c:\; \usr\ and parse through directory trees to save and retrieve the data. Why cant you specify in the beginning, Im working on my Restaurant Menu project and all open/save dialogs automatically default to that directory? Your desktop should automatically adjust to a new project. This idea was partially implemented in Apples openDOC.

#4 Where are my apps?
Ok, youre done for the day. Close/save all 20 windows and power down. When you return you need to open all those programs again and open all the project files in each program. Why cant you save the current project state and then shut down/open the project with a single command? (Another reason I like openDOC).

#5 Whats all this other crap?
Look at how much stuff you have to wade through just to find your data. Why have 100s of files for each app? If its not user data, directories/files should at very least be invisible to the user. The user should be able to see and copy ALL their data personal data an a harddrive with one simple step. Probably only 1% of directories on an average windows drive contains user data. UNIX does better with permissions, but its still not good enough. Also, why cant I copy an App in one step for use on another machine?

It seems to me that current windowing and file systems are extremely messy ways of handling data. They almost seem to reflect the multi-file complex code required to write them. It is interesting that the Palm OS is so popular even though it lacks a windowing system all together. Right now, the only alternative GUI concepts for the PC appear in games.

Does NASA itself want the space station?

On: NASA Consider "Demanning" Space Station

Is it possible the the people at NASA aren't so enthusiastic about ISS either? Maybe NASA's administration WANTS to shut it down. ISS has been a continuous drain on NASA for 15 years. I suspect that many people at NASA would like to move on to more interesting things.

IMHO, carefully allocated government support of the aerospace industry is a good investment since being a leader in any industry is good for the United States' ability to compete in a global economy. The shuttle, the hypersonic "space plane" (abandoned), other launch systems, and remote planetary exploration are examples of truly challenging projects. "We choose to go to the moon... and do other things, not because they are easy, but because they are hard."

ISS does not seems to capture the same sense of challenge.

The US seems to be losing its "edge" in the development of space related technologies that it worked so hard to acquire during the 60's. This has allowed Russia, ESA, and now even China, India, and Japan to gain significant ground. Similar things are happening to the US semiconductor, supercomputer, and aircraft industries. That is not good for "our" future economy.

Personally, I am very disappointed by NASA's decision to mostly abandon research on the air-breathing hypersonic "space plane" since it would have led to significant advances in materials, fluid dynamics, computational physics, aerospace engineering, and would ultimately lead to lowered launch costs. (It clearly had a significant utility for military purposes as well.)

ISS keeps many people employed, but a lot of those bright folks could find work on other projects.

What is the feeling about it inside NASA?

An ocean on Mars?

On: Hellish Vision of Mars Unveiled

There is convincing evidence that Mars once had an ENTIRE OCEAN covering most of the northern part of the planet, early-on during the period of extensive meteor bombardment. This theory was further substantiated using a sensitive altimeter onboard the Mars Global Surveyor. Just take a look at this map: (It speaks for itself!)

http://ltpwww.gsfc.nasa.gov/tharsis/shademap.html Pretty amazing, eh?

I am aware that there is considerable debate on the matter, but I haven't seen convincing evidence that could explain the lack of cratering and extremely flat terrain of this northern region. Perhaps someone here who knows more about planetary science than I could provide possible explanations?

Re:So what are options for Americans?

On: 239 MPG Car

I think insurance costs also factor into the mix. I'd very much like to get a small fuel efficient car for basic commutes and use my truck (old Chevy Blazer) on occassions when I need it. As it stands, it would cost me another $600+/year(liability) just to insure another vehicle which easily eliminates any fuel savings and doesn't make it worth it. For many people, owning two vehicles is a "luxury", though it would go a long way to reduce fuel consumption.

I'll keep the truck since it is more versitile. I never understood why insurance costs almost double for a single driver with two vehicles. Maybe someone can explain?

Simple, clever idea...

On: Using Sound To Test Internet Connections

Seems to me that the point of the article is not just determining latency, but rather doing more subtle cable quality tests, listening to the behavior of each "ping" similar to sonar/radar, by heterodyning to lower (audio) frequencies. If there are problems in the line or connected hardware, the ear can pick up the distortions and reflections that limit the quality of the signal path.

You should be able to pick up things like damaged cables, poorly made connectors, and faulty hardware that could lead to reliability problems.

(Hey, someone should try feeding a 100Mbit connection into an FM radio, that might make an interesting debugging tool!)

If you are wiring a hospital, you want to be sure that the local network is 100% reliable so you can just switch to a redundant service provider if the net "goes down" without having to debug the LAN as well. Sure, you could do this with a digital oscilloscope and spectrum analyzer, but this seems like a simpler (and perhaps more sensitive) method. - Nice application of a simple idea!

Re:Just keep on driving

On: Automakers to Make Diagnostic Codes Available

This is very bad advice. Hopefully you don't work at a nuclear power plant.

The "check engine" light is there to tell you that a sensor is malfunctioning/out of spec. Something IS wrong with your car and it will probably cost you more $$$ in the end. Just because a car "runs" doesn't mean that there are no problems. Often the sensor codes are triggered during specific events such as heavy acceleration/hot/cold engine that won't show up while idling so emmissions and other basic testing likely won't indicate any problems. Under some circumstances, the engine may even be operating under "open loop" which means that it is ignoring most sensor data and instead using factory established default values in order to keep running.

Considerable fault-tolerant design goes into the computer control of modern engines. For safety reasons, they will continue running as best they can despite significant problems. (except over-heat conditions which could cause fire)

By driving with an engine out of spec, you may quickly wear out your expensive catalytic converter... or worse. A car will most often pass state emmisions tests just so long as the catalytic converter is not used up. The garage could get true diagnostic readings at the O2 sensor, not at the tail pipe.

The code may also indicate a transmission fault that may be eating away fuel economy and cause transmission overheating such as if the "torque-converter clutch" isn't locking properly.

BTW, on most american cars, there is a "trick" to get the car to read back the fault code without an expensive scanner. These days, many parts stores such as "Advance Auto" will also read-out the codes for free.

The government's role here is to reduce pollution and improve highway/vehicle safety. By giving more people access to the diagnostic information(most of which has been "hacked" already), it means that on average, cars will operate better since fewer problems will be ignored.

My experiences with OCR... Scanfix+Textbridge

On: Accurate OCR?

I started a small document scanning service a few years ago. (I am no longer in that business). The biggest issue in OCR accuracy is pre-process. (in particular de-skew and grayscale removal). If the page is skewed even a couple of degrees OCR will fail miserably. I have had superb results using TMSSequoia Scanfix software which automatically cleans-up and straightens the page nicely. Its expensive but worth-it if you have a lot to scan. I believe that they still have a demo available.

My experience has been that the consumer OCR software is considerably MORE accurate than industrial versions that cost 20X as much. I obtained excellent OCR accuracy using Scansoft's Textbridge software which utilized the Xerox Textbridge engine. Scansoft appears to have purchased Omnipage OCR and discontinued the Textbridge OCR line. I found that I achieved much higher accuracy with Textbridge then with Omnipage after the document was processed by Scanfix. Textbridge did not have some of the features of Omnipage but Textbridge was faster and better at OCR. I would definately download the Textbridge 98 demo that is still floating around on the web.

Both Textbridge and Omnipage OCR were vastly superior to anything else I previewed, including Adobe's OCR engine. OCR can be surprisingly accurate but the source image needs to be free of distortion. Sometimes you will need to break up the page into several using photo-editing software since no OCR can inteterpret the structure of a document very well.

I suspect that you will be better off just typing in the mathematics in by hand. Maybe a visual LATEX editor like Scientific Workplace would be helpful. The LATEX output could be manipulated using a parser to put the equations into the simpler forms that you need while keeping the raw equation in a form that could be used for other purposes later on.

Honesty, 10pgs/hour is pretty good so it doesn't sound like you are doing all that much touch-up. I suspect that using Scanfix will provide the greatest boost in productivity.

Heavy pressure at national labs.

On: Ununoctium Wrapup

This guy was pushing his luck... But I can understand what might have driven him to do it.

I used to work at one of the national labs on the civilian research side. Funding sources are scarce and cutbacks are common. Funding for particle research is particularly difficult to obtain. Almost everyone at the national labs wishes that the cold war was still going on. Right now the strategy of the labs is to prove that they still have a purpose. There's a lot at stake: "laboratory reputation", "project manager repuation", "theorist reputation", and most of all $$$ to get successful results.

Once a project ends, you don't automatically get to work on another one. You usually don't have the luxury of lots of time to refine your experiment either. People DO fear for their jobs.

At the lab, I did observe some instances of "padding" experimental results although I am experienced enough to know that it goes on in most experimental research endeavors (public and private).

The motto at the national labs is "Publish or Perish". In practice, What percent of journal articles describe unsucessful experiments? Not many.

Often in particle physics you are merely validating what is strongly expected from theory so he probably felt he had a good chance to "get away with it" without having to invent physical constants. He went too far though. Honesty does matter in the end.

Re:G's

On: New Jersey Officially Limits G-Forces on Coasters

I wasn't trying to say that the amusement park companies are run by organized crime, rather I just implied that it is to their advantage to keep minor injuries (not involving death, maiming, or hospitalization) from becoming a big deal. There clearly exists financial incentive not to report such minor accidents even within industry organizations. Individuals are always willing to take calculated risks. People skydive, hangglide, rock climb, race motorcycles, etc. In those cases though, the equipment manufacturers don't tell people what to do with the equipment and the individual is mostly responsible for his/her own safety Amusement park rides are a little different in that the manufacturer/operator is responsible for BOTH equipment and the trip. There IS a limit to the number of G's a human body can withstand but it varies from person to person. Do you propose that the limit be such that 0.01% of riders suffer internal injuries? 0.001%? Exactly where do you set the standard? Clearly some people are already getting injured so limits of tolerance are being reached-which is why legislation is being passed. This sort of legislation prevents the amusement park companies from "experimenting" on the population to determine the amount of risk/injury that the public is willing to accept before profits suffer. Given the hundreds of people lined-up amusement companies realize that some of that the public is willing to accept A LOT more risk. Especially if they aren't made fully aware of the risks. Look, most rides are intended to look pretty tame from a spectator's POV until you actually ride them (that's the thrill). Without legislation, there would have to be a large number of casualties in the roller coaster "arms race" before limits would be set voluntarily by industry. There are federal crash/safety standards for cars and I don't see how putting G-force limits on coasters is any different. One can argue whether or not public safety should be a role of the government, but that is another issue. I used to work in electronics manufacturing for several years. From what I have observed, I am 100% convinced that without OSHA and EPA regulation, industry would be dumping their pollution into rivers and workers would be coming down with all kinds of preventable ailments. (as happened at the turn of the century) I think that most people who have worked in manufacturing industries would agree. I don't like the idea of depending on corporate altruism for public safety.

55% of dorm traffic at Cornell is from Kazaa!

On: USC To Students: No Sharing Files

P2P is an expensive problem for campuses. Here are some interesting statistics about network usage at Cornell University:

http://www.cit.cornell.edu/computer/students/bandw idth/charts.html

Over 55% of total dorm bandwidth was from Kazaa/Morpheus!

Re:G's

On: New Jersey Officially Limits G-Forces on Coasters

A few years ago I rode the "Superman" coaster in Darien lake, NY. For the next week my neck was so stiff that I felt like I had been in a car crash. I was in some serious pain. I'm fairly young, athletic, lift weights, average weight/height, and not a "sissy". I would never ride that coaster again.

I suspect that a lot of others come away from these rides with minor injuries like this and don't report it. This sort of legislation may seem silly until you experience an injury.

The G forces and heavy vibration on a coaster are nothing like those in car. (unless in a rollover!) Up till now, themeparks have "policed" themselves and I bet a lot of "minor" injuries reports are suppressed. I think that the head restrains and shock dampers on these things could be better designed.

UWB patents

On: Intel Promises UWB Products By 2006

The firm has applied for an array of patents to cover its wired UWB technology and apparently is the only company to use this approach

I've noticed a disturbing feeding frenzy for UWB patent applications similar to the internet business models. Somehow I can't believe that feeding the UWB signal down a cable TV transmission line instead of an antenna is a non-obvious patentable idea.

Here's a current list of some of the patents around UWB. Notice that only a few groups hold almost all of the cards:
http://www.aetherwire.com/CDROM/General/titles.htm l

Re:This is NOT P2P!

On: Peer-to-Peer Cell Phones

What I would like to see is the FCC to set aside a "limited license" band for establishing an open P2P digital wireless network. Problem with 802.11 WI-FI is that every other device/standard operating at 2.4Ghz can interfere with it. I'm not a wireless expert, but it is interesting that PCS and 802.11b use similar chunks of bandwidth but PCS can act as a long distance "network" while WIFI does not lend itself well to that since other devices/comm standards can interfere with it. PCS CDMA 1850..1910 Mhz w/ (40) 1.5Mhz channels @~200mW WIFI ~2.4Ghz w/(14) 25Mhz channels @ ~100mW (PCS delivers more power over much smaller channels to reduce noise for longer-distance communication) The standards for such a net could be developed by an industry consortium and/or govt. contracts. Power levels and channels could be dynamically adjusted to allow short and long distance P2P networking. (As the number of users go up, power levels/bandwidths would be brought down appropriately. Long distance routing would of be handled via hardwire "docking" to the internet. As the need for bandwidth increases, gradual "multi-mode" units could take advantage of higher frequency bands as they are allocated. The problem today is that we have two extremes: Big-Money'ed Corporate control of the licensed bands and anarchy on the unlicensed ones. It would seem to be in the public interest for the FCC to promote the implementation of such an "open" network. What do you think?

Read "Scientific American" Articles!

On: Physics Books for the Novice?

Read older "Scientific American" magazine articles between approx. 1950..1985. (They were truly excellent back then!) In my experience there is no better way to learn modern physics aside from taking university courses.

Having a physics degree, I've gone through practically all the standard books at one time or another. "Feynman Lectures" are well respected, but take a long time to read and absorb. Unfortunately, it is nearly impossible for most people to teach themselves this field with any serious rigor from textbooks without the benefit of a class. It requires intensive problem sets and LOTS of time. You probably aren't so interested in learning to solve differential equations, rather you want the core science. "Scientific American" provides that.

The 1950 and early 1970's "Scientific American" articles were written at a surprisingly high level. They'll provide introductory material appropriate for the "average" scientist/engineer and push into theoretical and experimental details that are appropriate for PhD's working in the field. The key equations are provided and explained without overly obscuring mathematics. The articles are written by the top scientists in their respected fields and often provide subtle understanding and real-world implementation of ideas that you never obtain from textbooks. The information density of the articles is much higher than from most other sources, so it is a very efficient method of learning. The older articles were definitely intended for professional scientists so they can be a challenging read. The 1950's and 60's articles are so theoretical, they could easily be used as basis for college courses.

You will learn the theory behind whatever physics topic interests you and can follow developments over several decades to understand how theory develops into experiments and get a sense of future directions in the field. The chronological development of the topic combined with re-enforcement of basic theories make this an extremely compelling method of teaching yourself.

For instance, although I've taken a number of courses in particle physics, most of my working knowledge of the field, including the "Standard Model" and "String Theory" has come directly from reading and re-reading old Scientific American's. I can say that I've developed very solid understanding of those topics from reading S.A. articles, something that none of my textbooks came close to providing.

You'll quickly find that "Scientific American" is an invaluable resource for learning physics. (And lots of other science.) Check it out!

Re:Tesla Coil, not snake oil

On: Perpetual Motion Delorean?

This might be part of the hoax. How much power can you recieve using a phased array microwave antenna? Enough to run a car after it has accelerated to speed?

Driving around a racetrack is convenient since the car won't need to start/stop and always have clear line of sight to a high power transmitter in a trailer. The power source will need to run the motor at ~20hp to keep the car moving at 55MPH so they would probably need to transmit ~30 Kilowatts taking into account inefficiencies in the transfer. (.75 KWatt/Horsepower)

Any microwave engineers know if this is possible using phased array antennas? (...seems like it might be, especially if the driver is protected by the "stainless steel construction of the Dalorean".)

FALSE ALARM! No X class flares today! (yet)

On: Today's Solar Flare

Nothing listed on www.spaceweather.com indicates an X class flare at 13:00 utc today nor do any of the other usual websites.

This is the information from SEC (NOAA): Only a few C class events are listed:
0221 UTC C2.8
0955 UTC C8.8
0930 UTC C6.7

You can easily see all the recent events from this plot of solar X-ray flux: (updated in real-time)

There was a minor X class flare last friday (as you can see from the plot!). This is what the poster may be refereing to:
8/30/02 1329 UTC X1.5 Sunspot 95
It was a limb event and isn't headed our way. The plot shows there were no X-ray flares today!

X class flares are fairly common (once every 2-3 weeks these days...) Usually they are near the limb or backside events and don't effect us other than a little radio interference. We might see an earth directed one from region 95 in the next few days. (But don't count on it!) If there is an X-ray event indicated by www.spaceweather.com, then check this alert page at Solar Terestrial Dispatch. These impact predictions are often very accurate! (Only updated if there is a strong earth directed event!)