Appliances of any kind are typically rated in volt-amps, which tells you the current they draw knowing the voltage, but under worst-case conditions without telling you the power factor to know the P (watts) instead of the |S| (volt-amps). Applicance rating plates are meant to tell you how much current draw and hence how big a circuit breaker or fuse you need -- they are not energy ratings.
My own house runs about 45 watts. The furnace alone has a microprocessor in it that takes a good 16 watts. Each GFI (ground fault interrupter circuit breaker that prevents you from getting shocked in wet places like kitchen, bathroom, outdoors) takes up a watt, but you can eliminate that draw by leaving them "popped." I have three motion detector lights -- they save energy, but they take about 2-3 watts each when the lights are off. The garage door opener has a radio receiver that draws about 4. We have a remote control TV that takes 6 watts. Phone answering machines are good for about 4-5 watts. Oh, and a PCI motherboard (it is always "on" when the computer is plugged in) is good for about 4 watts -- I have mine on a "power center", but I can't get my wife to put her computer on a "power center."
I know this info by using either a power meter that the local utility loans out through the public library or by counting turns on the outside electric meter (If you meter says 7.2 on it, it means it is 7.2 watt-hours for every turn. If it takes 10 minutes to make one full turn with everything turned of, it makes 3600/(10X60) turns or 6 turns per hour, or the house is using 6X7.2 or 43.2 watts -- instead of standing outside counting a full turn of the meter, you can turn on a light inside of known wattage to bias the reading higher so the meter turns faster. Also, you have to time a complete turn because there is runnout in the power meter rotor -- it goes faster and slower over different parts of a turn, but it is calibrated to read to better than 1 percent for a complete turn.).
I have never found Macintosh easy to use, not back in the day when it came out, and not recently when I tested some Java apps on a Mac.
It is said that the Linux window managers are imitating Windows. Could it be said that it was really Windows imitating X/Motif/Open look? Didn't windowing systems happen on Unix workstations before they happened on PCs, and wasn't Windows trying to be more like the workstations than like the Mac?
For starters, the Mac hangs on to the application program menu as this shared resource where the app that gets the focus also gets control over the single on-screen menu. That may have been fine back in the day of small screens and limited pixels, but in these days of monster displays and ever more pixels, for crying out loud, give each app its own menu as is done by the Linux window managers and by Windows. The Mac system of you have to think which app has control over the menu is too much a distraction. Interestingly, Java apps running under OS-X have their own menus along with a bare-bones Mac main menu.
I am beginning to think that Java/Swing has a lot going for it and its time has finally come.
Windows took a good 10 years until Windows 95 came out and Windows became the big deal it is today. Java has been out since 1995 -- so it is 2005 -- and I am beginning to think that Java is finally getting to the point where Swing has enough features (clipboard support, images where you can directly set the pixels, hardware graphics acceleration) that it is usable for serious applications. There is a ton of docs on Java/Swing on the Web. People used to talk about Internet time and how fast things moved in the industry, but it really takes a long, long time for many software products to mature.
Now Swing still has some rough edges -- why does a Files Open dialog take 3-10 seconds to come up on first activation, why does expanding a window cover up the Start bar on Windows (I guess there are some look-and-feel options).
But I was able to get an x-ray image viewer program developed under Windows to run under Solaris and Debian Linux without much fuss. I think Mr. Gates should be afraid, very afraid. Maybe not this year, or the year after, but there will be a time when you won't have to make a platform dependent GUI.
When you think of it, the Sun is really not all that effective of a fusion reactor.
While the Sun's core is really, really, really hot, and yes, fusion takes place in the Sun an accounts for the current temperature and physical state of the Sun, the reaction rates are really, really, really low. Think of it -- the Sun has lasted about 5 billion years in its present mode of fusion and is predicted to last another 5 billion years before it goes red giant. And it won't go red giant because it has exhausted all of its hydrogen -- because it has a non-convective core, it just needs to exhaust enough hydrogen in the core to start hydrogen shell burning, which turns it into a red giant.
Not only is the core of the Sun enormously hot and dense compared to even the inside of a Tokomak fusion reactor experiment, the Sun is so freaking huge and massive that even a very low reaction rate that allows it to stretch out its fuel for 10 billion years allows it to put out massive amounts of energy. Of course it is doing hydrogen fusion instead of deuterium or tritium, but when you think of a hydrogen bomb, the H-bomb is doing something quite unlike what happens in nature -- it is burning up its exotic fuel in the blink of any eye -- a hydrogen bomb is more supernova-like than star-like.
What got me thinking along those lines was supposedly the cold fusion thing got going when this Steve Jones fellow was working on some theory about very low rates of fusion happening inside the Earth to account of geothermal heat. I was wondering that these must be very low rates of fusion indeed, but I was going through an astronomy textbook talking about the Sun, and I thought, hey, wait a minute! The Sun is actually doing fusion at very low rates indeed, although whether fusion is taking place in the solid state inside the Earth is another matter to consider.
You think it is bad to admit watching Star Trek Voyager, I caught geek flack for admitting to watching Andromeda.
Andromeda -- no one understood the basic "Andromeda universe", the shows were all blaster fests with no premise or redeeming message, and the acting blew chunks. Why did we watch Andromeda? For the babes!
Andromeda (was the ship named Andromeda Ascending or some sort of thing?) itself/herself was this computer-generated babe. Then there was the blonde babe who piloted the ship who was one of those female-hetero-fitness-freak types who had the hots for the Rastafarian copilot, only Rasta-dude was some kind of alien who didn't understand the human sex thing, and it kind of went from there.
Huxley's Brave New World coined this term "pneumatic", and let me tell you, the redeeming quality of Andromeda was very pneumatic babes, very buff dudes -- it was all soap opera -- and who cared about writing or acting?
There is some fringe-science on whether the Plague was really Yersis Pestis with some speculation that it could have been a hemmoragic fever (like Ebola, many hemmoragic fevers are rodent-borne) or maybe Something Else. Yeah, yeah, I know that this is fringe science, but it is semi-plausible fringe science, not full-blown belief in flying saucers, etc.
As such, there is the fringe science regarding whether AIDS is really AIDS (i.e. HIV). I know people don't want to even talk about this, but it is helpful to find further confirmation of the HIV-AIDS link.
So, do people with this genetic HIV immunity not get AIDS?
Long after its original release, I went to see 2001 on the Harvard University campus "cheap movie" showing. About 30 seconds into the opening "Moonwatcher" scenes, some wise-acre yells out "I don't understand it!" I guess that is the culture where the Lampoon comes from.
There are a number of factors in considering to use an OS -- ease of use, cost, market share, performance, security, ease of developing/really neat developer tools.
My original point is that Windows never was particularly strong in performance. While Linux is strong in many of the above metrics, it does not really best Windows in terms of performance that a whole bunch of died-in-the-wool Windows developers are going to switch to Linux in pursuit of improved performance.
In many ways Windows is the bad squeezing out the good. I had heard that OS-2 (what Windows 95 effectively killed off even though MS was in partnership with IBM on OS-2) was pretty good on threads and soft real time. I had heard that BeOS was simply fantastic on this score.
What I am curious is whether the heavy use of threads came at a cost -- was it hard to wrap your mind around what you needed to do to write a BeOS app? I guess I need to follow the link and look at the online API docs to get some sense.
But the other thing I am saying is that if the goal is to compete/supplant Windows, there should be powerful incentive to switch. I do soft real time images and audio stuff, and BeOS sounded very interesting to me right at the time it started to crater.
What was BeOS really like, apart from there were no applications for it and no one used it?
One one hand there is Windows, which sucks monkey's balls when it comes to soft real time like multimedia with respect to issues such as control over task/thread scheduling, granularity of same, and Windows deciding to take siesta's for 10's of ms (we are not talking about any kind of hard guaranteed real time). On the other hand there is Linux, and I suppose there are real-time versions and frame buffer graphics and such, and yeah, yeah Linux is better in security and other aspects, but from what I have seen, Linux makes Windows look good on the gaming/soft realtime/attempt to be responsive front.
I had heard that BeOS was in a class by itself for soft real time. What was it like? I also heard that BeOS apps were quite thoroughly multi-threaded. What was it like writing a multi-threaded GUI app? Was it particularly hard?
I believe the minimum price point to do anything commercial with Borland is about a grand. The old $95 dollar standard versions, if they exist anymore, used to allow commercial use if the restricted features allowed what you wanted, but I understand they changed all of that.
Gee, what ever happened to the $49.95 Turbo Pascal? Borland, we hardly know you.
The worst part about the 3rd rail deal is that saying this is a myth may encourage someone to try the 3rd rail experiment (stranger things have happened).
Do anyone really, in the least bit, want to suggest that pissing on the 3rd rail is safe even though they found it safe under one trial? This is a setup for a Darwin award nomination.
There is a whole 'nother source of Java slowness -- the JVM startup and loading the classes used by your app. It is not just that some Java apps are slow, they take forever to get going.
The SSME is . . . a race car engine!
I guess there is not much wrong with the design apart from the fact that the engines are required to be very light weight for their thrust, operate at very high chamber pressure so they are efficient at both low and high altitudes with a fixed nozzle. These engines are operated on the edge of what the materials can handle, much like race car engines or the R-3350's on the WW-2 B-29's.
After seeing video of maintenance being done on the SSME's, these very high tech engines being serviced by techs with those ubiquitous red multi drawer tool cabinets, it dawned on me that we should contract the whole thing out to Roger Penske. He and his race car mechanics have a lot of experience doing tear downs and reassemblies of such motors.
By the way, if the SSMEs get used in heavy lift boosters, do these engines get thrown away? Do they even have production lines to make any more SSMEs?
We already have asteroid samples in the form of meteorites. I suppose we have comet samples in the form of meteoritic dust grains.
The funny thing is that the meteorites, stuff that falls to Earth, is judged to be mainly asteroids with the odd piece of Moon or Mars. Comets don't seem to generate meteorites, but they generate most of the meteors -- I guess comets are made of too small pieces or grains to make it all the way down without burning up.
Is there anything we will learn from this asteroid that we don't know from meteorites? Is this asteroid representative of a class of asteroids known from their spectra? It is believed that the known meteorites are a biased sample of the asteroids -- probably it is only a few asteroids that had recent collisions spalling off material that is feeding the current supply of meteorites.
The two types of asteroid I am most interested are the C-type -- a common type of asteroid but rare among the meteorites. It would be interesting to know the connection between these carbonaceous asteroids and the comets. The other type of asteroid of great interest would be the metallic ones on account of the mining potential.
I know it gets slammed as "just so 1980's", but cooperative multitasking based on responding to a message queue as a way of writing GUI's is one of the great ideas.
What I like about it is the granularity. When you are responding to a message, you are in control until you go back to the queue for the next message, effectively doing a yield to other processes until you are given the next message. That way you don't have to worry about locks, and semaphores, and protecting "this data structure" while worrying if it is OK to not protect "that data structure." Of course you still have to worry about callbacks into your code changing your state and resources in an unexpected way, but if you don't make a function call that triggers a callback, you won't have any preemption, deadlock, or race conditions to worry about. And if you make such a function call, the callback takes place at that call instead of any old place like with preemption.
Even when preemptive multitasking is added, all of the setups I have seen (mainly Windows and Java Swing, but I believe this true of Linux window managers), the GUI is still single-threaded so you don't need resource-protection locks out the wazoo for all of the resources used by a GUI (window object, graphic-contect (GC) object, font object, etc). If you run multiple threads, you sync with the GUI thread through PostMessage() and SendMessage(), which apply the proper synchronization locks to the GUI message queue. Java has the exact same thing only GUI objects have InvokeLater() and InvokeNow() (or something like that) synchronization methods which work exactly like PostMessage() and SendMessage().
When I first experimented with threads under Windows, I noticed that the granularity of time slices was much chunkier than with a well-tuned cooperative multitasking approach. I could never get the thread priorities to do what I wanted. I got the best result when I used the preemptive multitasking in a cooperative manner -- I made sure that a thread did some state update quickly and then did a Sleep() or did a Wait() on a signal object -- this works just like cooperative multitasking where you work quickly in response to a message and then do a yield when you dip back into the message queue for the next message. The Windows preemption scheduler is just too coarse grained and too clunky and the only way to get good performance with threads is to treat them like coroutines which yield to one another at explicit synchronization points.
Given that even with preemptive multitasking I was depending on cooperation of tasks (getting a signal, doing something quickly, and then blocking that thread waiting to be signaled again), the one stumbling block on Windows is disk I/O. The only reason disk IO has gone away as a problem is the computer and their disks have gotten so fast that you don't notice Windows being a hog on disk I/O. Yeah, yeah, Windows has asynchronous file I/O, but how does that help you with the "hidden I/O" of page swaps?
I wouldn't even need preemptive multi-threading if Microsoft would have gotten just one thing right. If you write your own message loop, you can do idle-time processing to do animations and updates -- essentially writing your own scheduler for a state machine model of those animations and updates. The dang trouble is that if you pop up a message box or even if the user holds the mouse button down on a menu selection, your state machine grinds to a halt because Microsoft patches in substitute message loops for message boxes and menus, even though they scold you if you write a multi-message loop modal application.
To get around this, I have used an "idle time" thread that keeps feeding the GUI thread with PostMessage(WM_USER) to do the animations and other updates -- this allows any message loop, including the unoverridable ones in menus and message boxes to run the animations. This takes a bit of work to get right -- your idle time messages have to be made lower priority than window messages so you don't gum up the
I don't mean cargo capacity, I mean gross vehicle weight. Was it Titan-III sized, Saturn V sized?
A lot of the Space Station supply missions are these Progress ships launched on a similar rocket that launched Sputnik. Those Progress ships are not very big -- they probably carry much less than the 18,000 lbs.
A reusable shuttle, whether two-stage or single-stage may have to be big to be able to carry any payload. On the other hand, if there is a way to build a small shuttle to handle a small payload, if you use a space station as a depot, it may be possible to do a lot by operating frequent flights.
I am wondering of the Shuttle fails on the capacity issue as much as the partially-reusable issue, that it was built way too big.
The original Von Braun space station idea was an orbiting depot for assembling space ships for travel to the Moon and beyond. If you have heavy-lift like Saturn V, you can go direct to the Moon without a space station. If you then take a step back and say lets develop a shuttle, why are you making it Saturn V sized and with a Titan-III sized payload (I know, the Air Force, Nixon, cancelling the space station).
Well, we now have a Space Station, and I guess it is used for various kinds of science-in-space kinds of things. I am wondering if a better use for it is as a depot, to experiment with bringing stuff up to the station and building bigger things from piece parts, and working out the details of storing and transfering fuel in zero-gee. Even if a replacement Shuttle is not in the works, I think they should be doing engineering experiments on whether a space station is practical as a depot and what it would take to assemble a large ship to go to the Moon and beyond out of small pieces.
Do you have a handle on the size of the Faget DC-3 proposal?
Part of the expense is making something fully reusable, and another part of the expense is making something big. Some things have to be big if they are to do any useful work at all.
I am wondering if the DC-3 was small enough to compensate for the cost of fully reusable. Of all of the should-haves and could-haves, there is the matter of blaming Nixon for the current Shuttle instead of a true reusable, but if the fully-reusable and full-payload Shuttle would have been a white elephant, you can't blame Nixon on that one. The decision to not go with the DC-3 was made much earlier on.
One of the questions is what you do with a shuttle. I guess the DC-3 ideas was that you hauled supplies up to a space station, and you don't need a 60,000 lb payload capacity to bring up milk, tea, water, and TV dinners.
Taking "Space Shuttle" by Dennis R Jenkins, Walsworth Publishing as a reference, the Faget DC-3 was indeed a side-payload piggyback design. The Orbiter was not stacked on top of the Booster.
The most controversial part of the DC-3 was the straight wing reentry design. Straight wing craft famously had stability problems at hypersonic speeds (fatalities in the Bell X-2, one of Chuck Yeager's more harrowing moments as a test pilot). Faget's take on the straight wing is that his vehicle would pancake in on reetry so the straight wing profile was like taking a cookie cutter to the Mercury heat shield profile, and Faget knew how to make that stable. The only scary part was making a transition from this pancake entry to gliding flight at subsonic speed.
As to the cost and the less-than-reusable compromises, I think there is a major chicken-and-egg problem with developing a reusable spacecraft. Jerry Pournelle and his DC-X friends notwithstanding, it would probably take 10-20 billion to develop such a craft if you were on a development path with no surprises -- similar costs are required to launch a new airliner. You then would have to amortize that development cost over x number of flights, and the problem is that the payloads aren't there. If you could drive the cost down, there may be a "if you build it, they will come effect" -- at a low enough cost, people would think of all kinds of creative uses for low-Earth orbit. But given that most of the commercial payloads go to geosynchronous orbit, you would need an expendable upper stage or perhaps an aero-braking transfer stage, and you are talking another 10-20 billion to develop that, and the volumes just aren't there.
The deal is that the Shuttle is a commercial failure, but if you read Jenkins book and see the size of the fully-reusable booster-orbiter designs, if we had gone ahead and built one of those, it may have been an even bigger failure.
Perhaps the most pragmatic proposal I have heard to drive launch costs down would be to decide on an expendable booster, order enough of them to keep a production line open, and just keep flying the things, boosting cement into orbit if you don't have any payloads (or perhaps you could boost stuff like water and steel into the orbit of a space station and make the purpose of the space station to see what you could do with all of this stuff in orbit). The idea is that you could drive the cost of low-Earth orbit enough that a market for payloads would develop which in turn would spark the development of a reusable system.
As it is, we can't seem to get over the hump of enough LEO payloads for economy of scale to kick in.
Is it cheating (according to Nevada Gaming Commission) to use anything besides your wits and senses (like a counter, a computer, a wire from a friend offering odds and advice)? Are those distinctions in the rules?
If so, a person barred for using their wits to play the games means the casinos are all a bunch of sore losers. On the other hand, bans against gadgets make sense -- you know, the baseball pitcher with the Vaseline in his hair that gets on his hand kind of cheat.
Given that you have to do such a big song and dance just to get the patches (yeah, yeah, it is at work at for a legal copy), what are the chances of getting zapped while you are downloading everything?
The other big hassle with Win2K patches is that some of the patches (835732 -- the Sasser patch -- and 889293 and some others) bolex up IE from working. So I am supposed to switch to Mozilla or whatever, but d'ya suppose Microsoft would like me to still use IE? Patching this one W2K machine is this big sifting and winnowing process of endless reboots to load/remove patches to find out which patches I can take and which ones I cannot.
Funny thing is that I had the problem on one W2K machine, and the problem was not so much IE as some Explorer component that I couldn't start Control Panel without a crash, but it required an IE reinstall to roll back. That machine is now fully patched because MS, bless their black little hearts, must have patched the patches.
This other machine however only had a problem with IE crashing on startup if I installed some patches and has been a pain to maintain. I have virus checked and spy checked and regsistry checked the darn thing all up and down to see if some malware is involved to no avail. Currently I am afraid to switch this machine on, although it is behind a firewall and it is SP4.
Sorry you got modded troll -- maybe you can get some justice in meta-moderation.
I guess I am going to "troll to your troll."
The point of the Mann "Hockey Stick" is not so much as that there were warm and cold periods in the past beyond the horizon of human history. The point is that the climate has been dead flat for 1400 years and only in the last 100 or even 50 years has the climate warmed, suggesting an anthropogenic source. The other part of the Hockey Stick is that the Medievel Warm Period and Viking Greenland colony days were local and not global effects.
Trouble with the Hockey Stick is that the times past data is from tree rings and other proxies and has really big error bars. The recent past data is from meteorolgical temperature records with all of the attendent problems of heat islands and the like. The reason you don't do proxies for the recent past is that the numbers would be all over the place and wouldn't show the blade of the Hockey Stick. The reason you smooth the heck out of the proxy data is that, well, they are so noisy.
Then you have those two Canadians who pointed out that the flat part of the Hockey Stick may be an artifact of data handling, but, wouldn't-cha-know-it, the Canadians are not climate scientists, so we can safely disregard everything they point out.
Apart from the Hockey Stick, there are the climate modelers. I believe that their deal is that yes, water vapor is the main greenhouse gas, but the tropics are at 100 percent humidity anyway, so you won't get any more of it, so CO2 is the "swing vote on the court" if you will. They are also assuming strong positive feedback mechanisms -- is it that warming will release more CO2 from reservoirs? Is it that warming will release more H2O at higher lattitudes. I am kind of shakey on that part.
You know, I have heard it argued that it is not entirely clear that the rise in CO2 is anthropogenic. Yes, CO2 is rising as we cut down the rain forests and burn fossil fuel, but at only half the rate, and the rest must be going into sinks. Or maybe the sources and sinks are the dominant effect, and the increase in CO2 with civilization may only be correlative and not causitive. But there are an awful lot of people who are sure of themselves.
Besides being wordy, Pascal has a quirky syntax with semicolon being a statement separator not terminator and the proliferation of begin-end's. Even Wirth left Pascal behind with Modula, Oberon, and now Component Pascal, although Wirth switched to case-sensitive and the newer Wirth languages have keywords in all caps which is so annoying when there are editors that can highlight keywords and a Component Pascal source listing reads like a bad troll.
The deal is that the Turbo Pascal's and later Delphi had such a fast compiler unlike anything else on the planet. The fastest of the bunch is still Delphi 2 which is still my fave, unless I require Delphi 6 or higher for ActiveX development. I always attributed the speed to Pascal, even with Delphi syntax -- the clean syntax made for fast parsers, and the fast compiles made it worthwhile to program in Pascal as opposed to the more modern Pascal-like languages such as Modula 2 or Adda.
But there is a dude who claims that the actual grammer of Delphi is one of life's little mysteries -- there is no simple railroad tracks diagram like the Jensen and Wirth report and the actual syntax may be a Borland trade secret. This is preventing third party tools for Delphi, but it is also scary in terms of the language never having an independent implementation -- Free Pascal makes a stab at Delphi compatibility, but code can take a lot of rework to make it through Free Pascal.
Anyway, if Delphi was really that clean, Delphi grammer could be considered for other uses -- like what we do with XML. While Pascal is not Lisp-simple, I was always thinking that it was much simpler than C/C++ or even Java, but the dude is saying that is not the case and the fast Delphi compiles may have to do more with Anders Hejlsberg than anything intrinsic in the language.
On this subject, I have always thought javac to be dog slow on compiles, but I have compiled stuff with recent versions of Eclipse and wow!, it does complete builds in a blink of an eye in comparison. What is Eclipse using for the Java compiler back end?
Milton Thompson in Edge of Space relates the bugs introduced by modifying the X-15 into the model with the extra drop fuel tanks.
They crumped an X-15 with a hard landing during a rocket-failed-to-light landing emergency -- this accident caused serious back injury to pilot Jack MacKay leading to long-term health problems. While the pilot was able to recover to return to flying the X-15, the powers that be decided to rebuild that X-15 with extra fuel tanks.
When they tested that X-15 in high-speed flight, the front landing gear popped out at speed, the tire burnt up, and they had a rough landing on the melted tire. They figured that the rebuilt aircraft was longer and there was no longer enough slack on the landing gear deploy cable so that when the X-15 heated up from high Mach flight, the cable stretched and popped the gear out. Mind you the deployment system for the gear was no more complex than the brake cable on a bicycle because they wanted this to be ultra rugged and reliable because they had not second chances putting the wheels down on an unpowered landing (the Shuttle adopted the X-15 style landing).
So they lengthened the cable to the point where there was almost too much slack to yank on the T handle and get the gear down, but the next time they flew, the wheel popped out again during high Mach flight.
One of the pilots suggested putting one of these "placards" in the X-15 cockpit saying "Do not deploy gear above Mach 5."
The Shuttle has a lot of the characteristics of a research rather than a production aircraft. Don't think there is anything intrinsically bad with it apart from what they are trying to get out of it (production space flight).
The Qt and GTK toolkits run under Windows, yes, but Qt famously is subject to the whining for the amount of developer tribute money required to use it on Windows and GTK is famous for having a rather weak Windows port. I don't have anything against anyone trying to recover their costs or make a small profit, but the Qt license model requires you to decide up front if you want to go commercial under Windows -- if you start out non-commercial you can't switch over.
If you are talking about the Linux GUI toolkits as being cross-platform, then the choices become GTK, Qt, wx (OK, layered on top of GTK on Linux but layered on top of Win32 on Windows rather than running a Windows GTK), Java, TK (way different and separate from GTK), Win32 (through Wine under Linux) and perhaps some others I have missed (Ok, Ok Mono, but that is a layer on to of either GTK or Win32/Wine).
In that universe, do GTK and Qt have any real "standout" features that merit developer attention? Are they that much better than Win32/GDI -- I am talking about as an API not the locked-into-Windows aspects? As to the complaints that Microsoft is a moving target and is frustrating Wine, you could say that Microsoft's attention is on.NET and beyond and that Win32/GDI is a legacy system with a ton of legacy software.
An analogy is that all of the Intel processors have RISC cores and run X86 in one or another kind of code translation or emulation and that X86 is almost a kind of byte code for a virtual machine just as much as Java uses its own virtual machine byte code. To push the analogy, X86 is a rather clunky byte code on account of nagging details relating to setting condition codes and the like.
So I am asking, are GTK and/or Qt that much cleaner abstractions than Win32/GDI -- I think they are about on the same level. Unless people can tell me what is particularly crufty and unabstractable about the Windows window manager and GDI, I am thinking that a lot of the criticism of Wine is by a reaction to Microsoft rather than the technical merits of providing a particular API, especially since the GUI layer is just one potentially interchangable piece of a larger operating system.
Slashdot Mirror
My own house runs about 45 watts. The furnace alone has a microprocessor in it that takes a good 16 watts. Each GFI (ground fault interrupter circuit breaker that prevents you from getting shocked in wet places like kitchen, bathroom, outdoors) takes up a watt, but you can eliminate that draw by leaving them "popped." I have three motion detector lights -- they save energy, but they take about 2-3 watts each when the lights are off. The garage door opener has a radio receiver that draws about 4. We have a remote control TV that takes 6 watts. Phone answering machines are good for about 4-5 watts. Oh, and a PCI motherboard (it is always "on" when the computer is plugged in) is good for about 4 watts -- I have mine on a "power center", but I can't get my wife to put her computer on a "power center."
I know this info by using either a power meter that the local utility loans out through the public library or by counting turns on the outside electric meter (If you meter says 7.2 on it, it means it is 7.2 watt-hours for every turn. If it takes 10 minutes to make one full turn with everything turned of, it makes 3600/(10X60) turns or 6 turns per hour, or the house is using 6X7.2 or 43.2 watts -- instead of standing outside counting a full turn of the meter, you can turn on a light inside of known wattage to bias the reading higher so the meter turns faster. Also, you have to time a complete turn because there is runnout in the power meter rotor -- it goes faster and slower over different parts of a turn, but it is calibrated to read to better than 1 percent for a complete turn.).
It is said that the Linux window managers are imitating Windows. Could it be said that it was really Windows imitating X/Motif/Open look? Didn't windowing systems happen on Unix workstations before they happened on PCs, and wasn't Windows trying to be more like the workstations than like the Mac?
For starters, the Mac hangs on to the application program menu as this shared resource where the app that gets the focus also gets control over the single on-screen menu. That may have been fine back in the day of small screens and limited pixels, but in these days of monster displays and ever more pixels, for crying out loud, give each app its own menu as is done by the Linux window managers and by Windows. The Mac system of you have to think which app has control over the menu is too much a distraction. Interestingly, Java apps running under OS-X have their own menus along with a bare-bones Mac main menu.
Windows took a good 10 years until Windows 95 came out and Windows became the big deal it is today. Java has been out since 1995 -- so it is 2005 -- and I am beginning to think that Java is finally getting to the point where Swing has enough features (clipboard support, images where you can directly set the pixels, hardware graphics acceleration) that it is usable for serious applications. There is a ton of docs on Java/Swing on the Web. People used to talk about Internet time and how fast things moved in the industry, but it really takes a long, long time for many software products to mature.
Now Swing still has some rough edges -- why does a Files Open dialog take 3-10 seconds to come up on first activation, why does expanding a window cover up the Start bar on Windows (I guess there are some look-and-feel options).
But I was able to get an x-ray image viewer program developed under Windows to run under Solaris and Debian Linux without much fuss. I think Mr. Gates should be afraid, very afraid. Maybe not this year, or the year after, but there will be a time when you won't have to make a platform dependent GUI.
While the Sun's core is really, really, really hot, and yes, fusion takes place in the Sun an accounts for the current temperature and physical state of the Sun, the reaction rates are really, really, really low. Think of it -- the Sun has lasted about 5 billion years in its present mode of fusion and is predicted to last another 5 billion years before it goes red giant. And it won't go red giant because it has exhausted all of its hydrogen -- because it has a non-convective core, it just needs to exhaust enough hydrogen in the core to start hydrogen shell burning, which turns it into a red giant.
Not only is the core of the Sun enormously hot and dense compared to even the inside of a Tokomak fusion reactor experiment, the Sun is so freaking huge and massive that even a very low reaction rate that allows it to stretch out its fuel for 10 billion years allows it to put out massive amounts of energy. Of course it is doing hydrogen fusion instead of deuterium or tritium, but when you think of a hydrogen bomb, the H-bomb is doing something quite unlike what happens in nature -- it is burning up its exotic fuel in the blink of any eye -- a hydrogen bomb is more supernova-like than star-like.
What got me thinking along those lines was supposedly the cold fusion thing got going when this Steve Jones fellow was working on some theory about very low rates of fusion happening inside the Earth to account of geothermal heat. I was wondering that these must be very low rates of fusion indeed, but I was going through an astronomy textbook talking about the Sun, and I thought, hey, wait a minute! The Sun is actually doing fusion at very low rates indeed, although whether fusion is taking place in the solid state inside the Earth is another matter to consider.
Andromeda -- no one understood the basic "Andromeda universe", the shows were all blaster fests with no premise or redeeming message, and the acting blew chunks. Why did we watch Andromeda? For the babes!
Andromeda (was the ship named Andromeda Ascending or some sort of thing?) itself/herself was this computer-generated babe. Then there was the blonde babe who piloted the ship who was one of those female-hetero-fitness-freak types who had the hots for the Rastafarian copilot, only Rasta-dude was some kind of alien who didn't understand the human sex thing, and it kind of went from there.
Huxley's Brave New World coined this term "pneumatic", and let me tell you, the redeeming quality of Andromeda was very pneumatic babes, very buff dudes -- it was all soap opera -- and who cared about writing or acting?
As such, there is the fringe science regarding whether AIDS is really AIDS (i.e. HIV). I know people don't want to even talk about this, but it is helpful to find further confirmation of the HIV-AIDS link.
So, do people with this genetic HIV immunity not get AIDS?
Long after its original release, I went to see 2001 on the Harvard University campus "cheap movie" showing. About 30 seconds into the opening "Moonwatcher" scenes, some wise-acre yells out "I don't understand it!" I guess that is the culture where the Lampoon comes from.
My original point is that Windows never was particularly strong in performance. While Linux is strong in many of the above metrics, it does not really best Windows in terms of performance that a whole bunch of died-in-the-wool Windows developers are going to switch to Linux in pursuit of improved performance.
In many ways Windows is the bad squeezing out the good. I had heard that OS-2 (what Windows 95 effectively killed off even though MS was in partnership with IBM on OS-2) was pretty good on threads and soft real time. I had heard that BeOS was simply fantastic on this score.
What I am curious is whether the heavy use of threads came at a cost -- was it hard to wrap your mind around what you needed to do to write a BeOS app? I guess I need to follow the link and look at the online API docs to get some sense.
But the other thing I am saying is that if the goal is to compete/supplant Windows, there should be powerful incentive to switch. I do soft real time images and audio stuff, and BeOS sounded very interesting to me right at the time it started to crater.
One one hand there is Windows, which sucks monkey's balls when it comes to soft real time like multimedia with respect to issues such as control over task/thread scheduling, granularity of same, and Windows deciding to take siesta's for 10's of ms (we are not talking about any kind of hard guaranteed real time). On the other hand there is Linux, and I suppose there are real-time versions and frame buffer graphics and such, and yeah, yeah Linux is better in security and other aspects, but from what I have seen, Linux makes Windows look good on the gaming/soft realtime/attempt to be responsive front.
I had heard that BeOS was in a class by itself for soft real time. What was it like? I also heard that BeOS apps were quite thoroughly multi-threaded. What was it like writing a multi-threaded GUI app? Was it particularly hard?
Gee, what ever happened to the $49.95 Turbo Pascal? Borland, we hardly know you.
Do anyone really, in the least bit, want to suggest that pissing on the 3rd rail is safe even though they found it safe under one trial? This is a setup for a Darwin award nomination.
There is a whole 'nother source of Java slowness -- the JVM startup and loading the classes used by your app. It is not just that some Java apps are slow, they take forever to get going.
The SSME is . . . a race car engine! I guess there is not much wrong with the design apart from the fact that the engines are required to be very light weight for their thrust, operate at very high chamber pressure so they are efficient at both low and high altitudes with a fixed nozzle. These engines are operated on the edge of what the materials can handle, much like race car engines or the R-3350's on the WW-2 B-29's. After seeing video of maintenance being done on the SSME's, these very high tech engines being serviced by techs with those ubiquitous red multi drawer tool cabinets, it dawned on me that we should contract the whole thing out to Roger Penske. He and his race car mechanics have a lot of experience doing tear downs and reassemblies of such motors. By the way, if the SSMEs get used in heavy lift boosters, do these engines get thrown away? Do they even have production lines to make any more SSMEs?
The funny thing is that the meteorites, stuff that falls to Earth, is judged to be mainly asteroids with the odd piece of Moon or Mars. Comets don't seem to generate meteorites, but they generate most of the meteors -- I guess comets are made of too small pieces or grains to make it all the way down without burning up.
Is there anything we will learn from this asteroid that we don't know from meteorites? Is this asteroid representative of a class of asteroids known from their spectra? It is believed that the known meteorites are a biased sample of the asteroids -- probably it is only a few asteroids that had recent collisions spalling off material that is feeding the current supply of meteorites.
The two types of asteroid I am most interested are the C-type -- a common type of asteroid but rare among the meteorites. It would be interesting to know the connection between these carbonaceous asteroids and the comets. The other type of asteroid of great interest would be the metallic ones on account of the mining potential.
What I like about it is the granularity. When you are responding to a message, you are in control until you go back to the queue for the next message, effectively doing a yield to other processes until you are given the next message. That way you don't have to worry about locks, and semaphores, and protecting "this data structure" while worrying if it is OK to not protect "that data structure." Of course you still have to worry about callbacks into your code changing your state and resources in an unexpected way, but if you don't make a function call that triggers a callback, you won't have any preemption, deadlock, or race conditions to worry about. And if you make such a function call, the callback takes place at that call instead of any old place like with preemption.
Even when preemptive multitasking is added, all of the setups I have seen (mainly Windows and Java Swing, but I believe this true of Linux window managers), the GUI is still single-threaded so you don't need resource-protection locks out the wazoo for all of the resources used by a GUI (window object, graphic-contect (GC) object, font object, etc). If you run multiple threads, you sync with the GUI thread through PostMessage() and SendMessage(), which apply the proper synchronization locks to the GUI message queue. Java has the exact same thing only GUI objects have InvokeLater() and InvokeNow() (or something like that) synchronization methods which work exactly like PostMessage() and SendMessage().
When I first experimented with threads under Windows, I noticed that the granularity of time slices was much chunkier than with a well-tuned cooperative multitasking approach. I could never get the thread priorities to do what I wanted. I got the best result when I used the preemptive multitasking in a cooperative manner -- I made sure that a thread did some state update quickly and then did a Sleep() or did a Wait() on a signal object -- this works just like cooperative multitasking where you work quickly in response to a message and then do a yield when you dip back into the message queue for the next message. The Windows preemption scheduler is just too coarse grained and too clunky and the only way to get good performance with threads is to treat them like coroutines which yield to one another at explicit synchronization points.
Given that even with preemptive multitasking I was depending on cooperation of tasks (getting a signal, doing something quickly, and then blocking that thread waiting to be signaled again), the one stumbling block on Windows is disk I/O. The only reason disk IO has gone away as a problem is the computer and their disks have gotten so fast that you don't notice Windows being a hog on disk I/O. Yeah, yeah, Windows has asynchronous file I/O, but how does that help you with the "hidden I/O" of page swaps?
I wouldn't even need preemptive multi-threading if Microsoft would have gotten just one thing right. If you write your own message loop, you can do idle-time processing to do animations and updates -- essentially writing your own scheduler for a state machine model of those animations and updates. The dang trouble is that if you pop up a message box or even if the user holds the mouse button down on a menu selection, your state machine grinds to a halt because Microsoft patches in substitute message loops for message boxes and menus, even though they scold you if you write a multi-message loop modal application.
To get around this, I have used an "idle time" thread that keeps feeding the GUI thread with PostMessage(WM_USER) to do the animations and other updates -- this allows any message loop, including the unoverridable ones in menus and message boxes to run the animations. This takes a bit of work to get right -- your idle time messages have to be made lower priority than window messages so you don't gum up the
Wasn't there just a story about how DIA has ditched their automatic luggage system?
A lot of the Space Station supply missions are these Progress ships launched on a similar rocket that launched Sputnik. Those Progress ships are not very big -- they probably carry much less than the 18,000 lbs.
A reusable shuttle, whether two-stage or single-stage may have to be big to be able to carry any payload. On the other hand, if there is a way to build a small shuttle to handle a small payload, if you use a space station as a depot, it may be possible to do a lot by operating frequent flights.
I am wondering of the Shuttle fails on the capacity issue as much as the partially-reusable issue, that it was built way too big.
The original Von Braun space station idea was an orbiting depot for assembling space ships for travel to the Moon and beyond. If you have heavy-lift like Saturn V, you can go direct to the Moon without a space station. If you then take a step back and say lets develop a shuttle, why are you making it Saturn V sized and with a Titan-III sized payload (I know, the Air Force, Nixon, cancelling the space station).
Well, we now have a Space Station, and I guess it is used for various kinds of science-in-space kinds of things. I am wondering if a better use for it is as a depot, to experiment with bringing stuff up to the station and building bigger things from piece parts, and working out the details of storing and transfering fuel in zero-gee. Even if a replacement Shuttle is not in the works, I think they should be doing engineering experiments on whether a space station is practical as a depot and what it would take to assemble a large ship to go to the Moon and beyond out of small pieces.
Part of the expense is making something fully reusable, and another part of the expense is making something big. Some things have to be big if they are to do any useful work at all.
I am wondering if the DC-3 was small enough to compensate for the cost of fully reusable. Of all of the should-haves and could-haves, there is the matter of blaming Nixon for the current Shuttle instead of a true reusable, but if the fully-reusable and full-payload Shuttle would have been a white elephant, you can't blame Nixon on that one. The decision to not go with the DC-3 was made much earlier on.
One of the questions is what you do with a shuttle. I guess the DC-3 ideas was that you hauled supplies up to a space station, and you don't need a 60,000 lb payload capacity to bring up milk, tea, water, and TV dinners.
The most controversial part of the DC-3 was the straight wing reentry design. Straight wing craft famously had stability problems at hypersonic speeds (fatalities in the Bell X-2, one of Chuck Yeager's more harrowing moments as a test pilot). Faget's take on the straight wing is that his vehicle would pancake in on reetry so the straight wing profile was like taking a cookie cutter to the Mercury heat shield profile, and Faget knew how to make that stable. The only scary part was making a transition from this pancake entry to gliding flight at subsonic speed.
As to the cost and the less-than-reusable compromises, I think there is a major chicken-and-egg problem with developing a reusable spacecraft. Jerry Pournelle and his DC-X friends notwithstanding, it would probably take 10-20 billion to develop such a craft if you were on a development path with no surprises -- similar costs are required to launch a new airliner. You then would have to amortize that development cost over x number of flights, and the problem is that the payloads aren't there. If you could drive the cost down, there may be a "if you build it, they will come effect" -- at a low enough cost, people would think of all kinds of creative uses for low-Earth orbit. But given that most of the commercial payloads go to geosynchronous orbit, you would need an expendable upper stage or perhaps an aero-braking transfer stage, and you are talking another 10-20 billion to develop that, and the volumes just aren't there.
The deal is that the Shuttle is a commercial failure, but if you read Jenkins book and see the size of the fully-reusable booster-orbiter designs, if we had gone ahead and built one of those, it may have been an even bigger failure.
Perhaps the most pragmatic proposal I have heard to drive launch costs down would be to decide on an expendable booster, order enough of them to keep a production line open, and just keep flying the things, boosting cement into orbit if you don't have any payloads (or perhaps you could boost stuff like water and steel into the orbit of a space station and make the purpose of the space station to see what you could do with all of this stuff in orbit). The idea is that you could drive the cost of low-Earth orbit enough that a market for payloads would develop which in turn would spark the development of a reusable system.
As it is, we can't seem to get over the hump of enough LEO payloads for economy of scale to kick in.
If so, a person barred for using their wits to play the games means the casinos are all a bunch of sore losers. On the other hand, bans against gadgets make sense -- you know, the baseball pitcher with the Vaseline in his hair that gets on his hand kind of cheat.
Given that you have to do such a big song and dance just to get the patches (yeah, yeah, it is at work at for a legal copy), what are the chances of getting zapped while you are downloading everything?
The other big hassle with Win2K patches is that some of the patches (835732 -- the Sasser patch -- and 889293 and some others) bolex up IE from working. So I am supposed to switch to Mozilla or whatever, but d'ya suppose Microsoft would like me to still use IE? Patching this one W2K machine is this big sifting and winnowing process of endless reboots to load/remove patches to find out which patches I can take and which ones I cannot.
Funny thing is that I had the problem on one W2K machine, and the problem was not so much IE as some Explorer component that I couldn't start Control Panel without a crash, but it required an IE reinstall to roll back. That machine is now fully patched because MS, bless their black little hearts, must have patched the patches.
This other machine however only had a problem with IE crashing on startup if I installed some patches and has been a pain to maintain. I have virus checked and spy checked and regsistry checked the darn thing all up and down to see if some malware is involved to no avail. Currently I am afraid to switch this machine on, although it is behind a firewall and it is SP4.
I guess I am going to "troll to your troll."
The point of the Mann "Hockey Stick" is not so much as that there were warm and cold periods in the past beyond the horizon of human history. The point is that the climate has been dead flat for 1400 years and only in the last 100 or even 50 years has the climate warmed, suggesting an anthropogenic source. The other part of the Hockey Stick is that the Medievel Warm Period and Viking Greenland colony days were local and not global effects.
Trouble with the Hockey Stick is that the times past data is from tree rings and other proxies and has really big error bars. The recent past data is from meteorolgical temperature records with all of the attendent problems of heat islands and the like. The reason you don't do proxies for the recent past is that the numbers would be all over the place and wouldn't show the blade of the Hockey Stick. The reason you smooth the heck out of the proxy data is that, well, they are so noisy.
Then you have those two Canadians who pointed out that the flat part of the Hockey Stick may be an artifact of data handling, but, wouldn't-cha-know-it, the Canadians are not climate scientists, so we can safely disregard everything they point out.
Apart from the Hockey Stick, there are the climate modelers. I believe that their deal is that yes, water vapor is the main greenhouse gas, but the tropics are at 100 percent humidity anyway, so you won't get any more of it, so CO2 is the "swing vote on the court" if you will. They are also assuming strong positive feedback mechanisms -- is it that warming will release more CO2 from reservoirs? Is it that warming will release more H2O at higher lattitudes. I am kind of shakey on that part.
You know, I have heard it argued that it is not entirely clear that the rise in CO2 is anthropogenic. Yes, CO2 is rising as we cut down the rain forests and burn fossil fuel, but at only half the rate, and the rest must be going into sinks. Or maybe the sources and sinks are the dominant effect, and the increase in CO2 with civilization may only be correlative and not causitive. But there are an awful lot of people who are sure of themselves.
The deal is that the Turbo Pascal's and later Delphi had such a fast compiler unlike anything else on the planet. The fastest of the bunch is still Delphi 2 which is still my fave, unless I require Delphi 6 or higher for ActiveX development. I always attributed the speed to Pascal, even with Delphi syntax -- the clean syntax made for fast parsers, and the fast compiles made it worthwhile to program in Pascal as opposed to the more modern Pascal-like languages such as Modula 2 or Adda.
But there is a dude who claims that the actual grammer of Delphi is one of life's little mysteries -- there is no simple railroad tracks diagram like the Jensen and Wirth report and the actual syntax may be a Borland trade secret. This is preventing third party tools for Delphi, but it is also scary in terms of the language never having an independent implementation -- Free Pascal makes a stab at Delphi compatibility, but code can take a lot of rework to make it through Free Pascal.
Anyway, if Delphi was really that clean, Delphi grammer could be considered for other uses -- like what we do with XML. While Pascal is not Lisp-simple, I was always thinking that it was much simpler than C/C++ or even Java, but the dude is saying that is not the case and the fast Delphi compiles may have to do more with Anders Hejlsberg than anything intrinsic in the language.
On this subject, I have always thought javac to be dog slow on compiles, but I have compiled stuff with recent versions of Eclipse and wow!, it does complete builds in a blink of an eye in comparison. What is Eclipse using for the Java compiler back end?
They crumped an X-15 with a hard landing during a rocket-failed-to-light landing emergency -- this accident caused serious back injury to pilot Jack MacKay leading to long-term health problems. While the pilot was able to recover to return to flying the X-15, the powers that be decided to rebuild that X-15 with extra fuel tanks.
When they tested that X-15 in high-speed flight, the front landing gear popped out at speed, the tire burnt up, and they had a rough landing on the melted tire. They figured that the rebuilt aircraft was longer and there was no longer enough slack on the landing gear deploy cable so that when the X-15 heated up from high Mach flight, the cable stretched and popped the gear out. Mind you the deployment system for the gear was no more complex than the brake cable on a bicycle because they wanted this to be ultra rugged and reliable because they had not second chances putting the wheels down on an unpowered landing (the Shuttle adopted the X-15 style landing).
So they lengthened the cable to the point where there was almost too much slack to yank on the T handle and get the gear down, but the next time they flew, the wheel popped out again during high Mach flight.
One of the pilots suggested putting one of these "placards" in the X-15 cockpit saying "Do not deploy gear above Mach 5."
The Shuttle has a lot of the characteristics of a research rather than a production aircraft. Don't think there is anything intrinsically bad with it apart from what they are trying to get out of it (production space flight).
If you are talking about the Linux GUI toolkits as being cross-platform, then the choices become GTK, Qt, wx (OK, layered on top of GTK on Linux but layered on top of Win32 on Windows rather than running a Windows GTK), Java, TK (way different and separate from GTK), Win32 (through Wine under Linux) and perhaps some others I have missed (Ok, Ok Mono, but that is a layer on to of either GTK or Win32/Wine).
In that universe, do GTK and Qt have any real "standout" features that merit developer attention? Are they that much better than Win32/GDI -- I am talking about as an API not the locked-into-Windows aspects? As to the complaints that Microsoft is a moving target and is frustrating Wine, you could say that Microsoft's attention is on .NET and beyond and that Win32/GDI is a legacy system with a ton of legacy software.
An analogy is that all of the Intel processors have RISC cores and run X86 in one or another kind of code translation or emulation and that X86 is almost a kind of byte code for a virtual machine just as much as Java uses its own virtual machine byte code. To push the analogy, X86 is a rather clunky byte code on account of nagging details relating to setting condition codes and the like.
So I am asking, are GTK and/or Qt that much cleaner abstractions than Win32/GDI -- I think they are about on the same level. Unless people can tell me what is particularly crufty and unabstractable about the Windows window manager and GDI, I am thinking that a lot of the criticism of Wine is by a reaction to Microsoft rather than the technical merits of providing a particular API, especially since the GUI layer is just one potentially interchangable piece of a larger operating system.