Ample? That was the whole point. There is nothing ample about Ms Jovovich, but there was especially nothing ample about what Ms. Jovovich was wearing.
Maybe the reason women vary all over the map in the ample department is that from an evolutionary standpoint, men can't make up their minds. When you think about it, ample is all about male sexual fetishism because it doesn't much matter when you switch off the lights. But ample and lots of curves suggests fertility in the style of -- hey, you have all seen those Stone Age "Ashtorath" clay figures looking like Rosanne's ancient ancestor.
But then not so ample and kind of skinny suggests a kind of adolescence -- OK don't get whigged out over the under-aged issue, but in ancient times women did all their child bearing before they reached modern legal age of consent because everyone was dead by age 30 anyway.
I guess I didn't know about the "Plain Old Text" option either, but that proves my point.
The topic, to get back on topic, is about C++0x, and any discussion of C++ promotes a flamefest. The parent post to this thread was flaming/whining/ranting about some aspect of C++ while neglecting to use any paragraph breaks, and people piled on to the parent post lecturing the dude/dudette about the need for paragraph breaks when posting.
I was trying to point out a parallel between C++ and Slashdot posting -- they are both feature-full and expressive at the risk of making a dumb mistake, and when the dumb mistake is made, people pile on telling you how dumb you are rather than considering whether C++/Slashdot could be more dumb resistant. And when a person dares suggest that something be made more dumb resistant, that suggestion triggers a response that a person is also dumb and a complainer to boot.
For more background on this, I defer to Cooper of "Inmates are Running the Asylum", anything by Donald Norman, and the infamous "Unix Hater's Handbook." There is this culture that any user-interface problems are the fault of the luser and that anyone pointing out such problems is a whining complainer. I was reading a book on Boost, and barely made it into the first chapter on auto pointers when my head started to spin on different flavors of auto pointer and how one worked in one situation but would result in a GP fault in another situation -- geez, Louise, in Java you just have garbage collection and don't have to wrap one's head around such issues. Of course, I am probably mentally deficient for not grasping Boost/C++ auto pointer flavors and I will go through life writing mentally-deficient Java.
It seems that Slashdot is the only place in the Web forum universe where you have to explicitly place the paragraph-break angle bracket-p-angle bracket. I guess I don't know enough HTML to get that symbol to show up in the posted text to show what I mean, but neither does anyone else around here it seems.
I also end up looking silly by placing paragraph break HTML on other sites I post. Everyone around Slashdot (just like -- name favorite feature -- C/C++) programmers gets in the habit of using proper paragraph breaks, but it is entertaining to watch a noob make a long post, or perhaps an old timer who has passed some brain gas, make a honkin run-on post.
I guess Slashdot could fix this and accept whitespace as a paragraph break, but I guess that would violate some deep principle. But I am not blaming Slashdot -- it is deep in the software developer culture to accept things the way they are and then holler at the newcomers about their mistakes brought on by the fractured syntax. How do you think the C syntax has become so viral?
Forget NASA. My plan is to play Powerball, and if I win a 100 million+ jackpot, to put a call through to Gilbert Levin to see what it would take to place a life-science payload on the surface of Mars. It is a long shot to win at Powerball and also a long shot to discover life on Mars, but if successful, Gilbert Levin will get a Nobel Prize and I will go down in history as that idiot who spent his lottery winnings on a Mars payload.
The one problem with life on Mars is the Gaia Hypothesis. We are supposing that life on Mars is hanging on by a thread -- the Gaia Hypothesis argues that life modifies its environment to perpetuate itself, and if there really were life on Mars, it would be pretty unmistakable.
Do you realize that the metric system is the last vestige of the French Revolution? Metric units -- heck -- bring back the metric calender! 10 day weeks and 10 month years. Bring back the Month of Thermidor!
My understanding is that Thomas Gold was arguing that the source of oil, gas, and possibly even coal (yeah, yeah, I know about how coal is dense with fossil imprints) is in the upper mantle, and the carbon got there by the cold accretion of carbonaceous meteorites during the formation of the Earth, the same way methane and other hydrocarbons to to Titan.
Two difficulties with that. One is that the Moon is supposed to have formed from a collision with a Mars-sized body over 4 billion years ago. That collision is believed to have melted everything down to a 1000 km. The other is that the mantle has abundant oxygen and heat, and plate tectonics says it is in convective motion mixing up the different layers (although very slowly) -- one would think all the carbon is now CO2 or CaCO3 or something by now.
However, there is still the matter of diamonds, which are not believed to be dead dinosaurs (diamonds are aged 1 billion years or more, typically) -- they had to have originated in the upper mantle (about 100 km down) and had to have been brought to the surface rapidly (to avoid reversion to graphite, although you get graphite pseudomorphs of diamond -- clumps of graphite shaped like diamond crystals which were probably diamonds brought up too slowly so they reverted to the stable graphite form). While diamonds are rare, and the Kiberlite pipe eruptions that brought them up are rare and maybe date only to an earier geologic epoch, there has to be something to produce reduced carbon down below.
J F Kenney and his Russian associates believe that starting with FeO, CaCO3 and H2O (stuff not hard to find in the mantle owing to limestone and water being subducted down and Fe being brought up by mantle convection) you can end up with CH4 plus higher chain hydrocarbons. The argument is that about 100 km down is the only place methane, octane, and above can form is that the thermodynamics works at those temps and pressures and that the thermodynamics don't work for turning plant/algae material into oil in the traditional "oil window" of about 1-2 miles down.
So, there you have it -- oil is created from the same place an process as engagement ring stones, not only does oil not come from dead dinosaurs but from rocks instead (although the subducted CaCO3 could have its origins in biology of reef building), but that oil is not latent solar energy (in the form of sequestered biomass) but that oil is in reality geothermal energy (geologic raw materials brought together by heat-driven mantle convection and endothermic reactions driven by mantle heat).
If oil is really geothermal instead of solar in origin, one could consider and advanced technological culture with the capability of somehow using the environment of 100 km down as a natural resource, and of establishing a closed-cycle renewable geothermal based energy economy based on -- oil! One could sequester CO2 deep below and get back reduced carbon, all driven by geothermal power, which has its origins in natural radioactive decay along with the latent heat of fusion of iron in the core.
I mean think about it. A lot of the speculation about advanced energy cultures for the far future look outward into space and of tapping the vast resource of solar energy on the Earth surface, in Earth orbit, and beyond -- think Dyson sphere. Has anyone speculated, either in popular science writing or science fiction, about an advanced energy culture fully utilizing the energy resource within a planet?
You may say drilling or tunneling 10 km is stretching it not to imagine 100 km? But who is to say drill. Some MIT dude suggested using a million tons of molton iron (some grant proposal) to melt and sink its way all the way to the core to carry some kind of probe to find out "what is down there." Who is to say that some related scheme may be able to both bring materials down to the mantle (say CO2) and bring back materials (oil and gas) in a closed loop? I am not saying it is practical with today's technology, but it is not anything violating
To your comment that you get aerobraking on the way down I add that on landing most of your fuel is gone, so you are using rocket thrust to brake something much lighter.
The main problem I have with rocket-thrust landing in the manner of the DC-X is what happens if a rocket engine conks out? Also, when you are coming in for a landing, the aero resistance of your nearly empty fuel tank and light weight means your terminal velocity is low, but to save on fuel, you can't be hovering on engine thrust for very long. As you approach the ground, you need a very carefully timed burst of rocket thrust to come in for a soft landing.
You could say that the Soyuz lands vertically on rocket thrust. Yes, they come down mainly on parachutes, but in the last few feet they fire these solid-fuel landing rockets attached to the parachute lanyards to land softly -- they need this for a dry land instead of a splashdown landing. I heard that they have had some landing rocket failures and there are cosmonauts missing some front teeth.
Can a landing rocket be reliable enough, or are you going to crump a few space ships on landing?
Arnold Scharzenegger showed the way out of the jam of that "President who was a Democrat" after the fact.
The President should have looked down, scuffed his feet, and said, "Without going into an specifics about what Ms. Jones has complained about, there are certain things I have done as a man that I am not proud of, and in the interest of putting this matter behind us and moving the country forward, I am offering Ms. Jones X dollars in settlement of this matter." That is the formula -- admit to no specific wrongdoing but admit to being a "man behaving badly", pay the dime (or make the public non-confession), and move on.
Quoting the late John Belushi, "but nooooooooo!" As much as there was Kenneth Starr plus every partisan attack lawyer going after the President, he had his own gaggle of hyper-partisan attack-dog lawyers and spokes-mouthpieces working everything from the courts to the Sunday talk shows, who made him stand his ground. He just couldn't do "rope-a-dope." When he made the admissions and apologies they were too little and too late, but maybe not, as they finally gave him the public sympathy to stay in office.
One could program to the Windows API in C after Charles Petzold's books or perhaps program in C++ and muck around with MFC with those forests of MESSAGE_MAP macros. Or one could just whip up a program in Visual Basic -- it would be bloated, it would be slow (at some level, but it would respond fast enough to mouse clicks), it would be visually crude like it was thrown together by someone who knew nothing of graphics design or of programming, but such a program would be there to do a job while the other guys were whining about how much work it took to develop a proper Windows app.
If you are talking cross-platform, you could learn the Qt or GTK APIs and muck around with getting Qt or GTK installed on Windows for your 98% of the world according to TFA Windows users. Or you could do it in Java. It would pretty much Just Work on Windows, Solaris, Linux, and OS X. Yes you have to get Java installed under Windows, but in most academic computing facilities, sysadmins putting Java on Windows machines is pretty much a given these days. You ask for Python and wxWidgets and you get a sullen stare -- as to Qt or GTK on Windows, fugetaboutit.
Java has this BufferedImage thing that pretty much lets you twiddle the bits of an RGB image to make any kind of graphical display you want under all of the mentioned OS's.
You may be programming to the Windows API, I may be programming to the Windows API, but talk to your friend, your sister-in-law, and the guy down the street developing a Windows app, and they are all doing Visual Basic. You may be proficient in Qt, I may be trying to learn Qt, but talk to the people doing graphical Linux apps in Comp Sci departments, and they will all tell you Java.
Java has a layout manager you say, true, and if you pick up NetBeans expecting it to be VB and you plunk a JButton down in the middle of a JFrame and find you can develop a nice app where a single button takes up the entire main window, well welcome to the club. Once you get over the layout manager learning curve, you will be using it like an old pro, and besides, things like wxPython are layout-manager oriented, now aren't they?
I can see myself switching my development of scientific visualization apps from Delphi to Java, and when I have enough of my stuff running in Java I could say to myself, "Hey, I could just as well be running Linux as I am running Windows -- if I get the hang of OpenOffice or AbiWord, maybe I don't even need Windows." I don't see myself investing in the learning curve to develop for Qt or GTK, especially since I am not prepared to ditch my Windows machine and figure out how to get Qt or GTK up on Windows.
Where are the Linux desktop apps? In a 98% Windows world it will come from people doing Windows with a cross-platform layer -- it could be Java, it could be RealBasic, it could be Python -- I really doubt it will be Qt or GTK. It may also get a boost from Longhorn/Vista and whatever new thing Microsoft is pushing and people deciding to stay with Windows but to go with one of these cross platform things rather than the Microsoft API du jour. When people have enough of these apps, people will say "Hey wait a minute, I could just as well be running Linux."
This is a blatent Demeter violation now, isn't it? Or is it only a Demeter violation of either the strong or the weak form of Demeter depending on whether invoking the function getContentPane() to return an internal object of the JFrame from a derived class is considered OK or not?
When structured programming came out, it was pretty obvious what kind of mess you had with GOTOs. When Glenford Meyers came out with his "coupling" and "module strength" criterea for function modules, it was pretty obvious what kind of mess you had with global variables or with bundling sets of variables into structs simply to cut down on the parameter lists of functions.
But with this Law of Demeter business, I am just not seeing what the fuss is, especially when you have to make programs a lot more complex with forwarding functions and convoluted applications of the Visitor pattern to avoid invoking a function on an object contained in another object.
I mean c'mon people. One of the big uses of object-oriented programming is GUIs, and everyone's GUI has this layered structure of frame object-graphics object-font object where you make some kind of call like myFrame.getGraphics().setFont(myFont). A lot of this is not that the Graphics object is some kind of mystery thing that you have to keep from users of the Frame object, but Frame and Graphics are most likely interfaces to some underlying implementation, and the reason for separate Frame and Graphics interfaces instead of giving Frame forwarding methods to every freakin' method of Graphics is to organize gobs of interface functions in some logical hierarchy.
The efficiency gains one is talking about can be obtained much easier by turbo-compounding -- do a Google and look up "turbo-compound" along with "R 3350." The R 3350 Turbo Compound was the engine that went into the DC-7, the Lockheed Constellation, and the Navy P2 Neptune sub hunter airplane (immediate predecessor of the P3 Orion). The engine was so fuel efficient they set unrefuelled distance records that are only now matched by A-340s and 777s. I also believe that same engine went into the Douglass Skyraider (single-engine single-seat prop plane that was the "Spad" of the Viet Nam war, but I believe that little plane had a comparable range and bomb load to a B-17).
The R 3350 TC put "blow down" exhaust turbines in the exhaust system, and it coupled those turbines to the main power shaft with automobile-style automatic transmissions.
The problem with turbo-compounding a car is that turbo-compounding works just great on airplanes that run on constant, high power settings, but a car cruises at very low power compared to the power needed to accelerate the car. The dominant loss in your car at highway cruise is engine friction, and going to fewer, smaller cylinders helps with highway mileage at the expense of pickup.
For example, the Ford Five Hundred and Freestyle use a small 6 cylinder to power a big car, and they use either a 6-speed or a continuous-variable transmission to have that engine turn over very few revs on the highway while allowing the motor to rev for pickup. The result is a car the size of a Crown Vic that gets better gas mileage than a Taurus (although owner gas mileage reports on these cars are all over the map -- they may be victims of the "hybrid" effect of suffering badly from lead-foot drivers). The result is also a car that every last auto columnist has been bitching, moaning, and whining is "badly underpowered" and "revs something fierce to accelerate" (I guess the Iraq war, ANWR, peak oil, etc. are of not concern to the writers shilling for cars). On the other hand, the Five Hundred is a big boat that gets great gas mileage if you drive it very gingerly, and I imagine Ford has the older-generation demographic sewn up with that car.
Anyway, one approach to hybrid gas mileage is a small engine geared tall with some kind of power booster -- simply improving the thermo efficiency of the cycle at part load won't do much because the add on power from the turbo-compound/steam engine/etc isn't needed at low load.
The steam engine could be a power booster -- you could have a boiler that could store enough steam to get a little boost. Problem is that while you could undersize the boiler, you couldn't undersize the condensor. If you are trying to get 3 litre power from a 1.5 litre engine, you have to make up fully half the peak engine power with the steam engine, and you are back to Doble/Bill Lear.
What about the condensor? Apart from the fact that it is a lot of plumbing that needs to not leak, you could say that it is no different than an auto engine radiator. The problem is that an auto engine radiator doesn't pass most of the heat -- most of the heat rejection is through the exhaust and only some heat leaks through the cylinder walls to be rejected by the radiator. A steam engine condensor has to reject all of the thermodynamic cycle heat at very low temperature differences -- unless you want to go open cycle like a Stanley or like a railroad steam locomotive.
1) What to you do/get for a GUI? Like Algol famously leaving "IO to the library" meaning they punted on IO and this really hurt Algol, Python leaves the GUI to the library, but there are two mainline choices -- TKInter and wxWidgets/wxPython. Does Lisp in a Box have no GUI, some homebrew Smalltalk/Squeek sandboxed GUI, or does it link to TK or wx?
2) Can you link to C/C++ to do drivers, low level stuff, reuse DLLs or SO modules, or is Lisp beyond connecting outside of the Lisp runtime? Even Java can reach outside with the much-maligned but really isn't too hard to use JNI -- I like JNI because it allows C/C++ to call Java and Java to call C/C++.
By the way, for all of the talk about.NET/C#,.NET has pretty easy ways for.NET to call down to C/C++, but for C/C++ to call.NET, you have to package the.NET module as a COM/ActiveX, and there are magic "metadata" keywords that you have to learn about that, and then you have to do all of this mystery "strong name" generation using a poorly understood process to register the.NET assembly in the Registry as a COM/ActiveX, and then you have to access a COM object through C/C++, which is far from pretty on account of Microsofts clunky use (hodgepodge of preprocessor, classes, templates, and opaque auto code generation) of the C++ language (MFC? ATL?) to do that. For all of the clunky nature of JNI, you can go both ways between Java and C/C++ without any of that Registry song and dance.
One of the strengths of Python is the ability to switch to C/C++ as the need arises, although the C/C++ side is still a little opaque to me because you have to go through some dance to represent Python's dynamic object and method name lookup on the C/C++ side. Anyway, I think part of Python's popularity is that you can drop down to C/C++ if you hit performance or feature roadblocks and the perception is that Lisp is like some Smalltalk implementations -- so highly sandboxed to prevent this sort of thing. What is the story on this?
Worked: Windows, Solaris, Debian; not worked: OS X
on
Java Is So 90s
·
· Score: 1
I have a Java imaging app that worked with no sweat in Windows, Solaris, and Debian Linux -- the only trouble was with OS X -- the image didn't show up.
Turns out OS X supports transparency, so the SetRGB() call of a BufferedImage needs the high byte set to all 1's (| 0xFF000000 you pix values) so that an image is not invisible -- the transparency byte is ignored on all other systems. This is kind of stupid -- usually "unused" fields are set to all zeroes, and all zeroes should mean a default of opaque instead of a default that the image is completely transparent (i.e. doesn't show up).
The other problem I ran into was that I could copy said images to the system clipboard on every system except for OS X. That was a fine how-do-you-do as the Mac is supposed to be so Java friendly according to the hype and the Mac is supposed to be so graphics friendly. Windows and Linux, supposedly much less Java friendly but for different reasons, didn't have any trouble. Turns out that this was a known bug, fixed by a revision to Java 5 (1.5) that came out a few weeks ago.
Oh, and the Mac had some different font width for label widgets compared with everyone else that required some tinkering with preferred panel sizes in the layout manager.
By the way, the biggest Java "platform" incompatibility is not between Windows-Solaris-Linux but between Java versions -- it seems you need to compile separate binaries for Java 4 and Java 5. This makes Java a lot like Delphi in that if you release visual widgets you have to have binaries for Delphi 3, 4, 5, 6, 7 or require everyone to use Delphi 7. Windows, bless their black little hearts, is famous for a great deal of upward and downward compatibility of Windows binaries among Windows versions apart from some problem with button colors if you used the wrong color defaulting mechanism.
I believe Carl Sagan made this point that scientists as he were criticizing the astrophysics of Velikovsky but were impressed with his grasp of archaelogy and ethnic myths, legends, and sacred writings, while the archaelogy community more or less assumed that the astrophysics was impressive but thought that Velikovsky's interepretations of ancient cultures was totally bogus.
If we are on the subject of ancient writings, myths, or oral traditions, the bit about the planet Venus being ejected as a flaming comet from Jupiter in Biblical times and showering manna down on the starving Israelites in the form of cometary hydrocarbons no longer interests me one way or the other.
The standard chronology is that you had the Trojan war, which supposedly came at the end of the Mycenean Greek culture about 1200 BCE, that culture collapsed, and then you had a reemergence of the Greek city state culture around 700, the height of Classical Greece around 500, the Peloponesian war which weekened them, the Macadonians taking over, Alexander, the superceding of the Alexander empire by Rome, etc. The interval between 1200 and 700 BCE is regarded as a Dark Age, a much deeper, darker stop backwards in the advance of culture than the supposed European Dark Age of the Middle Ages.
Jerry Pournelle writes often about the Greek Dark ages -- mysterious, unknowable things are always kind of cool, and ancient disasters involving regression from a more advanced culture, a kind of Atlantis, are even more cool. One of the big things supposedly lost in the Dark Ages was writing -- Homer's Iliad makes reference to writing borne by messengers in such a way to suggest that Homer didn't understand the concept of writing (Homer, by the way, is believed to be one source or perhaps a composite source, of an oral tradition). That a civilization and trading culture collapsed and regressed to cave man conditions so utterly and completely that they lost writing is fascinating -- if we had an atomic war or a global warming-triggered climate flip back into an ice age, you would think we may lose our automobiles, and central heat, but would hang on to painless dentistry through the cultivation of coca and poppy and would still know how to write.
While Velokovsky fascinates with cool tales of flaming comet planets, he takes a dump in the punch bowl of cool things to spin yarns about by claiming the Greek Dark Ages never happened, that the whole thing is a goof up from trying to connect the chronology of dynasties of Egypt to the Greeks? Anyone know anything more about this?
What a person needs to advertise is knowing Delphi at a deep level (component writing, ActiveX).
The people using Delphi out there are simply desperate to find anyone else who knows anything about it to help out.
Cleaning up after a VB programmer
on
Build a Program Now
·
· Score: 2, Informative
I don't dismiss the idea that there are language snobs, but part of the disdain for VB comes not from people developing quick-and-dirty applications in it but from people asked to maintain massive applications that started out as quick-and-dirty.
Part of the problem may be that VB enables people with not a lot of experience at program design in any language to generate these quick-and-dirty applications that morph into critical applications in some business that some person then has to maintain. I guess the ability to work with C++ is regarded as a kind of "union card." The C++ qualification weeds out a lot of people. But this idea of choice of language to control programming is not a new one: currently the idea is C++ as a kind of intelligence test; some 20 years ago the idea was forcing programming in Pascal as a kind of test for patience or perhaps obsessive/compulsive neatness as a qualification for good programming.
I also think there is some fault in Visual Basic itself. If you look over the GoF Design Patterns, Visual Basic is pretty much stuck on the Mediator Pattern, which may tend to flatten program structure and lead to spaghetti code.
The basic pattern in GUI programming is that you have a widget, say a button, that responds to user input, say the user clicking the button with a mouse. You then have some "business rules" or "program logic" that is supposed to be invoked in response to the button click. The Observer pattern states that you can register some Observer object with button object to be activated in response to clicks. The Mediator pattern states that if you have a collection of button objects, you register a single observer, the Mediator, with all of those buttons, to contain all of the "program logic."
In Visual Basic you have a form, and you place widgets on that form, and when something happens in those widgets, you have one or more functions for the main form object that respond to those events. Some of the widgets require a lot of hand holding -- they fire off a lot of events for a variety of conditions and the main form code has to figure out what to do in response. The consequence of this is that the main form can get complicated very quickly with numerous event functions and numerous state variables to keep track of what event got fired and which mode one is in.
In Java, you have these action listener and event listener interfaces that you can register with widgets that signal mouse clicks and other actions. You have the flexibility of having a separate listener object for each widget or having the main form Frame or JFrame object be the listener by implementing the listener interface. You have the design choice of having a flat hierarchy or of having a collection of objects respond to different events. Whether people programming Java are able to use this flexibility effectively is another matter, but Java programmers are not forced to stuffing all of the program logic into the main form Frame object.
For example, if you have a widget that fires a large number of events and requires specific responses to all of those events, and implementing a widget that way may be a way of make a widget very flexible and easy to customize, you could write a specific Java listener class to not only respond to all of those events but also to implement some filtering logic on those events. In Visual Basic, my understanding is that all of that filtering logic has to go into the main form and that there isn't a simple and obvious way to parcel off that logic to another object. It is for this reason that the language snobs consider that for all of its faults, Java is "industrial strength" while Visual Basic remains a "toy language."
Now there are language snobs who fault Java for not having closures, the ability to register functions with classes or objects on the fly, for a clunky way of inspecting an object for functions and function signatures called reflection, and so on. But the number of people who understand how to use those features effectively is more limited. There are a lot of people who know how to use objects effectively and are frustrated by VB 6 having a crippled object system.
COM-based, IE -- I guess you are talking about having IE host an ActiveX control, and anything you can do in Windows you can do from an ActiveX.
Yeah, yeah, ActiveX on IE as a security problem, and all of that. I like ActiveX, and it is a very powerful cross-language facility, but you are solidly restricted to Windows. This thing sounds like an other-than-Windows other-than-Java cross-platform way of doing graphics which is pretty interesting.
Python has its dependencies -- you have to figure out which GUI, there are other interesting libraries. You don't just install Python -- it is Python plus a bunch of supporting stuff.
Java on the other hand is Java. Unless you want to go the SWT route, if you are using Swing/Java2D, all the stuff is there.
I did some Web surfing on Java look-and-feel, and the idea is that while the Mac is really strict about human-interface guidelines on one extreme and Linux flies the Jolly Roger with regard to obeying any social forms or conventions, the notion of a Microsoft look-and-feel is a pretty loose notion too. Office famously has widgets not seen anywhere else and Borland had those ugly checkmark glyphs in buttons, and so on.
It depends. If you are a developer and considering Java to support a shrink-wrap product, one is perhaps fussy about the user experience. If you are developing apps for use by students in your own class at the U, for a small community of colleagues, or even developing a commercial app for a niche user community (consider all of the butt-ugly Visual Basic apps), precise look-and-feel is probably a minor issue.
The reason I am suddenly impressed with Java is that after 10 years of looking at it (OK, I developed some non-GUI apps in it), a requirement came along for an x-ray image viewer. I was under the impression that displaying a gray-scale image from computed pixels was a real big deal like it is in the Windows API with CreateDIBSection, but it turns out that Java now has this BufferedImage class that does the job simply and cleanly. Create a BufferedImage, set pixels wholesale with setRGB(), and drawImage() the BufferedImage to a Graphics object in the Swing paintComponent method. It also turns out that copying that image to the system clipboard is equally simple.
BufferedImage may have been around for a while, and imageFlavor (for the clipboard) has been around since Java 1.4, but these things haven't been around since day 1 sometime in 1995.
Anyway, the app my colleague and I coded up in a week (there was also a learning curve on layout managers) worked just fine on Windows and just peachily on Debian and Solaris. Oddly enough we hit some roadblocks with OS X, that product that is supposed to be much more Java, image, and clipboard friendly. One issue was setting the transparency values in setRGB() -- not an issue on OS's which don't think there is a need for transparent images, but our x-ray picture was invisible initially on OS X; a second issue was a mangled paste into a native OS X app -- turns out Java 1.5 Release 3 for OS X came out last week Monday PM and the version of Java 1.5 I was using was installed last week Monday AM.
Writing an image viewer where you compute your own pixels has always been considered an advanced topic in the GUI business only of interest to game programmers, and after two weeks I have such an app running on 4 OS's. It dawned on me that this could be the death of the Windows monopoly.
Yes there are hangups. My one pending UI gripe is the Java JFileChooser "Files Open" dialog widget. On Windows, it takes seemingly forever to come up the first time you do a Files Open -- on Debian it just pops right up, so much for Linux having second-class Java VMs. The Windows Files Open has text selection defaults that make it efficient to navigate directories and select files with the keyboards while the Java Files Open is just plain stupid and maddening.
But consider my 10 years rule. Java came out in 1995. Since there were other ways to do GUI, people didn't put up with Java Swing shortcoming and walked away. But 10 years later, I am telling you that Java Swing has had perhaps enough improvements made (there is an enormous amount of help on the Web getting up the layout manager and other aspects to Swing learning curve) that it deserves another look.
The OS may be the hardware abstraction layer and x86 may be the byte code, but the API could be independent of either.
Yes, Java-the-platform is in a way its own OS, but on the other hand, Java sits on top of whatever native OS. How about thinking about it this way: I don't have to clean format my hard disk and get all new apps if I install Java. Java may be its own API on top of an OS, but not only is it not tied to a particular native OS, it doesn't whine, fret, or threaten if it is not your exclusive API, and it plays with the other kids.
I can have Java, Python, Ruby on Rails, and even Mono coexisting in the same partition on top of the same OS. They can interoperate through files. While this is not always politically correct from SUN's way of thinking, I can also interoperate Java and C++ in both direction (Java calls C++ or C++ calls Java) using the JNI. I can interoperate Python and C/C++ and on Windows I have Python hosting ActiveX controls to do visuals which in turn can call Java signal processing plug-ins -- it is sort of like the Far Side Cartoon of the cops who find a man swallowed by a crocodile in turn being squeezed by a boa (perhaps a python?) and are required to sort that mess out.
While there is a thing such as "Internet time", I think there is a 10-year rule in effect for the maturation of major pieces of software. From the time of Windows introduction in the mid 1980's to the time of Windows 95, COM, and ActiveX, when Windows acquired the features that made it what it is today was about 10 years. Java is around its 10th anniversary in 2005. Maybe Java/Swing has acquired enough features and enough performance enhancements to be taken really seriously.
Um, volt-amps don't have units of watts. Or not if you wanted full credit on the exam problem I gave last week. A generally followed convention is that volt-amps is the voltage times current without taking into account the phase, watts is the in-phase voltage times current, and VARs (volt-amps, reactive) is the out-of-phase voltage time current. Watts plus VARS add up to VAs as vectors.
I don't want to suggest that anyone backfeed a powerline and get anyone hurt. However, there is strong evidence that the power company objections on safety grounds are a red herring for their economic concerns and unwillingness to fully comply with regulatory laws. To be screaming "safety, safety" when your real concern is "money, money" does a lot to hurt your credibility on both safety and money issues.
As to the safety issue, I haven't look into this, but suppose you feed your line with a solar panel to line inverter specially designed for grid-connect applications -- say an inverter that is excited from the power line and if the power line goes dead the inverter stops feeding power. The solar power hardware people may indeed have such inverters, and I even heard of solar panels with that sort of thing built right in.
The power companies could say, "yes, you can net meter and back feed, and you can only back feed up to a certain amount, and you have to use an approved inverter, and there is a nominal inspection charge for us to come out and give the OK to your equipment." But that is not what they say. They tell you you need an expensive install of a dual meter setup and that you get paid a lot less for the power generated than for the power consumed, which is not compliant with the net metering laws for consumers. When you politely ask about the net metering laws, they do the "CEO yelling at the TV reporter and threatening to shove the camera guy routine."
As to "why should the powercco give you the storage facility for free whenever you wan't it" the answer is because it is public policy and the law that's why. The same reason why the utilities have a monopoly and can get reg approval to jack up your electric bill to pay for their nuclear power plant problems or to pay off new power plants when you are using the same amount as you always used.
Why is the public policy and the law that way? Because elected representatives were convinced it was good public policy to have the regulated-monopoly power companies in effect subsidize the development of renewable and solar power for a whole variety of reasons, that's why.
The irony of this is our local powerco is putting up these wind mills and asking consumers to voluntarily fork over extra bucks to the power utility simply for the bragging rights of getting wind power. If you want to take your extra bucks, and instead of forking it over to the utility but build your own solar panel, and if you politely ask about what you need to comply with their safety regs and what they need to do to comply to laws on the books, and you do this asking at the Wisconsin Alternate Energy Fair where this same utility is crowing about how green they are, you get a burly utility representative in your face in a primate-species threat posture.
Why would I want to put up my own solar panels instead of forking bucks over to the utility to say I am green? It is an "ownership society" thing. I would rather make my own uneconomic green energy decisions rather than make what I see as a charitable donation to a for-profit company. Also, I own the freakin' solar panel, and if electric rates go through the roof, I benefit from already having paid for the solar panel.
If the power company offered me a deal "you could buy into wind power at 15 cents kwHr or you could purchase regular power at 10 cents a kwHr. Only thing is that we will guarantee the wind power for a fixed 15 cents for, say, 10 years while the 10 cent power you are taking your chances on natural gas getting scarce" I would snap that up. Their current wind power deal is merely writing a donation check to a for-profit company for green bragging rights.
But given the state of solar panels and members of genus Pan at the power companies, I am putting my money into more efficient wall warts. Tell me, where can I get a 1 watt phone answering machine?
Solar and energy conservation are not exclusive options. However, solar is about 5 dollars per peak watt, and that isn't even talking about finding a roofing contracter you can trust to put the panel up on the roof without introducing roof leaks. If you average 6 hours peak sun a day, you are talking 30 dollars per average watt for solar panels.
Suppose a transformer wall wart uses 4 watts and you can replace it with a solid-state ferrite switcher that uses.5 watts. It would take nearly 100 dollars of solar panel to do the same thing.
Oh, and about back feeding the line, you could probably get away with a small amount of back feed and just don't tell anyone about it. If you put up a serious solar panel setup and plan to back feed enough that the power company will notice, they get real, real huffy about that. In fact, they are supposed to by law buy back your power, but they really hate that. I was at an alternate energy fair where the local utility was touting their wind mills (you pay extra for the bragging rights of getting "green power"), and when I asked the utility dude about home solar panels and back feeds, he was telling me about all kinds of restrictions (two meter arrangements where you pay more for incoming and get back less on outgoing), and when I mentioned the laws regulating buyback, the fellow got in my face an I thought I would get punched. So much for committment to green power.
A PCI motherboard draws about 4 watts of power even when "off", but a phone answering machine draws as much.
I thought it to be a bargain to draw only 4 watts if I can could have instant boot-up, phone answering machine, and Fax, all on my PC. The idea was to enable hibernate and use wake-on-ring of the modem to bring the PC out of hibernate to answer the phone.
I have never, ever gotten hibernate to work properly on a Windows PC, and I have tried different PC's and motherboards. I think I could get wake-on-ring to answer the phone from sleep, which runs about 20 watts, but not from hibernate. In any event, recovery from hibernate does weird things like mess up the mouse driver, or put the video in some strange way, or just freeze the computer on restart.
Slashdot Mirror
They leave the West behind . . .
Ample? That was the whole point. There is nothing ample about Ms Jovovich, but there was especially nothing ample about what Ms. Jovovich was wearing.
Maybe the reason women vary all over the map in the ample department is that from an evolutionary standpoint, men can't make up their minds. When you think about it, ample is all about male sexual fetishism because it doesn't much matter when you switch off the lights. But ample and lots of curves suggests fertility in the style of -- hey, you have all seen those Stone Age "Ashtorath" clay figures looking like Rosanne's ancient ancestor.
But then not so ample and kind of skinny suggests a kind of adolescence -- OK don't get whigged out over the under-aged issue, but in ancient times women did all their child bearing before they reached modern legal age of consent because everyone was dead by age 30 anyway.
The topic, to get back on topic, is about C++0x, and any discussion of C++ promotes a flamefest. The parent post to this thread was flaming/whining/ranting about some aspect of C++ while neglecting to use any paragraph breaks, and people piled on to the parent post lecturing the dude/dudette about the need for paragraph breaks when posting.
I was trying to point out a parallel between C++ and Slashdot posting -- they are both feature-full and expressive at the risk of making a dumb mistake, and when the dumb mistake is made, people pile on telling you how dumb you are rather than considering whether C++/Slashdot could be more dumb resistant. And when a person dares suggest that something be made more dumb resistant, that suggestion triggers a response that a person is also dumb and a complainer to boot.
For more background on this, I defer to Cooper of "Inmates are Running the Asylum", anything by Donald Norman, and the infamous "Unix Hater's Handbook." There is this culture that any user-interface problems are the fault of the luser and that anyone pointing out such problems is a whining complainer. I was reading a book on Boost, and barely made it into the first chapter on auto pointers when my head started to spin on different flavors of auto pointer and how one worked in one situation but would result in a GP fault in another situation -- geez, Louise, in Java you just have garbage collection and don't have to wrap one's head around such issues. Of course, I am probably mentally deficient for not grasping Boost/C++ auto pointer flavors and I will go through life writing mentally-deficient Java.
I also end up looking silly by placing paragraph break HTML on other sites I post. Everyone around Slashdot (just like -- name favorite feature -- C/C++) programmers gets in the habit of using proper paragraph breaks, but it is entertaining to watch a noob make a long post, or perhaps an old timer who has passed some brain gas, make a honkin run-on post.
I guess Slashdot could fix this and accept whitespace as a paragraph break, but I guess that would violate some deep principle. But I am not blaming Slashdot -- it is deep in the software developer culture to accept things the way they are and then holler at the newcomers about their mistakes brought on by the fractured syntax. How do you think the C syntax has become so viral?
The one problem with life on Mars is the Gaia Hypothesis. We are supposing that life on Mars is hanging on by a thread -- the Gaia Hypothesis argues that life modifies its environment to perpetuate itself, and if there really were life on Mars, it would be pretty unmistakable.
Do you realize that the metric system is the last vestige of the French Revolution? Metric units -- heck -- bring back the metric calender! 10 day weeks and 10 month years. Bring back the Month of Thermidor!
Two difficulties with that. One is that the Moon is supposed to have formed from a collision with a Mars-sized body over 4 billion years ago. That collision is believed to have melted everything down to a 1000 km. The other is that the mantle has abundant oxygen and heat, and plate tectonics says it is in convective motion mixing up the different layers (although very slowly) -- one would think all the carbon is now CO2 or CaCO3 or something by now.
However, there is still the matter of diamonds, which are not believed to be dead dinosaurs (diamonds are aged 1 billion years or more, typically) -- they had to have originated in the upper mantle (about 100 km down) and had to have been brought to the surface rapidly (to avoid reversion to graphite, although you get graphite pseudomorphs of diamond -- clumps of graphite shaped like diamond crystals which were probably diamonds brought up too slowly so they reverted to the stable graphite form). While diamonds are rare, and the Kiberlite pipe eruptions that brought them up are rare and maybe date only to an earier geologic epoch, there has to be something to produce reduced carbon down below.
J F Kenney and his Russian associates believe that starting with FeO, CaCO3 and H2O (stuff not hard to find in the mantle owing to limestone and water being subducted down and Fe being brought up by mantle convection) you can end up with CH4 plus higher chain hydrocarbons. The argument is that about 100 km down is the only place methane, octane, and above can form is that the thermodynamics works at those temps and pressures and that the thermodynamics don't work for turning plant/algae material into oil in the traditional "oil window" of about 1-2 miles down.
So, there you have it -- oil is created from the same place an process as engagement ring stones, not only does oil not come from dead dinosaurs but from rocks instead (although the subducted CaCO3 could have its origins in biology of reef building), but that oil is not latent solar energy (in the form of sequestered biomass) but that oil is in reality geothermal energy (geologic raw materials brought together by heat-driven mantle convection and endothermic reactions driven by mantle heat).
If oil is really geothermal instead of solar in origin, one could consider and advanced technological culture with the capability of somehow using the environment of 100 km down as a natural resource, and of establishing a closed-cycle renewable geothermal based energy economy based on -- oil! One could sequester CO2 deep below and get back reduced carbon, all driven by geothermal power, which has its origins in natural radioactive decay along with the latent heat of fusion of iron in the core.
I mean think about it. A lot of the speculation about advanced energy cultures for the far future look outward into space and of tapping the vast resource of solar energy on the Earth surface, in Earth orbit, and beyond -- think Dyson sphere. Has anyone speculated, either in popular science writing or science fiction, about an advanced energy culture fully utilizing the energy resource within a planet?
You may say drilling or tunneling 10 km is stretching it not to imagine 100 km? But who is to say drill. Some MIT dude suggested using a million tons of molton iron (some grant proposal) to melt and sink its way all the way to the core to carry some kind of probe to find out "what is down there." Who is to say that some related scheme may be able to both bring materials down to the mantle (say CO2) and bring back materials (oil and gas) in a closed loop? I am not saying it is practical with today's technology, but it is not anything violating
The main problem I have with rocket-thrust landing in the manner of the DC-X is what happens if a rocket engine conks out? Also, when you are coming in for a landing, the aero resistance of your nearly empty fuel tank and light weight means your terminal velocity is low, but to save on fuel, you can't be hovering on engine thrust for very long. As you approach the ground, you need a very carefully timed burst of rocket thrust to come in for a soft landing.
You could say that the Soyuz lands vertically on rocket thrust. Yes, they come down mainly on parachutes, but in the last few feet they fire these solid-fuel landing rockets attached to the parachute lanyards to land softly -- they need this for a dry land instead of a splashdown landing. I heard that they have had some landing rocket failures and there are cosmonauts missing some front teeth.
Can a landing rocket be reliable enough, or are you going to crump a few space ships on landing?
The President should have looked down, scuffed his feet, and said, "Without going into an specifics about what Ms. Jones has complained about, there are certain things I have done as a man that I am not proud of, and in the interest of putting this matter behind us and moving the country forward, I am offering Ms. Jones X dollars in settlement of this matter." That is the formula -- admit to no specific wrongdoing but admit to being a "man behaving badly", pay the dime (or make the public non-confession), and move on.
Quoting the late John Belushi, "but nooooooooo!" As much as there was Kenneth Starr plus every partisan attack lawyer going after the President, he had his own gaggle of hyper-partisan attack-dog lawyers and spokes-mouthpieces working everything from the courts to the Sunday talk shows, who made him stand his ground. He just couldn't do "rope-a-dope." When he made the admissions and apologies they were too little and too late, but maybe not, as they finally gave him the public sympathy to stay in office.
If you are talking cross-platform, you could learn the Qt or GTK APIs and muck around with getting Qt or GTK installed on Windows for your 98% of the world according to TFA Windows users. Or you could do it in Java. It would pretty much Just Work on Windows, Solaris, Linux, and OS X. Yes you have to get Java installed under Windows, but in most academic computing facilities, sysadmins putting Java on Windows machines is pretty much a given these days. You ask for Python and wxWidgets and you get a sullen stare -- as to Qt or GTK on Windows, fugetaboutit.
Java has this BufferedImage thing that pretty much lets you twiddle the bits of an RGB image to make any kind of graphical display you want under all of the mentioned OS's.
You may be programming to the Windows API, I may be programming to the Windows API, but talk to your friend, your sister-in-law, and the guy down the street developing a Windows app, and they are all doing Visual Basic. You may be proficient in Qt, I may be trying to learn Qt, but talk to the people doing graphical Linux apps in Comp Sci departments, and they will all tell you Java.
Java has a layout manager you say, true, and if you pick up NetBeans expecting it to be VB and you plunk a JButton down in the middle of a JFrame and find you can develop a nice app where a single button takes up the entire main window, well welcome to the club. Once you get over the layout manager learning curve, you will be using it like an old pro, and besides, things like wxPython are layout-manager oriented, now aren't they?
I can see myself switching my development of scientific visualization apps from Delphi to Java, and when I have enough of my stuff running in Java I could say to myself, "Hey, I could just as well be running Linux as I am running Windows -- if I get the hang of OpenOffice or AbiWord, maybe I don't even need Windows." I don't see myself investing in the learning curve to develop for Qt or GTK, especially since I am not prepared to ditch my Windows machine and figure out how to get Qt or GTK up on Windows.
Where are the Linux desktop apps? In a 98% Windows world it will come from people doing Windows with a cross-platform layer -- it could be Java, it could be RealBasic, it could be Python -- I really doubt it will be Qt or GTK. It may also get a boost from Longhorn/Vista and whatever new thing Microsoft is pushing and people deciding to stay with Windows but to go with one of these cross platform things rather than the Microsoft API du jour. When people have enough of these apps, people will say "Hey wait a minute, I could just as well be running Linux."
In Java Swing, from a JFrame object, you say
myWidget = new MyWidget(this);
getContentPane().add(myWidget, BorderLayout.CENTER);
This is a blatent Demeter violation now, isn't it? Or is it only a Demeter violation of either the strong or the weak form of Demeter depending on whether invoking the function getContentPane() to return an internal object of the JFrame from a derived class is considered OK or not?
When structured programming came out, it was pretty obvious what kind of mess you had with GOTOs. When Glenford Meyers came out with his "coupling" and "module strength" criterea for function modules, it was pretty obvious what kind of mess you had with global variables or with bundling sets of variables into structs simply to cut down on the parameter lists of functions.
But with this Law of Demeter business, I am just not seeing what the fuss is, especially when you have to make programs a lot more complex with forwarding functions and convoluted applications of the Visitor pattern to avoid invoking a function on an object contained in another object.
I mean c'mon people. One of the big uses of object-oriented programming is GUIs, and everyone's GUI has this layered structure of frame object-graphics object-font object where you make some kind of call like myFrame.getGraphics().setFont(myFont). A lot of this is not that the Graphics object is some kind of mystery thing that you have to keep from users of the Frame object, but Frame and Graphics are most likely interfaces to some underlying implementation, and the reason for separate Frame and Graphics interfaces instead of giving Frame forwarding methods to every freakin' method of Graphics is to organize gobs of interface functions in some logical hierarchy.
The R 3350 TC put "blow down" exhaust turbines in the exhaust system, and it coupled those turbines to the main power shaft with automobile-style automatic transmissions.
The problem with turbo-compounding a car is that turbo-compounding works just great on airplanes that run on constant, high power settings, but a car cruises at very low power compared to the power needed to accelerate the car. The dominant loss in your car at highway cruise is engine friction, and going to fewer, smaller cylinders helps with highway mileage at the expense of pickup.
For example, the Ford Five Hundred and Freestyle use a small 6 cylinder to power a big car, and they use either a 6-speed or a continuous-variable transmission to have that engine turn over very few revs on the highway while allowing the motor to rev for pickup. The result is a car the size of a Crown Vic that gets better gas mileage than a Taurus (although owner gas mileage reports on these cars are all over the map -- they may be victims of the "hybrid" effect of suffering badly from lead-foot drivers). The result is also a car that every last auto columnist has been bitching, moaning, and whining is "badly underpowered" and "revs something fierce to accelerate" (I guess the Iraq war, ANWR, peak oil, etc. are of not concern to the writers shilling for cars). On the other hand, the Five Hundred is a big boat that gets great gas mileage if you drive it very gingerly, and I imagine Ford has the older-generation demographic sewn up with that car.
Anyway, one approach to hybrid gas mileage is a small engine geared tall with some kind of power booster -- simply improving the thermo efficiency of the cycle at part load won't do much because the add on power from the turbo-compound/steam engine/etc isn't needed at low load.
The steam engine could be a power booster -- you could have a boiler that could store enough steam to get a little boost. Problem is that while you could undersize the boiler, you couldn't undersize the condensor. If you are trying to get 3 litre power from a 1.5 litre engine, you have to make up fully half the peak engine power with the steam engine, and you are back to Doble/Bill Lear.
What about the condensor? Apart from the fact that it is a lot of plumbing that needs to not leak, you could say that it is no different than an auto engine radiator. The problem is that an auto engine radiator doesn't pass most of the heat -- most of the heat rejection is through the exhaust and only some heat leaks through the cylinder walls to be rejected by the radiator. A steam engine condensor has to reject all of the thermodynamic cycle heat at very low temperature differences -- unless you want to go open cycle like a Stanley or like a railroad steam locomotive.
1) What to you do/get for a GUI? Like Algol famously leaving "IO to the library" meaning they punted on IO and this really hurt Algol, Python leaves the GUI to the library, but there are two mainline choices -- TKInter and wxWidgets/wxPython. Does Lisp in a Box have no GUI, some homebrew Smalltalk/Squeek sandboxed GUI, or does it link to TK or wx?
2) Can you link to C/C++ to do drivers, low level stuff, reuse DLLs or SO modules, or is Lisp beyond connecting outside of the Lisp runtime? Even Java can reach outside with the much-maligned but really isn't too hard to use JNI -- I like JNI because it allows C/C++ to call Java and Java to call C/C++.
By the way, for all of the talk about .NET/C#, .NET has pretty easy ways for .NET to call down to C/C++, but for C/C++ to call .NET, you have to package the .NET module as a COM/ActiveX, and there are magic "metadata" keywords that you have to learn about that, and then you have to do all of this mystery "strong name" generation using a poorly understood process to register the .NET assembly in the Registry as a COM/ActiveX, and then you have to access a COM object through C/C++, which is far from pretty on account of Microsofts clunky use (hodgepodge of preprocessor, classes, templates, and opaque auto code generation) of the C++ language (MFC? ATL?) to do that. For all of the clunky nature of JNI, you can go both ways between Java and C/C++ without any of that Registry song and dance.
One of the strengths of Python is the ability to switch to C/C++ as the need arises, although the C/C++ side is still a little opaque to me because you have to go through some dance to represent Python's dynamic object and method name lookup on the C/C++ side. Anyway, I think part of Python's popularity is that you can drop down to C/C++ if you hit performance or feature roadblocks and the perception is that Lisp is like some Smalltalk implementations -- so highly sandboxed to prevent this sort of thing. What is the story on this?
Turns out OS X supports transparency, so the SetRGB() call of a BufferedImage needs the high byte set to all 1's (| 0xFF000000 you pix values) so that an image is not invisible -- the transparency byte is ignored on all other systems. This is kind of stupid -- usually "unused" fields are set to all zeroes, and all zeroes should mean a default of opaque instead of a default that the image is completely transparent (i.e. doesn't show up).
The other problem I ran into was that I could copy said images to the system clipboard on every system except for OS X. That was a fine how-do-you-do as the Mac is supposed to be so Java friendly according to the hype and the Mac is supposed to be so graphics friendly. Windows and Linux, supposedly much less Java friendly but for different reasons, didn't have any trouble. Turns out that this was a known bug, fixed by a revision to Java 5 (1.5) that came out a few weeks ago.
Oh, and the Mac had some different font width for label widgets compared with everyone else that required some tinkering with preferred panel sizes in the layout manager.
By the way, the biggest Java "platform" incompatibility is not between Windows-Solaris-Linux but between Java versions -- it seems you need to compile separate binaries for Java 4 and Java 5. This makes Java a lot like Delphi in that if you release visual widgets you have to have binaries for Delphi 3, 4, 5, 6, 7 or require everyone to use Delphi 7. Windows, bless their black little hearts, is famous for a great deal of upward and downward compatibility of Windows binaries among Windows versions apart from some problem with button colors if you used the wrong color defaulting mechanism.
If we are on the subject of ancient writings, myths, or oral traditions, the bit about the planet Venus being ejected as a flaming comet from Jupiter in Biblical times and showering manna down on the starving Israelites in the form of cometary hydrocarbons no longer interests me one way or the other.
The bit about Velikovsky that has my attention is "The Greek Dark Ages Never Happened." See http://en.wikipedia.org/wiki/Troy or http://www.varchive.org/dag/index.htm.
The standard chronology is that you had the Trojan war, which supposedly came at the end of the Mycenean Greek culture about 1200 BCE, that culture collapsed, and then you had a reemergence of the Greek city state culture around 700, the height of Classical Greece around 500, the Peloponesian war which weekened them, the Macadonians taking over, Alexander, the superceding of the Alexander empire by Rome, etc. The interval between 1200 and 700 BCE is regarded as a Dark Age, a much deeper, darker stop backwards in the advance of culture than the supposed European Dark Age of the Middle Ages.
Jerry Pournelle writes often about the Greek Dark ages -- mysterious, unknowable things are always kind of cool, and ancient disasters involving regression from a more advanced culture, a kind of Atlantis, are even more cool. One of the big things supposedly lost in the Dark Ages was writing -- Homer's Iliad makes reference to writing borne by messengers in such a way to suggest that Homer didn't understand the concept of writing (Homer, by the way, is believed to be one source or perhaps a composite source, of an oral tradition). That a civilization and trading culture collapsed and regressed to cave man conditions so utterly and completely that they lost writing is fascinating -- if we had an atomic war or a global warming-triggered climate flip back into an ice age, you would think we may lose our automobiles, and central heat, but would hang on to painless dentistry through the cultivation of coca and poppy and would still know how to write.
While Velokovsky fascinates with cool tales of flaming comet planets, he takes a dump in the punch bowl of cool things to spin yarns about by claiming the Greek Dark Ages never happened, that the whole thing is a goof up from trying to connect the chronology of dynasties of Egypt to the Greeks? Anyone know anything more about this?
The people using Delphi out there are simply desperate to find anyone else who knows anything about it to help out.
Part of the problem may be that VB enables people with not a lot of experience at program design in any language to generate these quick-and-dirty applications that morph into critical applications in some business that some person then has to maintain. I guess the ability to work with C++ is regarded as a kind of "union card." The C++ qualification weeds out a lot of people. But this idea of choice of language to control programming is not a new one: currently the idea is C++ as a kind of intelligence test; some 20 years ago the idea was forcing programming in Pascal as a kind of test for patience or perhaps obsessive/compulsive neatness as a qualification for good programming.
I also think there is some fault in Visual Basic itself. If you look over the GoF Design Patterns, Visual Basic is pretty much stuck on the Mediator Pattern, which may tend to flatten program structure and lead to spaghetti code.
The basic pattern in GUI programming is that you have a widget, say a button, that responds to user input, say the user clicking the button with a mouse. You then have some "business rules" or "program logic" that is supposed to be invoked in response to the button click. The Observer pattern states that you can register some Observer object with button object to be activated in response to clicks. The Mediator pattern states that if you have a collection of button objects, you register a single observer, the Mediator, with all of those buttons, to contain all of the "program logic."
In Visual Basic you have a form, and you place widgets on that form, and when something happens in those widgets, you have one or more functions for the main form object that respond to those events. Some of the widgets require a lot of hand holding -- they fire off a lot of events for a variety of conditions and the main form code has to figure out what to do in response. The consequence of this is that the main form can get complicated very quickly with numerous event functions and numerous state variables to keep track of what event got fired and which mode one is in.
In Java, you have these action listener and event listener interfaces that you can register with widgets that signal mouse clicks and other actions. You have the flexibility of having a separate listener object for each widget or having the main form Frame or JFrame object be the listener by implementing the listener interface. You have the design choice of having a flat hierarchy or of having a collection of objects respond to different events. Whether people programming Java are able to use this flexibility effectively is another matter, but Java programmers are not forced to stuffing all of the program logic into the main form Frame object.
For example, if you have a widget that fires a large number of events and requires specific responses to all of those events, and implementing a widget that way may be a way of make a widget very flexible and easy to customize, you could write a specific Java listener class to not only respond to all of those events but also to implement some filtering logic on those events. In Visual Basic, my understanding is that all of that filtering logic has to go into the main form and that there isn't a simple and obvious way to parcel off that logic to another object. It is for this reason that the language snobs consider that for all of its faults, Java is "industrial strength" while Visual Basic remains a "toy language."
Now there are language snobs who fault Java for not having closures, the ability to register functions with classes or objects on the fly, for a clunky way of inspecting an object for functions and function signatures called reflection, and so on. But the number of people who understand how to use those features effectively is more limited. There are a lot of people who know how to use objects effectively and are frustrated by VB 6 having a crippled object system.
Yeah, yeah, ActiveX on IE as a security problem, and all of that. I like ActiveX, and it is a very powerful cross-language facility, but you are solidly restricted to Windows. This thing sounds like an other-than-Windows other-than-Java cross-platform way of doing graphics which is pretty interesting.
Yeah, but how much of these libraries do you need for a game?
Java on the other hand is Java. Unless you want to go the SWT route, if you are using Swing/Java2D, all the stuff is there.
It depends. If you are a developer and considering Java to support a shrink-wrap product, one is perhaps fussy about the user experience. If you are developing apps for use by students in your own class at the U, for a small community of colleagues, or even developing a commercial app for a niche user community (consider all of the butt-ugly Visual Basic apps), precise look-and-feel is probably a minor issue.
The reason I am suddenly impressed with Java is that after 10 years of looking at it (OK, I developed some non-GUI apps in it), a requirement came along for an x-ray image viewer. I was under the impression that displaying a gray-scale image from computed pixels was a real big deal like it is in the Windows API with CreateDIBSection, but it turns out that Java now has this BufferedImage class that does the job simply and cleanly. Create a BufferedImage, set pixels wholesale with setRGB(), and drawImage() the BufferedImage to a Graphics object in the Swing paintComponent method. It also turns out that copying that image to the system clipboard is equally simple.
BufferedImage may have been around for a while, and imageFlavor (for the clipboard) has been around since Java 1.4, but these things haven't been around since day 1 sometime in 1995.
Anyway, the app my colleague and I coded up in a week (there was also a learning curve on layout managers) worked just fine on Windows and just peachily on Debian and Solaris. Oddly enough we hit some roadblocks with OS X, that product that is supposed to be much more Java, image, and clipboard friendly. One issue was setting the transparency values in setRGB() -- not an issue on OS's which don't think there is a need for transparent images, but our x-ray picture was invisible initially on OS X; a second issue was a mangled paste into a native OS X app -- turns out Java 1.5 Release 3 for OS X came out last week Monday PM and the version of Java 1.5 I was using was installed last week Monday AM.
Writing an image viewer where you compute your own pixels has always been considered an advanced topic in the GUI business only of interest to game programmers, and after two weeks I have such an app running on 4 OS's. It dawned on me that this could be the death of the Windows monopoly.
Yes there are hangups. My one pending UI gripe is the Java JFileChooser "Files Open" dialog widget. On Windows, it takes seemingly forever to come up the first time you do a Files Open -- on Debian it just pops right up, so much for Linux having second-class Java VMs. The Windows Files Open has text selection defaults that make it efficient to navigate directories and select files with the keyboards while the Java Files Open is just plain stupid and maddening.
But consider my 10 years rule. Java came out in 1995. Since there were other ways to do GUI, people didn't put up with Java Swing shortcoming and walked away. But 10 years later, I am telling you that Java Swing has had perhaps enough improvements made (there is an enormous amount of help on the Web getting up the layout manager and other aspects to Swing learning curve) that it deserves another look.
Yes, Java-the-platform is in a way its own OS, but on the other hand, Java sits on top of whatever native OS. How about thinking about it this way: I don't have to clean format my hard disk and get all new apps if I install Java. Java may be its own API on top of an OS, but not only is it not tied to a particular native OS, it doesn't whine, fret, or threaten if it is not your exclusive API, and it plays with the other kids.
I can have Java, Python, Ruby on Rails, and even Mono coexisting in the same partition on top of the same OS. They can interoperate through files. While this is not always politically correct from SUN's way of thinking, I can also interoperate Java and C++ in both direction (Java calls C++ or C++ calls Java) using the JNI. I can interoperate Python and C/C++ and on Windows I have Python hosting ActiveX controls to do visuals which in turn can call Java signal processing plug-ins -- it is sort of like the Far Side Cartoon of the cops who find a man swallowed by a crocodile in turn being squeezed by a boa (perhaps a python?) and are required to sort that mess out.
While there is a thing such as "Internet time", I think there is a 10-year rule in effect for the maturation of major pieces of software. From the time of Windows introduction in the mid 1980's to the time of Windows 95, COM, and ActiveX, when Windows acquired the features that made it what it is today was about 10 years. Java is around its 10th anniversary in 2005. Maybe Java/Swing has acquired enough features and enough performance enhancements to be taken really seriously.
Um, volt-amps don't have units of watts. Or not if you wanted full credit on the exam problem I gave last week. A generally followed convention is that volt-amps is the voltage times current without taking into account the phase, watts is the in-phase voltage times current, and VARs (volt-amps, reactive) is the out-of-phase voltage time current. Watts plus VARS add up to VAs as vectors.
As to the safety issue, I haven't look into this, but suppose you feed your line with a solar panel to line inverter specially designed for grid-connect applications -- say an inverter that is excited from the power line and if the power line goes dead the inverter stops feeding power. The solar power hardware people may indeed have such inverters, and I even heard of solar panels with that sort of thing built right in.
The power companies could say, "yes, you can net meter and back feed, and you can only back feed up to a certain amount, and you have to use an approved inverter, and there is a nominal inspection charge for us to come out and give the OK to your equipment." But that is not what they say. They tell you you need an expensive install of a dual meter setup and that you get paid a lot less for the power generated than for the power consumed, which is not compliant with the net metering laws for consumers. When you politely ask about the net metering laws, they do the "CEO yelling at the TV reporter and threatening to shove the camera guy routine."
As to "why should the powercco give you the storage facility for free whenever you wan't it" the answer is because it is public policy and the law that's why. The same reason why the utilities have a monopoly and can get reg approval to jack up your electric bill to pay for their nuclear power plant problems or to pay off new power plants when you are using the same amount as you always used.
Why is the public policy and the law that way? Because elected representatives were convinced it was good public policy to have the regulated-monopoly power companies in effect subsidize the development of renewable and solar power for a whole variety of reasons, that's why.
The irony of this is our local powerco is putting up these wind mills and asking consumers to voluntarily fork over extra bucks to the power utility simply for the bragging rights of getting wind power. If you want to take your extra bucks, and instead of forking it over to the utility but build your own solar panel, and if you politely ask about what you need to comply with their safety regs and what they need to do to comply to laws on the books, and you do this asking at the Wisconsin Alternate Energy Fair where this same utility is crowing about how green they are, you get a burly utility representative in your face in a primate-species threat posture.
Why would I want to put up my own solar panels instead of forking bucks over to the utility to say I am green? It is an "ownership society" thing. I would rather make my own uneconomic green energy decisions rather than make what I see as a charitable donation to a for-profit company. Also, I own the freakin' solar panel, and if electric rates go through the roof, I benefit from already having paid for the solar panel.
If the power company offered me a deal "you could buy into wind power at 15 cents kwHr or you could purchase regular power at 10 cents a kwHr. Only thing is that we will guarantee the wind power for a fixed 15 cents for, say, 10 years while the 10 cent power you are taking your chances on natural gas getting scarce" I would snap that up. Their current wind power deal is merely writing a donation check to a for-profit company for green bragging rights.
But given the state of solar panels and members of genus Pan at the power companies, I am putting my money into more efficient wall warts. Tell me, where can I get a 1 watt phone answering machine?
Suppose a transformer wall wart uses 4 watts and you can replace it with a solid-state ferrite switcher that uses .5 watts. It would take nearly 100 dollars of solar panel to do the same thing.
Oh, and about back feeding the line, you could probably get away with a small amount of back feed and just don't tell anyone about it. If you put up a serious solar panel setup and plan to back feed enough that the power company will notice, they get real, real huffy about that. In fact, they are supposed to by law buy back your power, but they really hate that. I was at an alternate energy fair where the local utility was touting their wind mills (you pay extra for the bragging rights of getting "green power"), and when I asked the utility dude about home solar panels and back feeds, he was telling me about all kinds of restrictions (two meter arrangements where you pay more for incoming and get back less on outgoing), and when I mentioned the laws regulating buyback, the fellow got in my face an I thought I would get punched. So much for committment to green power.
I thought it to be a bargain to draw only 4 watts if I can could have instant boot-up, phone answering machine, and Fax, all on my PC. The idea was to enable hibernate and use wake-on-ring of the modem to bring the PC out of hibernate to answer the phone.
I have never, ever gotten hibernate to work properly on a Windows PC, and I have tried different PC's and motherboards. I think I could get wake-on-ring to answer the phone from sleep, which runs about 20 watts, but not from hibernate. In any event, recovery from hibernate does weird things like mess up the mouse driver, or put the video in some strange way, or just freeze the computer on restart.
Anyone out there have better luck with hibernate?