You see, it is not the two month's salary. You should pose the choice, "sweetie, you would like this nice glittery stone OR whould you like 100 shares of stock in Procter and Gamble, which can be purchased without commission in a dividend reinvestment plan, which makes you wealthier anytime anyone anywhere in the world washes the dishes."
My wife bought the P&G thing, and we still haven't bought the stock because she is trying to "time" the market.
Does Mr. Cutler have any public rants against Unix? What is it about Unix that he didn't like? Was he a contributer to that "Unix Hater's Handbook" or was he a freelancer with his own take?
DOS had the very easy install -- you just popped in one disk and booted and from that one disk you could format and partition a hard disk and put a DOS image on it.
Windows had a very difficult install - a whole raft of disks in the pre-CD ROM days, and I believe there were some configuration issues. DOS was based on the principle that it was very simple to boot, very simple to write to-the-metal games programs right to the graphics card. Windows was based on the principle of getting it pre-installed on computers so users wouldn't even have to know how to install it.
The DOS/Windows dichotomy was such that for the longest time games would be DOS based. Microsoft made a big push to get the DOS games over to Windows with the WinG and later DirectX initiatives. There is a lot of Windows stuff to make it 2-D game friendly -- ScrollWindowEx() (hardware assisted scroll), CreateDIBSection() (allowed porting stuff that wrote to the frame buffer to Windows), and IDirectDraw::WaitForVerticalBlank() (allows flicker-free, tear-free scrolls and screen updates). I have a serious case of Windows lock-in because I depend on these calls and can't find them on other systems.
Steve Jobs got kicked out of Apple, he looked at Unix and saw its potential for layering on a GUI and giving Apple a kick in the backside, started Next, which was to do just that, Next foundered around, and things have come full circle where Apple brought Jobs back along with his NextStep and made it OS-X.
There is a community of people who think that some kind of 'nix underneath with a glitzy UI on top to satisfy the folks who require that kind of thing (OS-X and to a lesser extent the Linux distributions teamed up with either KDE or Gnome) is the best of both worlds. It is an article of faith that 'nix is the best foundation to build whatever kind of point-and-click thing is necessary to win over the end user, and even if Microsoft were not the Evil Empire, the Microsoft OS's are considerate second rate compared to Unix.
My understanding is that Bill Gates had not only looked at Unix but had a Xenix offering for the PC at one time or another (there is a James Burke style of "connections" to SCO and the SCO/Linux dustup). Microsoft put its toe in the Unix waters, but its customers were not clamoring for Unix on the PC (this was long before MS was the Evil Empire). Given their license arrangements, they could have gone the route of a 'nix for their advanced products (Windows 95 and beyond) but for some reason decided they weren't interested.
Was their some technical reason for Microsoft going with Unix early on, or was it a matter of preference or prejudice?
A friend of mine looked into the Mac Mini, and he rolled his eyes when he found out it had two USB jacks.
The whole idea behind the Mac Mini is that people already have monitor, mouse, and keyboard with their PC, so they just plug in a Mac Mini and switch over to Mac and OS-X. OK, the monitor plugs into the video port, the mouse and keyboard each plug into a USB port, so where do you put the printer?
I suggested, "get a hub" and my friend retorted that was the whole point. If the freakin' thing came out with 4 USB ports, he would get one, he would tell his friends to get one, but now he is just rolling his eyes.
Yeah, yeah, the thing is only $499, but that is just a gimmick because you have to accessorize it to do anything useful. It is kind of like the old Detroit gimmick that you could get a Chevy for some good price but that was without A/C, P/B, P/S, and a bunch of stuff to make the care usable (you could sweat out the no A/C, but there came a point when car designs were such that the "stripper" car that the dealer had one-of on his lot to get newspaper ad customers in the door was undrivable without P/B P/S).
So as to your notion that Apple has engineered the Mini to stay in certain market segments, you have a lot of agreement from my friend.
Coprocessors and "array processors" -- feh!
on
GPU Gems 2
·
· Score: 2, Interesting
Back in the day of 8086, 286, and even 386, I did stuff with audio processing on a DSP coprocessor board, machine language programming and all. It was good for its day, but the DSP coprocessor board became obsolete, the relentless march of 486, Pentium I, and Pentium II went on, and it just didn't pay to fiddle like that, especially when there were was good code optimization with VC++.
I guess we are stalled with a minor improvement with the Pentium 3, and step backward per clockrate with the Pentium 4 and limited availability of compilers that know how to optimize for it. With the pending multi cores per chip, I might play around with balancing DSP tasks between processors and threads, but I don't think I am going to muck around with some proprietary GPU or DSP -- life is too short and the "time to market" window is too narrow.
One of the main features of C is interchangable int and void*. Somehow I think you are better of biting the bullet and making int 64 bits just like when going from 16 to 32 bits int was so promoted. That way you only need to rewrite code where structs require 32-bit fields.
Linux can be happily 64 bit, and Windows may attempt to be 64 bit, but what are people going to do about Java?
Java is supposed to be platform independent, but the implicit assumption has always been a 32-bit platform of one sort or another. Yes, Java can run on a 64-bit processor, but the int is still 32 bits unless you want to change the behavior of an awful lot of Java code.
So will there be two Java's or are they going to come up with some kind of clever 64-bit Java extension or what?
Oh, and as to the comments that it takes a shockingly big Word document to bust a 32 bit address space, the big address space is not for Word, it is for video. The change to 32 bits and faster processors made CD-quality audio pretty much universal on desktop computers, but HD-quality video is not there yet. Sure you can stream from files or segment memory, but 4 Gig is still constraining with regard to high definition video files being handled in linear address space.
Most of your car purchase money goes into someone's health insurance. The difference between a land yacht and just basic transportation is not reflected in the sticker price. When someone buys the GM land yacht, they are keeping some old guy who worked for GM off Medicaid (Medicare doesn't pay for everything and you still need private insurance).
Those hybrid folks have some kind of MPG meter, although my wife's Chrysler has a similar readout.
For ordinary cars, you have the tank fill method, but I wonder if that is only accurate to about half a gallon. I try to get the same gas nozzle and park the car the same way because I figure if the car is tilted one way or the other you will get a different air bubble in the tank. Also I figure the shutoff is the same sensitivity. Even so, on a 10 gallon fill, my guess you are accurate to 5 percent (about 1 MPG in your case).
Do you get pretty much 25 MPG every time, or do you get 23, 24, 26, 25, 27, 24 -- that is more my experience (without changing any variables)?
The Green Hybrid site is slashdotted, but I wonder if they have data on non-hybrids -- comparing the Corolla or Camry against the Prius, comparing hybrid and nonhybrid Civics. In addition to the real-world data on the hybrids, it would be interesting to see real-world numbers on everything else and correlate it with EPA. That way I can see how I am doing with respect to drivers with other driving patterns and make some guesses regarding my real-world mileage from the EPA numbers of a car I am thinking of buying.
The other thing that could be done is that all cars have this OBD-II plug right under the steering column. Someone should come up with some kind of data logger that you could drive your car around -- to work, to Grandma's, whatever your personal drive cycle is -- and compare predicted mileage against actual mileage and make predictions of what you would get in another car model. I understand that if you are buying an 18-wheel truck, you can get fuel use predictions based on the truck route you want to cover, and this is a big, big factor in the economics of a truck purchase. I guess the technology exists in principle to do this with cars, but the usual people who sell cars don't even want you to know this.
I work pretty much in the 2-D world -- text, lines, textures, images, pixmaps. The basic arrangement with 2-D is that you have something called a graphics context -- this is pretty much the same across Windows, Java, Mac -- which allows you to do the operations I have indicated. If you want to output to a window, you either ask for a graphics context for that window or use the graphics context they give you when you get a Paint event. If you want to print, you ask for a graphics context to a printer and render to that.
I work in Windows but have read what I could about Quartz2D and Java to try and keep current. My understanding of Quartz2D is that they try and keep the graphics context coordinate system not bound to screen pixels so you have a uniform graphics context whether you are printing, plotting to the screen, or plotting to a PDF file. I am told that Quartz2D is not really PDF as much as a graphics context that tries to be as compatible with PDF as it can. Under Windows, a displayable object has to have different rendering methods to deal with the quirks of how graphics context coordinates vary between printer, screen, and metafile, while my understanding is that under Quartz, they try to make everything uniform, adding a layer to do all the scaling, so you need only one rendering method.
I don't have any 3-D experience, but I heard the buzzwords about triangles, vertices, shaders, and graphics pipelines. If Longhorn is going 3-D, is this what we are going to work with? If so, it will be radically different than Quartz, which appears to be pretty much conventional fonts-lines-shades-images-pixmaps 2-D graphics only done more uniformly and perhaps elegantly than other 2-D graphics.
And if Longhorn is indeed going 3-D, what am I going to have to learn to work in that world?
What's wrong with the.NET layer? Dunno -- the download for the target machine or having to distribute the.NET runtime with your app, features from the Win32 API missing from.NET. For example, Matlab makes a big deal that they can interoperate with all of the component software standards out there -- they do ActiveX, they do Java and recently they do Java widgets in Matlab figure windows. Do they do.NET widgets? Nope. Really don't know what is wrong with the.NET runtime, but I just kind of get this feeling that it hasn't caught fire, and if Microsoft has such confidence in it, why is it a download rather than coming pre-installed from Dell?
As to VC++/MFC being a substitute for Visual Basic, MFC may be a fine, fine framework and a great use of the features of C++, but it isn't even close in the VBish RAD department. The old Visual J++ was a Visual Basic substitute; VB.NET and C# are definitely VBishly RAD, although on the.NET runtime.
Delphi is a "buy a new tool" as well as a new language, but if you are developing for Win32 and using the Form Designer and a little bit of Pascal to glue things together, it is a pretty good VB replacement.
I guess the TIOBE survey has to be run through the filter that it may reflect what PHB's think they want to hire people for, but it is one of the few public data points we have.
What I find interesting is that Delphi is roaring back, and it is not just a temporary blip caused by the new Delphi 2005. My guess is its not so much that people love Pascal but some big problem with Microsoft's.NET and their stand on Visual Basic 6.
If you want to do what Visual Basic 6 used to do (and still does, thank you), which is support plain-old-Windows without the.NET layers, and do it in a VB sort of way with forms and widgets and property editors, it seems that Delphi has become the only game in town (although I should look into RealBasic). Yes, there are the legacy/spaghetti VB apps, and yes, there are VB coders who don't want to learn, but I think MS is missing the boat. It is not that people don't want to learn VB.NET: what they want is a VBish thing for the old Win32 API without the.NET layer, and they are willing to learn a new language to get it. The problem is not so much with VB.NET as it is with just plain.NET.
You have gotten a lot of responses, some of them diplomatic, others unnecessarily blunt, telling you that you don't know what you are doing if your C++ compile times are that long, and suggesting tweaks -- better make files, pre-compiled headers.
Well, you and your colleagues could bone up on the skills to compile large C++ programs, but you could also hire some people who are experts at IBM OS/360 JCL while you are at it.
Anyone here remember JCL? The idea was that you had a "separation of concerns" between abstract operations on files (Open, Read, Write, Seek, Close) and a whole panoply of file-specific parameters (disk, tape, random-access, sequential, block size, buffer sizes), and that you only had the abstract parts in actual compiled code while the connections between the abstract parts and the OS went into a kind of shell script that specified these connections. Then these guys in New Jersey came up with this Unix thingy which didn't have any of this stuff. Unix "solved" the problem JCL was addressing by not having that problem in the first place.
My point is not that Java is "better" than C++ -- there are many criteria in what language is suitable for what kind of problems -- but that C++ is so ancient history even within its own problem domain. Make files? Pre-compiled headers? The problems those features were meant to solve were obviated by Wirth and his gang back in the Modula 2 days just like Unix made JCL go away.
The suggestions you have rolling in to speed up your C++ compiles are all well and good, but the fact that people are calling you stupid. How about another analogy. The F-104 has some bad user-interface issues along with slim stall margins on landings. It killed a lot of pilots in its day, and the pilots who survived would all say that the pilots who didn't lacked "the Right Stuff", but I don't think there are any Air Force planes these days that are quite so unforgiving.
What problem does this aspects mechanism solve that isn't handled by a proper implementation of multiple inheritance. It seems you are trying to bolt on the same functions to a collection of classes (or perhaps methods on those classes) without having to make them part of the root class.
I don't have much experience with multiple inheritance, but I understand that introduces some pain of its own. With inheritance, you have some core functionality of the base class and then you add on functions. With mixins (adding a clump of functions), I suppose you could do those with multiple-inheritance, but with using object composition as a substitute, you have to write all of these stupid little forwarding methods (or confess to the sin of violating the Law of Demeter by having too many levels of the dot notation).
So what you are saying is that they reduce coupling and dependencies among the different piece parts by having them work independently and by having them reinvent a lot of stuff, and the bloat of duplication is much easier to deal with than a unified design.
I also suppose that the best model for software reuse to date is where features are part of a language or a standard library (STL lists, Python hashlists) rather than some dude in the Visual C++ parser team developing a regular expression tool and supporting its use across Microsoft.
How does a piece of bloatware like Windows XP even get written (disclaimer -- I am generally comfortable with Windows and not a flaming anti-MS person, but a person has to admit that XP is, well, pretty big)?
I don't mean from a process standpoint and PERT charts and waterfall diagrams and all of that. And I can understand things that at least started out as one-man operations for gifted persons (Cray computers, Linus and Linux, etc). But even if there is a head XP guy who has a roadmap, the dude has to farm out much of the coding to a vast army. How is that army organized?
Or maybe I have it all wrong -- maybe the Windows kernel is pretty compact and that much of Windows is built up out of apps -- you are responsible for this app, you are responsible for this other one. Even if the coders all work from detailed specifications, someone has to write the specifications, and I don't know if the specifications for that much code can fit within the span-of-control of one person let alone a small committee.
I will concede that Windows is a POS, but how does such a POS even boot?
I am pretty sure Gene Granz said, as spacecraft electrical systems were shutting down right and left, "OK people, let's figure out what's still good."
I use this saying when helping someone with a very initial debugging and it seems it is not possible to get a program running without crashing or lockup.
By the way, NOVA did an Apollo 13 show with extensive interviews of Jim Lovell and Gene Kranz -- much like the Beatles documentary that gave a lot of airtime to George Martin -- and I never saw the need to see the Ron Howard film, although Harrison Schmidt, the geology PhD who went to the Moon, tells audiences of his talk to go see it. Who am I to argue with a guy who actually went to the Moon whether the Ron Howard film is worth seeing?
I poked around the RealBasic site and didn't get too many leads on this.
What does it take to write extension modules (either GUI or non-GUI) for RealBasic? Do you write them in RealBasic or is there a way to write them in C/C++?
If you do a GUI extension module, does your application become dependent on Windows/OS-X/Linux or does it have some kind of abstraction of the GUI and abstraction of the graphics drawing surface accessible from C/C++?
This was some 20 years ago, but a colleague (we are not cops, just EEs) brought a Taser to a presentation -- his point that the Taser didn't really "stun" -- it just plain hurt. He thought the cops would find it wouldn't work, but would be mighty handy with "suspects in custody" if you know what I mean. He was warning us the Taser would be become the modern version of "rubber hoses."
Well anyway, the idea that you could have a weapon with two settings, stun and kill, gee, where did the cops get the idea that this would come in handy? Turns out that it is hard to stun someone without killing them. The old detective movie cliche of stunning someone by hitting them over the head? An MD writing in TV guide told readers that "if you hit someone over the head, most likely you just hurt them and make them mad. If you hit them harder to knock them out, chances are you kill someone from a hematoma."
I have seen films where they Taser a volunteer, and they start convulsing or they just faint and tip over and hit the mat in the gym they are doing this with a thud. I have seen a recent episode of "Cops" where they Tasered a whole bunch of different people, and it didn't seem to do anything. I suppose the electric shock causes intense, sudden pain, and that can cause someone to collapse, just like being shot with a bullet, only there is no tissue damage so the person doesn't bleed to death, but does it really work? I suppose a bullet doesn't always "work" in that a person can be fatally wounded but not always drop dead right away.
Anyway, I was the only person at the meeting willing to try the Taser (it was a model with two electrodes sticking out, and it didn't have the darts for remote action). I was going to zap my forearm and I was quickly advised "better try it on your leg. So I hold the Taser up to my leg and press the trigger. I was pretty embarrassed because I yelled out "Jesus Christ!" I was more startled than anything else, and it hurt real bad, but not the kind of hurt of someone drilling into your teeth -- more like the worst kind of muscle cramp -- it really hurt but I felt like my leg muscles were seizing up.
So what does a Taser do? Does it really knock a person out like on Star Trek, or does it merely cause a person to take the Lord's name in vain? Have they made the Taser more powerful in the last 20 years? More maybe because my finger was on the trigger, I let go when it first started to hurt, and a person has to deliver multiple jolts to get someone to collapse.
OK. A lot of the arguments are "you are not counting the utility power in the MPG" vs "I am getting my kWs from a solar cell."
I think that getting car propulsion from plug power opens up a whole lot of possibilities even if our current electric generation isn't that green, and people ranging from Burt Rutan who misses his EV-1 to an article in Tau Beta Pi magazine are talking about hybrids to get vehicles that are perhaps 80 percent plug power and only use gasoline to solve the battery range problem.
Missing from the proponents of the plug-power Prius, Burt Rutan, and Tau Beta Pi magazine is discussion of the kWHr cost of battery.
Let me just pull together some numbers, and people can tell me if I am at least in the ballpark-order-of-magnitude. A 10 kWHr lead acid battery pack (i.e. the EV-1 power pack) costs $3000 and allows for 300 charge/discharge cycles (you get many more cycles with shallow discharge, but I am not convinced that you get more total kWHr through the battery). A 10 kWHr pack delivers road power to replace, say, 1 gallon of gas. This is saying that the battery alone is the equivalent of gas at $10/gallon, and if $50/bbl oil is $2.50/gal gas, $10/gallon gas is $200/bbl oil, and at $200/bbl there is all kinds of crud (heavy oil, tar sands, oil shale) that you can turn into gasoline.
Batteries cost money, not just in initial cost but in operating cost because batteries wear out. There are batteries that have more cycles than lead acid, but they also cost more. The Prius uses a NiMH battery, and the consumer NiMH's are said to be good for about 500 cycles. The Prius battery holds much less than the EV-1 battery, and I heard figures of 7000-8000 dollars to replace a Prius battern.
That a battery only does so many charge/discharge cycles has to do with physical chemistry -- it is like you are electroplating and de-electroplating over an over again and how many cycles will a battery plate hold its shape and not crumble? Making oil from shale or tar sands is expensive, not because the operations are pilot-scale but because you have to process a lot of material. Storing energy in a battery is similarly expensive, not because battery production hasn't been ramped up but because just plain wear of the battery elements.
The reason for the Prius (or even Insight) style hybrid is that that batteries cost too much for economical substitution of plug power for gasoline, even for very expensive gasoline. The idea in these types of hybrids is to find where gasoline is used in the most wasteful modes (idling, coasting, stop-and-go-running) and to substitute those modes with battery-stored electricity. You know, if 80 percent of the gasoline does 20 percent of the work, replace that 20 percent with electricity, and even if battery electricity costs 4 times that of gasoline it works out.
Well, maybe it works out. The proof is whether you run a Prius (compared to a Corolla or a Camry, not an SUV!) until you wear out a battery pack and weigh the replacement cost of the battery pack against the cost of the gasoline difference. Even at $2.50/gallon gas I think I am better off with the Camry -- $5/gallon gas may be the crossover point.
But the weak point about batteries, whether you are talking EV-1, battery backup off-grid solar power, or a plug hybrid is that the battery itself is a consumable instead of a durable good, and electricity out of batteries costs several multiple of the cost of gasoline except in those "corner cases" where gasoline is used inefficiently compared to the electric power substitue.
Getting to be a whole lot like PBS Pledge Week
on
EU to Ban Macs
·
· Score: 4, Insightful
One April Fools stuck in among the regular stories would be funny, but I am looking for the one regular story stuck in among the April Fools.
It is a lot like a PBS station during Pledge Week. All of your regular programs are cancelled or moved around, there is maybe one program worth watching among the nonstop Daniel O'Donnell concerts, and that one program is chopped all to heck with those stupid people shilling for donations who don't know how to act in front of a camera.
Slashdot Mirror
My wife bought the P&G thing, and we still haven't bought the stock because she is trying to "time" the market.
Does Mr. Cutler have any public rants against Unix? What is it about Unix that he didn't like? Was he a contributer to that "Unix Hater's Handbook" or was he a freelancer with his own take?
Windows had a very difficult install - a whole raft of disks in the pre-CD ROM days, and I believe there were some configuration issues. DOS was based on the principle that it was very simple to boot, very simple to write to-the-metal games programs right to the graphics card. Windows was based on the principle of getting it pre-installed on computers so users wouldn't even have to know how to install it.
The DOS/Windows dichotomy was such that for the longest time games would be DOS based. Microsoft made a big push to get the DOS games over to Windows with the WinG and later DirectX initiatives. There is a lot of Windows stuff to make it 2-D game friendly -- ScrollWindowEx() (hardware assisted scroll), CreateDIBSection() (allowed porting stuff that wrote to the frame buffer to Windows), and IDirectDraw::WaitForVerticalBlank() (allows flicker-free, tear-free scrolls and screen updates). I have a serious case of Windows lock-in because I depend on these calls and can't find them on other systems.
There is a community of people who think that some kind of 'nix underneath with a glitzy UI on top to satisfy the folks who require that kind of thing (OS-X and to a lesser extent the Linux distributions teamed up with either KDE or Gnome) is the best of both worlds. It is an article of faith that 'nix is the best foundation to build whatever kind of point-and-click thing is necessary to win over the end user, and even if Microsoft were not the Evil Empire, the Microsoft OS's are considerate second rate compared to Unix.
My understanding is that Bill Gates had not only looked at Unix but had a Xenix offering for the PC at one time or another (there is a James Burke style of "connections" to SCO and the SCO/Linux dustup). Microsoft put its toe in the Unix waters, but its customers were not clamoring for Unix on the PC (this was long before MS was the Evil Empire). Given their license arrangements, they could have gone the route of a 'nix for their advanced products (Windows 95 and beyond) but for some reason decided they weren't interested.
Was their some technical reason for Microsoft going with Unix early on, or was it a matter of preference or prejudice?
The whole idea behind the Mac Mini is that people already have monitor, mouse, and keyboard with their PC, so they just plug in a Mac Mini and switch over to Mac and OS-X. OK, the monitor plugs into the video port, the mouse and keyboard each plug into a USB port, so where do you put the printer?
I suggested, "get a hub" and my friend retorted that was the whole point. If the freakin' thing came out with 4 USB ports, he would get one, he would tell his friends to get one, but now he is just rolling his eyes.
Yeah, yeah, the thing is only $499, but that is just a gimmick because you have to accessorize it to do anything useful. It is kind of like the old Detroit gimmick that you could get a Chevy for some good price but that was without A/C, P/B, P/S, and a bunch of stuff to make the care usable (you could sweat out the no A/C, but there came a point when car designs were such that the "stripper" car that the dealer had one-of on his lot to get newspaper ad customers in the door was undrivable without P/B P/S).
So as to your notion that Apple has engineered the Mini to stay in certain market segments, you have a lot of agreement from my friend.
I guess we are stalled with a minor improvement with the Pentium 3, and step backward per clockrate with the Pentium 4 and limited availability of compilers that know how to optimize for it. With the pending multi cores per chip, I might play around with balancing DSP tasks between processors and threads, but I don't think I am going to muck around with some proprietary GPU or DSP -- life is too short and the "time to market" window is too narrow.
One of the main features of C is interchangable int and void*. Somehow I think you are better of biting the bullet and making int 64 bits just like when going from 16 to 32 bits int was so promoted. That way you only need to rewrite code where structs require 32-bit fields.
Java is supposed to be platform independent, but the implicit assumption has always been a 32-bit platform of one sort or another. Yes, Java can run on a 64-bit processor, but the int is still 32 bits unless you want to change the behavior of an awful lot of Java code.
So will there be two Java's or are they going to come up with some kind of clever 64-bit Java extension or what?
Oh, and as to the comments that it takes a shockingly big Word document to bust a 32 bit address space, the big address space is not for Word, it is for video. The change to 32 bits and faster processors made CD-quality audio pretty much universal on desktop computers, but HD-quality video is not there yet. Sure you can stream from files or segment memory, but 4 Gig is still constraining with regard to high definition video files being handled in linear address space.
Most of your car purchase money goes into someone's health insurance. The difference between a land yacht and just basic transportation is not reflected in the sticker price. When someone buys the GM land yacht, they are keeping some old guy who worked for GM off Medicaid (Medicare doesn't pay for everything and you still need private insurance).
For ordinary cars, you have the tank fill method, but I wonder if that is only accurate to about half a gallon. I try to get the same gas nozzle and park the car the same way because I figure if the car is tilted one way or the other you will get a different air bubble in the tank. Also I figure the shutoff is the same sensitivity. Even so, on a 10 gallon fill, my guess you are accurate to 5 percent (about 1 MPG in your case).
Do you get pretty much 25 MPG every time, or do you get 23, 24, 26, 25, 27, 24 -- that is more my experience (without changing any variables)?
The Green Hybrid site is slashdotted, but I wonder if they have data on non-hybrids -- comparing the Corolla or Camry against the Prius, comparing hybrid and nonhybrid Civics. In addition to the real-world data on the hybrids, it would be interesting to see real-world numbers on everything else and correlate it with EPA. That way I can see how I am doing with respect to drivers with other driving patterns and make some guesses regarding my real-world mileage from the EPA numbers of a car I am thinking of buying.
The other thing that could be done is that all cars have this OBD-II plug right under the steering column. Someone should come up with some kind of data logger that you could drive your car around -- to work, to Grandma's, whatever your personal drive cycle is -- and compare predicted mileage against actual mileage and make predictions of what you would get in another car model. I understand that if you are buying an 18-wheel truck, you can get fuel use predictions based on the truck route you want to cover, and this is a big, big factor in the economics of a truck purchase. I guess the technology exists in principle to do this with cars, but the usual people who sell cars don't even want you to know this.
I work in Windows but have read what I could about Quartz2D and Java to try and keep current. My understanding of Quartz2D is that they try and keep the graphics context coordinate system not bound to screen pixels so you have a uniform graphics context whether you are printing, plotting to the screen, or plotting to a PDF file. I am told that Quartz2D is not really PDF as much as a graphics context that tries to be as compatible with PDF as it can. Under Windows, a displayable object has to have different rendering methods to deal with the quirks of how graphics context coordinates vary between printer, screen, and metafile, while my understanding is that under Quartz, they try to make everything uniform, adding a layer to do all the scaling, so you need only one rendering method.
I don't have any 3-D experience, but I heard the buzzwords about triangles, vertices, shaders, and graphics pipelines. If Longhorn is going 3-D, is this what we are going to work with? If so, it will be radically different than Quartz, which appears to be pretty much conventional fonts-lines-shades-images-pixmaps 2-D graphics only done more uniformly and perhaps elegantly than other 2-D graphics.
And if Longhorn is indeed going 3-D, what am I going to have to learn to work in that world?
As to VC++/MFC being a substitute for Visual Basic, MFC may be a fine, fine framework and a great use of the features of C++, but it isn't even close in the VBish RAD department. The old Visual J++ was a Visual Basic substitute; VB.NET and C# are definitely VBishly RAD, although on the .NET runtime.
Delphi is a "buy a new tool" as well as a new language, but if you are developing for Win32 and using the Form Designer and a little bit of Pascal to glue things together, it is a pretty good VB replacement.
What I find interesting is that Delphi is roaring back, and it is not just a temporary blip caused by the new Delphi 2005. My guess is its not so much that people love Pascal but some big problem with Microsoft's .NET and their stand on Visual Basic 6.
If you want to do what Visual Basic 6 used to do (and still does, thank you), which is support plain-old-Windows without the .NET layers, and do it in a VB sort of way with forms and widgets and property editors, it seems that Delphi has become the only game in town (although I should look into RealBasic). Yes, there are the legacy/spaghetti VB apps, and yes, there are VB coders who don't want to learn, but I think MS is missing the boat. It is not that people don't want to learn VB.NET: what they want is a VBish thing for the old Win32 API without the .NET layer, and they are willing to learn a new language to get it. The problem is not so much with VB.NET as it is with just plain .NET.
Well, you and your colleagues could bone up on the skills to compile large C++ programs, but you could also hire some people who are experts at IBM OS/360 JCL while you are at it.
Anyone here remember JCL? The idea was that you had a "separation of concerns" between abstract operations on files (Open, Read, Write, Seek, Close) and a whole panoply of file-specific parameters (disk, tape, random-access, sequential, block size, buffer sizes), and that you only had the abstract parts in actual compiled code while the connections between the abstract parts and the OS went into a kind of shell script that specified these connections. Then these guys in New Jersey came up with this Unix thingy which didn't have any of this stuff. Unix "solved" the problem JCL was addressing by not having that problem in the first place.
My point is not that Java is "better" than C++ -- there are many criteria in what language is suitable for what kind of problems -- but that C++ is so ancient history even within its own problem domain. Make files? Pre-compiled headers? The problems those features were meant to solve were obviated by Wirth and his gang back in the Modula 2 days just like Unix made JCL go away.
The suggestions you have rolling in to speed up your C++ compiles are all well and good, but the fact that people are calling you stupid. How about another analogy. The F-104 has some bad user-interface issues along with slim stall margins on landings. It killed a lot of pilots in its day, and the pilots who survived would all say that the pilots who didn't lacked "the Right Stuff", but I don't think there are any Air Force planes these days that are quite so unforgiving.
I don't have much experience with multiple inheritance, but I understand that introduces some pain of its own. With inheritance, you have some core functionality of the base class and then you add on functions. With mixins (adding a clump of functions), I suppose you could do those with multiple-inheritance, but with using object composition as a substitute, you have to write all of these stupid little forwarding methods (or confess to the sin of violating the Law of Demeter by having too many levels of the dot notation).
Only reason Swing runs without a companion distribution is that the required DLL's for Swing are part of the Java download for Windows, Mac, etc.
OK, so Swing is "lightweight" in that most of it is done in Java, but there still has to be a link to the native OS at some level.
I also suppose that the best model for software reuse to date is where features are part of a language or a standard library (STL lists, Python hashlists) rather than some dude in the Visual C++ parser team developing a regular expression tool and supporting its use across Microsoft.
In Soviet Russia, beta-test software reports on . . . (oh nevermind, it does the same thing over here too).
In Korea, only old people use production-release software (hey, someone had to say it).
I don't mean from a process standpoint and PERT charts and waterfall diagrams and all of that. And I can understand things that at least started out as one-man operations for gifted persons (Cray computers, Linus and Linux, etc). But even if there is a head XP guy who has a roadmap, the dude has to farm out much of the coding to a vast army. How is that army organized?
Or maybe I have it all wrong -- maybe the Windows kernel is pretty compact and that much of Windows is built up out of apps -- you are responsible for this app, you are responsible for this other one. Even if the coders all work from detailed specifications, someone has to write the specifications, and I don't know if the specifications for that much code can fit within the span-of-control of one person let alone a small committee.
I will concede that Windows is a POS, but how does such a POS even boot?
I use this saying when helping someone with a very initial debugging and it seems it is not possible to get a program running without crashing or lockup.
By the way, NOVA did an Apollo 13 show with extensive interviews of Jim Lovell and Gene Kranz -- much like the Beatles documentary that gave a lot of airtime to George Martin -- and I never saw the need to see the Ron Howard film, although Harrison Schmidt, the geology PhD who went to the Moon, tells audiences of his talk to go see it. Who am I to argue with a guy who actually went to the Moon whether the Ron Howard film is worth seeing?
What does it take to write extension modules (either GUI or non-GUI) for RealBasic? Do you write them in RealBasic or is there a way to write them in C/C++?
If you do a GUI extension module, does your application become dependent on Windows/OS-X/Linux or does it have some kind of abstraction of the GUI and abstraction of the graphics drawing surface accessible from C/C++?
Well anyway, the idea that you could have a weapon with two settings, stun and kill, gee, where did the cops get the idea that this would come in handy? Turns out that it is hard to stun someone without killing them. The old detective movie cliche of stunning someone by hitting them over the head? An MD writing in TV guide told readers that "if you hit someone over the head, most likely you just hurt them and make them mad. If you hit them harder to knock them out, chances are you kill someone from a hematoma."
I have seen films where they Taser a volunteer, and they start convulsing or they just faint and tip over and hit the mat in the gym they are doing this with a thud. I have seen a recent episode of "Cops" where they Tasered a whole bunch of different people, and it didn't seem to do anything. I suppose the electric shock causes intense, sudden pain, and that can cause someone to collapse, just like being shot with a bullet, only there is no tissue damage so the person doesn't bleed to death, but does it really work? I suppose a bullet doesn't always "work" in that a person can be fatally wounded but not always drop dead right away.
Anyway, I was the only person at the meeting willing to try the Taser (it was a model with two electrodes sticking out, and it didn't have the darts for remote action). I was going to zap my forearm and I was quickly advised "better try it on your leg. So I hold the Taser up to my leg and press the trigger. I was pretty embarrassed because I yelled out "Jesus Christ!" I was more startled than anything else, and it hurt real bad, but not the kind of hurt of someone drilling into your teeth -- more like the worst kind of muscle cramp -- it really hurt but I felt like my leg muscles were seizing up.
So what does a Taser do? Does it really knock a person out like on Star Trek, or does it merely cause a person to take the Lord's name in vain? Have they made the Taser more powerful in the last 20 years? More maybe because my finger was on the trigger, I let go when it first started to hurt, and a person has to deliver multiple jolts to get someone to collapse.
I think that getting car propulsion from plug power opens up a whole lot of possibilities even if our current electric generation isn't that green, and people ranging from Burt Rutan who misses his EV-1 to an article in Tau Beta Pi magazine are talking about hybrids to get vehicles that are perhaps 80 percent plug power and only use gasoline to solve the battery range problem.
Missing from the proponents of the plug-power Prius, Burt Rutan, and Tau Beta Pi magazine is discussion of the kWHr cost of battery.
Let me just pull together some numbers, and people can tell me if I am at least in the ballpark-order-of-magnitude. A 10 kWHr lead acid battery pack (i.e. the EV-1 power pack) costs $3000 and allows for 300 charge/discharge cycles (you get many more cycles with shallow discharge, but I am not convinced that you get more total kWHr through the battery). A 10 kWHr pack delivers road power to replace, say, 1 gallon of gas. This is saying that the battery alone is the equivalent of gas at $10/gallon, and if $50/bbl oil is $2.50/gal gas, $10/gallon gas is $200/bbl oil, and at $200/bbl there is all kinds of crud (heavy oil, tar sands, oil shale) that you can turn into gasoline.
Batteries cost money, not just in initial cost but in operating cost because batteries wear out. There are batteries that have more cycles than lead acid, but they also cost more. The Prius uses a NiMH battery, and the consumer NiMH's are said to be good for about 500 cycles. The Prius battery holds much less than the EV-1 battery, and I heard figures of 7000-8000 dollars to replace a Prius battern.
That a battery only does so many charge/discharge cycles has to do with physical chemistry -- it is like you are electroplating and de-electroplating over an over again and how many cycles will a battery plate hold its shape and not crumble? Making oil from shale or tar sands is expensive, not because the operations are pilot-scale but because you have to process a lot of material. Storing energy in a battery is similarly expensive, not because battery production hasn't been ramped up but because just plain wear of the battery elements.
The reason for the Prius (or even Insight) style hybrid is that that batteries cost too much for economical substitution of plug power for gasoline, even for very expensive gasoline. The idea in these types of hybrids is to find where gasoline is used in the most wasteful modes (idling, coasting, stop-and-go-running) and to substitute those modes with battery-stored electricity. You know, if 80 percent of the gasoline does 20 percent of the work, replace that 20 percent with electricity, and even if battery electricity costs 4 times that of gasoline it works out.
Well, maybe it works out. The proof is whether you run a Prius (compared to a Corolla or a Camry, not an SUV!) until you wear out a battery pack and weigh the replacement cost of the battery pack against the cost of the gasoline difference. Even at $2.50/gallon gas I think I am better off with the Camry -- $5/gallon gas may be the crossover point.
But the weak point about batteries, whether you are talking EV-1, battery backup off-grid solar power, or a plug hybrid is that the battery itself is a consumable instead of a durable good, and electricity out of batteries costs several multiple of the cost of gasoline except in those "corner cases" where gasoline is used inefficiently compared to the electric power substitue.
It is a lot like a PBS station during Pledge Week. All of your regular programs are cancelled or moved around, there is maybe one program worth watching among the nonstop Daniel O'Donnell concerts, and that one program is chopped all to heck with those stupid people shilling for donations who don't know how to act in front of a camera.