I am not offering any judgement on the biology part of Lee Hood's institute being IP'd up. There is a kind of gold rush fever in gene and other kinds of biotechnology, and this stuff gets IP'd by everybody who is involved in this, whether or not it has Bill and Melinda Gates funding.
What I am saying that is kind of interesting is that the software side of this institute is FOSS, although if you are Lee Hood, you might have enough street cred to stand up to Bill Gates on this. I believe Lee Hood emphasized this only applied to the desk-top software side of things -- the embedded software in whatever kind of custom analyzers and instruments did not get this treatment.
Lee Hood, the dude who developed the automated gene sequencer, was at the U at the end of May. He talked about what he is up to these days, which is running some kind of medical research foundation in Seattle using a lot of Bill and Melinda Gates money with some NIH support in the mix.
He mentioned that while all of the biology and engineering tech were all IP'd up, the software side was FOSS -- Google Cytoscape to look up their software project. Predictably, he mentioned that Bill Gates was against this arrangement, and Lee Hood mentioned it took a lot of upper-management pep talk and persuasion to get his in-house software people to be happy about it as well.
I didn't bother asking Lee Hood questions about the software aspects as it was a biology symposium and the grad students were more interested in the biology aspects of the project, but I looked up Cytoscape, and guess what, it is written and extendable in Java. And this is largely on Bill and Melinda's dime.
Dude, you may be on to something. Next time on a road trip, I will observe whether the person doing 55 in the left lane is driving GM-Ford-Chrysler or not.
But don't you think there may be Prius or perhaps Camry drivers in Complain About Other Drivers category? People buy Camrys because they don't want to even have to think about their car -- maybe a few people get them because they can be limo smooth, but mainly anti-car people drive them when they have to get a car. And Prius is a whole other league in this regard.
OK, what would it take to triangulate location within a 6" on a side cube? To within millimeter accuracy? We are currently using a multi-million dollar x-ray scanner to do this at low doses, and if this can be done for a few thousands of dollars, that would be big deal.
So a network is blocking American Gladiators and Medium. Gee, what is next, they will block recording of America's Funniest Home Videos, replete with groin kicks?
I guess the folks at Mathworks with their Matlab product are user-torturing psychos, because the UI to Matlab is Java Swing. Not just the Command Window and M-file Editor, a Matlab Figure Window where you do your pretty plots is a JFrame as well.
Ever notice the Matlab startup latency? It is the same such thing as the Java startup latency because that is what you are waiting on.
Come to think of it, I saw a Java Coffee Cup on the Toolbar once when Maple was slow to load. More psychos! Dweep! Dwepp! Dweep!
In Medvedev's Russia, women fail to understand you!
(trust me, that one works at more than one level)
If Ada, why not Delphi? Component Pascal?
on
The Return of Ada
·
· Score: 1
In some ways, Delphi has its own problems with the way Borland has decided to implement Object Pascal in a quirky way and all of the Windows-specific features that probably should be in a library -- the language is otherwise pretty much tied to Windows these days.
But of all of these Pascal-derived languages, why Ada? What about Wirth's Component Pascal? But how about Delphi?
The thing about Delphi is that it compiles like a bat-out-of-Hell while last I heard Ada compilers are these complex ponderous things. A lot had to do with Delphi preserving enough of the structure of Pascal that was compiler friendly; the word I heard is that Ada is far from compiler friendly.
Why worry that a thing is compiler friendly -- should it not be programmer friendly at leave the work to the computer. Yes, up to a point, but if something is not compiler friendly, how are we to know that the compilers don't have hidden bugs? Besides, fast compiles means something, even on Core-2 Duos and all of that. Have you compiled anything under Delphi lately and compared that experience to anything else?
You could say the ultimate in compiler friendly syntax is Lisp, but I think there has to be some middle ground some something more infix and readable. I think that Delphi, or at least Object Pascal if you are thinking other platforms (think Free Pascal) is that middle ground, not Ada.
Take a look at Belbruno, "Fly Me to the Moon" and come back and tell me that the process is obvious, prior art, or just plain physics. The dude has math journal articles on the subject, but that book is a good starting point.
I don't think anyone here would have a clue how to do those low-energy space rescues if Belbruno had not disclosed his method.
We can argue Intellectual Property, or the more politically-correct "patents, copyrights, and proprietary material" all we want, but one has to give the dude some respect for figuring out how to "hack" orbits. He also started out on the NASA/JPL payroll, and if the gummint had been willing to keep him on the payroll, this thing could have been public domain. A mathematician has to make a living, sometimes.
Here is what I ended up doing. I needed to create a set of custom widgets for a certain type of scientific data visualization, to use those widgets in my own programming, and to make those widgets available to students doing projects.
I rolled my own Windows API object framework -- very lean, the bare essentials, if it was missing a feature, I added it as needed rather than trying to make the base classes Swiss Army knives. I had a Delphi Object Pascal version and a C++ version of that object framework, and the C++ base class had a WndProc() with a big switch statement to intercept those Windows messages I was interested in -- I didn't see the message map macro as worth the bother for a tiny bit of speed advantage. My widgets were all done in Delphi, but without the Delphi VCL runtime library, making them very lean and giving very small exectutables for apps developed with them. I had the C++ version because at one time I thought my students would want to roll their own Win32 API programs using my widget library.
The next thing I did was to rewrite my object framework around a Model-View-Controller structure. The main insight was to put most of the functionality of the custom widget into the Controller object. I was able to use the same controller objects within the lean-and-mean Win32 apps as I used to create custom Delphi VCL widgets with them. Those custom VCL widgets are something I use "in house" when I need to rapidly customize an application using them. I then used the Delphi facility to make a set of ActiveX widgets out of those VCL widgets.
So I have a lean-and-mean, a Delphi VCL, and ActiveX libraries of the same widgets using the same code base. As far as the student projects are concerned, I kind of gave up on having non-CS students do anything with C++/Win32, even though engineering students say they want C++ experience as a line on their resume. They are completely sold on Matlab, and they do their projects in Matlab, embedding ActiveX controls in Matlab GUI apps. For my own work of producing customized apps, I pretty much go with the Delphi VCL controls -- I keep the lean-and-mean controls around for the big legacy app that I have done that way. But any change/bug fix/added feature to these widgets appears in all three places anyway.
Seeing the handwriting on the wall about Win32 and ActiveX controls, I also pretty much gave up on.NET and WinForms either. Why should I knock myself out doing WinForms when it 1) has less functionality than either Win32 or Java Swing, 2) who knows what MS is coming up with when they break legacy ActiveX support. These days I am pretty much doing Java Swing for graphical front ends of the completely new stuff I am doing, C++ for the numeric-intensive back end.
Java Swing is clumsy in a lot of ways, but at least I am not longer at the mercy of what MS decides. The performance is not native, but for my current projects and current generation of computers, it is good enough.
On the face of it, native Windows is not quite the worst API. It was object-oriented before that became popular. What the hey, Window handles are objects, no? You CreateWindow() them, DestroyWindow(), invoke various functions on them for SendMessage or PostMessage to them. What could be more object-oriented and easy?
I think the problem started when they went from the Petzold-style C programming (the WndProc()'s with the giant switch statement) to MFC (omigosh, omigoodness, whata mess!).
The minute the PHB comes along and says, none of the C stuff, we gotta get object oriented and program in C++ that the funny stuff starts.
The problem is that you have the native Windows object, referenced by an HWND, and then you have a C++ object as its proxy, stand-in, OO Overlord Protector. Window handles and C++ objects obey different kinds of rules for creation and especially destruction and lifetime management. A Window handle and its underlying window object can go Poof! in response to a DestroyWindow() call, and how is a C++ object to behave or respond when this happens. Not only that, a Window handle can go poof during CreateWindow -- WM_CREATE can return a value indicating that the Window handle can be destroyed.
Syncing the lifetime of the C++ object with the underlying Window handle is really hard. Another problem is that Windows messages are dispatched to a WndProc, which is not that hard to hook up to a C++ method through some function pointer and calling convention boojum, but the WndProc then has to dispatch to C++ object methods some how. The Microsoft C++ MFC message map macros are the ugliest looking kludges, or maybe I am sensitive to the message map macro statements IN ALL CAPS AND WHY IS THIS SOURCE CODE SHOUTING AT ME LIKE THAT!
Dunno, that MFC is so ugly is not so much evidence that the Win32 API is that bad but that the dudes at Microsoft are completely without clue as to how to use C++ as a developer's language. I disagree with you about Win32 being a bad API -- Charles Petzold pointed the way to know it, use it, and even love it, but he eschewed writing code or books about that goshdarnawful MFC and the MS way of C++ programming.
The other way to deal with the Win32 API is to wrap it so tightly that application programmers cannot get at the HWND to do something dastardly like DestroyWindow() on their own HWND. Java does that -- it presents Swing as an object framework on top of Win32 when hosted on Win32 and spanks your hand about reaching for the HWND. Delphi does it, and they modified the Object Pascal language enough to be friendly with WM_ messages being routed to object methods, and they let you touch the HWND, and the DC, and other things if you are inclined -- I do have to admit that the source codes to the Delphi RTL are this marvel of handling the corner cases of when things burp on HWNDs and other native resources, and the guys who wrote it as early in the history of Win32 as they did must have been geniuses.
Why.NET does that too -- yeah, yeah, various ways to "drill down to native", but by and large they wall of Win32, and they do an even worse job of exposing its features than Java does -- I want my ScrollWindowEx(), my CreateDIBSection(), where did they all go!
So you tell me that Trolltech has a passable C++ wrapper to the Win32 API in their QT4 when running on Windows. So, I heard something. Yeah? Well, just a little bit. Just a little bit. C++ GUI development can be halfway tolerable? Just a little bit.
Just because it has a central black hole doesn't make it a dwarf elliptical galaxy.
What distinguishes the Milky Way globular clusters is the the are all about the same, very old, almost as old as the Universe age. If there is reason to believe this is gravitationally bound to the Milky Way instead of some interloper, and if it has the same HR diagram turnoff point of other Milky Way globulars, there is no reason to think it is anything other than one of the bigger and fatter and closer of the globulars.
So did the ancients have a peer-review system that sucked camel balls?
Like a Transactions on the Geometry of Levers where they publish a paper where the logical conclusion is that a lever can apply a force without a fulcrum? And when a scribe tries to publish a paper pointing out that absurdity and corrects the flaw in the geometric reasoning, that paper languishes 7 months in review until some doofus of an anonymous reviewer complains that the author doesn't belong to the correct Guild of Scribes to publish in that journal and the associate editor acts all sanctimonious?
I know that others think that Joel Spolsky has a high opinion of himself, but he writes in a thoroughly entertaining manner and has some thought-provoking ideas.
Visual Basic is widely regarded as totally Night of the Living Dead zombified and is held in disdane by anyone who considers themselves a professional programmer, especially by anyone with any C/C++ experience. But what ol' Joel does is he hires who he considers to be the best C++/MFC programmers he can get, and he puts them to work using VB.
Why does he do this? Because VB, at least for things Joel is interested in, is by leaps and bounds more productive than C++/MFC. Just as you are leaps and bounds more productive using Matlab than the FORTRAN you might have started out with. But why subject C++ programmers to VB when they hate VB? A lot of the reason these dudes hate C++ is they have been told by PHBs to fix someone elses spaghetti VB app, but if you take C++ coders, i.e. people who know the rudiments of programming, the are absolute maestros with VB compared to some financial analyst type hacking together a mess of a VB program.
I am not contradicting anything you are saying. If you had any kind of background in FORTRAN or whatever the equivalent, Matlab is like the pleasure of taking off shoes that pinch. But if you haven't had that equivalent programming background, Matlab can be like any other programming language to a newbie -- endless debugging to get anything right.
Why gin up a Matlab substitute for undergrad teaching? One reason is that Python is FOSS while Matlab is not. I am not one of these hyper-partisan FOSS advocates, but on the other hand, I see a kind of conflict of interest situation for turning engineering colleges at public universities into wholly owned subsidiaries of anyone, whether it is Microsoft or Mathworks or whomever.
A second reason is that Python is a reasonable teaching language regarding OO, data structures, functional programming (to a limited extent) of what are the modern foundations of programming and programming languages and outside the numerical programming domain, Matlab is a dog's breakfast. Someone trained on Python can go off in your direction and be comfortable doing numeric calculations in Matlab or go off in another direction and write C code for an embedded computer system, based on learning some basic concepts that Python supports and Matlab does not.
Good question -- one should not automatically assume a guy likes gals and the other way around. But I think I was open enough with either of my parents that they had me geeky-shy towards dating women rather than geeky-shy about dating men.
What is to say I wasn't a guy who wanted to marry a woman doctor? Tried really hard, even went out with the reason you have to a take a lukewarm shower because of regulations limiting water heater temperatures to 120 degrees so kids don't get scalded -- I was too "defense lawyer" oriented for her sensibilities.
The trend we need to fight is not teaching programming using Java, C, or C++, but defaulting to Matlab as a programming language of choice. This is what is happening in Engineering colleges and other non-CS places at the U.
Granted, Engineering always went for things that CS considered "brain dead" -- Basic, PC's, DOS, Windows. But Matlab is more brain dead than most.
What happened is that a lot of the current generation of Engineering profs cut their teeth on FORTRAN -- their Intro to Programming was in FORTRAN, whatever industrial job they had before getting a PhD had them compute things in FORTRAN. Few of them were ever comfortable in it and most of them spent hours in the computing center debugging programs dumped to massive punch card decks.
When Matlab came around, it was numerical Nirvana. It had this massive numeric library that you didn't have to write your own Q-R linear equation solver or SVD subroutine, and you didn't have to go searching for this stuff either, it was all there. It had a command prompt to performed immediate execution along with reasonably friendly error messages. And it acquired a thoroughly feature-full graphics package.
Don't get me wrong, Matlab is a very capable numerical applications language and even turns out to be one of the better Java scripting languages of all things. But it really falls down in terms of extensibility of its type system, and as far as what Mathworks tacked on for object-oriented programming, fuggedaboutit. It is also the Swiss Army knife of software for a whole bunch of people, and forget about introducing them to a socket wrench and handle that can apply serious torque to a bolt when they think they can get by with the pliers tool.
While people who know what they are doing can benefit from the convenience of the numeric and graphics libraries, the immediate mode, the verbose error handling and rare instances of complete crashes, if you don't know what you are doing (i.e. you are just learning), it can lead to as many hour-gobbling skull-cracking debug sessions as anything else. Our required Numerical Methods course is in CS, it uses Matlab, our faculty is complaining that the students are complaining that they hate the course because they are spinning their wheels trying to get programs to run (in Matlab of all things), and we have guys in our department we want to teach Numerical Methods (in Matlab, of course), in the context of a watered-down Intro to Engineering offering.
What the community needs right now is a Python distro with enough of a numerics and graphics package rolled in to do 90 percent of what is in Matlab (Are the Python people still hashing out that Numerics/Numpy divide? Is there an engineering graphics library that is Numerics/Numpy compatible? 99 percent of what you do in Matlab is that you have a Leatherman Tool of a 2-D array type (Matlab, Matrix Lab) along with all of the libraries being compatible with that type.) CS departments could teach their Intro to Programming along with their Numerical Methods courses using that Python distro, and we can save a generation of engineers from brain damage.
I fail to see how Visual J++ was a dead end. Although they don't call it Visual J++ -- it is now called Visual Studio.NET C#. I had looked at Visual J++, and it is.NET version 0.8 in so many ways.
Let me get this straight, that some wieners at Purdue decided to teach Java instead of C++ in their Intro to Programming course, and that you had a wiener of a prof who was teaching you Java without adequate experience, this all put you on the path to dropping school, dropping your CS major, and taking courses at community college without finishing a degree?
I will grant you that if you or the parents are shelling out the Purdue tuition, maybe their CS department should find a better professor for their intro course. I am sorry to hear that this experience dissuaded you from completing a CS degree, and there is probably a lot more to your personal story than can be shared on Slashdot.
But I would like to communicate to others out there that you will have a few good teachers in your educational career who are really inspiring, a vast group of average teachers, and a number of who you consider to be really, really bad teachers. The "bad" teachers are that way (in your opinion) for a number of reasons -- they may be "nice guys or gals" who don't have enough preparation or smarts to teach, they may have admitted to you gaps in their preparation that you have taken upon yourself to hold them in disrespect for, or maybe they assign too much HW and work you too hard.
If one is going to take a passive approach, show up to class and demand, "Here, educate me", that is a good way to fail at getting a degree and also to fail at every other opportunity that presents itself down the road. If one is going to take an active approach, working as hard as one can at learning from all teachers, the good and the bad, supplementing gaps in instruction with self-study, working coding jobs, group study, one is going to be successful at college and everything else.
To suggest that a person can have one "bad" prof means that they are on the street drinking methyl antifreeze out of a jar wrapped in a paper bag, this suggests a very passive approach to not just education but life in all its aspects.
That men, and with compounded returns geek-men, misinterpret sexual interest that women may or may not have in them is what our entire culture revolves around. It is the central plot point of what we call romantic comedy in theatrical plays, novels, movies, and TV sit coms.
Being the geeky son of an even geekier Dad, my momma often took it upon herself to wack me about the side of the head with a clue bat about young women I would encounter. I remember this one time we were in the Walgreens (why I remember this episode, I don't know, maybe I need to prove to myself I wasn't raised totally stupid). A friendly enough woman clerk checks us out, and as we walk out, Momma does the stage whisper to tell her child he is an idiot to tell me that this young woman was doing everything short of giving out her home phone number.
I responded in my own stage whisper a) I may not have been totally clued in on the signals given, but b) this young lady is earning minimum wage running a checkout counter in a Walgreens -- I thought you had more higher aspirations for the station in life you expect from a daughter-in-law.
Yeah, yeah, I know what you are all thinking, that my momma took a lot of back-talk from her adult son, and what kind of elitist thinks that a woman who happens to work a counter is to be looked down upon for a date. OK, the lady was pretty enough, friendly enough, hot enough, and I was available enough, but how about if I was pushing 30 at the time and the clerk in question was still in high school?
Men are supposed to be mind readers about when women are interested or not interested in them, and it is all very funny that men are geeks or oafs when they are interested in a woman who is not turned on by them and even bigger geeks and oafs when they fail to pick on when a woman is hot of them. Big deal, sometimes a man has selection criterea, even if they are as stupidly pragmatic as non-jailbait.
My gripe with this multi-core let's program parallel fad is, where are these massively parallel machines?
Two cores? Big whoop! Four cores? Haven't seen one, and our computing center keeps current with the latest generation desktop machines.
But factors of (almost) two or (almost) four speedup are no big deal in the grand scheme of things. Wake me up when they are talking about 10 cores, or 100 cores.
But that is the problem. Our resident computer architecture dude tells us that maybe 10 or 16 cores is an upper limit on account of cache memory or other considerations. To go beyond that they will need another approach -- clusters on a chip? For the low order of parallelism of current multi-core multi-processor offerings, I can't see the payoff in investing a lot of effort into developing, testing, and debugging parallel code.
What is this business about helium fusion in the red giant phase?
Unless there are some revisions in the laws of physics, the nuclear processes throughout stellar evolution are well known based on computer models.
When helium "ash" accumulates in the core, helium fusion is not the next thing that happens. The core starts contracting and heating up, but that lights off H2 fusion in the shell surrounding the core. That phenomenon changes the luminosity and heat transfer rates of the star, causing the outer atmosphere to swell up into the red giant stage.
When shell burning runs its course, again the core contracts and heats up some more, resulting in the helium flash. Based on computer models, the helium flash is a major disruptive event caused by the sudden onset of helium fusion, it does not cause the star to go nova or anything, but it causes the star to change modes as it were, becoming somewhat bluer and smaller, but still more luminous than Main Sequence. From the computer models, it is believed that the upper-righthand HR diagram stars, red giants, are H2 shell burners while the horizontal branch above the Main Sequence represents He core burners.
For a massive enough star, exhaustion of core He will initiate shell He ignition, sending the star back into the red giant range, perhaps as a red supergiant for a massive star.
The red giant phase is only one phase of an evolved star. Everyone just kind of assumed that a star that goes supernova would be a red giant, but it seems like the star that popped off in Supernova 1987a in the LMS was blue.
Of all of the alternatives to Windows GDI, Java Swing looks to have the most features -- BufferedImage rocks. One of the limitations of X relative to Windows GDI is the inability to do hardware-assisted scrolls (ScrollWindowEx() in Windows is a legacy of the WinG thing) -- cannot do scrolls without redrawing the whole frame buffer in software -- but Swing under Linux can use OpenGL to get around that limitation.
The other good thing about Java is that the Java Native Interface (JNI) is reasonably well documented, and it lets Java call C/C++, and it lets C/C++ call back into Java. Maybe a little clunky and you have to understand the global and local locks thing on the native side or you can get some horrific bugs, but it is quite powerful once you get the hang of it.
My understanding of AJAX is that your programming is mainly in JavaScript to get code running on the client to do a lot of things that can only happen or happen efficiently on the client. Is there any way to call from JavaScript down to C/C++, either client side or server side?
It's that mandatory Circuits course that the mechanical engineers and others outside EE are required to take is what is breeding terrorists, I tell you. Requiring MEs to learn op amps is what is giving them that sour outlook on life.
Slashdot Mirror
What I am saying that is kind of interesting is that the software side of this institute is FOSS, although if you are Lee Hood, you might have enough street cred to stand up to Bill Gates on this. I believe Lee Hood emphasized this only applied to the desk-top software side of things -- the embedded software in whatever kind of custom analyzers and instruments did not get this treatment.
He mentioned that while all of the biology and engineering tech were all IP'd up, the software side was FOSS -- Google Cytoscape to look up their software project. Predictably, he mentioned that Bill Gates was against this arrangement, and Lee Hood mentioned it took a lot of upper-management pep talk and persuasion to get his in-house software people to be happy about it as well.
I didn't bother asking Lee Hood questions about the software aspects as it was a biology symposium and the grad students were more interested in the biology aspects of the project, but I looked up Cytoscape, and guess what, it is written and extendable in Java. And this is largely on Bill and Melinda's dime.
But don't you think there may be Prius or perhaps Camry drivers in Complain About Other Drivers category? People buy Camrys because they don't want to even have to think about their car -- maybe a few people get them because they can be limo smooth, but mainly anti-car people drive them when they have to get a car. And Prius is a whole other league in this regard.
Yeah, and the plutonium came from Hanford, Washington and was given on the Q-T to Tel Aviv.
OK, what would it take to triangulate location within a 6" on a side cube? To within millimeter accuracy? We are currently using a multi-million dollar x-ray scanner to do this at low doses, and if this can be done for a few thousands of dollars, that would be big deal.
So a network is blocking American Gladiators and Medium. Gee, what is next, they will block recording of America's Funniest Home Videos, replete with groin kicks?
Ever notice the Matlab startup latency? It is the same such thing as the Java startup latency because that is what you are waiting on.
Come to think of it, I saw a Java Coffee Cup on the Toolbar once when Maple was slow to load. More psychos! Dweep! Dwepp! Dweep!
(trust me, that one works at more than one level)
But of all of these Pascal-derived languages, why Ada? What about Wirth's Component Pascal? But how about Delphi?
The thing about Delphi is that it compiles like a bat-out-of-Hell while last I heard Ada compilers are these complex ponderous things. A lot had to do with Delphi preserving enough of the structure of Pascal that was compiler friendly; the word I heard is that Ada is far from compiler friendly.
Why worry that a thing is compiler friendly -- should it not be programmer friendly at leave the work to the computer. Yes, up to a point, but if something is not compiler friendly, how are we to know that the compilers don't have hidden bugs? Besides, fast compiles means something, even on Core-2 Duos and all of that. Have you compiled anything under Delphi lately and compared that experience to anything else?
You could say the ultimate in compiler friendly syntax is Lisp, but I think there has to be some middle ground some something more infix and readable. I think that Delphi, or at least Object Pascal if you are thinking other platforms (think Free Pascal) is that middle ground, not Ada.
I don't think anyone here would have a clue how to do those low-energy space rescues if Belbruno had not disclosed his method.
We can argue Intellectual Property, or the more politically-correct "patents, copyrights, and proprietary material" all we want, but one has to give the dude some respect for figuring out how to "hack" orbits. He also started out on the NASA/JPL payroll, and if the gummint had been willing to keep him on the payroll, this thing could have been public domain. A mathematician has to make a living, sometimes.
I rolled my own Windows API object framework -- very lean, the bare essentials, if it was missing a feature, I added it as needed rather than trying to make the base classes Swiss Army knives. I had a Delphi Object Pascal version and a C++ version of that object framework, and the C++ base class had a WndProc() with a big switch statement to intercept those Windows messages I was interested in -- I didn't see the message map macro as worth the bother for a tiny bit of speed advantage. My widgets were all done in Delphi, but without the Delphi VCL runtime library, making them very lean and giving very small exectutables for apps developed with them. I had the C++ version because at one time I thought my students would want to roll their own Win32 API programs using my widget library.
The next thing I did was to rewrite my object framework around a Model-View-Controller structure. The main insight was to put most of the functionality of the custom widget into the Controller object. I was able to use the same controller objects within the lean-and-mean Win32 apps as I used to create custom Delphi VCL widgets with them. Those custom VCL widgets are something I use "in house" when I need to rapidly customize an application using them. I then used the Delphi facility to make a set of ActiveX widgets out of those VCL widgets.
So I have a lean-and-mean, a Delphi VCL, and ActiveX libraries of the same widgets using the same code base. As far as the student projects are concerned, I kind of gave up on having non-CS students do anything with C++/Win32, even though engineering students say they want C++ experience as a line on their resume. They are completely sold on Matlab, and they do their projects in Matlab, embedding ActiveX controls in Matlab GUI apps. For my own work of producing customized apps, I pretty much go with the Delphi VCL controls -- I keep the lean-and-mean controls around for the big legacy app that I have done that way. But any change/bug fix/added feature to these widgets appears in all three places anyway.
Seeing the handwriting on the wall about Win32 and ActiveX controls, I also pretty much gave up on .NET and WinForms either. Why should I knock myself out doing WinForms when it 1) has less functionality than either Win32 or Java Swing, 2) who knows what MS is coming up with when they break legacy ActiveX support. These days I am pretty much doing Java Swing for graphical front ends of the completely new stuff I am doing, C++ for the numeric-intensive back end.
Java Swing is clumsy in a lot of ways, but at least I am not longer at the mercy of what MS decides. The performance is not native, but for my current projects and current generation of computers, it is good enough.
I think the problem started when they went from the Petzold-style C programming (the WndProc()'s with the giant switch statement) to MFC (omigosh, omigoodness, whata mess!).
The minute the PHB comes along and says, none of the C stuff, we gotta get object oriented and program in C++ that the funny stuff starts.
The problem is that you have the native Windows object, referenced by an HWND, and then you have a C++ object as its proxy, stand-in, OO Overlord Protector. Window handles and C++ objects obey different kinds of rules for creation and especially destruction and lifetime management. A Window handle and its underlying window object can go Poof! in response to a DestroyWindow() call, and how is a C++ object to behave or respond when this happens. Not only that, a Window handle can go poof during CreateWindow -- WM_CREATE can return a value indicating that the Window handle can be destroyed.
Syncing the lifetime of the C++ object with the underlying Window handle is really hard. Another problem is that Windows messages are dispatched to a WndProc, which is not that hard to hook up to a C++ method through some function pointer and calling convention boojum, but the WndProc then has to dispatch to C++ object methods some how. The Microsoft C++ MFC message map macros are the ugliest looking kludges, or maybe I am sensitive to the message map macro statements IN ALL CAPS AND WHY IS THIS SOURCE CODE SHOUTING AT ME LIKE THAT!
Dunno, that MFC is so ugly is not so much evidence that the Win32 API is that bad but that the dudes at Microsoft are completely without clue as to how to use C++ as a developer's language. I disagree with you about Win32 being a bad API -- Charles Petzold pointed the way to know it, use it, and even love it, but he eschewed writing code or books about that goshdarnawful MFC and the MS way of C++ programming.
The other way to deal with the Win32 API is to wrap it so tightly that application programmers cannot get at the HWND to do something dastardly like DestroyWindow() on their own HWND. Java does that -- it presents Swing as an object framework on top of Win32 when hosted on Win32 and spanks your hand about reaching for the HWND. Delphi does it, and they modified the Object Pascal language enough to be friendly with WM_ messages being routed to object methods, and they let you touch the HWND, and the DC, and other things if you are inclined -- I do have to admit that the source codes to the Delphi RTL are this marvel of handling the corner cases of when things burp on HWNDs and other native resources, and the guys who wrote it as early in the history of Win32 as they did must have been geniuses.
Why .NET does that too -- yeah, yeah, various ways to "drill down to native", but by and large they wall of Win32, and they do an even worse job of exposing its features than Java does -- I want my ScrollWindowEx(), my CreateDIBSection(), where did they all go!
So you tell me that Trolltech has a passable C++ wrapper to the Win32 API in their QT4 when running on Windows. So, I heard something. Yeah? Well, just a little bit. Just a little bit. C++ GUI development can be halfway tolerable? Just a little bit.
What distinguishes the Milky Way globular clusters is the the are all about the same, very old, almost as old as the Universe age. If there is reason to believe this is gravitationally bound to the Milky Way instead of some interloper, and if it has the same HR diagram turnoff point of other Milky Way globulars, there is no reason to think it is anything other than one of the bigger and fatter and closer of the globulars.
Like a Transactions on the Geometry of Levers where they publish a paper where the logical conclusion is that a lever can apply a force without a fulcrum? And when a scribe tries to publish a paper pointing out that absurdity and corrects the flaw in the geometric reasoning, that paper languishes 7 months in review until some doofus of an anonymous reviewer complains that the author doesn't belong to the correct Guild of Scribes to publish in that journal and the associate editor acts all sanctimonious?
Visual Basic is widely regarded as totally Night of the Living Dead zombified and is held in disdane by anyone who considers themselves a professional programmer, especially by anyone with any C/C++ experience. But what ol' Joel does is he hires who he considers to be the best C++/MFC programmers he can get, and he puts them to work using VB.
Why does he do this? Because VB, at least for things Joel is interested in, is by leaps and bounds more productive than C++/MFC. Just as you are leaps and bounds more productive using Matlab than the FORTRAN you might have started out with. But why subject C++ programmers to VB when they hate VB? A lot of the reason these dudes hate C++ is they have been told by PHBs to fix someone elses spaghetti VB app, but if you take C++ coders, i.e. people who know the rudiments of programming, the are absolute maestros with VB compared to some financial analyst type hacking together a mess of a VB program.
I am not contradicting anything you are saying. If you had any kind of background in FORTRAN or whatever the equivalent, Matlab is like the pleasure of taking off shoes that pinch. But if you haven't had that equivalent programming background, Matlab can be like any other programming language to a newbie -- endless debugging to get anything right.
Why gin up a Matlab substitute for undergrad teaching? One reason is that Python is FOSS while Matlab is not. I am not one of these hyper-partisan FOSS advocates, but on the other hand, I see a kind of conflict of interest situation for turning engineering colleges at public universities into wholly owned subsidiaries of anyone, whether it is Microsoft or Mathworks or whomever.
A second reason is that Python is a reasonable teaching language regarding OO, data structures, functional programming (to a limited extent) of what are the modern foundations of programming and programming languages and outside the numerical programming domain, Matlab is a dog's breakfast. Someone trained on Python can go off in your direction and be comfortable doing numeric calculations in Matlab or go off in another direction and write C code for an embedded computer system, based on learning some basic concepts that Python supports and Matlab does not.
What is to say I wasn't a guy who wanted to marry a woman doctor? Tried really hard, even went out with the reason you have to a take a lukewarm shower because of regulations limiting water heater temperatures to 120 degrees so kids don't get scalded -- I was too "defense lawyer" oriented for her sensibilities.
Granted, Engineering always went for things that CS considered "brain dead" -- Basic, PC's, DOS, Windows. But Matlab is more brain dead than most.
What happened is that a lot of the current generation of Engineering profs cut their teeth on FORTRAN -- their Intro to Programming was in FORTRAN, whatever industrial job they had before getting a PhD had them compute things in FORTRAN. Few of them were ever comfortable in it and most of them spent hours in the computing center debugging programs dumped to massive punch card decks.
When Matlab came around, it was numerical Nirvana. It had this massive numeric library that you didn't have to write your own Q-R linear equation solver or SVD subroutine, and you didn't have to go searching for this stuff either, it was all there. It had a command prompt to performed immediate execution along with reasonably friendly error messages. And it acquired a thoroughly feature-full graphics package.
Don't get me wrong, Matlab is a very capable numerical applications language and even turns out to be one of the better Java scripting languages of all things. But it really falls down in terms of extensibility of its type system, and as far as what Mathworks tacked on for object-oriented programming, fuggedaboutit. It is also the Swiss Army knife of software for a whole bunch of people, and forget about introducing them to a socket wrench and handle that can apply serious torque to a bolt when they think they can get by with the pliers tool.
While people who know what they are doing can benefit from the convenience of the numeric and graphics libraries, the immediate mode, the verbose error handling and rare instances of complete crashes, if you don't know what you are doing (i.e. you are just learning), it can lead to as many hour-gobbling skull-cracking debug sessions as anything else. Our required Numerical Methods course is in CS, it uses Matlab, our faculty is complaining that the students are complaining that they hate the course because they are spinning their wheels trying to get programs to run (in Matlab of all things), and we have guys in our department we want to teach Numerical Methods (in Matlab, of course), in the context of a watered-down Intro to Engineering offering.
What the community needs right now is a Python distro with enough of a numerics and graphics package rolled in to do 90 percent of what is in Matlab (Are the Python people still hashing out that Numerics/Numpy divide? Is there an engineering graphics library that is Numerics/Numpy compatible? 99 percent of what you do in Matlab is that you have a Leatherman Tool of a 2-D array type (Matlab, Matrix Lab) along with all of the libraries being compatible with that type.) CS departments could teach their Intro to Programming along with their Numerical Methods courses using that Python distro, and we can save a generation of engineers from brain damage.
I fail to see how Visual J++ was a dead end. Although they don't call it Visual J++ -- it is now called Visual Studio .NET C#. I had looked at Visual J++, and it is .NET version 0.8 in so many ways.
I will grant you that if you or the parents are shelling out the Purdue tuition, maybe their CS department should find a better professor for their intro course. I am sorry to hear that this experience dissuaded you from completing a CS degree, and there is probably a lot more to your personal story than can be shared on Slashdot.
But I would like to communicate to others out there that you will have a few good teachers in your educational career who are really inspiring, a vast group of average teachers, and a number of who you consider to be really, really bad teachers. The "bad" teachers are that way (in your opinion) for a number of reasons -- they may be "nice guys or gals" who don't have enough preparation or smarts to teach, they may have admitted to you gaps in their preparation that you have taken upon yourself to hold them in disrespect for, or maybe they assign too much HW and work you too hard.
If one is going to take a passive approach, show up to class and demand, "Here, educate me", that is a good way to fail at getting a degree and also to fail at every other opportunity that presents itself down the road. If one is going to take an active approach, working as hard as one can at learning from all teachers, the good and the bad, supplementing gaps in instruction with self-study, working coding jobs, group study, one is going to be successful at college and everything else.
To suggest that a person can have one "bad" prof means that they are on the street drinking methyl antifreeze out of a jar wrapped in a paper bag, this suggests a very passive approach to not just education but life in all its aspects.
Being the geeky son of an even geekier Dad, my momma often took it upon herself to wack me about the side of the head with a clue bat about young women I would encounter. I remember this one time we were in the Walgreens (why I remember this episode, I don't know, maybe I need to prove to myself I wasn't raised totally stupid). A friendly enough woman clerk checks us out, and as we walk out, Momma does the stage whisper to tell her child he is an idiot to tell me that this young woman was doing everything short of giving out her home phone number.
I responded in my own stage whisper a) I may not have been totally clued in on the signals given, but b) this young lady is earning minimum wage running a checkout counter in a Walgreens -- I thought you had more higher aspirations for the station in life you expect from a daughter-in-law.
Yeah, yeah, I know what you are all thinking, that my momma took a lot of back-talk from her adult son, and what kind of elitist thinks that a woman who happens to work a counter is to be looked down upon for a date. OK, the lady was pretty enough, friendly enough, hot enough, and I was available enough, but how about if I was pushing 30 at the time and the clerk in question was still in high school?
Men are supposed to be mind readers about when women are interested or not interested in them, and it is all very funny that men are geeks or oafs when they are interested in a woman who is not turned on by them and even bigger geeks and oafs when they fail to pick on when a woman is hot of them. Big deal, sometimes a man has selection criterea, even if they are as stupidly pragmatic as non-jailbait.
Two cores? Big whoop! Four cores? Haven't seen one, and our computing center keeps current with the latest generation desktop machines.
But factors of (almost) two or (almost) four speedup are no big deal in the grand scheme of things. Wake me up when they are talking about 10 cores, or 100 cores.
But that is the problem. Our resident computer architecture dude tells us that maybe 10 or 16 cores is an upper limit on account of cache memory or other considerations. To go beyond that they will need another approach -- clusters on a chip? For the low order of parallelism of current multi-core multi-processor offerings, I can't see the payoff in investing a lot of effort into developing, testing, and debugging parallel code.
Unless there are some revisions in the laws of physics, the nuclear processes throughout stellar evolution are well known based on computer models.
When helium "ash" accumulates in the core, helium fusion is not the next thing that happens. The core starts contracting and heating up, but that lights off H2 fusion in the shell surrounding the core. That phenomenon changes the luminosity and heat transfer rates of the star, causing the outer atmosphere to swell up into the red giant stage.
When shell burning runs its course, again the core contracts and heats up some more, resulting in the helium flash. Based on computer models, the helium flash is a major disruptive event caused by the sudden onset of helium fusion, it does not cause the star to go nova or anything, but it causes the star to change modes as it were, becoming somewhat bluer and smaller, but still more luminous than Main Sequence. From the computer models, it is believed that the upper-righthand HR diagram stars, red giants, are H2 shell burners while the horizontal branch above the Main Sequence represents He core burners.
For a massive enough star, exhaustion of core He will initiate shell He ignition, sending the star back into the red giant range, perhaps as a red supergiant for a massive star.
The red giant phase is only one phase of an evolved star. Everyone just kind of assumed that a star that goes supernova would be a red giant, but it seems like the star that popped off in Supernova 1987a in the LMS was blue.
Soviet Russia joke, . . . , tired of you!
The other good thing about Java is that the Java Native Interface (JNI) is reasonably well documented, and it lets Java call C/C++, and it lets C/C++ call back into Java. Maybe a little clunky and you have to understand the global and local locks thing on the native side or you can get some horrific bugs, but it is quite powerful once you get the hang of it.
My understanding of AJAX is that your programming is mainly in JavaScript to get code running on the client to do a lot of things that can only happen or happen efficiently on the client. Is there any way to call from JavaScript down to C/C++, either client side or server side?
It's that mandatory Circuits course that the mechanical engineers and others outside EE are required to take is what is breeding terrorists, I tell you. Requiring MEs to learn op amps is what is giving them that sour outlook on life.