Show me what Linux can do for a business, and I'll show you how Microsoft does it 20 times better.
My business is molecular modeling.
I need to do a lot of coding in C, C++, F77, and F90, along with some csh, ksh, bash, and perl scripting. I need to test the same code on my PC that runs on the 128-way SMP boxes in the high-performance computing facility, so I need compilers that support a POSIX-ish C api and MPICH, and I'll also need good (scriptable) connectivity ala ssh, scp, and rsync. Oh yeah, one of the data centers uses Kerberos. I also need reasonable data analysis tools like Matlab (though Octave will do in a pinch) and Maple. I need visualization tools like PyMOL, viewmol, vaspview, and GaussView, but also an X server so I can run beefier packages like Cerius2 directly off the big machines. I need to be able to write both small reports for quick printing and large (50+ page) papers with lots of mathematical formulas and endnotes/footnotes, and of course I need to output PDF. I also need virtual desktops to keep my workflow organized: desktop 1 is development, desktop 2 is remote terminals, desktop 3 is data analysis, and desktop 4 is general purpose desktop. Finally, I need to be able to back up my work easily, preferably with just a simple file copy, and all of my file formats will need to readable for 20+ years.
So far my needs are met at near zero cost with Debian Linux plus two commercial packages (GaussView and Maple). I have ssh, scp, rsync, perl, csh, ksh, bash, gcc/g++, g77, gfortran, MPICH, MPICH2, X11, LaTeX, Emacs, Octave, KMail, and OOo. And as a nice bonus with Debian my PC both plays DVDs (and ignores the UOP flag allowing me to skip directly to the menu) and browses the 'Net with ease, and so far I have had no problems with viruses.
I'm very interested in how a Microsoft solution will be 20 times better. Please tell us more!
I'm not a statistics expert, but I can sort of believe it. Essentially everyone who does NOT ingest caffeine does so for reasons that imply they do some other things differently too. (Anecdote: The only caffeine-free people I know are serious into either exercise or restricted diet so they wouldn't be appropriate for a general study.) It would be like finding a control group to measure the neurological effects of television exposure.
That was the opinion of an actual doctor I found via google. I wish I could find the link again.
I gradually came off caffeine last year after having had sodas, coffee, espresso, and tea off and on for 14 years. Two days after switching from small caffeinated sodas to small decaf sodas, the caffeine withdrawal headache began.
The headache was 5 days long and required 800 mg ibuprofen (4 Advil) + 440 mg naproxen sodium (2 Aleve) every 3-6 hours to relieve the pain. I could not sleep without painkillers, I could not think or go into work without painkillers. On the last day I actually went to a doctor seeking a better painkiller because I was down to 3-hour cycles on the Advil+Aleve and was worried about screwing my kidneys and liver. Fortunately that day the pain finally eased off and over the next three days the "memory" of the headache faded. However, I still felt sluggish and "off" for about eight weeks.
I discovered some interesting things during this time. First, some argue that no good studies about caffeine addiction have been produced in the Western world because over 90% of the population is regularly exposed to caffeine and a fully-caffeine-free control group cannot be assembled. Second, caffeine crosses through the placenta easily so most newborns are already exposed. Finally, one class of migraine headaches might be entirely explained as caffeine withdrawal, especially since caffeine is a major component of migraine treatment. I won't go so far as to say that all things non-"natural" are evil, but caffeine is definitely an insidious influence in American society. We don't need it like electricity, it provides zero value if taken daily, yet if most people went a few days without it they would be surprised at the intensity of the withdrawal effects.
Since getting off caffeine, I have noticed that I sleep much deeper and in general learn and retain information better, but my thinking speed is sometimes noticeably slower. I've had caffeine a few times to speed up when I need it, but I also taste the caffeine now clearly in sodas.
For me, I have to be careful with caffeine. I do NOT want that five-day migraine again, and the benefits are only rarely needed. My brother had a similar experience getting off, and my mother still cannot get off caffeine after using it for close to thirty years.
Though I know where you're coming from (having read The Screwing of the Average Man many years ago), you might get further traction if you pick your numbers more carefully.
The bottom edge of the top 5% sorted by income isn't obscene -- I think that is somewhere around $300-$500K USD per year. It's hard to dismiss the value of a $500K CEO whose wheeling and dealing keeps two hundred other people employed at $40K+ each -- that's $500K propping up $8 million in salaries.
Now, the bottom edge of the top 1% is getting there -- I believe that is near $1-$3 million. At $1 million you're talking near $500 per hour or $8 per minute. That means it costs a Fortune 500 company $50 every time a VP takes a piss.
The trick is to get the IRS figures and find good values to delineate the range between "making a lot of money by producing a lot of value for others" and "making so much it literally is obscene". I used to have the numbers handy back when I liked to argue about things, but I've forgotten them now.
Can I ask you something slightly OT? Since your sig indicates that "conservatives explain why they think you're wrong", I will assume that you are a conservative. If you aren't, please disregard.
I'm curious: what, exactly, IS the liberal agenda? What is the goal of the left-wing Democrats? What do Pelosi, Dean, Feingold, et al, actually want?
This may seem harsh, but maybe it would have been a "smart decision" not to borrow so much money. If you defer gratification until you have the means to afford it then you won't be paying interest to anyone.
That depends on a variety of criteria. Borrowing money to attend university that doubles one's annual earnings is a smart investment. Borrowing money to purchase a house such that the monthly mortgage and upkeep is very close to the cost of an apartment in the same area is also a good investment, as the money spent on apartment rent can instead be spent building home equity (and reducing income tax to boot).
Assuming the OP is referring to college and house, they likely made the right decisions to both maximize their earning potential and reduce sunk rent costs. And their point about still spending about half their lifetime income on interest may still be true if they live in an expensive area such as the California Bay Area.
The only criticism I would lay is in their decision not to move to a cheaper area such as the Midwest or into a rural area, but that too has its complications.
KDE has autorun? Do you mean the actions available under the "Peripherals -> Storage Media" item in Control Center? If so, then you must have a different KDE than me: all the actions listed in my menu are for media playback, media encoding, and opening in Konqueror. There is no action for executing code unless I add it myself. BTW this is Debian testing.
or shell rc files.
Could you elaborate? My first thought is that you mean the.login/.bashrc/.profile/etc files, but if those are set to fork-bomb then the user was already compromised before they logged in.
With the optimizer cranked up, the loop containing a bounds check was faster than the loop with the bounds check removed.
That actually makes sense to me. If your bounds check was very simple and the only loop outcome was breaking out (throw an exception, exit the loop, exit the function, etc., without altering the loop index), the optimizer could move it out of the loop entirely and alter the loop index check to incorporate the effect of the bounds check. Result is a one-time bounds check before entering the loop and a simplified loop, hence faster execution.
I remember in the discussion on the D compiler someone pointed this out.
Actually it is bad, so bad that Microsoft and a bunch of other keyboard manufactures moved the 'Insert' key which switches between those modes and made it only available via weird Fn-Key combinations.
So that's why new keyboards are so f'd up.
Rather than change the hardware, Microsoft might have considered implementing the software-only solution of 'joe': insert key inserts a space character, but insert/overwrite mode is selected via the editor toggle settings. Then their keyboard remains standard but people don't get confused.
One notable company that does is IBM - and their management staff is appreciably better than average, as most workers there will attest.
As a former IBMer who witnessed firsthand the official lies about the pension program (which screwed a whole generation of IBMers), the outsourcing program (which has eliminated many junior-level development positions in the USA), and the variable pay program (which is rigged to be a zero-sum game that has little bearing on actual performance), and the jockeying for position among 2nd-line managers that led quality staff to leave IBM entirely, and heard the horror stories from older IBMers about the internal culture during the anti-trust trial back in the 80's, I shudder to think that other companies are significantly worse.
I expect you know all this already, but I feel I should post it for others following the thread...
with one decent variant that is used and supported by everyone. Common things like threads, sockets, threads, connecting to databases, calling foreign functions...
Real threads are in SBCL, Allegro, and (I think) ECL. Multiprocessing ala mcclim tends to work everywhere else. The Common Lisp Cookbook should have code for sockets, databases is generally CLSQL, and foreign functions is generally CFFI (the evolved version of UFFI).
a less elistist and knee-jerk defensive community would help too.
comp.lang.lisp is showing many signs of evolution away from the flamewars that engulfed it over the last decade or so. In the last two years a new gamut of sites dedicated to both doing exactly what you are asking for above (see the Lisp gardeners project) and making a lot of newbie-friendly documentation available have come up. I won't say it's perfect, but I think the message has gotten across that a new generation of Lispers are ready to do their part and the existing community is opening to them.
I personally think Lisp is ready for another book that includes what you are asking for: network code, threads, FFI, SQL, ASDF, GUI, webapps, XML, crypto, and a section on specifically installing on both Windows and Unix at least the currently-major implementations (CLISP, CMUCL, SBCL, ECL, Allegro). Call it "_Really_ Practical Common Lisp" and have it produce one typical desktop application and one web application. Given that book, CL would probably become much more popular quickly.
frustrated that someone doesn't have the balls to start a company dedicated to making an absolutely, positively 100%-compatible Windows clone based on a Unix-like operating system.
First, a 100%-compatible clone of Windows based on Unix is not technically possible. Such a system would have to understand Windows filesystem semantics which do not translate into Unix semantics such as locking, being unable to delete open files, and Windows ACLs which cannot map to user/group/other. (Unix has ACLs too, but I don't think they work exactly as Windows ACLs do.) It would also need to do the full (enormous) Windows C API, including heavyweight processes, the "desktop" and "security context" stuff, asynchronous IO (much more complicated and some argue superior to Unix's select()/poll()), DirectX, and a heck of a lot of other stuff.
Second, the next best thing has already been done, FOUR times. 1) Whole-processor emulation ala QEMU/VmWare. 2) Windows API emulation ala Wine. 3) ReactOS, a direct clone of Windows NT. 4) A whole host of compability efforts with Windows networking (Samba), file formats (Linux kernel NTFS, FUSE NTFS driver), and programming languages (Mono).
If you want Windows, you can use Windows, emulation, or (soon) ReactOS. If you want Windows applications, you've got Wine. Plenty of people have shown they've got the "balls" to do everything short of outright purchasing you a copy of Windows.
I'm reporting what actually happened. Go build yourself a time machine, travel to IBM Building 500 in RTP circa spring 2001, and talk to the people in the WebSphere Applications group in room 240. My teammates at that time will be able to outline the various bugs encountered in the JVMs. Here is the OutOfMemoryError we saw: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id =4390238 . We didn't report the bug, we just found it after having encountered it independently.
Also, check the Top 25 Bugs at bugs.sun.com. Of those, here are the platform-specific bugs that are currently OPEN:
For anyone writing non-trivial applications, it is NOT HARD to encounter a situation where the JVM either behaves differently on platforms or has an outright bug. bugs.sun.com is filled with them.
If you can't do real programming, maybe it's best you stay away from professional's tools and stick with what you can do.
Ah of course, I must not be a "real programmer" because the 25-person team I was working with in the IBM WebSphere Applications group (*cough*) managed to discover some bugs in the JVMs. And all those people reporting JVM bugs at bugs.sun.com must be idiots too.
I don't care about MMPORGs so much since those tend to require huge fanbases and a Microsoft operating system. However, I think it would be awesome to have a F/OSS VR engine ala the "Metaverse" of Snowcrash or the "Matrix" of Neuromancer.
Some applications of such:
1. Training simulations. Emergency response, medical procedures, aviation, driving. All of these require learning how to coordinate with other people.
2. A VR space for business meetings.
3. A permanent VR space (ala Metaverse). Transfer files by "picking up a card" from another user.
4. A VR space grafted onto the real world as "enhanced vision".
A F/OSS engine + some initial usable artwork that could grow with more CC-licensed artwork could get the ball rolling for the Internet as sci-fi envisaged it twenty years ago.
Back in the days when JDK levels 1.2.2 and 1.3.1 were the bleeding edge of commercial application servers, it was actually quite easy to find 100% Pure Java code that would barf on one JVM/platform combo but work quite well elsewhere. Off the top of my head, here are some examples:
1. Rendering double-byte character sets on AIX and Solaris crashed the JVM for various combinations of language (ja and ko were notoriously fragile), JVM, and OS level.
2. Code that spawned processes worked correctly in Windows, AIX 4.3.3, and Linux, but failed under AIX 5L and Solaris.
3. Solaris would throw an erroneous OutOfMemoryError when lots of EJBs were loaded, but it worked fine under AIX 4.3.3, AIX 5L, Linux, and Windows.
Our project was a relatively simple and classic J2EE webapp with a few EJBs for adminstration, some JDBC for large dataset handling, and JSPs for the frontend. It had an "InstallShield Multi-Platform"(tm) Java-based installer. We generally found JVM errors at the rate of once a month. Most could be worked around, but a few had to be documented to the user, and some (such as running the installer in a particular double-byte character set) could not be solved at all given the existing necessary workarounds for other JVM bugs.
Perhaps the JVMs have improved in reliability and work now as you say they do. I moved away from Java after that project and I doubt I'll ever switch back now.
Ayn Rand would most certainly be a fan of Microsoft, a company that has wildly succeeded in the market and gained such a firm foothold that people cannot imagine that a PC can even run without Windows. And Microsoft has done it without government help and in spite of the efforts of a collectivist uprising to replace it with zero-cost software.
Then again, perhaps you have read Rand and already moved beyond Objectivism and are trying in a roundabout way to show that following Rand's ideas leads to the kind of market distortions that Microsoft demonstrates.
I want an OS that isn't in my face - I want it to run tasks without me having to KNOW i am running an OS... MS is the least shoddy there (Mac may be better - i can't get hold of it).
Funny, for me Linux/etc. with KDE are far superior to Windows in the "don't interrupt me while I'm doing work" business.
KDE notifications don't steal keyboard/mouse focus; desktop icons stay where I put them; inserting a DVD puts a DVD icon on the desktop that I can right-click->Play, and then the DVD player will actually allow me to skip ads and go straight to the movie; virtual desktops work without hosing the applications; the system monitor applet shows me CPU and memory usage in near real time so I know immediately if that loading that next application will cause swapping; clicking a PDF in Konqueror loads kpdf to view it rather than a whole separate Adobe application with its own splash screen; similarly, downloading a file in Konqueror loads Kget which will resume transfers automatically if I run out of disk space halfway through. I could go on for quite a while, the KDE folks have clearly spent a lot of time listening to their users.
I generally find that things are so much smoother under Linux (Debian Etch) that Windows feels really painful and in-your-face in comparison. I have even started moving away from Excel by using Octave and Maple instead, just so that I don't have to keep booting to Windows (Excel + Wine + read-only NTFS don't work well).
I'm going to buck a trend and say that C is actually a GREAT first language to learn:
1. It has a simple way to do console and file IO (scanf, printf) that will enable students to write and test simple programs.
2. It is used everywhere. As students learn more of C, they will be able to access and use the vast array of libraries available to C and go as deep as they want. When I was 16, it was common to learn C in order to start writing games, and then assembly language to make the fast loops faster. Another student learned assembly language in order to debug DOS viruses.
3. It is completely deterministic when run in one thread.
4. Its syntax is very simple. There are very few gotchas, most of which can be resolved with extra parenthesis.
Everyone else saying "pointers suck", "use a simpler language", etc. are forgetting two things about that first programming language. First, no matter what language you start with, it will be difficult to begin programming. The learning curve for Lisp/Python/Java/Ruby/etc. will be at least as steep as that for C for new programmers, and additionally there is little point in teaching OOP to people unfamiliar with structured programming and the problems OOP is meant to address. Second, it is frustrating to learn a language that has little use elsewhere, especially when it is because someone else claimed that it was better to know general concepts than to know a useful language.
First, by support the article refers to technical support contracts, not whether or not the software will actually run on Debian. And MySQL has decided that they will provide technical support only for a very limited subset of the popular Linux distros. As far is this issue is concerned, Debian is in the same boat as a lot of other distros and was not singled out for special treatment.
Second, the Mozilla trademark issue was at its core unavoidable. Debian has to be able to say to its derivative distros that everything in "main" is really free, Mozilla had copyrighted images that were NOT free, so Debian couldn't use them and Mozilla responded by saying they had to rename the browser. So they did, and the Mozilla-branded browser remains in "non-free" due to the copyrighted images. Everyone accusing Debian of hypocrisy on the trademark issue because they have an official logo is (to be blunt) wrong. Debian has an official logo (that they hardly ever use) to provide legal recourse to stop anyone else claiming to be Debian. It is otherwise of no use in the project and does nothing to prevent derivative distros from doing their own thing when they want to.
Incidentally, the Mozilla trademark dispute has caused me to reinvestigate my use of ALL software from Mozilla. I'm finding that KDE software is far more user-friendly and powerful than the Mozilla software across a number of applications. KMail can be made (rather easily) to store mail in ~/Mail in mbox format, its mail filters execute much faster, I can right-click -> "Create Filter" -> "Filter on From" in seconds, and in dozens of other ways it kicks mozilla-mail's ass. Likewise KNode, Konqueror, and Kontact.
Re:We have a bigger problem...
on
Saving U.S. Science
·
· Score: 4, Interesting
Science doesn't "surrport" manufacturing. High-level science and engineering invent things that are high-tech for a while, and are manufactured in the US as long as those things require a high-tech work force. Later they become commoditized and are moved offshore. By then we've moved on to something else.
That sounds well and good, but I don't think history supports it. Yes, VCRs started here and ended in Japan. Yes, computers started here and ended in Taiwan. But these days both manufacturing and research are rapidly moving. Taiwanese manufacturers are doing their own research and creating products with USA labels that have generated essentially no new USA expertise in the problem domain. Japanese cars are far ahead of USA cars in innovation and quality. In my major (chemical engineering), new plants are simply not being built anymore in the USA, so research in plant efficiency (which is an umbrella idea encompassing most of the traditional major) is beginning to stagnate; a number of experienced engineers have said that the future of the major lies in China now.
My major is transitioning to new fields, particularly bioengineering and nanotechnology design and fabrication, which may be supporting your point, but OTOH the new stuff is very bleeding edge and will not be ready for productizing for a decade or four. I don't see a good skills transition in the interim, because all the other nations are starting at about the same place we are and unlike USA they have the manufacturing facilities to experiment with.
I agree that we don't want Flint, MI, to be the entire USA. But we also shouldn't want USA to become uber-specialized Chiba City surrounded by a vast wasteland of poor serfs.
If I understand your logic, one could reasonably argue that *all* money in the U.S. is blood money of one sort or another. At some point, one must choose an arbitrary point at which to say "nothing before this counts". What point should that be?
Truth is, it IS all blood money going back millenia. Personally, I have an idea about what to do: use a nonlinear curve to tax assets. In essence it goes like this:
1. Pick an arbitrarily large point to be the ideal maximum asset. Let's call this X, and initially set it to $10 billion.
2. Assess a progressive tax for all assets beyond X such that after 30 years, ALL assets beyond X are eliminated.
3. For all assets below X, this tax does not apply.
4. Once every 50 years, X can be moved by popular vote.
The idea is that someone can be a self-made billionaire and enjoy the fruits of their labor and even pass on all of that wealth to their descendants. If the passed on wealth is ginormous (much larger than X), the descendants get an entire generation to spend that excess however they want to. If they do nothing, the same amount of wealth is passed down to the next generation. If they wish to grow their fortune really further beyond X, they can work to do so but then the next generation has to work much harder to keep that personal fortune so vast. And for fortunes valued near X, the interest gained on the assets can be spent every year as an unearned income salary.
One can make it rich and top out, or keep fighting for more wealth. Eventually many more people could top out and the definition of "middle class" might move upward. If (somehow) too many people reach X and prices get out of control for people far below X, X can move or the currency can be re-valued.
I think it would be a neat idea to try out.
An unworthy end to an otherwise well-argued post.
The phrase "point of a gun" (or "end of a gun") is a coded phrase for Libertarianism/Objectivism. The vast majority of Libertarians/Objectivists encountered online are not actually open for debate; for them debating means showing the world that they are right. If you are not one of those, sorry.
I read Ayn Rand at an impressionable time, liked it, tried it, and quickly discovered that it didn't work so well in personal life. Since then I've found numerous critiques of Libertarianism/Objectivism (for all practical purposes the same thing, though purists will differ) and remain committed to avoiding accepting any "one truth" that explains it all. People are complicated, but fundamentally human nature is constant regardless of social background. Some people steal, rich and poor. Some people work hard, rich and poor.
I think this discussion has illustrated clearly how differently people come at the issue depending on their background, and unfortunately it DOES seem to be breaking down into oversimplified categories. Here on/., people who encountered circumstances of rapid upward mobility assume everyone else can get rich if they only try hard enough. Others who experienced barriers trying to move up assume that very few must be able to make the big time. I'm only recently seeing in my past some opportunities to advance that I hadn't seen at the time, but I still don't see any dramatic rags-to-riches chances I passed up. So I tend to think that the problem is still complicated and ultimately a question of psychology and education rather than economics.
Slashdot Mirror
Show me what Linux can do for a business, and I'll show you how Microsoft does it 20 times better.
My business is molecular modeling.
I need to do a lot of coding in C, C++, F77, and F90, along with some csh, ksh, bash, and perl scripting. I need to test the same code on my PC that runs on the 128-way SMP boxes in the high-performance computing facility, so I need compilers that support a POSIX-ish C api and MPICH, and I'll also need good (scriptable) connectivity ala ssh, scp, and rsync. Oh yeah, one of the data centers uses Kerberos. I also need reasonable data analysis tools like Matlab (though Octave will do in a pinch) and Maple. I need visualization tools like PyMOL, viewmol, vaspview, and GaussView, but also an X server so I can run beefier packages like Cerius2 directly off the big machines. I need to be able to write both small reports for quick printing and large (50+ page) papers with lots of mathematical formulas and endnotes/footnotes, and of course I need to output PDF. I also need virtual desktops to keep my workflow organized: desktop 1 is development, desktop 2 is remote terminals, desktop 3 is data analysis, and desktop 4 is general purpose desktop. Finally, I need to be able to back up my work easily, preferably with just a simple file copy, and all of my file formats will need to readable for 20+ years.
So far my needs are met at near zero cost with Debian Linux plus two commercial packages (GaussView and Maple). I have ssh, scp, rsync, perl, csh, ksh, bash, gcc/g++, g77, gfortran, MPICH, MPICH2, X11, LaTeX, Emacs, Octave, KMail, and OOo. And as a nice bonus with Debian my PC both plays DVDs (and ignores the UOP flag allowing me to skip directly to the menu) and browses the 'Net with ease, and so far I have had no problems with viruses.
I'm very interested in how a Microsoft solution will be 20 times better. Please tell us more!
I'm not a statistics expert, but I can sort of believe it. Essentially everyone who does NOT ingest caffeine does so for reasons that imply they do some other things differently too. (Anecdote: The only caffeine-free people I know are serious into either exercise or restricted diet so they wouldn't be appropriate for a general study.) It would be like finding a control group to measure the neurological effects of television exposure.
That was the opinion of an actual doctor I found via google. I wish I could find the link again.
I gradually came off caffeine last year after having had sodas, coffee, espresso, and tea off and on for 14 years. Two days after switching from small caffeinated sodas to small decaf sodas, the caffeine withdrawal headache began.
The headache was 5 days long and required 800 mg ibuprofen (4 Advil) + 440 mg naproxen sodium (2 Aleve) every 3-6 hours to relieve the pain. I could not sleep without painkillers, I could not think or go into work without painkillers. On the last day I actually went to a doctor seeking a better painkiller because I was down to 3-hour cycles on the Advil+Aleve and was worried about screwing my kidneys and liver. Fortunately that day the pain finally eased off and over the next three days the "memory" of the headache faded. However, I still felt sluggish and "off" for about eight weeks.
I discovered some interesting things during this time. First, some argue that no good studies about caffeine addiction have been produced in the Western world because over 90% of the population is regularly exposed to caffeine and a fully-caffeine-free control group cannot be assembled. Second, caffeine crosses through the placenta easily so most newborns are already exposed. Finally, one class of migraine headaches might be entirely explained as caffeine withdrawal, especially since caffeine is a major component of migraine treatment. I won't go so far as to say that all things non-"natural" are evil, but caffeine is definitely an insidious influence in American society. We don't need it like electricity, it provides zero value if taken daily, yet if most people went a few days without it they would be surprised at the intensity of the withdrawal effects.
Since getting off caffeine, I have noticed that I sleep much deeper and in general learn and retain information better, but my thinking speed is sometimes noticeably slower. I've had caffeine a few times to speed up when I need it, but I also taste the caffeine now clearly in sodas.
For me, I have to be careful with caffeine. I do NOT want that five-day migraine again, and the benefits are only rarely needed. My brother had a similar experience getting off, and my mother still cannot get off caffeine after using it for close to thirty years.
I don't think anyone uses Maple anymore.
I know of several chemical and mechanical engineers who do (including me). For symbolic manipulation Maple rocks.
Though I know where you're coming from (having read The Screwing of the Average Man many years ago), you might get further traction if you pick your numbers more carefully.
The bottom edge of the top 5% sorted by income isn't obscene -- I think that is somewhere around $300-$500K USD per year. It's hard to dismiss the value of a $500K CEO whose wheeling and dealing keeps two hundred other people employed at $40K+ each -- that's $500K propping up $8 million in salaries.
Now, the bottom edge of the top 1% is getting there -- I believe that is near $1-$3 million. At $1 million you're talking near $500 per hour or $8 per minute. That means it costs a Fortune 500 company $50 every time a VP takes a piss.
The trick is to get the IRS figures and find good values to delineate the range between "making a lot of money by producing a lot of value for others" and "making so much it literally is obscene". I used to have the numbers handy back when I liked to argue about things, but I've forgotten them now.
Can I ask you something slightly OT? Since your sig indicates that "conservatives explain why they think you're wrong", I will assume that you are a conservative. If you aren't, please disregard.
I'm curious: what, exactly, IS the liberal agenda? What is the goal of the left-wing Democrats? What do Pelosi, Dean, Feingold, et al, actually want?
I'm genuinely curious, I don't want a flame war.
This may seem harsh, but maybe it would have been a "smart decision" not to borrow so much money. If you defer gratification until you have the means to afford it then you won't be paying interest to anyone.
That depends on a variety of criteria. Borrowing money to attend university that doubles one's annual earnings is a smart investment. Borrowing money to purchase a house such that the monthly mortgage and upkeep is very close to the cost of an apartment in the same area is also a good investment, as the money spent on apartment rent can instead be spent building home equity (and reducing income tax to boot).
Assuming the OP is referring to college and house, they likely made the right decisions to both maximize their earning potential and reduce sunk rent costs. And their point about still spending about half their lifetime income on interest may still be true if they live in an expensive area such as the California Bay Area.
The only criticism I would lay is in their decision not to move to a cheaper area such as the Midwest or into a rural area, but that too has its complications.
or KDE's equivalent feature,
.login/.bashrc/.profile/etc files, but if those are set to fork-bomb then the user was already compromised before they logged in.
KDE has autorun? Do you mean the actions available under the "Peripherals -> Storage Media" item in Control Center? If so, then you must have a different KDE than me: all the actions listed in my menu are for media playback, media encoding, and opening in Konqueror. There is no action for executing code unless I add it myself. BTW this is Debian testing.
or shell rc files.
Could you elaborate? My first thought is that you mean the
With the optimizer cranked up, the loop containing a bounds check was faster than the loop with the bounds check removed.
That actually makes sense to me. If your bounds check was very simple and the only loop outcome was breaking out (throw an exception, exit the loop, exit the function, etc., without altering the loop index), the optimizer could move it out of the loop entirely and alter the loop index check to incorporate the effect of the bounds check. Result is a one-time bounds check before entering the loop and a simplified loop, hence faster execution.
I remember in the discussion on the D compiler someone pointed this out.
Actually it is bad, so bad that Microsoft and a bunch of other keyboard manufactures moved the 'Insert' key which switches between those modes and made it only available via weird Fn-Key combinations.
So that's why new keyboards are so f'd up.
Rather than change the hardware, Microsoft might have considered implementing the software-only solution of 'joe': insert key inserts a space character, but insert/overwrite mode is selected via the editor toggle settings. Then their keyboard remains standard but people don't get confused.
If they discipline anyone except a white male directly, it is very likely they will be sued.
Care to back that up?
One notable company that does is IBM - and their management staff is appreciably better than average, as most workers there will attest.
As a former IBMer who witnessed firsthand the official lies about the pension program (which screwed a whole generation of IBMers), the outsourcing program (which has eliminated many junior-level development positions in the USA), and the variable pay program (which is rigged to be a zero-sum game that has little bearing on actual performance), and the jockeying for position among 2nd-line managers that led quality staff to leave IBM entirely, and heard the horror stories from older IBMers about the internal culture during the anti-trust trial back in the 80's, I shudder to think that other companies are significantly worse.
I expect you know all this already, but I feel I should post it for others following the thread...
with one decent variant that is used and supported by everyone. Common things like threads, sockets, threads, connecting to databases, calling foreign functions...
Real threads are in SBCL, Allegro, and (I think) ECL. Multiprocessing ala mcclim tends to work everywhere else. The Common Lisp Cookbook should have code for sockets, databases is generally CLSQL, and foreign functions is generally CFFI (the evolved version of UFFI).
a less elistist and knee-jerk defensive community would help too.
comp.lang.lisp is showing many signs of evolution away from the flamewars that engulfed it over the last decade or so. In the last two years a new gamut of sites dedicated to both doing exactly what you are asking for above (see the Lisp gardeners project) and making a lot of newbie-friendly documentation available have come up. I won't say it's perfect, but I think the message has gotten across that a new generation of Lispers are ready to do their part and the existing community is opening to them.
I personally think Lisp is ready for another book that includes what you are asking for: network code, threads, FFI, SQL, ASDF, GUI, webapps, XML, crypto, and a section on specifically installing on both Windows and Unix at least the currently-major implementations (CLISP, CMUCL, SBCL, ECL, Allegro). Call it "_Really_ Practical Common Lisp" and have it produce one typical desktop application and one web application. Given that book, CL would probably become much more popular quickly.
frustrated that someone doesn't have the balls to start a company dedicated to making an absolutely, positively 100%-compatible Windows clone based on a Unix-like operating system.
First, a 100%-compatible clone of Windows based on Unix is not technically possible. Such a system would have to understand Windows filesystem semantics which do not translate into Unix semantics such as locking, being unable to delete open files, and Windows ACLs which cannot map to user/group/other. (Unix has ACLs too, but I don't think they work exactly as Windows ACLs do.) It would also need to do the full (enormous) Windows C API, including heavyweight processes, the "desktop" and "security context" stuff, asynchronous IO (much more complicated and some argue superior to Unix's select()/poll()), DirectX, and a heck of a lot of other stuff.
Second, the next best thing has already been done, FOUR times. 1) Whole-processor emulation ala QEMU/VmWare. 2) Windows API emulation ala Wine. 3) ReactOS, a direct clone of Windows NT. 4) A whole host of compability efforts with Windows networking (Samba), file formats (Linux kernel NTFS, FUSE NTFS driver), and programming languages (Mono).
If you want Windows, you can use Windows, emulation, or (soon) ReactOS. If you want Windows applications, you've got Wine. Plenty of people have shown they've got the "balls" to do everything short of outright purchasing you a copy of Windows.
Wow, did I run over your dog or something?
d =4390238 . We didn't report the bug, we just found it after having encountered it independently.
d =4957990 (Solaris)d =4171239 (Windows)d =4787931 (Windows)d =4673298 (Linux)d =4049083 (Windows/Solaris)
I'm reporting what actually happened. Go build yourself a time machine, travel to IBM Building 500 in RTP circa spring 2001, and talk to the people in the WebSphere Applications group in room 240. My teammates at that time will be able to outline the various bugs encountered in the JVMs. Here is the OutOfMemoryError we saw: http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
Also, check the Top 25 Bugs at bugs.sun.com. Of those, here are the platform-specific bugs that are currently OPEN:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
For anyone writing non-trivial applications, it is NOT HARD to encounter a situation where the JVM either behaves differently on platforms or has an outright bug. bugs.sun.com is filled with them.
OutofMemory is an user error.
d =4390238 . Note that WE did not report the bug, we discovered the writeup only after having encountered it independently.
No it's not. Here is the actual bug report: http://bugs.sun.com/bugdatabase/view_bug.do?bug_i
If you can't do real programming, maybe it's best you stay away from professional's tools and stick with what you can do.
Ah of course, I must not be a "real programmer" because the 25-person team I was working with in the IBM WebSphere Applications group (*cough*) managed to discover some bugs in the JVMs. And all those people reporting JVM bugs at bugs.sun.com must be idiots too.
I don't care about MMPORGs so much since those tend to require huge fanbases and a Microsoft operating system. However, I think it would be awesome to have a F/OSS VR engine ala the "Metaverse" of Snowcrash or the "Matrix" of Neuromancer.
Some applications of such:
1. Training simulations. Emergency response, medical procedures, aviation, driving. All of these require learning how to coordinate with other people.
2. A VR space for business meetings.
3. A permanent VR space (ala Metaverse). Transfer files by "picking up a card" from another user.
4. A VR space grafted onto the real world as "enhanced vision".
A F/OSS engine + some initial usable artwork that could grow with more CC-licensed artwork could get the ball rolling for the Internet as sci-fi envisaged it twenty years ago.
Back in the days when JDK levels 1.2.2 and 1.3.1 were the bleeding edge of commercial application servers, it was actually quite easy to find 100% Pure Java code that would barf on one JVM/platform combo but work quite well elsewhere. Off the top of my head, here are some examples:
1. Rendering double-byte character sets on AIX and Solaris crashed the JVM for various combinations of language (ja and ko were notoriously fragile), JVM, and OS level.
2. Code that spawned processes worked correctly in Windows, AIX 4.3.3, and Linux, but failed under AIX 5L and Solaris.
3. Solaris would throw an erroneous OutOfMemoryError when lots of EJBs were loaded, but it worked fine under AIX 4.3.3, AIX 5L, Linux, and Windows.
Our project was a relatively simple and classic J2EE webapp with a few EJBs for adminstration, some JDBC for large dataset handling, and JSPs for the frontend. It had an "InstallShield Multi-Platform"(tm) Java-based installer. We generally found JVM errors at the rate of once a month. Most could be worked around, but a few had to be documented to the user, and some (such as running the installer in a particular double-byte character set) could not be solved at all given the existing necessary workarounds for other JVM bugs.
Perhaps the JVMs have improved in reliability and work now as you say they do. I moved away from Java after that project and I doubt I'll ever switch back now.
You installed Linux and it automagically recognised all you USB peripherals,
... yeah, right.
/etc .
Yes.
it installed nice drivers for your printer (no, I'm not talking about those crapy GIMP drivers)
The printer was a network printer, so I had to tell CUPS what IP it was on and what make it was. After that, yes it worked well.
and it also configured 3D acceleration to your graphic card.
I had to download the nVidia driver myself, exit X, and run the installer. After that it worked well.
You didn't had to configure fstab to mount your windows or other external partitions,
The Debian installer found all the other partitions and asked which ones to put in fstab.
nor to configure your bluethoot device, and you TV card worked right out of the box
Don't have those, so I don't know.
Also:
* My sound card and modem were autodetected at boot time and setup correctly.
* DVDs play right out of the box with Kaffeine. (No need to install a DVD codec like Windows.)
* K3b located my CD and DVD burners automatically.
None of these required messing with stuff in
Ayn Rand would most certainly be a fan of Microsoft, a company that has wildly succeeded in the market and gained such a firm foothold that people cannot imagine that a PC can even run without Windows. And Microsoft has done it without government help and in spite of the efforts of a collectivist uprising to replace it with zero-cost software.
Then again, perhaps you have read Rand and already moved beyond Objectivism and are trying in a roundabout way to show that following Rand's ideas leads to the kind of market distortions that Microsoft demonstrates.
I want an OS that isn't in my face - I want it to run tasks without me having to KNOW i am running an OS ... MS is the least shoddy there (Mac may be better - i can't get hold of it).
Funny, for me Linux/etc. with KDE are far superior to Windows in the "don't interrupt me while I'm doing work" business.
KDE notifications don't steal keyboard/mouse focus; desktop icons stay where I put them; inserting a DVD puts a DVD icon on the desktop that I can right-click->Play, and then the DVD player will actually allow me to skip ads and go straight to the movie; virtual desktops work without hosing the applications; the system monitor applet shows me CPU and memory usage in near real time so I know immediately if that loading that next application will cause swapping; clicking a PDF in Konqueror loads kpdf to view it rather than a whole separate Adobe application with its own splash screen; similarly, downloading a file in Konqueror loads Kget which will resume transfers automatically if I run out of disk space halfway through. I could go on for quite a while, the KDE folks have clearly spent a lot of time listening to their users.
I generally find that things are so much smoother under Linux (Debian Etch) that Windows feels really painful and in-your-face in comparison. I have even started moving away from Excel by using Octave and Maple instead, just so that I don't have to keep booting to Windows (Excel + Wine + read-only NTFS don't work well).
I'm going to buck a trend and say that C is actually a GREAT first language to learn:
1. It has a simple way to do console and file IO (scanf, printf) that will enable students to write and test simple programs.
2. It is used everywhere. As students learn more of C, they will be able to access and use the vast array of libraries available to C and go as deep as they want. When I was 16, it was common to learn C in order to start writing games, and then assembly language to make the fast loops faster. Another student learned assembly language in order to debug DOS viruses.
3. It is completely deterministic when run in one thread.
4. Its syntax is very simple. There are very few gotchas, most of which can be resolved with extra parenthesis.
Everyone else saying "pointers suck", "use a simpler language", etc. are forgetting two things about that first programming language. First, no matter what language you start with, it will be difficult to begin programming. The learning curve for Lisp/Python/Java/Ruby/etc. will be at least as steep as that for C for new programmers, and additionally there is little point in teaching OOP to people unfamiliar with structured programming and the problems OOP is meant to address. Second, it is frustrating to learn a language that has little use elsewhere, especially when it is because someone else claimed that it was better to know general concepts than to know a useful language.
First, by support the article refers to technical support contracts, not whether or not the software will actually run on Debian. And MySQL has decided that they will provide technical support only for a very limited subset of the popular Linux distros. As far is this issue is concerned, Debian is in the same boat as a lot of other distros and was not singled out for special treatment.
Second, the Mozilla trademark issue was at its core unavoidable. Debian has to be able to say to its derivative distros that everything in "main" is really free, Mozilla had copyrighted images that were NOT free, so Debian couldn't use them and Mozilla responded by saying they had to rename the browser. So they did, and the Mozilla-branded browser remains in "non-free" due to the copyrighted images. Everyone accusing Debian of hypocrisy on the trademark issue because they have an official logo is (to be blunt) wrong. Debian has an official logo (that they hardly ever use) to provide legal recourse to stop anyone else claiming to be Debian. It is otherwise of no use in the project and does nothing to prevent derivative distros from doing their own thing when they want to.
Incidentally, the Mozilla trademark dispute has caused me to reinvestigate my use of ALL software from Mozilla. I'm finding that KDE software is far more user-friendly and powerful than the Mozilla software across a number of applications. KMail can be made (rather easily) to store mail in ~/Mail in mbox format, its mail filters execute much faster, I can right-click -> "Create Filter" -> "Filter on From" in seconds, and in dozens of other ways it kicks mozilla-mail's ass. Likewise KNode, Konqueror, and Kontact.
Science doesn't "surrport" manufacturing. High-level science and engineering invent things that are high-tech for a while, and are manufactured in the US as long as those things require a high-tech work force. Later they become commoditized and are moved offshore. By then we've moved on to something else.
That sounds well and good, but I don't think history supports it. Yes, VCRs started here and ended in Japan. Yes, computers started here and ended in Taiwan. But these days both manufacturing and research are rapidly moving. Taiwanese manufacturers are doing their own research and creating products with USA labels that have generated essentially no new USA expertise in the problem domain. Japanese cars are far ahead of USA cars in innovation and quality. In my major (chemical engineering), new plants are simply not being built anymore in the USA, so research in plant efficiency (which is an umbrella idea encompassing most of the traditional major) is beginning to stagnate; a number of experienced engineers have said that the future of the major lies in China now.
My major is transitioning to new fields, particularly bioengineering and nanotechnology design and fabrication, which may be supporting your point, but OTOH the new stuff is very bleeding edge and will not be ready for productizing for a decade or four. I don't see a good skills transition in the interim, because all the other nations are starting at about the same place we are and unlike USA they have the manufacturing facilities to experiment with.
I agree that we don't want Flint, MI, to be the entire USA. But we also shouldn't want USA to become uber-specialized Chiba City surrounded by a vast wasteland of poor serfs.
If I understand your logic, one could reasonably argue that *all* money in the U.S. is blood money of one sort or another. At some point, one must choose an arbitrary point at which to say "nothing before this counts". What point should that be?
/., people who encountered circumstances of rapid upward mobility assume everyone else can get rich if they only try hard enough. Others who experienced barriers trying to move up assume that very few must be able to make the big time. I'm only recently seeing in my past some opportunities to advance that I hadn't seen at the time, but I still don't see any dramatic rags-to-riches chances I passed up. So I tend to think that the problem is still complicated and ultimately a question of psychology and education rather than economics.
Truth is, it IS all blood money going back millenia. Personally, I have an idea about what to do: use a nonlinear curve to tax assets. In essence it goes like this:
1. Pick an arbitrarily large point to be the ideal maximum asset. Let's call this X, and initially set it to $10 billion.
2. Assess a progressive tax for all assets beyond X such that after 30 years, ALL assets beyond X are eliminated.
3. For all assets below X, this tax does not apply.
4. Once every 50 years, X can be moved by popular vote.
The idea is that someone can be a self-made billionaire and enjoy the fruits of their labor and even pass on all of that wealth to their descendants. If the passed on wealth is ginormous (much larger than X), the descendants get an entire generation to spend that excess however they want to. If they do nothing, the same amount of wealth is passed down to the next generation. If they wish to grow their fortune really further beyond X, they can work to do so but then the next generation has to work much harder to keep that personal fortune so vast. And for fortunes valued near X, the interest gained on the assets can be spent every year as an unearned income salary.
One can make it rich and top out, or keep fighting for more wealth. Eventually many more people could top out and the definition of "middle class" might move upward. If (somehow) too many people reach X and prices get out of control for people far below X, X can move or the currency can be re-valued.
I think it would be a neat idea to try out.
An unworthy end to an otherwise well-argued post.
The phrase "point of a gun" (or "end of a gun") is a coded phrase for Libertarianism/Objectivism. The vast majority of Libertarians/Objectivists encountered online are not actually open for debate; for them debating means showing the world that they are right. If you are not one of those, sorry.
I read Ayn Rand at an impressionable time, liked it, tried it, and quickly discovered that it didn't work so well in personal life. Since then I've found numerous critiques of Libertarianism/Objectivism (for all practical purposes the same thing, though purists will differ) and remain committed to avoiding accepting any "one truth" that explains it all. People are complicated, but fundamentally human nature is constant regardless of social background. Some people steal, rich and poor. Some people work hard, rich and poor.
I think this discussion has illustrated clearly how differently people come at the issue depending on their background, and unfortunately it DOES seem to be breaking down into oversimplified categories. Here on