I absolutely agree with your claim that languages should abstract away machine-level details.
However, representing bit-vectors as integers is an extremely common practice in C and other languages, and any competent programmer will recognize this idiom. (See for example the "flags" argument for the UNIX open(2) call.) With a little bit of (tasteful) preprocessor use, you can basically get the same sort of code as the example in your initial post. Finally, while you must put in some extra effort if data are to be shared between machines of different architectures, this practice is basically portable as well.
BTW, I know that IHBT, I figured that if someone stumbles across this thread, I didn't want them to be misinformed. FOAD, troll.
There's no reason to flame, and I certainly wasn't trolling you. (You may read my prior comments if you believe me to be a troll.)
My point (which, in retrospect, could have been a lot clearer) was that it seems extremely unlikely that Intel will retool what is admittedly a good evolution of a good microarchitecture to compete with chips based on a newer microarchitecture that Intel is currently pushing.
Your petition statement does acknowledge that business realities may make repackaging the Pentium M for desktop use prohibitively costly; I maintain that it would probably be politically difficult (both internally and from a marketing perspective) for Intel to do so as well -- because to do so would be to admit that the Pentium 4 core does not currently perform as well as the last revision of the P6 core. This would be like marketing Diet Coke against Coke -- when Coca-Cola did that, sales for both dropped off. Since there was a huge investment in the new microarchitecture, it doesn't make sense to not put marketing muscle behind it.
Finally, the P4 microarchitecture has a lot of potential; trace caches and the limited SMT alone are huge potential wins, and Intel basically has a pretty large percentage of the world's total population of competent architects since acquiring the Alpha group in 2001. If you're offended by me equating the Pentium M to the PPro, then you have to admit that the advancements in the P4 microarchitecture could really pay performance dividends after a few evolutionary revisions.
I'd be interested in seeing your movie, or in hearing the soundtrack.
A good music theory book is Tonal Harmony by Stefan Kostka and Dorothy Payne. (and older editions are cheap!) It's an undergrad-level text, but assumes no real prior knowledge of music. I think most schools use it over several semesters (my "borderline conservatory" undergrad music program took four semesters to go through the book, while augmenting it with acoustics, formal analysis, jazz theory, and more post-tonal theory), but you can pick and choose what you're interested in after the first few chapters.
If you're more interested in form than in harmony, there are probably better books, but most will presuppose knowledge similar to that contained in the Kostka/Payne book. One good book about form (not for the faint-of-heart) is Schoenberg's Structural Functions of Harmony.
Yes, but you're a programmer, not a computer scientist. Programming isn't an engineering discipline (if it were, many programmers would be in prison!). Rather, it's more of a weird amalgam of art, craft, and design, that may bear some resemblence to engineering, but without the licensure, best practices, or liability of engineering.
Tony Hoare's 1980 Turing award lecture describes these phenomena pretty well: get it here.
Of course, the indictment is not merely on professional programmers. Computer scientists could do their part by spending more time researching program understanding tools and reliability-enhancers like model checkers (check out SLAM, for example, from MS research) -- but a lot of PL research is going in that direction now. (Unfortunately, there's a social problem even more serious than the technical problem: getting tools and safe languages accepted outside of the ivory tower is hard as well.) Dijkstra said it pretty well in a CACM article for the millenial hoopla (this is a paraphrase): "I would posit that the central challenge of computing science -- 'to not make a mess of it' -- has not been met."
Why couldn't you just put a big static array of zeros in the code, and supply a secondary (closed source) program which overwrites those zeros with an actual key? You couldn't distribute pre-keyed binaries (since they wouldn't correspond to the source) but you can allow the end user to run a quick command to insert the keys into the binary him/herself.
Yeah, but then the modified binary would fail the signature check; you've just suggested nothing. If you send out a signed e-mail, someone can trivially edit it to say something unfavorable about one of your female relatives -- but then the recipient will know the message (as they received it) didn't come from you.
What Linus is proposing would allow people to run Linux on platforms that require the OS or binaries to be signed by some centralized authority. Presumably, to get a signature from a "trusted authority", the Linux kernel would be neutered to remove ptrace() and similar "circumvention device"/debugging system calls, and perhaps to add hooks into other system calls to acquire some kind of authorization before accessing the disk or network (i.e. so only a binary signed by the trusted authority could access files on a certain part of your disk) -- all of which is within the GPL so long as they share their neutering with the public under the GPL.
That's the real danger here -- consumer hardware that will only run Windows or a Linux kernel that delegates OS policy to whomever corporate America appoints as "trusted". Say goodbye to everything described in all the man pages in section 2.
Whoa, check out Intuem. It's $80 (with a trial), and worth it. I use it as a MIDI sketchpad, but it's functional enough to use for arranging and editing sequences, too. (Logic is the way to go for real work, IMHO, but I've been too busy hacking on AudioUnits lately to make much music.)
Furthermore, the "logic big box" is only like $250 -- an astonishing deal when you consider what it includes. There is nothing like that on Windows anymore.
Your statement that sequencer manufacturers have to "include separate drivers from each MIDI adapter manufacturer" is FUD, and wrong. Even in pre-OS X versions of MacOS, there was robust MIDI support in the form of OMS. So, the lack of MIDI Manager didn't really affect Mac MIDI stuff.
In OS X, there is the CoreMIDI API, which is the best MIDI API available on any platform (and I've written MIDI apps for most platforms that matter).
I think that there is a paucity of cheap sequencers for the Mac (as the original poster laments) because people who use the Mac for audio are typically doing professional work and need something like Logic, not Cakewalk or some other toy. It's also possible (and floated around the rumor mills) that Apple will come out with a music iApp soon, given their acquisition of emagic, which would take care of the low end of the market.
In any case, the CoreMIDI API includes a great sequencing engine, and it's just a matter of tacking a good GUI on to get a functional sequencer. As I state in another post, this has been done.
You can learn a lot about CS by looking at pre-1930 math and philosophy.
You'll see a lot of the genesis of proto-computer-science in Frege and Wittgenstein. Of course, it comes out full bore in Church and Turing. Any graduate-level logic text should cover combinatory logic (basically, with the lambda-calculus, the basis of functional programming). One such text (that I can recommend) is _Computability and Logic_, by Boolos.
Of course, SICP (as recommended elsewhere) is excellent, and the Sipser book (likewise) is a readable introduction to theory of computation.
What are you interested in? If you're interested in models for computation, check out the logic books. If you're interested in types (a la Russell), contrast logical types with PL types by reading Luca Cardelli's "Type Systems" (available online free from citeseer, use google). If you're interested in algorithms, Sedgewick or the massive MIT Press algorithms book is good.
You can't be hurt by reading the fundamentals, either: Church, Turing, Rosser, anything by Burstall or Hoare, and (if you're in for some fun), PL semantics (the seminal work in this area is Strachey/Scott). There is a good book available online called "Semantics with Applications" or something similar; I believe that the authors are named Nielsen.
Be careful, though. I was an undergrad philosophy major, and now I'm a grad student in programming languages.
You may want to call yourself a "computational natural scientist", especially if you're competent with parallel or distributed code and other facets of HPC. However, you say "computer programming" and not "computer science", so my guess is that you don't have the necessary CS background to do computational science.
However, it is unclear from your question why you spent seven years as an undergrad pursuing an ill-defined major (did you do it because you like botany, physics, and programming? was it just cool? or is there some thread that ties them together?). It is also unclear why it took you seven years to get what seems, from your description, to be basically a double major plus computer programming. You will probably want to focus your objectives before distilling how you spent seven years of your life into a major title, especially if you're hoping to impress grad school admissions committees or employers.
In fact, you'd probably be better off emphasizing one or two aspects of your education when selling yourself. There may be jobs available that demand equal parts of physics, botany, and programming, but I would guess that there are more jobs that require physics and programming or that require botany and programming. Emphasizing all three will paint you as an overspecialized goofball.
If, on the other hand, you're just pleased with yourself for graduating with what you think is an offbeat major, then do whatever you want.:-)
RedHat 6.2 uses a different glibc than the 7.x series. glibc 2.2, which is included with 7.x, is much better on 2.4 kernels, but applications compiled against an earlier glibc won't run on it. OTOH, using glibc 2.2 statically linked in on a 2.4 system imposes a huge performance hit IME.
Has anyone here used the VV dictation software on Linux? I'm interested in hearing how good it is, and whether or not it works with arbitrary applications (i.e. could I use it to dictate LaTeX to emacs).
Also, does it work with a reasonably modern distro? All the info I can find about it refers to RH 6.2.
You need to have the alternate resolutions defined in your XF86Config (or XF86Config-4) -- Here's the relevant section from mine (the "modes" should all be on one line):
Slashdot Mirror
I absolutely agree with your claim that languages should abstract away machine-level details.
However, representing bit-vectors as integers is an extremely common practice in C and other languages, and any competent programmer will recognize this idiom. (See for example the "flags" argument for the UNIX open(2) call.) With a little bit of (tasteful) preprocessor use, you can basically get the same sort of code as the example in your initial post. Finally, while you must put in some extra effort if data are to be shared between machines of different architectures, this practice is basically portable as well.
Not to be snide, but if you have a bit-vector of truth values, can't you just use ints?
There's no reason to flame, and I certainly wasn't trolling you. (You may read my prior comments if you believe me to be a troll.)
My point (which, in retrospect, could have been a lot clearer) was that it seems extremely unlikely that Intel will retool what is admittedly a good evolution of a good microarchitecture to compete with chips based on a newer microarchitecture that Intel is currently pushing.
Your petition statement does acknowledge that business realities may make repackaging the Pentium M for desktop use prohibitively costly; I maintain that it would probably be politically difficult (both internally and from a marketing perspective) for Intel to do so as well -- because to do so would be to admit that the Pentium 4 core does not currently perform as well as the last revision of the P6 core. This would be like marketing Diet Coke against Coke -- when Coca-Cola did that, sales for both dropped off. Since there was a huge investment in the new microarchitecture, it doesn't make sense to not put marketing muscle behind it.
Finally, the P4 microarchitecture has a lot of potential; trace caches and the limited SMT alone are huge potential wins, and Intel basically has a pretty large percentage of the world's total population of competent architects since acquiring the Alpha group in 2001. If you're offended by me equating the Pentium M to the PPro, then you have to admit that the advancements in the P4 microarchitecture could really pay performance dividends after a few evolutionary revisions.
Hey, there already is a "desktop pentium m" -- it's called the Pentium Pro.
Did you change your password with the "passwd" command? Be sure to change it in NetInfo (or in System Preferences -- I can't recall) also.
I'd be interested in seeing your movie, or in hearing the soundtrack.
A good music theory book is Tonal Harmony by Stefan Kostka and Dorothy Payne. (and older editions are cheap!) It's an undergrad-level text, but assumes no real prior knowledge of music. I think most schools use it over several semesters (my "borderline conservatory" undergrad music program took four semesters to go through the book, while augmenting it with acoustics, formal analysis, jazz theory, and more post-tonal theory), but you can pick and choose what you're interested in after the first few chapters.
If you're more interested in form than in harmony, there are probably better books, but most will presuppose knowledge similar to that contained in the Kostka/Payne book. One good book about form (not for the faint-of-heart) is Schoenberg's Structural Functions of Harmony.
Tony Hoare's 1980 Turing award lecture describes these phenomena pretty well: get it here.
Of course, the indictment is not merely on professional programmers. Computer scientists could do their part by spending more time researching program understanding tools and reliability-enhancers like model checkers (check out SLAM, for example, from MS research) -- but a lot of PL research is going in that direction now. (Unfortunately, there's a social problem even more serious than the technical problem: getting tools and safe languages accepted outside of the ivory tower is hard as well.) Dijkstra said it pretty well in a CACM article for the millenial hoopla (this is a paraphrase): "I would posit that the central challenge of computing science -- 'to not make a mess of it' -- has not been met."
Yeah, but then the modified binary would fail the signature check; you've just suggested nothing. If you send out a signed e-mail, someone can trivially edit it to say something unfavorable about one of your female relatives -- but then the recipient will know the message (as they received it) didn't come from you.
What Linus is proposing would allow people to run Linux on platforms that require the OS or binaries to be signed by some centralized authority. Presumably, to get a signature from a "trusted authority", the Linux kernel would be neutered to remove ptrace() and similar "circumvention device"/debugging system calls, and perhaps to add hooks into other system calls to acquire some kind of authorization before accessing the disk or network (i.e. so only a binary signed by the trusted authority could access files on a certain part of your disk) -- all of which is within the GPL so long as they share their neutering with the public under the GPL.
That's the real danger here -- consumer hardware that will only run Windows or a Linux kernel that delegates OS policy to whomever corporate America appoints as "trusted". Say goodbye to everything described in all the man pages in section 2.
check here: http://developers.slashdot.org/~woggo/journal/
Dude, Jack Valenti is older than Strom Thurmond. He doesn't have any young grandchildren!
Furthermore, the "logic big box" is only like $250 -- an astonishing deal when you consider what it includes. There is nothing like that on Windows anymore.
In OS X, there is the CoreMIDI API, which is the best MIDI API available on any platform (and I've written MIDI apps for most platforms that matter).
I think that there is a paucity of cheap sequencers for the Mac (as the original poster laments) because people who use the Mac for audio are typically doing professional work and need something like Logic, not Cakewalk or some other toy. It's also possible (and floated around the rumor mills) that Apple will come out with a music iApp soon, given their acquisition of emagic, which would take care of the low end of the market.
In any case, the CoreMIDI API includes a great sequencing engine, and it's just a matter of tacking a good GUI on to get a functional sequencer. As I state in another post, this has been done.
You can learn a lot about CS by looking at pre-1930 math and philosophy.
You'll see a lot of the genesis of proto-computer-science in Frege and Wittgenstein. Of course, it comes out full bore in Church and Turing. Any graduate-level logic text should cover combinatory logic (basically, with the lambda-calculus, the basis of functional programming). One such text (that I can recommend) is _Computability and Logic_, by Boolos.
Of course, SICP (as recommended elsewhere) is excellent, and the Sipser book (likewise) is a readable introduction to theory of computation.
What are you interested in? If you're interested in models for computation, check out the logic books. If you're interested in types (a la Russell), contrast logical types with PL types by reading Luca Cardelli's "Type Systems" (available online free from citeseer, use google). If you're interested in algorithms, Sedgewick or the massive MIT Press algorithms book is good.
You can't be hurt by reading the fundamentals, either: Church, Turing, Rosser, anything by Burstall or Hoare, and (if you're in for some fun), PL semantics (the seminal work in this area is Strachey/Scott). There is a good book available online called "Semantics with Applications" or something similar; I believe that the authors are named Nielsen.
Be careful, though. I was an undergrad philosophy major, and now I'm a grad student in programming languages.
Try audacity. http://audacity.sf.net
You may want to call yourself a "computational natural scientist", especially if you're competent with parallel or distributed code and other facets of HPC. However, you say "computer programming" and not "computer science", so my guess is that you don't have the necessary CS background to do computational science.
:-)
However, it is unclear from your question why you spent seven years as an undergrad pursuing an ill-defined major (did you do it because you like botany, physics, and programming? was it just cool? or is there some thread that ties them together?). It is also unclear why it took you seven years to get what seems, from your description, to be basically a double major plus computer programming. You will probably want to focus your objectives before distilling how you spent seven years of your life into a major title, especially if you're hoping to impress grad school admissions committees or employers.
In fact, you'd probably be better off emphasizing one or two aspects of your education when selling yourself. There may be jobs available that demand equal parts of physics, botany, and programming, but I would guess that there are more jobs that require physics and programming or that require botany and programming. Emphasizing all three will paint you as an overspecialized goofball.
If, on the other hand, you're just pleased with yourself for graduating with what you think is an offbeat major, then do whatever you want.
s/betyr/heter
Unnskyld foer min daarlig norsk. (Jeg bor i USA, men jeg longe studerte norsk i universitet, men jeg har glemte mange ord naa.)
Jeg synes at du snakke engelsk veldig godt!
---
Hva betyr en person som snakke tre spraak paa engelsk?
"Trilingual"
Hva betyr en person som snakke to spraak paa engelsk?
"Bilingual"
Hva betyr en person som snakke en spraak paa engelsk?
"American"
Where are you located? I seem to remember a "Record and Tape Traders" outside of DC, where I grew up.
Does anjuta allow you to use it with emacs?
check out this on OpenFirmware and this on partitioning your Mac.
RedHat 6.2 uses a different glibc than the 7.x series. glibc 2.2, which is included with 7.x, is much better on 2.4 kernels, but applications compiled against an earlier glibc won't run on it. OTOH, using glibc 2.2 statically linked in on a 2.4 system imposes a huge performance hit IME.
Also, does it work with a reasonably modern distro? All the info I can find about it refers to RH 6.2.
Is there an apt-rpm repository for Ximian? I'd sure like to be able to use apt-get instead of redcarpet.
So the student membership does qualify you for the hardware discount? That is pretty unclear from the site.
You need to have the alternate resolutions defined in your XF86Config (or XF86Config-4) -- Here's the relevant section from mine (the "modes" should all be on one line):
Section "Screen"
Identifier "Screen0"
Device "Matrox Millennium G400"
Monitor "Sony CPD-200ES"
DefaultDepth 16
Subsection "Display"
Depth 16
Modes "1152x864" "1024x768" "800x600" "640x480" "352x288" "352x240" "320x240" "320x200" EndSubSection
EndSection