Considering he's the inventor of this *exact* idea -- that the server not retain information until the ACK (and that it do so by hashing various qualities of the initial SYN into an initial sequence number for the responding SYN+ACK.
It's a great concept. Read more about how it was created at http://cr.yp.to/syncookies.html .
Really, I'd like to see this implemented in the BSDs. Currently, they still create a pcb on the SYN and just drop random ones (during a SYN flood, random ones are most likely to be fake anyway) if there are too many (more than 4096, iirc). SYN cookies are a much better solution. I'm curious why they haven't been implemented already.
Please, check your facts before you complain--it's as simple as going to the ports tree and doing a 'make'.
'less' is licensed as follows (from LICENSE):
Less Copyright (C) 1984-2000 Mark Nudelman Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer. 2. Redistributions in binary form must reproduce the above copyright notice in the documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
and 'tcsh' is licensed as follows (from README):
Feel free to use it. These changes to csh may only be included in a commercial product if the inclusion or exclusion does not change the purchase price, level of support, etc. Please respect the individual authors by giving credit where credit is due (in other words, don't claim that you wrote portions that you haven't, and don't delete the names of the authors from the source code or documentation).
I just swung over there in my favorite text only browser (I've been holding off on X this install until 4.0 came out. It's out now, and it's compiling already:-)) w3m, and it was a *very* nice looking site. It's nice to see a company respecting those of us who still don't browse in pictures.
This kind of act could spell the end of commercial software's support in business, as no business wants to be stopped at the whim of another company. This could definitely be good for free software, since that will be the only software that aims to *preserve* consumer rights.
UCITA doesn't bother me in the least. It can't touch me.
The answer to any operation you do on an HP calculator is put on the first level of the stack. So basically, this is how it goes: Push as many operands until you get an operator. Then pop however many the operator needs, then push the answer back on the stack.
Whereas on a TI calculator, you type "( x + 1 ) / ( x + 2 )" for a total of 11 keystrokes, on an HP calculator, you would simply type "x 1 + x 2 +/" for a total of 7 keystrokes. And yes, i do find the second method more natural; as I said before, it's a case of environment.
As I mentioned before, the HP48 has had "pretty print" for years. Therefore what I see on my calculator does look exactly like what I see in my books.
This, however, is not how I see and think of equations in my head. I don't think "1+1" I think "1 1 +" because that's the way I am trained. I still maintain that this "interface from brain to calculator" is not inborn, but a learned behavior. I interface in RPN. In your words, my internal representation of a mathematical function is in RPN.
There's no arguing it, RPN is more efficient, keystroke-wise. Under some circumstances, RPN has fewer keystrokes than algebraic notation, however under no circumstances is the opposite true. Learn RPN and you'll be more efficient, it's that simple. Your brain will learn.
Does anyone else find it ironic that the company that makes the video with the absolute WORST 3d in linux is supposedly working on a fully compliant OpenGL subsystem?
I find that hard to believe given their apparent apathy to the current linux market.
Does your TI calculator use the Risch algorithm to do indefinite integrals? No. Does Maple, Mathematica, Mupad, or Derive use Risch? No. Does the hp48/49? Yes.
Does your TI calculator do Groebner bases? No. Does the hp48/49? Yes.
Can you do arbitrary precision calculation on TI calculators? No -- even with the TI89/92, you're capped at 10^612.:-) With the hp48/49, there is no artificial limit.
You say the majority of the world uses TI calculators. The majority of the world also uses windows. What's your point? Prevalence does not determine quality.
The HP48 uses open hardware to communicate with an open protocol. The TI calculators use closed hardware to communicate with a closed protocol.
Linux won't be ported to the TI calculator series because there's not enough memory and TI won't release enough specs for someone with enough interest to develop the standard C libraries on the calculator. So scratch that idea.
As far as durability is concerned, when an HP user speaks of HP durability, he's not talking about a year and half. Heck, he's probably not talking about 5 years. He's talking about professors still using their HP12s that they bought in the seventies. They're talking about people using their original HP48SX (bought in 1986) throughout high school and college and graduate studies. That's durability. An anti-slip pad coming loose after only 1.5 years of use is not durability.
I have an HP48GX, and I am in Calc IV. It works nicely for that. I used to use a TI-82, but it definately had limitations. I used it as a tool for basically two things: graphing functions and doing petty arithmetic. My upgrade to the 48 has given me a tool with significantly more capabilities, such as (especially) symbolic manipulation. Now I can let the calc do the dirty work while I concentrate on (a) putting the problem into an equation and (b) interpreting the results. Note that in this situation, I am actually doing most of the work. However, this work is what my mind can handle well. By adding the calculator to the picture, I have removed the necessity of doing all of the work in my head, but (here is the important part for this discussion) I have added the need to interface my brain with the calculator. On one level, I concentrate on what I can do, and on the other level, I concentrate on what the calc can do. There are thus two important requirements that I make of my calculator:
1.that what it can do and what I can do complement nicely, and 2.that how it works with things and how I work with things complement nicely.
With an TI and Algebraic Notation, I have to go through much more mental rigor to put what I'm thinking into something that it can understand. My brain does not work in a string; it works in a stack. My brain is postfix, the TI is infix. To have to constantly change around the two would require too much work for me to use it as a tool in the completion of a problem. I would only use it for arithmetic.
Ok, enough mimicry. The point here is that it's all a matter of environment. RPN was strange to me for the first week I used my calculator; after that, it became second nature, and in fact, became the way *I* think. Now, when I'm required to use my TI30X for tests that don't allow graphing calculators, I have problems doing even simple math because I think in RPN.
Think about what you have to do when you see a complicated equation in your math book and you want to put it into your calculator. You have to parse through the entire equation, looking for the places where the parentheses need to be in your calculator, and then input it. Most likely, in a truly complex equation, you'll miss a parenthese somewhere, press enter, and have your calculator tell you you've got mismatched parentheses. Then you have to parse through what you put in your calculator and debug your parenthese placement. Even in the best case, where you make no mistakes, there is quite a bit of thought and keystroking that goes into putting that equation in your calculator.
Contrast this with the way I input that equation into my HP48. I first look at the equation and zero in on the innermost expression. I then start building my equation subexpression by subexpression from the inside out. I never have to worry about parentheses, because my hp48 puts them in when necessary. I simply punch in variables and operators, steadily increasing the complexity of the equation as I go along, never having to think any more at one point than I would on another.
In computer terms, entering an equation printed in a book into an algebraic calculator would likely be an O(n) operation, since the complexitiy of entering the equation increases with the complexity of the equation. On the other hand, entering that same equation into an RPN calculator is likely an O(1) operation, since the complexity of entering the equation wouldn't change with the increasing complexity of the equation.
The point is, once you step away from your environmental training, RPN is no more complex, and is in fact easier than Algebraic notation. It just takes that first step.
As a side note, notice that this conversation followed a slightly LIFO-like structure. When I converse, I interrupt myself a lot. The unique thing that my friends have noted that I almost *always* return to previous topics of conversation in a LIFO manner; a stack.:-)
Algebraic is no more "natural" a way to do math than RPN; it's simply the method we grow up with. In computers, stack based math rules, as it should; it requires far less computational complexity. It's the same with humans--once someone steps away from their training, he finds that stack based math (and hence, RPN) is far superior.
I think (from your perspective) that you screwed up that proportion. As you wrote it, it is perfectly correct, though.
If you walk into an exam not knowing how to do a definite integral on your calculator, you're up a creek no matter what calculator you're using. Knowing the arbitrary ordering of variables in the definite integral function in a TI is *no different* than knowing an arbitrary stack ordering of those same variables in the definite integral command in an HP.
But you do neglect and incredibly important part about the HP: it has an equation editor. On my hp, I don't have to remember the order of the variables on the stack (though I do, regardless). All I have to do is open the equation editor and edit like I would if I was using a commcercial CAS like Maple.
Also worth pointing out is that the hp offers the exact same functional notation for its commands as the TI. Which means that if you can't figure out how to input a definite integral on the hp, you wouldn't be able to figure it out on a TI. In which case, you shouldn't be in a class that requires knowledge of definite integrals
Contrary to your statement, the original poster's first assertion is true. You can do more on the hp48 than you can on a TI calculator. It gives you more options than a simplified "Ti Basic" and an ASM command. No one ever claimed it was easy to learn linux, and yet people do, and they are far more happy with linux that they ever were with DOS. An hp48 may have a steeper learning curve, but it's certainly more rewarding when you reach a much higher summit than you could've on a TI calculator.
And as far as the original analogy goes, an hp calculator *does* use a standard communication interface (kermit or xmodem through an rs232 connector) unlike the TI, which uses both proprietary link hardware and protocols.
If you want games, stick with a TI, it has many more of them and does them better. But if you want mathematical programs, written by geniuses in both mathematics and computer programming, then you'll get an HP calculator.
Check out the archive at http://www.hp48.bismarck.k12.nd.us and compare it to any TI archive. What you'll see is that there are many more useful programs (as opposed to games) on the hp48 site.
The only calculator that even compares to an hp calculator is the TI89, and even that has been equalled/surpassed by the hp49.
The hp offers a rich programming environment, and an even richer selection of already programmed tools. The TI calculators can't even compare.
if you've lived with a TI calculator all your life.
First, let me start off by referring you to the premiere hp calculator website out there (whose domain is currently fscked up) www.hpcalc.org, currently known as www.hp48.bismarck.k12.nd.us .
Also, the hp49 is available but unlikely to be found in any stores; you'd have to buy it online, but its symbolic capabilities have been pumped up to the level (and some would say beyond) of a ti89. Having been created in 1986, the hp48 series' symbolic capabilities *out of box* are not quite up to the level of the ti89. There are numerous (free, gpl'ed) third party programs to help, though. Since I've no experience with the hp49, (my hp48 has served me well all the way through calc IV) I can't vouch for it's quality, but I know the programmers who wrote the software in the hp49 are the same programmers whose free software you'll be using if you get an hp48--you can pretty much consider the hp48 software a beta for the the hp49 software.
Anyway, on to your real question. The hp48 is incredibly programable. Basically, there are 3 languages: userRPL, sysRPL, and saturn ASM. RPL here stands for "Reverse Polish Lisp". I've never used lisp before, so I can't say how "lisp-like" it is, but that's what it's called. userRPL is the language you'll read about in the manual that you get with the calculator. It's where most user's programming begins and ends. Even so, despite the fact that it's the highest level (and slowest) of the three languages, it's still a very powerful tool for automation and various other tasks you might program for. This language would be the equivalent of TiBasic, but it (from all accounts I've heard--I've never used TiBasic myself) is more powerful.
If you're willing (and have the memory) to download Mika Heiskanen's excellent Jazz library, you can program in sysRPL, a lower level language than userRPL. sysRPL is the language created by the original hp48 architects (some of which still post to comp.sys.hp48) to create the operating system, mathematical functions, and utilities of the calculator. It's basically a set of memory addresses of assembly language routines that accomplish certain things; what Jazz does for you is give you mnemonics for those addresses, and compile your code into a machine readable format. sysRPL is what you'll be using most of the time when you decide to code anything of complexity. Anything you can do in userRPL, you can do in sysRPL, except in sysRPL it will probably be about 10 times faster. SysRPL also give you access to the low level graphics routines necessary for doing graphical programming on the hp48.
And, of course, beyond sysRPL is the saturn assembly. The saturn processor, which the hp48/9 series is based on, is a unique processor, and therefore has a unique assembly language. The saturn processer processes 4 bits per cycle (it's about a 3.76mhz processor), has 4 64bit working registers, 5 64bit saving registers, 2 20bit memory pointers and an 8 level return stack. This all adds up to a very unique assembly language
Beyond all that programming on the software side, the hardware side of the hp48 is also very well known and accessible. For the hp48G, there are numerous tutorials on the internet as to how to upgrade the ram from the original 32kb to up to 4mb. If you're on a budget and don't mind some soldering, this is what I suggest you do. The hp48GX, which can accept up to two expansion cards, had had sound cards, radio transmitters, and various other cards created for it--not to mention the ability to make your own memory cards rather than having to buy one (i have a homebuilt ram card in my 48 right now, in fact). You can find various schematics on the internet for the hp48, making it one of the most customizable calculators out there.
In all, your "geek experience" will be much greater with the hp48. You'll learn more about computer and electronics from it. And you'll enjoy it.
Jeremy
P.S. I didn't go into the "holy war" topic between TI users and HP users (RPN) for two reasons: the answer is self evident (RPN is superior) and because the hp49, which I would suggest you get if you get an hp calculator, supports both RPN and algebraic notation. And it adds HPBasic, yet another programming language to write in.
It's a perfectly logical statement. I think you may have just misinterpreted it.
The point here is that if I don't believe in ET, I'm not going to use my cpu cycles to look for him. If I know that the box over there in the corner is empty, I'm not going to go and look in the box just to make sure.
I believe strongly enough that there *isn't* extra terrestrial intelligence that I wouldn't feel productive looking for it. Saying that I don't belive in ET is simply my reasoning for not being involved in seti@home.
Believe it or not, there are some people who don't believe in extra terrestrial intelligence. And those people aren't all psycho fundies either. We choose to work on Distributed.net because to us, it's far more purposeful than seti at home, which would be like looking in an empty box for...something.
We could also run Prime95 but I would rather discover how easy (or hard) it is to crack encryption than waste my cpu cycles looking for a bigger prime number.
So, given the choice between looking for something in an empty box, or finding the next big prime number (which does, um, nothing for number theory) we choose to crack codes, which might prove useful (primarily in discovering what kind of computing power is necessary to do it).
Yeah, that's about the idea. An example is the difference in the linux kernels. 2.0 just put the entire kernel in a lock, so only one processor could be in the kernel at a time. 2.2 Added locks to the major subsystems, so one processor could be sheduling while another processor did some memory management. 2.4 promises even finer grained locks. For instance, 2.2 locks all wait queues as a whole, while 2.4 will lock each wait queue individually.
You'll have to pardon me here, but I don't see the purpose of this "workstation $DISPLAY swap" thing. Either way, you're running both an X server and a client on the same CPU, so you're still gonna be context switching quite a bit. Except after the swap, you've got network overhead too.
Someone offered to trade me his dual head G400 for my Voodoo3 3000. Even though it meant pretty much giving up quake3 (sure, I could still play, but it would be at about half the speed, which would be intolerable for me) I decided to go ahead with the deal because I knew that even though 3dfx does make decent linux drivers, I felt that Matrox's method, "open the specs and let the hacking begin" was more in line with my beliefs.
I'm happy to see that Matrox is not only meeting my expectations, but pleasantly surpassing them by going the extra mile to make sure their hardware is supported in linux. I'll make sure Matrox gets my money next time I actually buy a video card.
Slashdot Mirror
Considering he's the inventor of this *exact* idea -- that the server not retain information until the ACK (and that it do so by hashing various qualities of the initial SYN into an initial sequence number for the responding SYN+ACK.
It's a great concept. Read more about how it was created at http://cr.yp.to/syncookies.html .
Really, I'd like to see this implemented in the BSDs. Currently, they still create a pcb on the SYN and just drop random ones (during a SYN flood, random ones are most likely to be fake anyway) if there are too many (more than 4096, iirc). SYN cookies are a much better solution. I'm curious why they haven't been implemented already.
Jeremy
Perl is far faster than Java is, for every imagineable task. It's slightly slower than C, and in general will outpace even C++ with the STL.
You obviously haven't used Perl.
Jeremy
Um, the HP49 has had a translucent case before TI even considered the idea.
Not that it matters.
Jeremy
Please, check your facts before you complain--it's as simple as going to the ports tree and doing a 'make'.
'less' is licensed as follows (from LICENSE):
and 'tcsh' is licensed as follows (from README):
Please, check your facts before posting.
Jeremy
The X axis is the usefulness of what you can do in the language; the Y axis is is the amount of knowledge of the language you have learned thus far.
A subject with a steep learning curve, therefore, is a subject where much learning is involved to do relatively simple tasks.
Jeremy
I just swung over there in my favorite text only browser (I've been holding off on X this install until 4.0 came out. It's out now, and it's compiling already :-)) w3m, and it was a *very* nice looking site. It's nice to see a company respecting those of us who still don't browse in pictures.
Jeremy
I, personally, don't use commercial software.
This kind of act could spell the end of commercial software's support in business, as no business wants to be stopped at the whim of another company. This could definitely be good for free software, since that will be the only software that aims to *preserve* consumer rights.
UCITA doesn't bother me in the least. It can't touch me.
Jeremy
The answer to any operation you do on an HP calculator is put on the first level of the stack. So basically, this is how it goes: Push as many operands until you get an operator. Then pop however many the operator needs, then push the answer back on the stack.
/" for a total of 7 keystrokes. And yes, i do find the second method more natural; as I said before, it's a case of environment.
Whereas on a TI calculator, you type "( x + 1 ) / ( x + 2 )" for a total of 11 keystrokes, on an HP calculator, you would simply type "x 1 + x 2 +
As I mentioned before, the HP48 has had "pretty print" for years. Therefore what I see on my calculator does look exactly like what I see in my books.
This, however, is not how I see and think of equations in my head. I don't think "1+1" I think "1 1 +" because that's the way I am trained. I still maintain that this "interface from brain to calculator" is not inborn, but a learned behavior. I interface in RPN. In your words, my internal representation of a mathematical function is in RPN.
There's no arguing it, RPN is more efficient, keystroke-wise. Under some circumstances, RPN has fewer keystrokes than algebraic notation, however under no circumstances is the opposite true. Learn RPN and you'll be more efficient, it's that simple. Your brain will learn.
Jeremy
Does anyone else find it ironic that the company that makes the video with the absolute WORST 3d in linux is supposedly working on a fully compliant OpenGL subsystem?
I find that hard to believe given their apparent apathy to the current linux market.
Jeremy
The link for hpcalc.org is down. Use this instead:
http://www.hp48.bismarck.k12.nd.us
This link has enough adovcacy to make a decision about a calculator.
Jeremy
And that dictionary will help you on the math section how? I assume you don't know that the verbal section does not allow a calculator.
The SAT is simple. You don't need complex math nor a dictionary to score highly on it.
Jeremy
Palmtops are not legal on standardized tests, nor are they legal for math tests in most schools.
:-)
Have you ever looked into the multitude of PIMs and databases available for the HP48?
Jeremy
The TI programmers are smarter? Laughable.
:-) With the hp48/49, there is no artificial limit.
Does your TI calculator use the Risch algorithm to do indefinite integrals? No. Does Maple, Mathematica, Mupad, or Derive use Risch? No. Does the hp48/49? Yes.
Does your TI calculator do Groebner bases? No. Does the hp48/49? Yes.
Can you do arbitrary precision calculation on TI calculators? No -- even with the TI89/92, you're capped at 10^612.
You say the majority of the world uses TI calculators. The majority of the world also uses windows. What's your point? Prevalence does not determine quality.
Jeremy
The HP48 uses open hardware to communicate with an open protocol. The TI calculators use closed hardware to communicate with a closed protocol.
Linux won't be ported to the TI calculator series because there's not enough memory and TI won't release enough specs for someone with enough interest to develop the standard C libraries on the calculator. So scratch that idea.
As far as durability is concerned, when an HP user speaks of HP durability, he's not talking about a year and half. Heck, he's probably not talking about 5 years. He's talking about professors still using their HP12s that they bought in the seventies. They're talking about people using their original HP48SX (bought in 1986) throughout high school and college and graduate studies. That's durability. An anti-slip pad coming loose after only 1.5 years of use is not durability.
Jeremy
I have an HP48GX, and I am in Calc IV. It works nicely for that. I used to use a TI-82, but it definately had limitations. I used it as a tool for basically two things: graphing functions and doing petty arithmetic. My upgrade to the 48 has given me a tool with significantly more capabilities, such as (especially) symbolic manipulation. Now I can let the calc do the dirty work while I concentrate on (a) putting the problem into an equation and (b) interpreting the results. Note that in this situation, I am actually doing most of the work. However, this work is what my mind can handle well. By adding the calculator to the picture, I have removed the necessity of doing all of the work in my head, but (here is the important part for this discussion) I have added the need to interface my brain with the calculator. On one level, I concentrate on what I can do, and on the other level, I concentrate on what the calc can do. There are thus two important requirements that I make of my calculator:
:-)
1.that what it can do and what I can do complement nicely, and
2.that how it works with things and how I work with things complement nicely.
With an TI and Algebraic Notation, I have to go through much more mental rigor to put what I'm thinking into something that it can understand. My brain does not work in a string; it works in a stack. My brain is postfix, the TI is infix. To have to constantly change around the two would require too much work for me to use it as a tool in the completion of a problem. I would only use it for arithmetic.
Ok, enough mimicry. The point here is that it's all a matter of environment. RPN was strange to me for the first week I used my calculator; after that, it became second nature, and in fact, became the way *I* think. Now, when I'm required to use my TI30X for tests that don't allow graphing calculators, I have problems doing even simple math because I think in RPN.
Think about what you have to do when you see a complicated equation in your math book and you want to put it into your calculator. You have to parse through the entire equation, looking for the places where the parentheses need to be in your calculator, and then input it. Most likely, in a truly complex equation, you'll miss a parenthese somewhere, press enter, and have your calculator tell you you've got mismatched parentheses. Then you have to parse through what you put in your calculator and debug your parenthese placement. Even in the best case, where you make no mistakes, there is quite a bit of thought and keystroking that goes into putting that equation in your calculator.
Contrast this with the way I input that equation into my HP48. I first look at the equation and zero in on the innermost expression. I then start building my equation subexpression by subexpression from the inside out. I never have to worry about parentheses, because my hp48 puts them in when necessary. I simply punch in variables and operators, steadily increasing the complexity of the equation as I go along, never having to think any more at one point than I would on another.
In computer terms, entering an equation printed in a book into an algebraic calculator would likely be an O(n) operation, since the complexitiy of entering the equation increases with the complexity of the equation. On the other hand, entering that same equation into an RPN calculator is likely an O(1) operation, since the complexity of entering the equation wouldn't change with the increasing complexity of the equation.
The point is, once you step away from your environmental training, RPN is no more complex, and is in fact easier than Algebraic notation. It just takes that first step.
As a side note, notice that this conversation followed a slightly LIFO-like structure. When I converse, I interrupt myself a lot. The unique thing that my friends have noted that I almost *always* return to previous topics of conversation in a LIFO manner; a stack.
Algebraic is no more "natural" a way to do math than RPN; it's simply the method we grow up with. In computers, stack based math rules, as it should; it requires far less computational complexity. It's the same with humans--once someone steps away from their training, he finds that stack based math (and hence, RPN) is far superior.
Jeremy
As a note, since the release of the hp49, the hp48 has been available online for as low as $129. I don't have a link, but you can check around.
Jeremy
I think (from your perspective) that you screwed up that proportion. As you wrote it, it is perfectly correct, though.
If you walk into an exam not knowing how to do a definite integral on your calculator, you're up a creek no matter what calculator you're using. Knowing the arbitrary ordering of variables in the definite integral function in a TI is *no different* than knowing an arbitrary stack ordering of those same variables in the definite integral command in an HP.
But you do neglect and incredibly important part about the HP: it has an equation editor. On my hp, I don't have to remember the order of the variables on the stack (though I do, regardless). All I have to do is open the equation editor and edit like I would if I was using a commcercial CAS like Maple.
Also worth pointing out is that the hp offers the exact same functional notation for its commands as the TI. Which means that if you can't figure out how to input a definite integral on the hp, you wouldn't be able to figure it out on a TI. In which case, you shouldn't be in a class that requires knowledge of definite integrals
Contrary to your statement, the original poster's first assertion is true. You can do more on the hp48 than you can on a TI calculator. It gives you more options than a simplified "Ti Basic" and an ASM command. No one ever claimed it was easy to learn linux, and yet people do, and they are far more happy with linux that they ever were with DOS. An hp48 may have a steeper learning curve, but it's certainly more rewarding when you reach a much higher summit than you could've on a TI calculator.
And as far as the original analogy goes, an hp calculator *does* use a standard communication interface (kermit or xmodem through an rs232 connector) unlike the TI, which uses both proprietary link hardware and protocols.
Jeremy
More available to TI users? Whatever!
If you want games, stick with a TI, it has many more of them and does them better. But if you want mathematical programs, written by geniuses in both mathematics and computer programming, then you'll get an HP calculator.
Check out the archive at http://www.hp48.bismarck.k12.nd.us and compare it to any TI archive. What you'll see is that there are many more useful programs (as opposed to games) on the hp48 site.
The only calculator that even compares to an hp calculator is the TI89, and even that has been equalled/surpassed by the hp49.
The hp offers a rich programming environment, and an even richer selection of already programmed tools. The TI calculators can't even compare.
Jeremy
if you've lived with a TI calculator all your life.
First, let me start off by referring you to the premiere hp calculator website out there (whose domain is currently fscked up) www.hpcalc.org, currently known as www.hp48.bismarck.k12.nd.us .
Also, the hp49 is available but unlikely to be found in any stores; you'd have to buy it online, but its symbolic capabilities have been pumped up to the level (and some would say beyond) of a ti89. Having been created in 1986, the hp48 series' symbolic capabilities *out of box* are not quite up to the level of the ti89. There are numerous (free, gpl'ed) third party programs to help, though. Since I've no experience with the hp49, (my hp48 has served me well all the way through calc IV) I can't vouch for it's quality, but I know the programmers who wrote the software in the hp49 are the same programmers whose free software you'll be using if you get an hp48--you can pretty much consider the hp48 software a beta for the the hp49 software.
Anyway, on to your real question. The hp48 is incredibly programable. Basically, there are 3 languages: userRPL, sysRPL, and saturn ASM. RPL here stands for "Reverse Polish Lisp". I've never used lisp before, so I can't say how "lisp-like" it is, but that's what it's called. userRPL is the language you'll read about in the manual that you get with the calculator. It's where most user's programming begins and ends. Even so, despite the fact that it's the highest level (and slowest) of the three languages, it's still a very powerful tool for automation and various other tasks you might program for. This language would be the equivalent of TiBasic, but it (from all accounts I've heard--I've never used TiBasic myself) is more powerful.
If you're willing (and have the memory) to download Mika Heiskanen's excellent Jazz library, you can program in sysRPL, a lower level language than userRPL. sysRPL is the language created by the original hp48 architects (some of which still post to comp.sys.hp48) to create the operating system, mathematical functions, and utilities of the calculator. It's basically a set of memory addresses of assembly language routines that accomplish certain things; what Jazz does for you is give you mnemonics for those addresses, and compile your code into a machine readable format. sysRPL is what you'll be using most of the time when you decide to code anything of complexity. Anything you can do in userRPL, you can do in sysRPL, except in sysRPL it will probably be about 10 times faster. SysRPL also give you access to the low level graphics routines necessary for doing graphical programming on the hp48.
And, of course, beyond sysRPL is the saturn assembly. The saturn processor, which the hp48/9 series is based on, is a unique processor, and therefore has a unique assembly language. The saturn processer processes 4 bits per cycle (it's about a 3.76mhz processor), has 4 64bit working registers, 5 64bit saving registers, 2 20bit memory pointers and an 8 level return stack. This all adds up to a very unique assembly language
Beyond all that programming on the software side, the hardware side of the hp48 is also very well known and accessible. For the hp48G, there are numerous tutorials on the internet as to how to upgrade the ram from the original 32kb to up to 4mb. If you're on a budget and don't mind some soldering, this is what I suggest you do. The hp48GX, which can accept up to two expansion cards, had had sound cards, radio transmitters, and various other cards created for it--not to mention the ability to make your own memory cards rather than having to buy one (i have a homebuilt ram card in my 48 right now, in fact). You can find various schematics on the internet for the hp48, making it one of the most customizable calculators out there.
In all, your "geek experience" will be much greater with the hp48. You'll learn more about computer and electronics from it. And you'll enjoy it.
Jeremy
P.S. I didn't go into the "holy war" topic between TI users and HP users (RPN) for two reasons: the answer is self evident (RPN is superior) and because the hp49, which I would suggest you get if you get an hp calculator, supports both RPN and algebraic notation. And it adds HPBasic, yet another programming language to write in.
It's a perfectly logical statement. I think you may have just misinterpreted it.
The point here is that if I don't believe in ET, I'm not going to use my cpu cycles to look for him. If I know that the box over there in the corner is empty, I'm not going to go and look in the box just to make sure.
I believe strongly enough that there *isn't* extra terrestrial intelligence that I wouldn't feel productive looking for it. Saying that I don't belive in ET is simply my reasoning for not being involved in seti@home.
Jeremy
Believe it or not, there are some people who don't believe in extra terrestrial intelligence. And those people aren't all psycho fundies either. We choose to work on Distributed.net because to us, it's far more purposeful than seti at home, which would be like looking in an empty box for...something.
We could also run Prime95 but I would rather discover how easy (or hard) it is to crack encryption than waste my cpu cycles looking for a bigger prime number.
So, given the choice between looking for something in an empty box, or finding the next big prime number (which does, um, nothing for number theory) we choose to crack codes, which might prove useful (primarily in discovering what kind of computing power is necessary to do it).
Jeremy
Yeah, that's about the idea. An example is the difference in the linux kernels. 2.0 just put the entire kernel in a lock, so only one processor could be in the kernel at a time. 2.2 Added locks to the major subsystems, so one processor could be sheduling while another processor did some memory management. 2.4 promises even finer grained locks. For instance, 2.2 locks all wait queues as a whole, while 2.4 will lock each wait queue individually.
Jeremy
That variable is r_smp.
Type r_smp 1 at the console, and restart the game, and you've enabled q3's smp support..
Also important to note is that 3dfx cards do not support q3 in SMP mode.
Jeremy
You'll have to pardon me here, but I don't see the purpose of this "workstation $DISPLAY swap" thing. Either way, you're running both an X server and a client on the same CPU, so you're still gonna be context switching quite a bit. Except after the swap, you've got network overhead too.
Jeremy
Someone offered to trade me his dual head G400 for my Voodoo3 3000. Even though it meant pretty much giving up quake3 (sure, I could still play, but it would be at about half the speed, which would be intolerable for me) I decided to go ahead with the deal because I knew that even though 3dfx does make decent linux drivers, I felt that Matrox's method, "open the specs and let the hacking begin" was more in line with my beliefs.
I'm happy to see that Matrox is not only meeting my expectations, but pleasantly surpassing them by going the extra mile to make sure their hardware is supported in linux. I'll make sure Matrox gets my money next time I actually buy a video card.
Jeremy