Well, ASM in general isn't (IA-32... well, Intel screwed up. Learn something easier, say MIPS or 68K or HC11 or PPC or StrongARM).
Also, C compiler optimization is a limit of knowledge of the optimization programmers (i.e., how well do they know the architecture, and what language construct modifications can reduce the architecture instructions [taking advantage of instructions that do more than one thing at the same time {integer division often returns both the quotient and the remainder - save several hundred cycles by storing the remainder if its used later}]). While *in general* a C compiler can probably generate pretty good assembly, nothing does beat handcrafted ASM (with proper comments), rather than looking at the compiler output before assembly.
Likewise, from the Art of Assembly (book available online, don't have URL) by Randall Hyde, it's possible to create lousy assembly that runs slower than its C counterpart, but it's also possible to make asm much faster than a C compiler can ever generate. Plus, if you know the processor well, you can take advantage of architectural efficiencies (like helping branch prediction) by reordering asm code without having to look too convoluted. Additionally, you get some raw power that is difficult to do with C (bit twiddling is much easier with asm than C, plus, it's easy to initialize a large data memory allocation in asm to some specific value (initialize pointer. Write value using largest processor can do, using a post-increment addressing mode. Is value of pointer > end of data block [which can be of various types]? If so, end, else repeat). Of course, this has serious alignment issues.
Wow. so spacious! (keeping in line with this current thread)
Wait until you program for a system with either: 2k of ROM and a whopping *36* bytes of RAM (Basic Stamp II) or a nice 1k [= 1024 instructions] rom, 68 bytes of RAM (common PIC, I belive the 16F84). Of course, those who know PIC programming know we have a whopping 37 instructions with which to play with. >GG. Amazingly, the project fit in 5k of RAM (not counting the features we needed to add.)
Then again, having to do a processor in VHDL on a FPGA must be fun [my friends designed a processor (~10 instructions, 16 bits each), with "rom" of 256 instructions, and ram of 256 bytes... Even more fun.].
That I've noticed. When I did a program in VB (yes, yes I know, but it was easy and speed wasn't an issue [no one will notice if it took 1 second or.01 seconds to do what it had to do]), I spent far more time managing the windows than trying to do the actual code...
Something like 10% did the actual work, 90% just to manage the windows/buttons/etc portion.
Well, assembly is *NOT* hard to code with. Only the horrendous x86 architecture makes it seem so (Looking at all the exceptions and modes of the x86 makes my head spin). Basically, the x86 has a register set which is probably similar to a pocket calculator. [Honestly, certain instructions faster in say, AX rather than BX/CX/DX? Trying to optimize this is quite difficult. The 386 is popular because it generalized a few addressing modes across more registers (very helpful)]. Of course, MMX is a real mess (using the floating point register/stack for MMX manipulation, and having to reinitialize the registers if you want to do FP instructions (Intel tutorials)).
There *WERE* processors of the era that have a much nicer architecture (like the 68000), and many more registers [wow. 8 32-bit data registers, and 7+1 32 bit address registers...]. Yes, I'm biased. I love coding in assembly on a 68k processor. And I've found out how to code in assembly on a nice HC11 microcontroller [great fun! Register deprived, yes, but nothing too complicated, and very simple to code for].
Intel recognized this, which is why the later processors (Pentium II, Pentium III, etc) have more hardware-mapped registers that the hardware remaps to the x86 registers. (These registers can't be altered from software - they for out of order execution, etc).
People do program in assembly still, like Steve Gibson, and a link off of him has someone who *TRIED* to do DDE, and gave up after seeing its complexity. Still amazes me to see that SpinRite 5 is only 100k or so.
Re:Not all stupid, not all true
on
Dumb Laws
·
· Score: 1
Don't forget, some of these laws have been repealed a long time back. (There's a column in the paper that deals with such within-city items, and "Dumb Laws" sometimes appear occasionally. Usually they call up the local PD to verify it's truth - sometimes it's true, othertimes it's an urban legend, and usually, it's an outdated law that has long been struck off the books)
There is the occasional dumb law that slips through, though. But it's usually so stupid and obscure no one is ever caught under it. Gotta find it.
Heh. IMHO, I don't like mice, even when they worked.
However, I've hated trackballs worse (even though I prefer these), because all the gunk they collected really ruined scrolling/mousing[trackballing?].
Then I tried a touchpad. Nice, but not "precise" enough for me. Trackpoints were OK, but the tips kept wearing out (until you're pushing a mushy piece of rubber. When new, the IBM Trackpoints are really neat and easy to use [nice grippy texture until it wears down]). So I stuck using the mouse.
Then I found the excellent Trackman Marble FX (optical trackball). No problems mousing, and less frustration. Might get the optical M$ mouse, but would be kind of hard to get me to convert back to mice >G
Actually, it turns out that a lot of PC games have CD copyprotection, and it's not the kind where "Insert CD to continue." [Well, you need the CD in it to play the game... so much bloat...]
Basically, they do lots of standard fudging techniques (680 MB CDs, invalid TOC, etc), plus a few unique ones - unreadable tracks (tracks burned such that they don't conform to ISO spec, or tracks with bad checksums [similar to PlayStation]) that are checked at run time, or digital signatures/keys that checksum the entire CD in such a way that copies will always fail (how, I do not know).
Standard disclaimers apply. (BTW, making image copies of CDs, even your backup copy, can take ages because of these protections [CD-ROM drive just grinds to a halt reading errors on a CD]).
Hell. I always pick up the local *FREE* computer mags (which are 90% ads) *for* the ads. Heck, there's a few others computer mags that are free, with less ads, and I always skip over them.
Why? The local computer companies advertise a lot in there (and because of the competitiveness, the prices tend to be better than having to shop online for it [shipping, taxes...]). So, I just use the 'net for the few things I can't get locally.
Re:But, there was no year 1 on the calender ether,
on
Happy Odd Day!
·
· Score: 1
Millenium? I thought that started on 2001.
I thought people partied this year because of the Y2K BUG! [Yay! A break from all these computers] ;-)
1's and l's? tsk. A common irritant I have are those pesky l (lowercase L), and I's (uppercase i's). An amazing amount of fonts I use don't make the distinction (one vertical stroke).
PITA when you get passwords in email, and therefore can't guess if it's l or I.
(Hint to make more unmaintainable code: Choose a nice font that has this property. Lots do, don't worry, and make it the default font for all coders ).
I'm pretty sure the 386 (i386) didn't *HAVE* the SX/DX markings until much later in the production run (say, around the time that 486's got them).
Now, what you said about the busses, etc may be true, but the big difference (and noticable) of the 486SX and 486DX was that the DX's co-pro wasn't disabled. (In fact, the 487 co-pro for a 486SX was just a 486DX that took over after being handed control from the SX [Intel swapped a few pins to make it not pin-to-pin compatible, but the SX processor wasn't used except for initial booting - and acting as a space heater]).
I'm going to get one of these (after being recommended by someone else).
Omni*'s are *digital* switches, so the switch won't wear out (like the old mechanical one I've had to use. Doesn't switch properly at times... Avoid at all costs!).
Plus, they have keyboard/mouse emulators, so you don't have to boot the computer with the switch set to it (annoying as hell, and makes it near-impossible to utilize auto-shutdown/watchdog/autostartup programs).
Omnicubes are nice because they use standard cables (unlike omniviews, where you have to buy omniview cables only).
As standard, the Omni*'s have adaptor cables to utilize computers with serial mice, Sun machines (!), and Apple Macs, all from the same PC monitor, keyboard, and mouse.
Yay, but it's also a big pain to lug around a laptop with a 21" diagonal profile... (assuming that they leave half an inch around the edges for stuff).
Ugh, large displays are nice, but they're really getting in the way of portability (especially since the panel is weakest at the center, all the more so fragile). Soon we won't be able to bring them in as carry on baggage (hmm. A laptop with a 20" display... that will allow what, everyone in the same row to play games with you?).
I dunno, but what's the *resonant* frequency of the heads? (they're suspended over with a piece of metal... the classic mass-on-a-spring SHM problem). Car suspensions are good, but they're horrendous for high-frequency damping (the seat cushion absorbs most of the vibrations).
I really don't like the idea of the harddrive starting/stopping all the time (I dunno, with a hard drive that big, I don't think I'll rip many CDs at 128kbps... probably 256kbps or higher.)
Well, I've done it through telnet to SMTP servers, and seen email clients do it, but I'm sure a 2 second limit isn't much of a deterrent, because spammers probably load up the RCPT TO: lines up to the limit of the server. (Great fun. Leave the (B)CC'ing to the server).
Perhaps a combination of each connection only allowing a very low number (say, 10 at most) RCPT TO: commands, and each non-mailing list connection to be handled with a few extra seconds between connections (as well as limiting simulataneous connections to oh, 2 per machine?).
Well, can we all convince Creative Labs to release their nifty AWE64 Gold card in PCI format?
That, and I want at least *7* PCI slots in my next computer, to replace the 2 ISA slots that I'll lose (2 sound cards... 1 ISA for backwards compat, 1 PCI for modern sound).
Question is, how friggin' big are the heatsinks (and external fans) - assuming that you'll have to cover up the power supply fan slots as well (even though the slots are still smaller than what speeds a modern computer runs at, doesn't mean the waves stop, they just spread out more).
Hmm, with Intel building more and more space heaters, we'll need power supplies that cool without fans (heatsinks to external fans?). Or, having to put signs up - "CAUTION: HOT SURFACE" all over the computer, just to dissipate the megawatts of heat the next chip produces [Could it be that *intel* is intentionally trying to spy on people by making it impossible to shield a computer completely?
2B Systems makes RA2Wav, converts RA streams to WAV, and for all those pesky pnm:// stream servers, X-FileGet will get pnm:// streams (as well as the usual FTP/HTTP transfers).
Heh. Wince on a remote. Not only will you have to find a $300 remote, but every time you're finished, you have to charge it (lest the next day it'll leave you stuck on a pro-MS channel). I dunno, but having to nightly charge/change batteries weekly defeats the purpose of a remote.
Some people get seriously ill if they see or smell a bird. So, by having a bird in the window in my house, I might harm someone passing by. You simply can't make laws agoind [sic] something as vague as "harming other people". No nation has such a law.
IANAL, but I do know a homeowner is liable if anyone slips/falls/get hurt/etc on their sidewalk (even though the sidewalk isn't necessarily owned by said owner). I'm not clear if this extends onto the road (I doubt it), but it's quite possible if your bird causes someone to get ill, you'd be liable for it, *if* it happened on your side and stretch of the sidewalk (and they decide to press charges).
A reason why most people tend to shovel the snow off of their part of the sidewalk when it snows (patch of ice underneath snow is especially dangerous). I'm wary of people who don't shovel the snow (I end up walking on the road... much safer, plus I hope to keep out of the hospital).
Star Wars fans, in particular, can't seem to comprehend the idea of sci-fi without constant fight scenes and explosions. So, I guess what I want to say is, no, just because you don't get it doesn't mean it sucks.
Well, not a star wars fan myself, but if you look deeper at eps 4, 5, 6 (A New Hope, Empire Strikes Back, and Return of the Jedi), they actually had a cohesive subtext. It's a basic plot, yes, but the symbolism is amazing - look at the Storm Troopers for an example. Their uniforms are white for a reason (and basically similar) - if you associate white with purity, it can be thought of purity, but with bad associates (i.e., the Force gone wrong, aka the Dark side).
Only problem is the need to strip away the effects to get at the subtext takes a lot of the willingness to watch away. (And yes, I learned this through a class called Religion 12b. Nice course [had to take it]. Watched a stripped down version of 2001, which also has a similar subtext hidden away).
Back to this parent article, I wholeheartedly agree, that Trek is more than Trek - it actually is a reflection of the state of the world as it is/was (TOS: A *RUSSIAN* on board... with *AMERICANS*. Working *together*.) In fact, most of past sci-fi actually did reflect the state of the world (but now, movies are now 1 1/2 hours of effects stiched together with one of the 14 Standard Hollywood Plots(tm)).
Slashdot Mirror
ASM != arcane and difficult.
Well, ASM in general isn't (IA-32... well, Intel screwed up. Learn something easier, say MIPS or 68K or HC11 or PPC or StrongARM).
Also, C compiler optimization is a limit of knowledge of the optimization programmers (i.e., how well do they know the architecture, and what language construct modifications can reduce the architecture instructions [taking advantage of instructions that do more than one thing at the same time {integer division often returns both the quotient and the remainder - save several hundred cycles by storing the remainder if its used later}]). While *in general* a C compiler can probably generate pretty good assembly, nothing does beat handcrafted ASM (with proper comments), rather than looking at the compiler output before assembly.
Likewise, from the Art of Assembly (book available online, don't have URL) by Randall Hyde, it's possible to create lousy assembly that runs slower than its C counterpart, but it's also possible to make asm much faster than a C compiler can ever generate. Plus, if you know the processor well, you can take advantage of architectural efficiencies (like helping branch prediction) by reordering asm code without having to look too convoluted. Additionally, you get some raw power that is difficult to do with C (bit twiddling is much easier with asm than C, plus, it's easy to initialize a large data memory allocation in asm to some specific value (initialize pointer. Write value using largest processor can do, using a post-increment addressing mode. Is value of pointer > end of data block [which can be of various types]? If so, end, else repeat). Of course, this has serious alignment issues.
Wow. so spacious! (keeping in line with this current thread)
Wait until you program for a system with either:
2k of ROM and a whopping *36* bytes of RAM (Basic Stamp II) or a nice 1k [= 1024 instructions] rom, 68 bytes of RAM (common PIC, I belive the 16F84). Of course, those who know PIC programming know we have a whopping 37 instructions with which to play with. >GG. Amazingly, the project fit in 5k of RAM (not counting the features we needed to add.)
Then again, having to do a processor in VHDL on a FPGA must be fun [my friends designed a processor (~10 instructions, 16 bits each), with "rom" of 256 instructions, and ram of 256 bytes... Even more fun.].
That I've noticed. When I did a program in VB (yes, yes I know, but it was easy and speed wasn't an issue [no one will notice if it took 1 second or .01 seconds to do what it had to do]), I spent far more time managing the windows than trying to do the actual code...
Something like 10% did the actual work, 90% just to manage the windows/buttons/etc portion.
Well, assembly is *NOT* hard to code with. Only the horrendous x86 architecture makes it seem so (Looking at all the exceptions and modes of the x86 makes my head spin). Basically, the x86 has a register set which is probably similar to a pocket calculator. [Honestly, certain instructions faster in say, AX rather than BX/CX/DX? Trying to optimize this is quite difficult. The 386 is popular because it generalized a few addressing modes across more registers (very helpful)]. Of course, MMX is a real mess (using the floating point register/stack for MMX manipulation, and having to reinitialize the registers if you want to do FP instructions (Intel tutorials)).
There *WERE* processors of the era that have a much nicer architecture (like the 68000), and many more registers [wow. 8 32-bit data registers, and 7+1 32 bit address registers...]. Yes, I'm biased. I love coding in assembly on a 68k processor. And I've found out how to code in assembly on a nice HC11 microcontroller [great fun! Register deprived, yes, but nothing too complicated, and very simple to code for].
Intel recognized this, which is why the later processors (Pentium II, Pentium III, etc) have more hardware-mapped registers that the hardware remaps to the x86 registers. (These registers can't be altered from software - they for out of order execution, etc).
People do program in assembly still, like Steve Gibson, and a link off of him has someone who *TRIED* to do DDE, and gave up after seeing its complexity. Still amazes me to see that SpinRite 5 is only 100k or so.
Don't forget, some of these laws have been repealed a long time back. (There's a column in the paper that deals with such within-city items, and "Dumb Laws" sometimes appear occasionally. Usually they call up the local PD to verify it's truth - sometimes it's true, othertimes it's an urban legend, and usually, it's an outdated law that has long been struck off the books)
There is the occasional dumb law that slips through, though. But it's usually so stupid and obscure no one is ever caught under it. Gotta find it.
*YES* Good book. I've read it (borrowed it from physics teacher in grade 12). BTW, doesn't chapters.ca ship to Canada only?
FWIW, I'll throw in "The Elegant Universe" by Brian Greene. A book all about superstring theory (and ever so interesting).
Hmm. How did the parent of this thread get scored Score 5: Troll?
IANAPL, but...
Well, that's true everywhere, except the US (And possibly Canada).
In the US, you're allowed to patent something that you've published, as long as the patent is filed *within a year* of publication.
Wierd, yes, considering no one else in the world does it (except possibly Canada).
Heh. IMHO, I don't like mice, even when they worked.
However, I've hated trackballs worse (even though I prefer these), because all the gunk they collected really ruined scrolling/mousing[trackballing?].
Then I tried a touchpad. Nice, but not "precise" enough for me. Trackpoints were OK, but the tips kept wearing out (until you're pushing a mushy piece of rubber. When new, the IBM Trackpoints are really neat and easy to use [nice grippy texture until it wears down]). So I stuck using the mouse.
Then I found the excellent Trackman Marble FX (optical trackball). No problems mousing, and less frustration. Might get the optical M$ mouse, but would be kind of hard to get me to convert back to mice >G
Actually, it turns out that a lot of PC games have CD copyprotection, and it's not the kind where "Insert CD to continue." [Well, you need the CD in it to play the game... so much bloat...]
Basically, they do lots of standard fudging techniques (680 MB CDs, invalid TOC, etc), plus a few unique ones - unreadable tracks (tracks burned such that they don't conform to ISO spec, or tracks with bad checksums [similar to PlayStation]) that are checked at run time, or digital signatures/keys that checksum the entire CD in such a way that copies will always fail (how, I do not know).
Check out: GameCopyWorld (www.gamecopyworld.com) for more info.
Standard disclaimers apply. (BTW, making image copies of CDs, even your backup copy, can take ages because of these protections [CD-ROM drive just grinds to a halt reading errors on a CD]).
Hell. I always pick up the local *FREE* computer mags (which are 90% ads) *for* the ads. Heck, there's a few others computer mags that are free, with less ads, and I always skip over them.
Why? The local computer companies advertise a lot in there (and because of the competitiveness, the prices tend to be better than having to shop online for it [shipping, taxes...]). So, I just use the 'net for the few things I can't get locally.
Millenium? I thought that started on 2001.
I thought people partied this year because of the Y2K BUG! [Yay! A break from all these computers]
;-)
1's and l's? tsk. A common irritant I have are those pesky l (lowercase L), and I's (uppercase i's). An amazing amount of fonts I use don't make the distinction (one vertical stroke).
PITA when you get passwords in email, and therefore can't guess if it's l or I.
(Hint to make more unmaintainable code: Choose a nice font that has this property. Lots do, don't worry, and make it the default font for all coders ).
*Excellent* book!
Read that book, and loved all of it (certainly made me more knowledable in the subject).
GUT was a subject that wasn't covered in my physics courses }}:-(.
I'm pretty sure the 386 (i386) didn't *HAVE* the SX/DX markings until much later in the production run (say, around the time that 486's got them).
Now, what you said about the busses, etc may be true, but the big difference (and noticable) of the 486SX and 486DX was that the DX's co-pro wasn't disabled. (In fact, the 487 co-pro for a 486SX was just a 486DX that took over after being handed control from the SX [Intel swapped a few pins to make it not pin-to-pin compatible, but the SX processor wasn't used except for initial booting - and acting as a space heater]).
I'm going to get one of these (after being recommended by someone else).
Omni*'s are *digital* switches, so the switch won't wear out (like the old mechanical one I've had to use. Doesn't switch properly at times... Avoid at all costs!).
Plus, they have keyboard/mouse emulators, so you don't have to boot the computer with the switch set to it (annoying as hell, and makes it near-impossible to utilize auto-shutdown/watchdog/autostartup programs).
Omnicubes are nice because they use standard cables (unlike omniviews, where you have to buy omniview cables only).
As standard, the Omni*'s have adaptor cables to utilize computers with serial mice, Sun machines (!), and Apple Macs, all from the same PC monitor, keyboard, and mouse.
Yay, but it's also a big pain to lug around a laptop with a 21" diagonal profile... (assuming that they leave half an inch around the edges for stuff).
Ugh, large displays are nice, but they're really getting in the way of portability (especially since the panel is weakest at the center, all the more so fragile). Soon we won't be able to bring them in as carry on baggage (hmm. A laptop with a 20" display... that will allow what, everyone in the same row to play games with you?).
But, I won't mind a 20" LCD panel on my desk.
I dunno, but what's the *resonant* frequency of the heads? (they're suspended over with a piece of metal... the classic mass-on-a-spring SHM problem). Car suspensions are good, but they're horrendous for high-frequency damping (the seat cushion absorbs most of the vibrations).
I really don't like the idea of the harddrive starting/stopping all the time (I dunno, with a hard drive that big, I don't think I'll rip many CDs at 128kbps... probably 256kbps or higher.)
Well, I've done it through telnet to SMTP servers, and seen email clients do it, but I'm sure a 2 second limit isn't much of a deterrent, because spammers probably load up the RCPT TO: lines up to the limit of the server. (Great fun. Leave the (B)CC'ing to the server).
Perhaps a combination of each connection only allowing a very low number (say, 10 at most) RCPT TO: commands, and each non-mailing list connection to be handled with a few extra seconds between connections (as well as limiting simulataneous connections to oh, 2 per machine?).
Well, can we all convince Creative Labs to release their nifty AWE64 Gold card in PCI format?
That, and I want at least *7* PCI slots in my next computer, to replace the 2 ISA slots that I'll lose (2 sound cards... 1 ISA for backwards compat, 1 PCI for modern sound).
Question is, how friggin' big are the heatsinks (and external fans) - assuming that you'll have to cover up the power supply fan slots as well (even though the slots are still smaller than what speeds a modern computer runs at, doesn't mean the waves stop, they just spread out more).
Hmm, with Intel building more and more space heaters, we'll need power supplies that cool without fans (heatsinks to external fans?). Or, having to put signs up - "CAUTION: HOT SURFACE" all over the computer, just to dissipate the megawatts of heat the next chip produces [Could it be that *intel* is intentionally trying to spy on people by making it impossible to shield a computer completely?
Does that cover RM G2 formats, too?
(BTW, here's anothe rprogram for Win* users):
2B Systems makes RA2Wav, converts RA streams to WAV, and for all those pesky pnm:// stream servers, X-FileGet will get pnm:// streams (as well as the usual FTP/HTTP transfers).
Heh. Wince on a remote. Not only will you have to find a $300 remote, but every time you're finished, you have to charge it (lest the next day it'll leave you stuck on a pro-MS channel). I dunno, but having to nightly charge/change batteries weekly defeats the purpose of a remote.
It's off to the Ministry of Love for you.
Oceania has always been at war with Eurasia.
Some people get seriously ill if they see or smell a bird. So, by having a bird in the window in my house, I might harm someone passing by. You simply can't make laws agoind [sic] something as vague as "harming other people". No nation has such a law.
IANAL, but I do know a homeowner is liable if anyone slips/falls/get hurt/etc on their sidewalk (even though the sidewalk isn't necessarily owned by said owner). I'm not clear if this extends onto the road (I doubt it), but it's quite possible if your bird causes someone to get ill, you'd be liable for it, *if* it happened on your side and stretch of the sidewalk (and they decide to press charges).
A reason why most people tend to shovel the snow off of their part of the sidewalk when it snows (patch of ice underneath snow is especially dangerous). I'm wary of people who don't shovel the snow (I end up walking on the road... much safer, plus I hope to keep out of the hospital).
Well, not a star wars fan myself, but if you look deeper at eps 4, 5, 6 (A New Hope, Empire Strikes Back, and Return of the Jedi), they actually had a cohesive subtext. It's a basic plot, yes, but the symbolism is amazing - look at the Storm Troopers for an example. Their uniforms are white for a reason (and basically similar) - if you associate white with purity, it can be thought of purity, but with bad associates (i.e., the Force gone wrong, aka the Dark side).
Only problem is the need to strip away the effects to get at the subtext takes a lot of the willingness to watch away. (And yes, I learned this through a class called Religion 12b. Nice course [had to take it]. Watched a stripped down version of 2001, which also has a similar subtext hidden away).
Back to this parent article, I wholeheartedly agree, that Trek is more than Trek - it actually is a reflection of the state of the world as it is/was (TOS: A *RUSSIAN* on board... with *AMERICANS*. Working *together*.) In fact, most of past sci-fi actually did reflect the state of the world (but now, movies are now 1 1/2 hours of effects stiched together with one of the 14 Standard Hollywood Plots(tm)).