I've learned quite a bit from newer games such as Medieval: Total War and older games such as Seven Cities of Gold (which wasn't geographically accurate, but had a lot of historically accurate information), as well as historical sims.
Flight Simulators often have real world locations in them, and relative directions can be learned. Even combat flight sims often have real world locations.
Simulators in general often have real world historical ships in them - the now fairly ancient sub simulator GATO comes to mind (I haven't played a lot of recent sims, so I'm not up on the latest:)
It's possible to make something basically unhackable - make it so no outside input is taken and then stick it in a solid box that can't be accessed physically. You could make a system unhackable also by only allowing the system to evaluate and interpret data and burn the entire thing into a ROM-processor combo (no RAM and no external access to the CPU - therefore, you can't do anything with the actual program itself - without replacing the ROM-processor, that is).
This is close to what is proposed, but the proposed still has RAM (I think) and therefore technically can still be hacked (but not necessarily easily). Since the softwall is changes to the plane's controlling software, someone familiar with how this software is written could just change it back to the older version - provided they had access to it, such as through a mechanic or plane takeover. I'm sure 'rev A' software is probably archived by terrorists somewhere.
It does kinda remind me of the "uncrackable" MS flight simulator, though. Cracked in 3 hours and distributed nationwide within 12. Ah, nostalgia:)
Second button is control-click, usually, but I'm quibbling:)
I was happy to see that replacing the hockey puck mouse on my G3 with a $15 3 button + wheel optical mouse worked flawlessly in MacOSX, even way back at X.0. I needed to install drivers for both Linux and macOS 9 (both had non-functional wheel although 9 in emulation mode worked).
This is true, but basically it gives them a reason to sue you and the competing company - which will probably hurt you if you're footing your own legal.
I seem to remember some people from my company having this problem.
There are other factors as well - you may be able to boost the performance on a mac (or UNIX box)simply by renicing the process. I do this occasionally on compiles that generally take several hours and shave off a few minutes.
Windows has, at least in the past, been notorious for bad multiprocessing, as well, leaving the second processor idle a large percentage of the time. I don't know if that has been fixed, as I don't use any dual processor PCs. That alone could skew dual processor benchmarks, especially with applications like Photoshop, that don't write directly to the hardware (where a SPEC test probably would). In reality, that's a real world test of performance, but you may get better scores running Linux (or some other OS) on the same machine.
Yeah - in addition, most motherboard builders target Microsoft Windows, which has a 2 processor kernel limitation in the license agreement (unless you pay a lot extra for the 3 or more processor version, which scales to the number of processors you have), so it's very hard to find more than 2 processor mobos, and when you do, they cost a small fortune. This is probably why Intel and AMD primarily target single and dual processors in their mobo designs and often don't even bother with more than that (except for high margin servers, which may have one new mobo every other cycle). Come to think of it, restricting number of processors also forces upgrades sooner, since you can't just add processors to handle the load, and it encourages purchases of high end, high margin chips and motherboards that add to their bottom line, so it really is in their best interests, as well.
nah, not yet, but it should be made an option, soon.
I've bought several PC motherboards in the last year, and only the high end one for my gaming box came with serial ATA (and that board supports Parallel, as well, as most do). Once low-end manufacturers jump on the bandwagon, parallel ATA will start trickling away, but not until then. Low end usually doesn't need the performance, anyhow.
Programming is also about creating and creative process - a programmer with no creativeness isn't going to get far in the industry. The objective may be to solve a problem, but how you get there is 99% of the battle.
I also don't fully agree with your 'simpler is better' arguement - I started coding around 9-10 years old, and moved quickly to assembly programming (the only alternative to BASIC) because BASIC was too slow for the stuff I wanted to do - write games. The one thing I wanted to learn as a 10 year old was how to move a sprite around the screen and make it blow up. Later, when I'd learned how to do that by writing Space Invaders, I wanted to know how to scroll the entire screen to one-up my friend's Space Invaders (plus, we'd seen it in Ultima and some newer side-scrollers).
The kicker was when I was asked for help in scrolling the screen by a 13 year old on IRC last week, so I guess times haven't changed much:)
you can't really build macosx from scratch... unless you want to completely re-write compatible code open-source for us...:)
On the other hand, you can build Darwin, and there are tutorials on how to do this online. I wouldn't recommend trying to build all of Darwin and then put the GUI back on top, as I suspect you'll need to rebuild the GUI and can't (for instance, the gcc that looks like it will be used for the next macosx is 3.3, which isn't binary compatible with the current one being used - 3.1, so you'll likely have failures there). I've built pieces of darwin using the 3.1 compiler (and earlier on older releases) and the latest sources successfully and used them, so it's not a complete bust.
I think you're wrong - you can pretty much do anything you want to with the music as long as you don't share it.
The trick is to guarantee that others don't share the stream. Password protection on the connecting web page should be enough - even if someone cracks your password to steal the music, you've done your part and they're in violation of the DMCA, not you.
I doubt the RIAA could win a case against this if they decided to pursue one - you're the owner of the copyrighted material, and the only listener. Any other listener has violated the DMCA by cracking your encryption (probably even the RIAA, themselves, if they found out about it - I'd sue 'em back for DMCA violations). Honestly, this is no different than downloading your owned mp3 files from home and playing them at work.
That'd be handy... I can name a couple hundred off the top of my head...
Akalabeth (aka Ultima 0) Airheart (ooh - double hi-res) Apple Panic Ali Baba Archon (chess... sorta) Aztec (aka Indiana Jones without the license) Bard's Tale Beer Run (...Olympus beer?) Below the Root (loved the game... don't know why) Captain Goodnight (a strangely familiar landscape) Choplifter (video game based on it) Crypt of Medea Drol (nice graphics for Apple ][) Eamon (text adventure maker) Forbidden Castle Gemstone Warrior... - several hours later -... Lode Runner Masquerade... - a couple more hours -... Questron I & II Rescue Raiders (...choplifter deluxe) Sabotage (first Apple game I played) Space Vikings Starfox (or spacefox? - action flight game) Sneakers (...elementary school) Star Blazer Sun Dog (fun RPG/action game - sorta Elite-ish) Syzygy Ultima I-IV (second computer RPG I played) Wizardry (first computer RPG I played) Wizard's Crown (skill based RPG) Xevious (vidgame port) Zork I-III
> Was it, or was it not, designed to be used in this way? If it was not, why does the system > let you try it?
The answer depends on your OS.
Windows pre NT based (Win 1,2.x,3.x,95,98,ME), like MacOS pre X were designed initially to be single user, single application systems, so no, they weren't designed that way. Multitasking was hacked in later, because that's what users wanted. The newer kernels (macosx, NT/2000/XP) are built around multitasking cores, so they are designed for it.
NT based Windows is primarily unstable when in low memory or low diskspace conditions, which may happen if you're running too many apps. I've found that Windows and Windows apps usually handle these situations poorly - especially with disk space, where the errors sometimes are meaningless or confusing.
I have almost no crashing problems on my Windows XP box, my Linux box or my MacOSX box. Actually, I've _never_ had any of them crash unless I did it myself (like writing kernel drivers... poorly, or running the 3 line crash Windows XP program:)
I am an obsessive updater, so my uptime is rarely more than 2 weeks - that may be one reason. I also don't usually leave the Windows XP or Linux machines running (since they're on the same machine and I usually flip-flop from wanting to play games or wanting to write code, which I don't do for Windows)
My Windows 2000 machine at work has blue screened on occasion, but that appears to have something to do with running tail from a network drive to watch a log file (which is certainly a app problem, but the app MAY actually be the OS). I stopped doing that about 2 months ago, and haven't seen a bluescreen since. The only reason I ever boot that machine now is for patches.
bah - the commercial software I work on during my day job is a _HUGE_ improvement over what's available via open source, at least at the moment. Some products like Linux and OpenOffice are getting to the point of acceptance and close to (if not better than) commercial competition, but for every one of those, you have something that's not even close from either a functional or usability standpoint, or has been abandoned by its maker. Linux and OpenOffice still are't accepted in most major businesses, which is why I say "close to" acceptance.
Industry demands support for most products, and open-source's un-guaranteed support isn't good enough for them and they don't want to pay someone to learn all about a product just to fix minor bugs. A few companies are offering general product Linux support, but try to convince a manager to use them - "Did this guy write the software? No? Then we're paying this guy to learn it and fix it? Why don't we just pay our own guy to do it?... hmm, I don't like the idea of supporting other people's software." I don't just say that - I've heard it from managers, sales, and marketing types talking about _RED_HAT_ for God's sake, which is basically a supported commercial version of Linux.
If open-source is producing new-innovative games, I have yet to see it. I work on three sourceforge games (mostly maintaining ports) myself, none of which is really all that original or anywhere near completion. If some of these games hold onto their good developers, someday, maybe, there'll be a finished product. I won't hold my breath, but if it happens, great.
um, no. CISC has been proven to be slower than RISC, but RISC needs to run faster because it essentially extends its smaller amount of actions to do the same things as CISC.
Both AMD and Intel essentially use Hybrid processors that take longer CISC instructions and break them down in microcode to RISC instructions. RISC processors rely on several instructions to perform certain actions a CISC processor does in one (read/write, for example), but by doing this, pipeline stalls can be reduces, as well as many other optimizations. Intel uses a fancy name instead of hybrid-RISC (which I forget), but it really breaks down to the processor itself running RISC instructions.
Altivec originally had a serious advantage over SSE, in that it could be run in parallel with the FPU and Integer Units. This is no longer true since, as you said, SSEII offers this. If I remember correctly, though, Altivec can also be used as a separate FPU, as well, which essentially amounts to having a second FPU. I don't actually have an Altivec mac, and I'm not sure if this was ever implemented (it was in a white paper I read about Altivec years ago).
As for the backward compatible 286 instructions, they're essentially being emulated in hardware. Intel tried to ditch them with the Pentium 1, remember? The 286/386 compatibility has had a high cost - fewer general purpose registers (compared to most modern processors) and a deeper pipeline (to support the longer instruction set).
I almost always noticed the exact opposite comparing my old 7500/150 (604e) and my old work 200MHz Pentium Pro (Both had 1GB Fast SCSI drives, I think the mac was a Quantum Fireball and the PC was a Matrox of some kind).
The thing is, unless you've got a pristine, unfragmented drive with no bad sectors and a clean install of the OS, performance can easily be different using the exact same hard disk. Macs running OS 9 or earlier, like PCs running pretty much any OS, get bloated and slow in time due to massively sized registry or finder entries (the bigger it is, the longer it takes to search). For Macs running OS 9, you usually end up rebuilding the desktop (hold cmd-opt on boot) and defragging the drive (using a third party product like Norton Utilities). For Windows PCs, you use registry cleaning (regclean) and disk defragment tools (and eventually reinstall the OS, in my experience). Old disks get bad sectors, which are often found in write verification, so the computer may be spending several minutes testing and eventually invalidating a particular sector in the disk.
btw, it sounds like the mac has some extension conflicts, or is possibly running too many (basically a driver conflict). There are programs to deal with that, but I haven't used OS 7/8/9 in so long, I have no idea what they are, anymore. I seriously doubt OS-X would have that problem, but OS-X isn't supported for 8600s.
My current PC smokes my mac (OS-X box) in copy tests by about 3-1 (it feels about that at least - I haven't done any time tests or anything), but it also has over a GHz process advantage and the disk has an 8Mb cache (as opposed to the 1MB cache on my mac disk) - not to mention the drive being about a year newer. Both are 80GB/7200RPM drives. Interestingly enough, I get even faster (feeling) copies running Linux on either...
Honestly, it's not about faster and cheaper, though - if it were, you'd be using Linux or FreeBSD on an Intel-based platform (heck, I usually do - unless I have to do some word/excel crap for work or want to play a game:). How do you justify a $300 OS (OEM ~$129, but off the shelf around ~$300 - assuming the Pro/Business editions)? I've made sure I get a new processor and motherboard with that $300, just to get the OEM version (and still save money). I usually need Word and Excel, so save another $200-300 by getting the OEM, as well (no, OpenOffice won't do - my manager likes to add features to the forms that aren't supported by OpenOffice yet).
Now I'll defend commercial OSes a bit - the way I justify it, to a degree, is that Windows and MacOS are easier to use than Linux/FreeBSD/others. The other way I justify it is that many of the cool/useful games/applications can only be run on Windows or Mac, which requires purchasing that or those OSes. I think Apple has done by far the best job of abstraction and simplification of the OS objects (applications are self contained and not the collection of files you see on Windows or UN*X, for example), which is why many find it the easiest to learn. Microsoft, on the other hand, is easier to use than most UN*Xes, and fakes a lot of the application abstraction by using desktop and start menu shortcuts (and therefore, isn't _too_ much harder to use than macs). Microsoft did a wonderful job of wooing developers in the mid-90s and has built an impressive software base. Linux (and UN*X in general) pretty much fragmented into a bunch of interfaces (so lacks consistency), has a mixed software base, and lacks a consistent upgrade path (RedHat and Debian have made progress in fixing this, but neither is perfect, IMO). It also lacks the ease of use, especially for set up. I can train my mom to use KDE, but train her to install linux? Forget it. Some packages have improved, but I still get asked questions like how I want my disk partitioned, and most of my computer savvy friends (from a user standpoint) don't know what that means, so do you think my mom could figure it out (she's not exactly a computer idiot, either)? We're talking fundamental stuff, but it's still beyond the scope of the average user.
As for hacking, it's a lot harder to do this kind of thing now - most companies have an inkling of what security is, at least.
Back in the late '70s and early '80s, you could set your modem on autodial, try a series of common names and (dictionary) passwords and hack in to many sites. Finding sites was a matter of wardialing phone numbers (anyone remember answering those faxlike calls?). Some software used by some sites still used the default password, and others had well known backdoors.
Yes, there were some good hackers, but there were just as many bad hackers. Bad hackers then (like myself) are the equivalent of script kiddies today.
You say there was no fraud or vandalism? I remember one kid in jr. high school who hacked a chemical company and ordered 26000 tons of suphuric and nitric acids to be delivered to a day care facility AS A JOKE. Yeah, we laughed at the time, but I can't imagine the company or the day care found it funny at all. I think age and maturity have a LOT to do with vandalism with hacking. I would guess the majority of these hacker vandals are between 13 and 18 years old - at least in my experience, that's the age most seem to be.
since '83, phreaking hasn't been the same... it's more hacking and stealing now (but some people argue that's what it was then, as well). Before that, the phone company used tonal verification for coin drops into pay phones (which is what the whistle, and later black/red/orange/etc boxes spoof). Nowadays, phones mostly handle this internally.
I remember some pirate/hacker (child/teen-hood) friends using different colored boxes for their long distance downloads and uploads, but that ended before I actually built one myself. I was even saving allowance money for parts when one of those guys told me they no longer worked.
nowadays, I suspect it's easier and more common to steal phone and/or credit cards than to hack the phone company to steal long distance. After having credit cards stolen, myself, though, I definitely have a VERY negative view of this practice.
I actually fall in the middle of this, somewhere - I love both Irish and German beers, as well as a lot of microbrews. Guiness is yummy. Spaten is yummy (esp Oktoberfest, from a keg, although it's probably better from a keg in Germany, which I'll have to try someday). Oechsner wasn't bad for a pilsner, and since I dislike pilsners as a rule, that's saying a lot. The other ones you've listed I've never heard of, much less tried. Someday I'll hit my (Great-grandparent's) motherland, Germany (actually Prussia at the time they left) and Austria, and hopefully try some of these.
Germany makes the only wheat beers I'll drink. American "wheat beers" (the ones showing pictures of wheat on 'em) are mostly rice, which gives 'em a nasty aftertaste (my apologies to the Bud, Coors, Michelob, etc. drinkers - it's my opinion and I'm sticking with it:)
It was ported to Mac, so maybe there's hope that some C files are lying around. Maybe the mac porter will release their code (probably still in Pascal:P
It wasn't uncommon to write code in C, and then profile and convert heavily time dependent code to assembly back then. If the original C files exist, then the port wouldn't be too hard. If not, I agree - yuck.
As much as I like(d) writing assembly, I hate reverse engineering it. Assembly is pretty much write once, use once (although many of the ideas can be reused). Assembly is also very processor dependent, because it is usually written to avoid processor stalls. Change one pipeline step, and your loops don't unroll the same, your processor stalls more often, etc.
a) - a bunch of hooey. The clone market failed because Steve Jobs wanted it to fail (he wanted an "appliance" and not all the driver problems in the Windows world). If Apple was worried about profits on clones, they could just jack up the prices of licenses, which would be a lot cheaper than manufacturing the hardware themselves - they didn't even make this an option.
b) As for vertically integrated, yes, Apple is (for the most part), but so are Sun (Solaris), SGI (IRIX), IBM (AIX), et. al. (Runix, DUX...), none of which are doing well or have died. Sun, SGI and IBM all are trying to move to Linux on their hardware, which is probably a better solution, because then they can control hardware and at least partially control software, without having to rewrite everything themselves.
Controlling everything in a large market isn't always a good thing - actually, in many ways it's a bad thing. It stifles innovation and change - look at Microsoft (software), Apple (hard/sofware), Betamax (both), Bell telephone (service/phones), Cable TV (service/boxes). One thing it does usually do is offer stability and often better quality materials and workmanship (but not always - especially over time - Bell telephone is an excellent example).
George W Bush today named himself the Good bit, and anything Arab the Evil bit. Everyone else will be sorted on a case-by-case basis. The NSA hailed this development, stating that it will help packet sorting by at least threefold, as well as aiding in priority sorting for Carnivore.
Slashdot Mirror
How about history and geography?
:)
I've learned quite a bit from newer games such as Medieval: Total War and older games such as Seven Cities of Gold (which wasn't geographically accurate, but had a lot of historically accurate information), as well as historical sims.
Flight Simulators often have real world locations in them, and relative directions can be learned. Even combat flight sims often have real world locations.
Simulators in general often have real world historical ships in them - the now fairly ancient sub simulator GATO comes to mind (I haven't played a lot of recent sims, so I'm not up on the latest
It's possible to make something basically unhackable - make it so no outside input is taken and then stick it in a solid box that can't be accessed physically. You could make a system unhackable also by only allowing the system to evaluate and interpret data and burn the entire thing into a ROM-processor combo (no RAM and no external access to the CPU - therefore, you can't do anything with the actual program itself - without replacing the ROM-processor, that is).
:)
This is close to what is proposed, but the proposed still has RAM (I think) and therefore technically can still be hacked (but not necessarily easily). Since the softwall is changes to the plane's controlling software, someone familiar with how this software is written could just change it back to the older version - provided they had access to it, such as through a mechanic or plane takeover. I'm sure 'rev A' software is probably archived by terrorists somewhere.
It does kinda remind me of the "uncrackable" MS flight simulator, though. Cracked in 3 hours and distributed nationwide within 12. Ah, nostalgia
Second button is control-click, usually, but I'm quibbling :)
I was happy to see that replacing the hockey puck mouse on my G3 with a $15 3 button + wheel optical mouse worked flawlessly in MacOSX, even way back at X.0. I needed to install drivers for both Linux and macOS 9 (both had non-functional wheel although 9 in emulation mode worked).
This is true, but basically it gives them a reason to sue you and the competing company - which will probably hurt you if you're footing your own legal.
I seem to remember some people from my company having this problem.
mine's 1 year no-compete.
:)
not that I have any desire to continue in the industry I'm in
There are other factors as well - you may be able to boost the performance on a mac (or UNIX box)simply by renicing the process. I do this occasionally on compiles that generally take several hours and shave off a few minutes.
Windows has, at least in the past, been notorious for bad multiprocessing, as well, leaving the second processor idle a large percentage of the time. I don't know if that has been fixed, as I don't use any dual processor PCs. That alone could skew dual processor benchmarks, especially with applications like Photoshop, that don't write directly to the hardware (where a SPEC test probably would). In reality, that's a real world test of performance, but you may get better scores running Linux (or some other OS) on the same machine.
read the very bottom of this:
Verizon
apparently there is still a bill in congress that may delay the number change date.
Yeah - in addition, most motherboard builders target Microsoft Windows, which has a 2 processor kernel limitation in the license agreement (unless you pay a lot extra for the 3 or more processor version, which scales to the number of processors you have), so it's very hard to find more than 2 processor mobos, and when you do, they cost a small fortune. This is probably why Intel and AMD primarily target single and dual processors in their mobo designs and often don't even bother with more than that (except for high margin servers, which may have one new mobo every other cycle). Come to think of it, restricting number of processors also forces upgrades sooner, since you can't just add processors to handle the load, and it encourages purchases of high end, high margin chips and motherboards that add to their bottom line, so it really is in their best interests, as well.
nah, not yet, but it should be made an option, soon.
I've bought several PC motherboards in the last year, and only the high end one for my gaming box came with serial ATA (and that board supports Parallel, as well, as most do). Once low-end manufacturers jump on the bandwagon, parallel ATA will start trickling away, but not until then. Low end usually doesn't need the performance, anyhow.
Programming is also about creating and creative process - a programmer with no creativeness isn't going to get far in the industry. The objective may be to solve a problem, but how you get there is 99% of the battle.
:)
I also don't fully agree with your 'simpler is better' arguement - I started coding around 9-10 years old, and moved quickly to assembly programming (the only alternative to BASIC) because BASIC was too slow for the stuff I wanted to do - write games. The one thing I wanted to learn as a 10 year old was how to move a sprite around the screen and make it blow up. Later, when I'd learned how to do that by writing Space Invaders, I wanted to know how to scroll the entire screen to one-up my friend's Space Invaders (plus, we'd seen it in Ultima and some newer side-scrollers).
The kicker was when I was asked for help in scrolling the screen by a 13 year old on IRC last week, so I guess times haven't changed much
you can't really build macosx from scratch... unless you want to completely re-write compatible code open-source for us... :)
On the other hand, you can build Darwin, and there are tutorials on how to do this online. I wouldn't recommend trying to build all of Darwin and then put the GUI back on top, as I suspect you'll need to rebuild the GUI and can't (for instance, the gcc that looks like it will be used for the next macosx is 3.3, which isn't binary compatible with the current one being used - 3.1, so you'll likely have failures there). I've built pieces of darwin using the 3.1 compiler (and earlier on older releases) and the latest sources successfully and used them, so it's not a complete bust.
I think you're wrong - you can pretty much do anything you want to with the music as long as you don't share it.
The trick is to guarantee that others don't share the stream. Password protection on the connecting web page should be enough - even if someone cracks your password to steal the music, you've done your part and they're in violation of the DMCA, not you.
I doubt the RIAA could win a case against this if they decided to pursue one - you're the owner of the copyrighted material, and the only listener. Any other listener has violated the DMCA by cracking your encryption (probably even the RIAA, themselves, if they found out about it - I'd sue 'em back for DMCA violations). Honestly, this is no different than downloading your owned mp3 files from home and playing them at work.
That'd be handy... I can name a couple hundred off the top of my head...
... ... ... ...
Akalabeth (aka Ultima 0)
Airheart (ooh - double hi-res)
Apple Panic
Ali Baba
Archon (chess... sorta)
Aztec (aka Indiana Jones without the license)
Bard's Tale
Beer Run (...Olympus beer?)
Below the Root (loved the game... don't know why)
Captain Goodnight (a strangely familiar landscape)
Choplifter (video game based on it)
Crypt of Medea
Drol (nice graphics for Apple ][)
Eamon (text adventure maker)
Forbidden Castle
Gemstone Warrior
- several hours later -
Lode Runner
Masquerade
- a couple more hours -
Questron I & II
Rescue Raiders (...choplifter deluxe)
Sabotage (first Apple game I played)
Space Vikings
Starfox (or spacefox? - action flight game)
Sneakers (...elementary school)
Star Blazer
Sun Dog (fun RPG/action game - sorta Elite-ish)
Syzygy
Ultima I-IV (second computer RPG I played)
Wizardry (first computer RPG I played)
Wizard's Crown (skill based RPG)
Xevious (vidgame port)
Zork I-III
> Was it, or was it not, designed to be used in this way? If it was not, why does the system
> let you try it?
The answer depends on your OS.
Windows pre NT based (Win 1,2.x,3.x,95,98,ME), like MacOS pre X were designed initially to be single user, single application systems, so no, they weren't designed that way. Multitasking was hacked in later, because that's what users wanted. The newer kernels (macosx, NT/2000/XP) are built around multitasking cores, so they are designed for it.
NT based Windows is primarily unstable when in low memory or low diskspace conditions, which may happen if you're running too many apps. I've found that Windows and Windows apps usually handle these situations poorly - especially with disk space, where the errors sometimes are meaningless or confusing.
I have almost no crashing problems on my Windows XP box, my Linux box or my MacOSX box. Actually, I've _never_ had any of them crash unless I did it myself (like writing kernel drivers... poorly, or running the 3 line crash Windows XP program :)
:P
I am an obsessive updater, so my uptime is rarely more than 2 weeks - that may be one reason. I also don't usually leave the Windows XP or Linux machines running (since they're on the same machine and I usually flip-flop from wanting to play games or wanting to write code, which I don't do for Windows)
My Windows 2000 machine at work has blue screened on occasion, but that appears to have something to do with running tail from a network drive to watch a log file (which is certainly a app problem, but the app MAY actually be the OS). I stopped doing that about 2 months ago, and haven't seen a bluescreen since. The only reason I ever boot that machine now is for patches.
Now that I've said that, I better knock on wood
bah - the commercial software I work on during my day job is a _HUGE_ improvement over what's available via open source, at least at the moment. Some products like Linux and OpenOffice are getting to the point of acceptance and close to (if not better than) commercial competition, but for every one of those, you have something that's not even close from either a functional or usability standpoint, or has been abandoned by its maker. Linux and OpenOffice still are't accepted in most major businesses, which is why I say "close to" acceptance.
Industry demands support for most products, and open-source's un-guaranteed support isn't good enough for them and they don't want to pay someone to learn all about a product just to fix minor bugs. A few companies are offering general product Linux support, but try to convince a manager to use them - "Did this guy write the software? No? Then we're paying this guy to learn it and fix it? Why don't we just pay our own guy to do it?... hmm, I don't like the idea of supporting other people's software." I don't just say that - I've heard it from managers, sales, and marketing types talking about _RED_HAT_ for God's sake, which is basically a supported commercial version of Linux.
If open-source is producing new-innovative games, I have yet to see it. I work on three sourceforge games (mostly maintaining ports) myself, none of which is really all that original or anywhere near completion. If some of these games hold onto their good developers, someday, maybe, there'll be a finished product. I won't hold my breath, but if it happens, great.
um, no. CISC has been proven to be slower than RISC, but RISC needs to run faster because it essentially extends its smaller amount of actions to do the same things as CISC.
Both AMD and Intel essentially use Hybrid processors that take longer CISC instructions and break them down in microcode to RISC instructions. RISC processors rely on several instructions to perform certain actions a CISC processor does in one (read/write, for example), but by doing this, pipeline stalls can be reduces, as well as many other optimizations. Intel uses a fancy name instead of hybrid-RISC (which I forget), but it really breaks down to the processor itself running RISC instructions.
Altivec originally had a serious advantage over SSE, in that it could be run in parallel with the FPU and Integer Units. This is no longer true since, as you said, SSEII offers this. If I remember correctly, though, Altivec can also be used as a separate FPU, as well, which essentially amounts to having a second FPU. I don't actually have an Altivec mac, and I'm not sure if this was ever implemented (it was in a white paper I read about Altivec years ago).
As for the backward compatible 286 instructions, they're essentially being emulated in hardware. Intel tried to ditch them with the Pentium 1, remember? The 286/386 compatibility has had a high cost - fewer general purpose registers (compared to most modern processors) and a deeper pipeline (to support the longer instruction set).
I almost always noticed the exact opposite comparing my old 7500/150 (604e) and my old work 200MHz Pentium Pro (Both had 1GB Fast SCSI drives, I think the mac was a Quantum Fireball and the PC was a Matrox of some kind).
:). How do you justify a $300 OS (OEM ~$129, but off the shelf around ~$300 - assuming the Pro/Business editions)? I've made sure I get a new processor and motherboard with that $300, just to get the OEM version (and still save money). I usually need Word and Excel, so save another $200-300 by getting the OEM, as well (no, OpenOffice won't do - my manager likes to add features to the forms that aren't supported by OpenOffice yet).
The thing is, unless you've got a pristine, unfragmented drive with no bad sectors and a clean install of the OS, performance can easily be different using the exact same hard disk. Macs running OS 9 or earlier, like PCs running pretty much any OS, get bloated and slow in time due to massively sized registry or finder entries (the bigger it is, the longer it takes to search). For Macs running OS 9, you usually end up rebuilding the desktop (hold cmd-opt on boot) and defragging the drive (using a third party product like Norton Utilities). For Windows PCs, you use registry cleaning (regclean) and disk defragment tools (and eventually reinstall the OS, in my experience). Old disks get bad sectors, which are often found in write verification, so the computer may be spending several minutes testing and eventually invalidating a particular sector in the disk.
btw, it sounds like the mac has some extension conflicts, or is possibly running too many (basically a driver conflict). There are programs to deal with that, but I haven't used OS 7/8/9 in so long, I have no idea what they are, anymore. I seriously doubt OS-X would have that problem, but OS-X isn't supported for 8600s.
My current PC smokes my mac (OS-X box) in copy tests by about 3-1 (it feels about that at least - I haven't done any time tests or anything), but it also has over a GHz process advantage and the disk has an 8Mb cache (as opposed to the 1MB cache on my mac disk) - not to mention the drive being about a year newer. Both are 80GB/7200RPM drives. Interestingly enough, I get even faster (feeling) copies running Linux on either...
Honestly, it's not about faster and cheaper, though - if it were, you'd be using Linux or FreeBSD on an Intel-based platform (heck, I usually do - unless I have to do some word/excel crap for work or want to play a game
Now I'll defend commercial OSes a bit - the way I justify it, to a degree, is that Windows and MacOS are easier to use than Linux/FreeBSD/others. The other way I justify it is that many of the cool/useful games/applications can only be run on Windows or Mac, which requires purchasing that or those OSes. I think Apple has done by far the best job of abstraction and simplification of the OS objects (applications are self contained and not the collection of files you see on Windows or UN*X, for example), which is why many find it the easiest to learn. Microsoft, on the other hand, is easier to use than most UN*Xes, and fakes a lot of the application abstraction by using desktop and start menu shortcuts (and therefore, isn't _too_ much harder to use than macs). Microsoft did a wonderful job of wooing developers in the mid-90s and has built an impressive software base. Linux (and UN*X in general) pretty much fragmented into a bunch of interfaces (so lacks consistency), has a mixed software base, and lacks a consistent upgrade path (RedHat and Debian have made progress in fixing this, but neither is perfect, IMO). It also lacks the ease of use, especially for set up. I can train my mom to use KDE, but train her to install linux? Forget it. Some packages have improved, but I still get asked questions like how I want my disk partitioned, and most of my computer savvy friends (from a user standpoint) don't know what that means, so do you think my mom could figure it out (she's not exactly a computer idiot, either)? We're talking fundamental stuff, but it's still beyond the scope of the average user.
As for hacking, it's a lot harder to do this kind of thing now - most companies have an inkling of what security is, at least.
Back in the late '70s and early '80s, you could set your modem on autodial, try a series of common names and (dictionary) passwords and hack in to many sites. Finding sites was a matter of wardialing phone numbers (anyone remember answering those faxlike calls?). Some software used by some sites still used the default password, and others had well known backdoors.
Yes, there were some good hackers, but there were just as many bad hackers. Bad hackers then (like myself) are the equivalent of script kiddies today.
You say there was no fraud or vandalism? I remember one kid in jr. high school who hacked a chemical company and ordered 26000 tons of suphuric and nitric acids to be delivered to a day care facility AS A JOKE. Yeah, we laughed at the time, but I can't imagine the company or the day care found it funny at all. I think age and maturity have a LOT to do with vandalism with hacking. I would guess the majority of these hacker vandals are between 13 and 18 years old - at least in my experience, that's the age most seem to be.
since '83, phreaking hasn't been the same... it's more hacking and stealing now (but some people argue that's what it was then, as well). Before that, the phone company used tonal verification for coin drops into pay phones (which is what the whistle, and later black/red/orange/etc boxes spoof). Nowadays, phones mostly handle this internally.
I remember some pirate/hacker (child/teen-hood) friends using different colored boxes for their long distance downloads and uploads, but that ended before I actually built one myself. I was even saving allowance money for parts when one of those guys told me they no longer worked.
nowadays, I suspect it's easier and more common to steal phone and/or credit cards than to hack the phone company to steal long distance. After having credit cards stolen, myself, though, I definitely have a VERY negative view of this practice.
lamer.
:P
e s/uti lity/misc/visicalc.zip
use the ORIGINAL Apple ][ version in an emulator, at least. You're obviously not a well trained Apple bigot yet
here's the archive of the rom
ftp://ftp.apple.asimov.net/pub/apple_II/imag
find your emulator at someplace like emulation.net
mmm...keystone... ngah ngah drool (kidding :)
:)
I actually fall in the middle of this, somewhere - I love both Irish and German beers, as well as a lot of microbrews. Guiness is yummy. Spaten is yummy (esp Oktoberfest, from a keg, although it's probably better from a keg in Germany, which I'll have to try someday). Oechsner wasn't bad for a pilsner, and since I dislike pilsners as a rule, that's saying a lot. The other ones you've listed I've never heard of, much less tried. Someday I'll hit my (Great-grandparent's) motherland, Germany (actually Prussia at the time they left) and Austria, and hopefully try some of these.
Germany makes the only wheat beers I'll drink. American "wheat beers" (the ones showing pictures of wheat on 'em) are mostly rice, which gives 'em a nasty aftertaste (my apologies to the Bud, Coors, Michelob, etc. drinkers - it's my opinion and I'm sticking with it
It was ported to Mac, so maybe there's hope that some C files are lying around. Maybe the mac porter will release their code (probably still in Pascal :P
It wasn't uncommon to write code in C, and then profile and convert heavily time dependent code to assembly back then. If the original C files exist, then the port wouldn't be too hard. If not, I agree - yuck.
As much as I like(d) writing assembly, I hate reverse engineering it. Assembly is pretty much write once, use once (although many of the ideas can be reused). Assembly is also very processor dependent, because it is usually written to avoid processor stalls. Change one pipeline step, and your loops don't unroll the same, your processor stalls more often, etc.
a) - a bunch of hooey. The clone market failed because Steve Jobs wanted it to fail (he wanted an "appliance" and not all the driver problems in the Windows world). If Apple was worried about profits on clones, they could just jack up the prices of licenses, which would be a lot cheaper than manufacturing the hardware themselves - they didn't even make this an option.
b) As for vertically integrated, yes, Apple is (for the most part), but so are Sun (Solaris), SGI (IRIX), IBM (AIX), et. al. (Runix, DUX...), none of which are doing well or have died. Sun, SGI and IBM all are trying to move to Linux on their hardware, which is probably a better solution, because then they can control hardware and at least partially control software, without having to rewrite everything themselves.
Controlling everything in a large market isn't always a good thing - actually, in many ways it's a bad thing. It stifles innovation and change - look at Microsoft (software), Apple (hard/sofware), Betamax (both), Bell telephone (service/phones), Cable TV (service/boxes). One thing it does usually do is offer stability and often better quality materials and workmanship (but not always - especially over time - Bell telephone is an excellent example).
George W Bush today named himself the Good bit, and anything Arab the Evil bit. Everyone else will be sorted on a case-by-case basis. The NSA hailed this development, stating that it will help packet sorting by at least threefold, as well as aiding in priority sorting for Carnivore.