If the code does get a fault ("exception or error" - this could be an exception without being an error, e.g. a page fault), then anything that code did "speculatively" and that wouldn't have been done by the untranslated code had it gotten that exception hasn't made any permanent state change, so the fault cancels/backs out any uncommitted state changes and presumably traps to software that would do whatever is necessary to do what the untranslated code would have done.
That's exactly what I got out of that part. And it sounds pretty cool. This particulary would have applications in multiprocessing systems. On the Alpha, we handle exceptions using something called trap barriers, which is a software method of handling this sort of thing. What happens is a fault appears to issue from the trapb and you are left to your own devices(from a compiler perspective) to discover where the exception occurs. It isolates the exception down to what's called a "trap shadow". This translates to a pain in the ass because we don't know precisely where a fault issued from, only the "shadow". Multiproc's complicate this mess further. This makes for interesting, but complicated compiler development.
Moving this to hardware, OTOH, would greatly simplify things, especially when emulation adds a layer of obfuscation.
That's why the exception handling part is what I zero'ed in on. It sounds really neat.
Maybe I'm just forgetful, but I don't remember anyone ever gambling over a Tickle-Me-Elmo..
You don't gamble with Pokemon cards. The closest to that was Magic where you have the option of playing for "ante". I don't think anyone ever does that anymore though. (When I played, I did, But then again I was playtesting with prototype magic decks before the game was released.) I gave away all my cards years ago (it got boring), I got curious a while ago and discovered that the cards I got rid of are worth thousands of dollars now. (shrug)
Thank you... I'm amazed when people get this elitist viewpoint, then wonder why Microsoft stomps all over them. You want Microsoft to go away? First you have to understand what Microsoft does right. They know what average people want. They know that many people don't care to recompile their kernel when they add a piece of hardware. Unfortunately, they also know that, given ease of use, people will tolerate buggy software.
What I got from his comments regarding upgrades is that he was referring to software "staples". It's an interesting concept. I mean think about it, what value can they add to office? It already does just about everything. The paperclip's about as sickeningly user friendly as you can get. I probably haven't used 10% of the features in Office. Why should I upgrade? Bug fixes? Sorry...not worth hundreds of dollars.
Anyway, what I'm saying is that this concept probably won't apply everywhere. As new technologies and concepts develop, there is always a price curve. But the dramatic crash in prices for a mature piece of software like a word proccessor or spreadsheet will be pretty unique. And may teach certain huge software companies a lesson!
That's silly...Your phone lines are just older, and above ground. I've got buried lines, so neither is likely to go out. Phone companies are burying thier lines as quickly as possible. It's in thier own best interest.
They generally make a practice of grossly overselling bandwidth.
That's the thing though. It isn't so much an issue when you can switch ISP's (a la DSL). With cable modems, you're stuck. In the case of DSL, the pipe from the CO to your ISP is generally huge. The problem comes when your ISP oversells it's bandwidth. At which time you can switch. No such options with Cable. Now obviously, if you can get your hands on business grade SDSL, that is ideal. But impractical if you're po' like me.
From what I've seen, broadband mileage varies greatly depending on the area. For example, here in Seattle, we've got US West and GTE (depending on the area you live in) offering DSL. I use US West, I know people with GTE. Both DSL offerings are excellent. And I pay ~$50 a month for my service (726/256). (I get to chose ISP's to boot). I know several people who have TCI @Home and have had nothing but trouble. Apparently TCI hasn't prepared well enough for the large increase in customers and they keep running out of bandwidth. TCI in general has also been notoriously unreliable here. (The WA state utilities commision has actually fined them several times.)
So the moral is: check with people in your area first. Until the dust settles, there is no way of knowing what your best bet is without checking locally.
I can't say anything either way about the Mindspring Engineering people. But the customer service department needs to be hit by a couple times by a clue-by-four. Especially the billing department. I never had any problems with either one relating to the actual service (though I may have had a few less busy signals once Mindspring took over). But man, my experiences with the customer service department has soured the whole thing for me.
Also, a disclamer: Please don't take this the wrong way, I'm sure Mindspring has a wonderful engineering department.
Sorry I don't have any alternitive suggestions (other than a local ISP), but I'd strongly recommend NOT having anything to do with Mindspring. They've caused me more trouble than I care to think about.
Ever since Mindspring gobbled up my ISP, I've had nothing but trouble. Luckily, I have DSL now. I wouldn't recommend that anyone have anything to do with Mindspring. And any Earthlink customers should leave while they still can. Mindspring made me wait on hold for 30 minutes(!!!) to cancel my account. That's the only way they said they allowed cancellations.
First off, let me say that I use DSL (from US West), and wouldn't trade it for anything. A couple things to remember about cable modems: The cable companies tend to be VERY clueless about non-Windows OS's. Good luck getting them to even install service on a Linux box. Or a home network. Or anything but a single Windows machine. US West, on the other hand, gave me the option (which I took) of installing the modem myself. They just sent me the router. It's not brain surgery. Plug router into phone line. Plug router into hub. Plug router into wall outlet.
The second thing, is about choice. I can chose any ISP I want. I found one that gave me a few static IP addresses, has a very liberal AUP (no warez), and costs $20 a month! Add this to the fact that US West hasn't turned on the bandwidth caps yet:)
A key feature of LinuxOpen is its ability to run on the most advanced PC workstations with devices that increase communication bandwidth, such as ADSL and cable modem. LinuxOne will support these new technologies with its sophisticated proprietary device drivers (software that provides an interface between an operating environment and its associated hardware).
You caught me. I've glanced at them(we actually printed them out the first day they became publicly available), but certainly haven't given them enough attention to critique them. I understand your point now.
We're actually a little smug about Merced. Probably a couple years until they even ship, and we've had 64 bit RISC for about 7 years now.:) However, I'm afraid that Intel, being the 600 pound gorrila that it is, may wipe that smug smile off our faces when their marketing machine kicks in.
It's a programmer's nightmare because they decided it would be easier to let the programmer/compiler do the optimizations than to actually design them into the chip.
If I understand what you are saying, that's silly because that's the whole point of a RISC architecture. It's not a matter of it being "easier". (I'll resist the urge to go into in-depth processor design theory) Optimization requires assumptions. If you go with a CISC architecture, you're doing general purpose optimizations, while with RISC you get what you need, at the price of code complexity. Complexity can be hidden by good compiler optimizations, giving the best of both worlds. Most code isn't written in assembly anyway, so who cares? If something in particular needs to be hand optimized, you can do it, with out too much of a hit.
BTW, I work on for Compaq/Digital on one of the Alpha compilers.
Break up on a team/group level (50-200 people or so)
Do you have any idea how many people work at Microsoft?! 30,000/100=300 small companies!! Besides, the infrastructure of a small company is necesaraly much different from a large one. The logistics of that sort of arbitrary breakup would be a nightmare from every perspective.
That's neato. As a side note, I don't know if any of you have had a chance to check out www.douglasadams.com, but it's a really nice site. Also there's a web-board that Douglas reads. He's really a pretty nice guy and makes himself as available as possible, considering his busy schedule, although he can get kind of short with blithering idiots. At one point he requested (publicly on the message forum) that a new "stupid crap" forum be created, and that all posts relating to: 1) tech support for Starship Titanic 2) "I'm writing a report for school and was wondering if you could answer this involved question which was not very well thought out in the first place." 3) Anything similar to some of the lesser AC posts here. Also note that Mr. Adams has officially retired from answering any questions relating to the number 42.
Nope he was wrong...both the logon screen and screen servers default to the absolute lowest priority("low" in NT parlance). They don't count as "foreground terminal applications". That game of solitare you are playing, however, does.
Why isn't Corel putting Corel LINUX on an FTP site for download? The first version of Corel LINUX is still in the development and testing phase and is not ready for general distribution. Beta testing is part of the development process that Corel is following to ensure that Corel LINUX reaches a suitable level of completeness and stability before it becomes widely available.
How many people will be testing Corel LINUX? In order to ensure that the beta testing process can be properly administered, we will be limiting the number of registered beta testers for this first round of testing. The exact number of testers has not yet been determined.
Is this testing style in keeping with the spirit of Open Source software distribution? Yes. The recruitment of outside Beta testers is part of the initial development process of Corel LINUX and will allow Corel to release the first version of Corel LINUX to the general public at an earlier date than would be possible if Corel relied only on its internal testing resources. Once Corel LINUX is ready for general distribution, it will be distributed in accordance with the criteria for Open Source software.
Ignores extra spaces(I'm getting carried away)
on
Nitrozac Answers
·
· Score: 1
int x,y=8,z;void main(){while((x=getc(stdin))!=-1&&y==y--)z|=x==49? 1<<y:x-48?y!=7?putc(z,stdout),y=8,z= 0:y++,z=0:0;}
This got me thinking about that, so I obscurified it some more: int x,y=8,z;void main(){while((x=getc(stdin))!=-1&&y==y--)z|=x==49? 1<<y:x-48?putc(z,stdout),y=8,z=0:0;}
Slashdot Mirror
If the code does get a fault ("exception or error" - this could be an exception without being an error, e.g. a page fault), then anything that code did "speculatively" and that wouldn't have been done by the untranslated code had it gotten that exception hasn't made any permanent state change, so the fault cancels/backs out any uncommitted state changes and presumably traps to software that would do whatever is necessary to do what the untranslated code would have done.
That's exactly what I got out of that part. And it sounds pretty cool. This particulary would have applications in multiprocessing systems.
On the Alpha, we handle exceptions using something called trap barriers, which is a software method of handling this sort of thing. What happens is a fault appears to issue from the trapb and you are left to your own devices(from a compiler perspective) to discover where the exception occurs. It isolates the exception down to what's called a "trap shadow". This translates to a pain in the ass because we don't know precisely where a fault issued from, only the "shadow". Multiproc's complicate this mess further. This makes for interesting, but complicated compiler development.
Moving this to hardware, OTOH, would greatly simplify things, especially when emulation adds a layer of obfuscation.
That's why the exception handling part is what I zero'ed in on. It sounds really neat.
Maybe I'm just forgetful, but I don't remember anyone ever gambling over a Tickle-Me-Elmo..
You don't gamble with Pokemon cards. The closest to that was Magic where you have the option of playing for "ante". I don't think anyone ever does that anymore though. (When I played, I did, But then again I was playtesting with prototype magic decks before the game was released.) I gave away all my cards years ago (it got boring), I got curious a while ago and discovered that the cards I got rid of are worth thousands of dollars now. (shrug)
i mean whats next? Suing Id software for promoting school shootings?
Actually, it's been done. I believe the parents of the victims of some school shooting or other sued just about everyone...including Id I believe.
Thank you... I'm amazed when people get this elitist viewpoint, then wonder why Microsoft stomps all over them. You want Microsoft to go away? First you have to understand what Microsoft does right. They know what average people want. They know that many people don't care to recompile their kernel when they add a piece of hardware. Unfortunately, they also know that, given ease of use, people will tolerate buggy software.
What I got from his comments regarding upgrades is that he was referring to software "staples". It's an interesting concept. I mean think about it, what value can they add to office? It already does just about everything. The paperclip's about as sickeningly user friendly as you can get. I probably haven't used 10% of the features in Office. Why should I upgrade? Bug fixes? Sorry...not worth hundreds of dollars.
Anyway, what I'm saying is that this concept probably won't apply everywhere. As new technologies and concepts develop, there is always a price curve. But the dramatic crash in prices for a mature piece of software like a word proccessor or spreadsheet will be pretty unique. And may teach certain huge software companies a lesson!
That's silly...Your phone lines are just older, and above ground. I've got buried lines, so neither is likely to go out. Phone companies are burying thier lines as quickly as possible. It's in thier own best interest.
They generally make a practice of grossly overselling bandwidth.
That's the thing though. It isn't so much an issue when you can switch ISP's (a la DSL). With cable modems, you're stuck. In the case of DSL, the pipe from the CO to your ISP is generally huge. The problem comes when your ISP oversells it's bandwidth. At which time you can switch. No such options with Cable.
Now obviously, if you can get your hands on business grade SDSL, that is ideal. But impractical if you're po' like me.
From what I've seen, broadband mileage varies greatly depending on the area. For example, here in Seattle, we've got US West and GTE (depending on the area you live in) offering DSL. I use US West, I know people with GTE. Both DSL offerings are excellent. And I pay ~$50 a month for my service (726/256). (I get to chose ISP's to boot). I know several people who have TCI @Home and have had nothing but trouble. Apparently TCI hasn't prepared well enough for the large increase in customers and they keep running out of bandwidth. TCI in general has also been notoriously unreliable here. (The WA state utilities commision has actually fined them several times.)
So the moral is: check with people in your area first. Until the dust settles, there is no way of knowing what your best bet is without checking locally.
I can't say anything either way about the Mindspring Engineering people. But the customer service department needs to be hit by a couple times by a clue-by-four. Especially the billing department. I never had any problems with either one relating to the actual service (though I may have had a few less busy signals once Mindspring took over). But man, my experiences with the customer service department has soured the whole thing for me.
Also, a disclamer: Please don't take this the wrong way, I'm sure Mindspring has a wonderful engineering department.
Sorry I don't have any alternitive suggestions (other than a local ISP), but I'd strongly recommend NOT having anything to do with Mindspring. They've caused me more trouble than I care to think about.
Ever since Mindspring gobbled up my ISP, I've had nothing but trouble. Luckily, I have DSL now. I wouldn't recommend that anyone have anything to do with Mindspring. And any Earthlink customers should leave while they still can. Mindspring made me wait on hold for 30 minutes(!!!) to cancel my account. That's the only way they said they allowed cancellations.
He obviously haven't tried US West's DSL service, which is tremendous
Exactly. I've never had much good to say about US West, but their DSL has been excellent.
First off, let me say that I use DSL (from US West), and wouldn't trade it for anything. A couple things to remember about cable modems: The cable companies tend to be VERY clueless about non-Windows OS's. Good luck getting them to even install service on a Linux box. Or a home network. Or anything but a single Windows machine. US West, on the other hand, gave me the option (which I took) of installing the modem myself. They just sent me the router. It's not brain surgery. Plug router into phone line. Plug router into hub. Plug router into wall outlet.
:)
The second thing, is about choice. I can chose any ISP I want. I found one that gave me a few static IP addresses, has a very liberal AUP (no warez), and costs $20 a month! Add this to the fact that US West hasn't turned on the bandwidth caps yet
From their web site:
A key feature of LinuxOpen is its ability to run on the most advanced PC workstations with devices that increase communication bandwidth, such as ADSL and cable modem. LinuxOne will support these new technologies with its sophisticated proprietary device drivers (software that provides an interface between an operating environment and its associated hardware).
Proprietary?! On a Linux distro? Hmmmm....
But have you read the specs on IA-64?
:) However, I'm afraid that Intel, being the 600 pound gorrila that it is, may wipe that smug smile off our faces when their marketing machine kicks in.
You caught me. I've glanced at them(we actually printed them out the first day they became publicly available), but certainly haven't given them enough attention to critique them. I understand your point now.
We're actually a little smug about Merced. Probably a couple years until they even ship, and we've had 64 bit RISC for about 7 years now.
Tell me about it. In my office, I've got 15 Miata's (Alpha 500's) and one Rawhide (a Dual processor Alpha Server). It gets like an oven in here.
It's a programmer's nightmare because they decided it would be easier to let the programmer/compiler do the optimizations than to actually design them into the chip.
If I understand what you are saying, that's silly because that's the whole point of a RISC architecture. It's not a matter of it being "easier". (I'll resist the urge to go into in-depth processor design theory) Optimization requires assumptions. If you go with a CISC architecture, you're doing general purpose optimizations, while with RISC you get what you need, at the price of code complexity. Complexity can be hidden by good compiler optimizations, giving the best of both worlds. Most code isn't written in assembly anyway, so who cares? If something in particular needs to be hand optimized, you can do it, with out too much of a hit.
BTW, I work on for Compaq/Digital on one of the Alpha compilers.
Break up on a team/group level (50-200 people or so)
Do you have any idea how many people work at Microsoft?! 30,000/100=300 small companies!! Besides, the infrastructure of a small company is necesaraly much different from a large one. The logistics of that sort of arbitrary breakup would be a nightmare from every perspective.
That's neato. As a side note, I don't know if any of you have had a chance to check out www.douglasadams.com, but it's a really nice site. Also there's a web-board that Douglas reads. He's really a pretty nice guy and makes himself as available as possible, considering his busy schedule, although he can get kind of short with blithering idiots.
At one point he requested (publicly on the message forum) that a new "stupid crap" forum be created, and that all posts relating to: 1) tech support for Starship Titanic 2) "I'm writing a report for school and was wondering if you could answer this involved question which was not very well thought out in the first place." 3) Anything similar to some of the lesser AC posts here.
Also note that Mr. Adams has officially retired from answering any questions relating to the number 42.
GnrcMan
Nope he was wrong...both the logon screen and screen servers default to the absolute lowest priority("low" in NT parlance). They don't count as "foreground terminal applications". That game of solitare you are playing, however, does.
The following is taken from the Corel Linux Beta FAQ...The third one is my favorite
Why isn't Corel putting Corel LINUX on an FTP site for download?
The first version of Corel LINUX is still in the development and testing phase and is not ready for general distribution. Beta testing is part of the development process that Corel is following to ensure that Corel LINUX reaches a suitable level of completeness and stability before it becomes widely available.
How many people will be testing Corel LINUX?
In order to ensure that the beta testing process can be properly administered, we will be limiting the number of registered beta testers for this first round of testing. The exact number of testers has not yet been determined.
Is this testing style in keeping with the spirit of Open Source software distribution?
Yes. The recruitment of outside Beta testers is part of the initial development process of Corel LINUX and will allow Corel to release the first version of Corel LINUX to the general public at an earlier date than would be possible if Corel relied only on its internal testing resources. Once Corel LINUX is ready for general distribution, it will be distributed in accordance with the criteria for Open Source software.
int x,y=8,z;void main(){while((x=getc(stdin))!=-1&&y==y--)z|=x==49? 1<<y:x-48?y!=7?putc(z,stdout),y=8,z= 0:y++,z=0:0;}
This got me thinking about that, so I obscurified it some more: int x,y=8,z;void main(){while((x=getc(stdin))!=-1&&y==y--)z|=x==49? 1<<y:x-48?putc(z,stdout),y=8,z=0:0;}
Well, my C is just fine but my HTML leaves something to be desired. Why on earth does previewing the comment remove the lt codes?
#include <stdio.h>
= (c=='1')?(1<<i):(isspace(c ))?putc(t,stdout),i=8,t=0:0;}
#include <ctype.h>
void main(){int c,i=8,t=0;while(((c=getc(stdin))!=EOF)&&i==i--)t|