If ease of use is an issue for your data entry clerks because they don't have much computer experience, then MS is the worst thing to give them.
Unless the apps you are using are legacy Windows (and sometimes even DOS) apps that have already been used for years... It's all about the apps... not the OS.
Of course, we could just not all work ourselves into a tizzy and wait for some other benchmark sites to do a better job (AnandTech usually is much better). If the test is invalid because of XYZ reasons, then it is invalid. What's the use of arguing over an invalid test? Kind of a waste of time, IMO. There will be other reviews.
It almost is. Back in the day, I was surprised at the response to the Palm machines. Today, almost everyone has some kind of geek tendency (spend time on message boards, play online games, chats, etc.) I think everyone will evolve into geeks at some point if they aren't born that way:)
Yes... It does say that. But it also says further down in the Remarks section:
The function modifies the state of some types of synchronization objects. Modification occurs only for the object or objects whose signaled state caused the function to return. For example, the count of a semaphore object is decreased by one. When bWaitAll is FALSE, and multiple objects are in the signaled state, the function chooses one of the objects to satisfy the wait; the states of the objects not selected are unaffected.
So, you can do it either way - use the WFMO to wait for one, then poll the rest every time you get control back, or you can just put it in a loop. You will pay basically the same penalty either way as you'll have to call WFSO inside a loop inside the WFMO to get each event. IMO, it's best to just use the simple WFMO in a loop alone.
Polling typically sacrifices cycles for latency. Async I/O sacrifices latency for cycles (for doing other things). Each has its advantages/disadvantages and problems that work well for each method. For example, if you have a tightly coupled problem that is data dependent, you have nothing else to do but burn cycles waiting on data. However, if you have a much more coarse grained app (like some client/server or master/slave) then the async can work very well.
Yeah, in the past, I've basically put a while loop around the WFMO and just loop. Inside the loop is a multiplexor based on the event signaled if you need it. The code is very simple to do this.
I've use the Asynchronous I/O with windows sockets and got very nice results. It was very hard to get nearly the same results on Linux back then (1998-ish) but things may have changed since then on the Linux side. I haven't had to try to get the same levels of performance out of subsystems since those projects because the things I've worked on since then haven't required it.
In my opinion, the POSIX thread/synchronization stuff is bad. It think it's ugly and clunky. I had a friend who was on the POSIX committee for it and I told him the same.
*yawn* Ellison has been yapping this stuff for the past 5 years or something. Every few months he says the same thing *again* and someone writes it up *again* and posts it as news.
Personally, I couldn't care less what Ellison thinks about Linux/Windows. I'll believe that he stands behind his statements when he discontinues development/support for Oracle for Windows. Until then, he is just another yapper.
Well... first, this is very old news. I read in on other sites over a week ago.
Second, (I'm not saying anything of the case here, but) two of the most notorious bugs I have ever tracked down in software were flaws in algorithms which resulted in incorrect results. In each case, the algorithms made assumptions that were simply incorrect, and terminated the algorithm early or "optimized out" cases that should not have been. In both cases, fixing the bug did result in a slowdown. So, the choice was to take a performance hit and have code that actually worked, or leave the faster, buggy code in place.
"So tell me. Do you go through the pains of examining the source of every application you use?
Not the point. The point is: At least with OSS I *can*. "
No... the point is exactly what he said. You see so many folks touting that they can review the code to make sure that it doesn't do anything they don't like so they trust it. Do they? More often not. So... they blindly trust it assuming that *someone else* has reviewed it and hasn't found such a problem. This is no different than not having the source or using closed source. In either case, you haven't a clue as to what the software is really doing, you just trust that it doesn't do anything bad and then crow that simply having the source code in their possession somehow makes it more trustworthy.
Let me draw a quick parallel. The Windows folks get tricked by having files named ReadThis.bat sent to them via email. They clearly have the source available because it is a BAT file (script). They don't review the code but run it anyway and bad things happen.
I have a number of boxes that run Windows that run 24/7 (except for the patch reboots) used for development, SQLServer, etc. and I haven't seen a BSOD in around 2 years.
Very important rule for Windows: Buy and use only name-brand, quality hardware. If you buy cheap crap hardware that you've never heard of then you get what you deserve.
"Christianity, Islam, Judaism, and any religion has absolutely nothing to do with reality."
This statement in itself is completely false. Regardless of what you believe, the fact that other people believe these things, some of which dictate the policy of whole nations, make these things have everything to do with reality. It is real and a fact that the religious rules of Islam dictate how people dress and act in some countries, for instance. That makes it reality.
I understand what you are saying completely and I agree with you. I'm not arguing that what I'm saying is "right". I'm saying it is simply reality (thankfully a reality that is changing every day).
I'm not one of the folks I described either. I'm 34, jeans/t-shirt/polo-type, sneakers as well. I don't have any piercings, mostly because I don't want to deal with the maintenance and I think it would be distracting (I don't even wear a watch). I do have tatoos but none are visible if I'm wearing a shirt/pants. I've been a senior level programmer for around 8 years or so now. Computers have been my hobby, diversion, entertainment, and pretty much most of what I do since the late 70s. Computers have also been my profession since the late 80s. This is not my point.
Stand in the shoes of a suit who knows little/nothing about computers for a few. What do you think of when you hear of "hobbyist programmer"? Sure, they stereotype us just like any other type of person is stereotyped. Is it fair? Who cares. It happens. They've seen pictures of alternative types who are UN*X/Linux users/programmers and they assume that is what we look like (although that perception is changing with the weight of companies like IBM/HP/etc. talking about Linux).
Now... think of this: "Hey boss, here is some new software we can start using. It was written by hobbyist programmers. It's better software than that stuff we buy for big $$$$$ from CompanyX." Boss thinks: "hmm... hobbyist programmers... write stuff in their spare time better than folks who do it professionally for a living... ya... i'll believe that one".
The reason I bring up this point is because I have *witnessed* it firsthand on more than one occassion in more than one organization/company. Companies like IBM, HP, RedHat, Oracle, and the others add serious credibility to the claims that suits hear. They have definitely heard of IBM and Oracle before (and have most likely seen things about Linux in the various business rags). Knowing that Linux has that kind of professional, high-profile backing lessens their fears. Combine this with IT/Programmers who have earned their trust giving praises to Linux greatly helps convince them that the risk of switching from whatever they have to Linux isn't that big.
Again... I'm not saying that all Linux people are freaky. I'm not saying that stereotyping people is "right". I'm saying that regardless of what you and I think, there are people who think this way and that certainly has influence on their decisions. The term "hobby", courtesy of http://www.m-w.com means:
Main Entry: 2hobby Function: noun Inflected Form(s): plural hobbies Etymology: short for hobbyhorse Date: 1816 : a pursuit outside one's regular occupation engaged in especially for relaxation
has the connotation of something that is done not as a profession. Even the term "enthusiast" doesn't carry seriousness with it. The basic idea that I'm trying to get across is that these terms do not add strength to the perception of Linux or anything that you want to use seriously (in essence, betting the farm on). You have hobbies and things you are enthusiastic about and then you have things that you do professionally. Would you go to a physician who practiced medicine as a hobby or was simply an enthusiast about medicine? Would you hire someone who only practiced law as a hobby to represent you in court?
Sure Linux and the associated applications were started and have been maintained by hobbyists for quite a while. Today, more and more professionals are getting into Linux. The professionalism and professionals are what sell Linux and get it to expand into the realms where you have to put money where your mouth is. Sure, lots of smaller companies bought into Linux a long time ago, particularly as a cost saving measure or because they were already familiar with UN*X.
How many Fortune500 companies were using Linux back in the pre-1.0 kernel days? in the pre-2.0 kernel days? There are many reasons why, one of which I am proposing as being the image that Linux is/was created/maintained by "hobbyists" and, therefore, wasn't a serious thing. The risk to adopt the platform were correspondingly high.
I'm not saying that hobbyists don't contribute good code, make things better, yada yada yada.
If you look around, most of the threads you see are about Linux "taking over the desktop" and "taking over the server world" type thing.
What I was saying is that to "take over the desktop" and to "take over the server world" you have to show that the stuff isn't maintained by a bunch of tatood, body pierced, wild haired bunch of teenagers. That image does not inspire confidence to folks in suits. You won't see an exec saying "sure, I'll trust my 4TB customer database to something that was written by a hack who lives in his mother's basement living off of bawlz and pizza" regardless of whether you or I are actually one of those folks or whether we know that folks like that can produce good code. They will be worried about the guy putting back doors into the program, stability issues, etc.
The inroads that have been made into those areas have been because there *are* companies who give the stuff a professional air and folks who have earned the trust of those they work for who are in positions inside those companies to assuage any fears or doubts about the software.
The path works both ways. Hobbyists who have a good reputation at home can promote Linux at work. Folks who know little/nothing about Linux can be exposed to it at work and bring it home (Joe Sixpack, for instance).
I used to run Linux because it was kewlz0r. I grew up. I run Linux now because it does what I want it to do.
No... I was in my robe. There was no reason for him to kill me (got nothing from it) and when he went into fighting stance, I just stayed out of his reach. If he had magic skill, it would have been more difficult.
Heh... In Ultima Online once, I was PK'd by a griefer. Instead of letting him grief me, I just followed him around in my robes asking him questions like "does it make you feel big?" and any time someone else came on the screen, I'd let them know he was a PK griefer. Every time he would hide so he could ambush someone else, I'd make sure everyone knew where he was and that he was going to PK them. He couldn't kill me (again) but he couldn't shake me either. Eventually he got a friend to come over to help kill me and I was able to lure that one close to the guards who killed him.
I had a very fun time doing this actually, after about 30 minutes or so, the guy got so pissed off at me he logged out.
"3. Most people developing for Linux are HOBBYISTS. This means, they use the tools they enjoy using. And there's NOTHING wrong with this, either. Different people, using different tools and following their own muses, make life interesting."
This is a major part of the problem that Linux has to overcome now. Businesses do not like "hobby" tools or the idea that their business will be letting a "hobby" be in control of its data/etc. Linux and Linux apps must present, at least an appearance of, being professional quality. The "hobby" stigma of Linux is one of the things holding it back from taking over even faster, IMO. Views are changing and Linux is getting wider acceptance daily, but it is most definitely *not* due to the perception that it is a "hobby" that is created/propagated by "hobbyists". It is because it is taking on the appearance of (and becoming) a professional quality OS with professional quality apps.
Slashdot Mirror
Sounds like the rest of your class had no business being in a compiler course. They probably wouldn't have been able to do anything on Linux either.
If ease of use is an issue for your data entry clerks because they don't have much computer experience, then MS is the worst thing to give them.
Unless the apps you are using are legacy Windows (and sometimes even DOS) apps that have already been used for years... It's all about the apps... not the OS.
Not to mention that the latency of IP is staggering.
Of course, we could just not all work ourselves into a tizzy and wait for some other benchmark sites to do a better job (AnandTech usually is much better). If the test is invalid because of XYZ reasons, then it is invalid. What's the use of arguing over an invalid test? Kind of a waste of time, IMO. There will be other reviews.
Windows has also had windows with varying transparency for at least 3 years now. I remember thinking it was neat back then and playing with it.
It almost is. Back in the day, I was surprised at the response to the Palm machines. Today, almost everyone has some kind of geek tendency (spend time on message boards, play online games, chats, etc.) I think everyone will evolve into geeks at some point if they aren't born that way :)
Yes... It does say that. But it also says further down in the Remarks section:
The function modifies the state of some types of synchronization objects. Modification occurs only for the object or objects whose signaled state caused the function to return. For example, the count of a semaphore object is decreased by one. When bWaitAll is FALSE, and multiple objects are in the signaled state, the function chooses one of the objects to satisfy the wait; the states of the objects not selected are unaffected.
So, you can do it either way - use the WFMO to wait for one, then poll the rest every time you get control back, or you can just put it in a loop. You will pay basically the same penalty either way as you'll have to call WFSO inside a loop inside the WFMO to get each event. IMO, it's best to just use the simple WFMO in a loop alone.
Polling typically sacrifices cycles for latency. Async I/O sacrifices latency for cycles (for doing other things). Each has its advantages/disadvantages and problems that work well for each method. For example, if you have a tightly coupled problem that is data dependent, you have nothing else to do but burn cycles waiting on data. However, if you have a much more coarse grained app (like some client/server or master/slave) then the async can work very well.
Yeah, in the past, I've basically put a while loop around the WFMO and just loop. Inside the loop is a multiplexor based on the event signaled if you need it. The code is very simple to do this.
I've use the Asynchronous I/O with windows sockets and got very nice results. It was very hard to get nearly the same results on Linux back then (1998-ish) but things may have changed since then on the Linux side. I haven't had to try to get the same levels of performance out of subsystems since those projects because the things I've worked on since then haven't required it.
In my opinion, the POSIX thread/synchronization stuff is bad. It think it's ugly and clunky. I had a friend who was on the POSIX committee for it and I told him the same.
How many RedHat patches have you gotten in the past month? I'm averaging about one every three days.
heh... but were you working on these patches before reading the post you responded to?
Except that when I've installed Windows XP, it takes less time than it predicts. /shrug
heh... Why would you log into your home machine remotely just to remotely use a web browser?
I think WINE/WINEX runs maybe one of the games that I play regularly.
Anyway, there's no greater zealot than the converted.
Both OSs leave a lot to be desired. Each is useful depending on what I want to do. I use both, daily, to get my work done.
*yawn* Ellison has been yapping this stuff for the past 5 years or something. Every few months he says the same thing *again* and someone writes it up *again* and posts it as news.
Personally, I couldn't care less what Ellison thinks about Linux/Windows. I'll believe that he stands behind his statements when he discontinues development/support for Oracle for Windows. Until then, he is just another yapper.
Well... first, this is very old news. I read in on other sites over a week ago.
Second, (I'm not saying anything of the case here, but) two of the most notorious bugs I have ever tracked down in software were flaws in algorithms which resulted in incorrect results. In each case, the algorithms made assumptions that were simply incorrect, and terminated the algorithm early or "optimized out" cases that should not have been. In both cases, fixing the bug did result in a slowdown. So, the choice was to take a performance hit and have code that actually worked, or leave the faster, buggy code in place.
Heh, no kidding. I could buy at least a dozen roses per week all year-round for what I spend in computer parts each year.
"So tell me. Do you go through the pains of examining the source of every application you use?
Not the point. The point is: At least with OSS I *can*.
"
No... the point is exactly what he said. You see so many folks touting that they can review the code to make sure that it doesn't do anything they don't like so they trust it. Do they? More often not. So... they blindly trust it assuming that *someone else* has reviewed it and hasn't found such a problem. This is no different than not having the source or using closed source. In either case, you haven't a clue as to what the software is really doing, you just trust that it doesn't do anything bad and then crow that simply having the source code in their possession somehow makes it more trustworthy.
Let me draw a quick parallel. The Windows folks get tricked by having files named ReadThis.bat sent to them via email. They clearly have the source available because it is a BAT file (script). They don't review the code but run it anyway and bad things happen.
I agree.
:) We could have milliBug, centiBug, Bug, kiloBug, etc!
Should we have a new metric measurement? The Bug?
I have a number of boxes that run Windows that run 24/7 (except for the patch reboots) used for development, SQLServer, etc. and I haven't seen a BSOD in around 2 years.
Very important rule for Windows: Buy and use only name-brand, quality hardware. If you buy cheap crap hardware that you've never heard of then you get what you deserve.
heh... anyone else getting annoyed at RedHat having patches every day for like the last month?
For being wrong.
"Christianity, Islam, Judaism, and any religion has absolutely nothing to do with reality."
This statement in itself is completely false. Regardless of what you believe, the fact that other people believe these things, some of which dictate the policy of whole nations, make these things have everything to do with reality. It is real and a fact that the religious rules of Islam dictate how people dress and act in some countries, for instance. That makes it reality.
I understand what you are saying completely and I agree with you. I'm not arguing that what I'm saying is "right". I'm saying it is simply reality (thankfully a reality that is changing every day).
I'm not one of the folks I described either. I'm 34, jeans/t-shirt/polo-type, sneakers as well. I don't have any piercings, mostly because I don't want to deal with the maintenance and I think it would be distracting (I don't even wear a watch). I do have tatoos but none are visible if I'm wearing a shirt/pants. I've been a senior level programmer for around 8 years or so now. Computers have been my hobby, diversion, entertainment, and pretty much most of what I do since the late 70s. Computers have also been my profession since the late 80s. This is not my point.
Stand in the shoes of a suit who knows little/nothing about computers for a few. What do you think of when you hear of "hobbyist programmer"? Sure, they stereotype us just like any other type of person is stereotyped. Is it fair? Who cares. It happens. They've seen pictures of alternative types who are UN*X/Linux users/programmers and they assume that is what we look like (although that perception is changing with the weight of companies like IBM/HP/etc. talking about Linux).
Now... think of this: "Hey boss, here is some new software we can start using. It was written by hobbyist programmers. It's better software than that stuff we buy for big $$$$$ from CompanyX." Boss thinks: "hmm... hobbyist programmers... write stuff in their spare time better than folks who do it professionally for a living... ya... i'll believe that one".
The reason I bring up this point is because I have *witnessed* it firsthand on more than one occassion in more than one organization/company. Companies like IBM, HP, RedHat, Oracle, and the others add serious credibility to the claims that suits hear. They have definitely heard of IBM and Oracle before (and have most likely seen things about Linux in the various business rags). Knowing that Linux has that kind of professional, high-profile backing lessens their fears. Combine this with IT/Programmers who have earned their trust giving praises to Linux greatly helps convince them that the risk of switching from whatever they have to Linux isn't that big.
Again... I'm not saying that all Linux people are freaky. I'm not saying that stereotyping people is "right". I'm saying that regardless of what you and I think, there are people who think this way and that certainly has influence on their decisions. The term "hobby", courtesy of http://www.m-w.com means:
Main Entry: 2hobby
Function: noun
Inflected Form(s): plural hobbies
Etymology: short for hobbyhorse
Date: 1816
: a pursuit outside one's regular occupation engaged in especially for relaxation
has the connotation of something that is done not as a profession. Even the term "enthusiast" doesn't carry seriousness with it. The basic idea that I'm trying to get across is that these terms do not add strength to the perception of Linux or anything that you want to use seriously (in essence, betting the farm on). You have hobbies and things you are enthusiastic about and then you have things that you do professionally. Would you go to a physician who practiced medicine as a hobby or was simply an enthusiast about medicine? Would you hire someone who only practiced law as a hobby to represent you in court?
Sure Linux and the associated applications were started and have been maintained by hobbyists for quite a while. Today, more and more professionals are getting into Linux. The professionalism and professionals are what sell Linux and get it to expand into the realms where you have to put money where your mouth is. Sure, lots of smaller companies bought into Linux a long time ago, particularly as a cost saving measure or because they were already familiar with UN*X.
How many Fortune500 companies were using Linux back in the pre-1.0 kernel days? in the pre-2.0 kernel days? There are many reasons why, one of which I am proposing as being the image that Linux is/was created/maintained by "hobbyists" and, therefore, wasn't a serious thing. The risk to adopt the platform were correspondingly high.
I'm not saying to ignore heritage.
I'm not saying that hobbyists don't contribute good code, make things better, yada yada yada.
If you look around, most of the threads you see are about Linux "taking over the desktop" and "taking over the server world" type thing.
What I was saying is that to "take over the desktop" and to "take over the server world" you have to show that the stuff isn't maintained by a bunch of tatood, body pierced, wild haired bunch of teenagers. That image does not inspire confidence to folks in suits. You won't see an exec saying "sure, I'll trust my 4TB customer database to something that was written by a hack who lives in his mother's basement living off of bawlz and pizza" regardless of whether you or I are actually one of those folks or whether we know that folks like that can produce good code. They will be worried about the guy putting back doors into the program, stability issues, etc.
The inroads that have been made into those areas have been because there *are* companies who give the stuff a professional air and folks who have earned the trust of those they work for who are in positions inside those companies to assuage any fears or doubts about the software.
The path works both ways. Hobbyists who have a good reputation at home can promote Linux at work. Folks who know little/nothing about Linux can be exposed to it at work and bring it home (Joe Sixpack, for instance).
I used to run Linux because it was kewlz0r. I grew up. I run Linux now because it does what I want it to do.
No... I was in my robe. There was no reason for him to kill me (got nothing from it) and when he went into fighting stance, I just stayed out of his reach. If he had magic skill, it would have been more difficult.
Heh... In Ultima Online once, I was PK'd by a griefer. Instead of letting him grief me, I just followed him around in my robes asking him questions like "does it make you feel big?" and any time someone else came on the screen, I'd let them know he was a PK griefer. Every time he would hide so he could ambush someone else, I'd make sure everyone knew where he was and that he was going to PK them. He couldn't kill me (again) but he couldn't shake me either. Eventually he got a friend to come over to help kill me and I was able to lure that one close to the guards who killed him.
I had a very fun time doing this actually, after about 30 minutes or so, the guy got so pissed off at me he logged out.
"3. Most people developing for Linux are HOBBYISTS. This means, they use the tools they enjoy using. And there's NOTHING wrong with this, either. Different people, using different tools and following their own muses, make life interesting."
This is a major part of the problem that Linux has to overcome now. Businesses do not like "hobby" tools or the idea that their business will be letting a "hobby" be in control of its data/etc. Linux and Linux apps must present, at least an appearance of, being professional quality. The "hobby" stigma of Linux is one of the things holding it back from taking over even faster, IMO. Views are changing and Linux is getting wider acceptance daily, but it is most definitely *not* due to the perception that it is a "hobby" that is created/propagated by "hobbyists". It is because it is taking on the appearance of (and becoming) a professional quality OS with professional quality apps.