Surely there's a synchronization cost if both cpus are accessing the same resource.
You have synchronization in dual cache too because there must be away for them to tell each other when it is being modified.
More, with dual cache, synchronization may actually be worse because it's two "distant" units that need to be synchronized instead of one cache telling itself to not give the data to the other processor yet. You can tell something to yourself at least as fast as (and hopefully faster than) you can tell someone else.
Yea, if 3 out of 5 machines failed to come back up, it needs some polishing.
Yes, until 5 out of 5 machines fail to come back up, there will be security issues. The thing I don't understand is why it takes so long to prevent all machines from booting....
It helped me react to the situation better, and more quickly than I think I otherwise would have
And how many people will just freeze in place or panic to the point of hindering the rescue team? I don't have numbers, but from what I read, it seems that people are more subject to freeze/panic than anything else in such situation.
The problem is not about desensitization in itself, it's about not understanding what violence really is and what the consequences are. Not being afraid of blood is good, not being afraid of shoutout is good, thinking that it's normal to beat someone up because you don't like his shirt color is not good.
And it's also very complicated to use. I still don't understand how this thing works. More often than not, when I want it to do something, I end up doing it. Worse, sometimes I don't ask for anything and I end up doing something still:(
Wrong, you don't want a RAID 0+1, you want RAID 1+0. In the RAID 0+1, you are more likely to lose your data if two disks fails. If you have four drives A, B, C and D. RAID 0+1 is about building two RAID 0 (A+B and C+D) which you then combine in a RAID 1. RAID 1+0 is two RAID 1 which you them combine in a RAID 0.
In the case of RAID 0+1, if A fails, A+B isn't usable anymore and you get your data from C+D. So if C or D fails, you're screwed. So, if a second drive fails, 2/3 of the time, you lose all your data. In the case of RAID 1+0, if A fails, you get your data for B and (C+D). So you get screw only if B fails. If a second drive fails, you lose all your data only 1/3 of tihe time.
What about the unfair advantage that C++ is natively compiled for a specific processor (and thus optimized for it) while Java is compiled for a virtual machine? What about the unfair advantage that C++ programmers can do tricks with pointers to make their application run faster while Java programmers can't?
The two languages, while similar, are not identical. So you can always find the benchmark biased if you want to. It all depend on the "rules" you give yourself. If the benchmark is about coding an application that does something specific, it doesn't matter how the thing is done. If the program use a time machine to give the answer faster, that fine with me even if the real processing times is 100x slower than when using another language/environment.
Don't confuse C++ and "Managed C++". Managed C++ is a C++ like language that is compiled to run on a.NET virtual machine. Regular C++ doesn't need a virtual machine to run.
Visual Studio.NET supports both regular C++ and managed C++. If you see C++ and C# used together to describe a framework/compiler/..., chances are that it's really about managed C++. I didn't look at Mono but if they say that they have a C++ compiler, it's very likely that they actually have a managed C++ compiler.
Hmmm... Let's suppose, for just a moment, that you will take everyone's guns away. Make it a crime to carry a gun. Ok, now take a deep breath and think about this next one before answering it: Will the thieves also give up their guns? Or will they be empowered because suddenly they are the only ones with guns? Think this one through very carefully.
Did you think that one through? If we follow through your reasonning, what you are basically suggesting is to banish all laws because criminals don't obey them anyway.
it only requires 64 MB of RAM Where did you see that? I read that the choice device has 64MB of physical memory. AFAIK, in most such devices, this includes the files on the "disk", the OS on the "disk", the runtime data for both OS and applications,... And if you RTFA, you'll see that Minimo uses only 25MB.
And even then, "formatting pictures" can use quite a bit a memory (see various 3D games)
No, no, you don't get it, it's because of all the filters. It fixes posts. So now, pe0ple kan wreite az badi az the want & get there pausts fyxt too a corekt form@t. This is a big help for spam filtering, no more v.iagr@, v1 gra, and stuff.
I wonder if this technology would work for/. to spelling.
Slashdot Mirror
More, with dual cache, synchronization may actually be worse because it's two "distant" units that need to be synchronized instead of one cache telling itself to not give the data to the other processor yet.
You can tell something to yourself at least as fast as (and hopefully faster than) you can tell someone else.
You must be new here, don't you that know *BSD are dead?
From your link:
Am I missing something here?
The thing I don't understand is why it takes so long to prevent all machines from booting....
Yes, it's surprising, I'm more used to scooting poopers
No, no, please, I'm just bleeding, my shirt is really white.
It helped me react to the situation better, and more quickly than I think I otherwise would have
And how many people will just freeze in place or panic to the point of hindering the rescue team? I don't have numbers, but from what I read, it seems that people are more subject to freeze/panic than anything else in such situation.
The problem is not about desensitization in itself, it's about not understanding what violence really is and what the consequences are.
Not being afraid of blood is good, not being afraid of shoutout is good, thinking that it's normal to beat someone up because you don't like his shirt color is not good.
I don't know about that particular case (I'm not even sure that it's not an hoax) but the thing is that you can file a frivolous lawsuit and win
And it's also very complicated to use. I still don't understand how this thing works. More often than not, when I want it to do something, I end up doing it. Worse, sometimes I don't ask for anything and I end up doing something still :(
You don't want to use RAID 0+1 (aka 01) you want RAID 1+0 (aka 10).
See my other post
If you have four drives A, B, C and D. RAID 0+1 is about building two RAID 0 (A+B and C+D) which you then combine in a RAID 1. RAID 1+0 is two RAID 1 which you them combine in a RAID 0.
In the case of RAID 0+1, if A fails, A+B isn't usable anymore and you get your data from C+D. So if C or D fails, you're screwed. So, if a second drive fails, 2/3 of the time, you lose all your data.
In the case of RAID 1+0, if A fails, you get your data for B and (C+D). So you get screw only if B fails. If a second drive fails, you lose all your data only 1/3 of tihe time.
See that page
What about the unfair advantage that C++ is natively compiled for a specific processor (and thus optimized for it) while Java is compiled for a virtual machine?
What about the unfair advantage that C++ programmers can do tricks with pointers to make their application run faster while Java programmers can't?
The two languages, while similar, are not identical. So you can always find the benchmark biased if you want to. It all depend on the "rules" you give yourself.
If the benchmark is about coding an application that does something specific, it doesn't matter how the thing is done. If the program use a time machine to give the answer faster, that fine with me even if the real processing times is 100x slower than when using another language/environment.
Don't confuse C++ and "Managed C++". Managed C++ is a C++ like language that is compiled to run on a .NET virtual machine. Regular C++ doesn't need a virtual machine to run.
.NET supports both regular C++ and managed C++.
Visual Studio
If you see C++ and C# used together to describe a framework/compiler/..., chances are that it's really about managed C++.
I didn't look at Mono but if they say that they have a C++ compiler, it's very likely that they actually have a managed C++ compiler.
No, it's not built on .NET, it's regular C++. It was programmed using Visual Studio .NET and uses MFC for the GUI.
(admire my l337 sills in ASCII arts!)
it only requires 64 MB of RAM
Where did you see that? I read that the choice device has 64MB of physical memory. AFAIK, in most such devices, this includes the files on the "disk", the OS on the "disk", the runtime data for both OS and applications,...
And if you RTFA, you'll see that Minimo uses only 25MB.
And even then, "formatting pictures" can use quite a bit a memory (see various 3D games)
I was going to mod you up, but given your signature, I'm not going to.
And if I didn't feel like warning others, I would mod you Troll +1.
Every time you breath, you inhale pollen, dust mites, various chemical vapors, and all sorts of organic detritus.
[...]
take a deep breath
Yeah, right!
No, no, you don't get it, it's because of all the filters. It fixes posts. So now, pe0ple kan wreite az badi az the want & get there pausts fyxt too a corekt form@t. This is a big help for spam filtering, no more v.iagr@, v1 gra, and stuff.
/. to spelling.
I wonder if this technology would work for