Doesn't multiple competing standards for single sign-on basically defeat the purpose of single sign-on since all the sites you visit will have all hop on the same bandwagon?
This is a community of dorks and nerds. Who the heck thought "Director of Titanic" was going to get more attention than "Director of the Terminator movies." Hello! Titanic and is the kind of movie you go see with your girl friend and I don't think I need to discuss how few of us here have those.....
So I guess what you're saying is that we're all born with the same potential. It just matters where we spend our time. That seems a bit idealistic to me. You want to explain why some people are more athletic than others? Why some people have higher iq's than others? Do you want to somehow argue that this is purely and totally environment and has nothing at all to do with genetics?
I have a real hard time accepting that as I suspect that most people on/. would.
There's an old expression that says "the world needs ditch diggers too". That is to say that some people just aren't good at some things and there's nothing wrong with that. Not everyone has the intellect to be college material no matter how hard they study or practice. Likewise, some people just don't have the knack for art and some don't for programming.
Likewise, there is little any book can teach you about which colors just look good together or how movement should flow across a graphic. They can teach you basics like complimentary colors and vanishing points and different things. There are guidelines that they'll teach you in basic art courses but these are just that. Guidelines. True artistic talent comes from having the judgement of when and how to break these and when to embrace these and there is little any book can do to quantify that.
The problem with art is that its all based on opinion and you can't quantify opinion. I know it frustrates us logical types to think there are things that can't be quantified, but those things are out. Our culture and our human nature are full of them.
Now what you're pointing out with that book is good. It is possible to develop the artistic talent that we have and make the most of what we have. But making the most of what we have isn't the same as being top notch.
Likewise, if you don't have artistic talent, you're opinion of how awesome your friends are at art may not be accurate. You may be easy to impress with mediocre works. Other recognized artists may say they suck. On the other hand, you might be some sort of a prodigy and your friends too. Who knows? It's all opinion either way.
Point is, everything is relative and subject to majority rule as far as opinions go. It's just unusual that people are relatively high in both logic and popularly accepted art talent.
You make one key (and very false) assumption. Art generally isn't intellectual. That's what makes it hard for most of us tech people to pick up. In fact people who demonstrate high skill in both art and programming are so rare that even in these jobless times, they're almost guarenteed a job in the gaming industry if they want it.
Art is one of those things that you generally have to have a knack for or all the practice in the world isn't going to help. This is not to say the originally poster doesn't have the knack. I don't know. But if he doesn't, he should put his money toward hiring an artist instead of buying art books.
I went to college with quite a few guys that were ace programmers but couldn't sketch their way out of a wet paper bag. At the same time, a tremendous number of the greatest artists (even digital artists) barely have enough technical knowledge to turn on a pc let alone program one.
It's just that these two activities use very different parts of the brain and not many people have both parts sufficiently developed to be good at both.
The issue here was that hardware raid is no less recoverable than software raid. Bad sectors tend to creap up a little at a time. If you have a drive showing bad sectors (that is the drive has run out of spare sectors to remap) you damn well should replace that drive ASAP. There is no real excuse for the partial failure of multiple drives that you discribe other than incompetence on the part of the admin. It doesn't matter if it's hardware or software raid.
As for total failure on two drives, it's conceivable that some files would not span across more than one failed disk. However, then there's no way to check the integrity of that file either. If you're looking at total failure on more than one disk, more often than not it's either a heat or power problem which likely means that other drives are suffering or not far behind and likewise cannot be trusted. This iswhere we start talking about things like raid 50 across seperate enclusres (in case of heat problems from ventilation failure and certainly disperate powersupplies), but the issue here is SW raid which is quite impracticle for any reasonable sized raid 50 array in software.
Likewise, files that small probably should not be on a raid 5 array anyway since they're going to cause all sorts of "small write" problems. Something that noone seems to understand is that raid 5 is not an all perfect solution. If you do not write a file the whole way across the array, that is hit every disk except parity, you have to read the other cluster that you didn't write to in order to calculate the new parity.
Of course, small writes are a significantly smaller problem on hardware raid than software since much of the redundancy from the reads and checksumming invovled with small writes is contained locally to the card and does no require the intervention of the cpu thereby keeping quite a bit of unecessary data off the system bus.
If you have quite a few small files and more than about 3 maybe 4 drives modern fast drives tops in your raid 5 array, you're going to kill your array performance on SW raid 5. Perhaps you should look at raid 10. as a more small file friendly solution as there are less disks to read and no checksumming to perform.
As for "forcing" the sytem to mount a damaged disk, total drive failure implies that the disk is well....totally failed. Let's stop and think carefully about this. That means it can't be read. I am not aware of any raid controllers that won't continue operating with a disk that has bad sectors in order for you to continue revovering what you can (virtually the whole array). In the case of total failure, most cards have some form of maintenance or forced mode in which they will continue to operate in an unrecoverable state so that you can salvage what little is there.
Now, lets discuss raid 0 and recovering files. Let's assume a striped set of just two disks since you'd be insane to stripe more than that. Too many points of failure, data integrity and all those problems. I just checked. A completely blank MS Word file (I know, I'm a heretic for not using Open Office) and that comes to 23.5 and an even 24k on disk. Now, golly, wonder what the cluster size is on my disk. (Counts on fingers) I reckon it must be 8K. Now I have a realatively small 60gig disk. Lets say that 16k clusters are more on the average for most people now. That means that even for a totally blank Word document, you need two full clusters. Hopefully I don't need to explain that the way striping (raid 0) works is that every other cluster goes on the other disk. Just how many many modern document types do you think won't span the striped set. Certainly no office documents and there aren't many source files either.
Raid 5 makes things a little more complicated since the most basic raid 5 set would be similar in situation with 2 data disks and 1 parity disk. One thing to point out, as the number of disks in a raid 5 set go up, so does the chance to recover small files. However, so does the overall volatility of the set as there a
If the machine is a dedicated file server this is conceivable. But in that case, you're wasting your time worrying about performance on the array anyway since your network will be a super bottleneck anyway. In this situation, the only purpose of raid 5 is for data integrity.
The real place where the performance side of raid 5 matters is for things like application (especially in the web sense) and database servers. That is, where the data consumed by the server is far more significant than the data dispatched to the clients. In cases like this, CPU cycles are generally of high value and any ability to offload work to other devoted hardware is a necessity.
I got a newsflash for you. If you're using raid 5 and you drop more than one drive at the same time, you're done period. It doesn't matter if you're running hardware or software raid. It's the way raid 5 works. RAID 5 can recreate the info from any one drive. But if it loses two drives, it can't determine both variables. There is no convicing about it. If two drives die, so does your array. Do not pass go. Do not collet $200.
It scares me that they let people like you play with the sort of computing resources that have 50TB of disk space.
I don't think it's going to make much difference. It seems new and novel now, but within a couple years, all the schools will have competing programs. And it will be like trying to shop for a cell phone now. All the schools will have different plans that are nearly impossible to compare to one another to really be objective.
Besides, some schools have been making things like laptops part of the tuition for years. This really isn't that new when you look at it this way. It's just the next evolution in perks one gets from going to college. It used to be access to a great library. Then it was great discounts when you flashed your student ID. The last few years, it's been very high speed internet access. Now it's free music downloads.
Ummmm....is this where we should discuss the piss poor dacs that virtually all computer sound cards whether on board or add-on use? There are exceptions, but those are few and far between and are certainly not the rule. Should we also discuss the problems with rf/emi inside a computer case and what that will do to analog signals?
Anyone that claims to be serious about pro audio production on their computer better be running digital outside their box into an external dac, especially if you think you really need mackies!
Thinking parent poster must be deaf to not hear any problems with that setup.
Find a responsible roommate ASAP. Locking the door when you leave won't do crap if your roommate doesn't do the same. This may not help for your first semester as most universities insist that you suffer with your randomly selected roommate for at least one semester, but should improve there after.
Also, find a roommate you can trust. I've heard plenty of horror stories about the roommate being the ultimate thief when everything was settled. Again, this may not help your first sememster/year, but make it a priority. At least then you only have a small peice of your college life to suffer with minimal effects in your room for fear of theivery.
The first thing to realize is that a web service no longer indicates internet anymore. It's a shift in paradigm. It's time for a lot of you guys to throw out your blue polyester shirts and suspenders. Web simply means it's delivered from a central location and typically works through a browser or derivative of a browser. This doesn't mean 2 billion users will be streaming office from microsoft.com. It doesn't mean that when MS's servers go down that the entire world will be without Office.
This is simply the realization of the thin client paradigm. As corporate environments go, it's about time.
And before anyone panics about all those stand alone machines out there (like us developers are all so fond of), there are a bunch of appies out there that are essentially written this way already. VS.NET is web driven. That front end is all xml/html driven. We see it with the MS management console and MSC snappins for it too. This is the sort of thing we're looking at with the future of office.
The front end will be web based. The back end will likely have a few different options and standalone on the local machine I would wager will still be one of them. But at the same time, the back end could be centralized greatly simplifying mangement. I wouldn't be suprised if the next incarnation of Visual Studio can be set up to compile on a central server.
This should in theory simplify development of the Office software and reduce all versions of Office to a single codebase once.NET reaches full maturity and is available for other platforms. Realistically, Linux might be running MS office in the near future.
Glad to see you keep up and base your opinions on current versions of software. Clippy was killed in Office 2k3. Try living in the present. Or at least within the last year.
Ok, I'm not too much older than you if you're still in college, but I'm going to play old curmudgeon anyway.
you can put together a decent solution for computing now for around a grand. Kick in another $250 for needing a good workstation board to get the right slots and say $600 ($300x2) for the two cards and you're still just under $2000. THIS IS CHEAP. I'm sorry. I know how many lawns I had to mow as a youngin to buy my first pentium 60. That was $2k for JUST the computer and monitor. That included a baseline 1 meg video card, no cdrom and no sound. The cdrom and sound card cost me another $400 a couple months later.
So cry me a freaking river. Get a weekend job. Stop spending so much money on booz. If this is a priority for you, then you'll find the money. If it's not a priority, then quit your pissing and moaning.
Don't confuse the CPU a system uses for the entire performance value of the system. There are different bus and memory architectures that can do a lot to differentiate the performance of a "pricey" Sun with an AMD and the "value" machine you'd assemble from commodity parts
SGI did this with Pentiums (II's or III's if I remember correctly), though a lot depends on marketing which has not beeb SGI's strong point as of late so don't site SGI as an anecdote to predict Suns failure also.
Universities set these arbitrary requirements simply so there computer centers aren't supporting decade old hardware cause you know there would be some student show up next year with his family's old 486 running windows 3.0 and demanding they get him connected to the network in his dorm room.
I promise you, I can take the original parent post's notebook (the 500) and put it on Tech's dorm network. I can surf the web, I can take it to class and the library, write papers and code on it and Tech would never know the difference.
If anyone is silly enough to actually beleive they need a particular clock speed (especially in the multiple ghz range), hard drive capacity, etc to simply have their computer at college and presumably connected to the campus' network, then quite frankly they shouldn't be attending Virginia Tech(at least not for computer science or a related feild).
Well, perhaps you are the visionary here, but I see that convergence as very far off.If ever. We all know how slow standards move. HDTV anyone?
I look at my Sony DB930 Reciever (not super high end by anyone's definition, was $700 new) and it's pretty chock full at component width and some of the actuall highend models from B&K, Denon, Bob Carver, etc are even taller (at the same component width) and just as full or fuller. If I must have one component in my system that big, I'd rather waste a little space in the other components so they stack nice and neat.
But not all the other components are wasting space. My 400 disc CD carasel is pretty chock full (once it has CD's in it) as would be one of those spiffy DVD carasels. Yeah, my single disc tray loading DVD player could be thinner. I wish it was actually, but I don't want it any narrower.
These types of devices will never make it into my home entertainment system until they at east fit in my AV cabinet with my other components. 17 inches people. It's not that hard.
Slashdot Mirror
Doesn't multiple competing standards for single sign-on basically defeat the purpose of single sign-on since all the sites you visit will have all hop on the same bandwagon?
This is a community of dorks and nerds. Who the heck thought "Director of Titanic" was going to get more attention than "Director of the Terminator movies." Hello! Titanic and is the kind of movie you go see with your girl friend and I don't think I need to discuss how few of us here have those.....
So I guess what you're saying is that we're all born with the same potential. It just matters where we spend our time. That seems a bit idealistic to me. You want to explain why some people are more athletic than others? Why some people have higher iq's than others? Do you want to somehow argue that this is purely and totally environment and has nothing at all to do with genetics?
/. would.
I have a real hard time accepting that as I suspect that most people on
There's an old expression that says "the world needs ditch diggers too". That is to say that some people just aren't good at some things and there's nothing wrong with that. Not everyone has the intellect to be college material no matter how hard they study or practice. Likewise, some people just don't have the knack for art and some don't for programming.
Likewise, there is little any book can teach you about which colors just look good together or how movement should flow across a graphic. They can teach you basics like complimentary colors and vanishing points and different things. There are guidelines that they'll teach you in basic art courses but these are just that. Guidelines. True artistic talent comes from having the judgement of when and how to break these and when to embrace these and there is little any book can do to quantify that.
The problem with art is that its all based on opinion and you can't quantify opinion. I know it frustrates us logical types to think there are things that can't be quantified, but those things are out. Our culture and our human nature are full of them.
Now what you're pointing out with that book is good. It is possible to develop the artistic talent that we have and make the most of what we have. But making the most of what we have isn't the same as being top notch.
Likewise, if you don't have artistic talent, you're opinion of how awesome your friends are at art may not be accurate. You may be easy to impress with mediocre works. Other recognized artists may say they suck. On the other hand, you might be some sort of a prodigy and your friends too. Who knows? It's all opinion either way.
Point is, everything is relative and subject to majority rule as far as opinions go. It's just unusual that people are relatively high in both logic and popularly accepted art talent.
"It's just downright anti-intellectual"
You make one key (and very false) assumption. Art generally isn't intellectual. That's what makes it hard for most of us tech people to pick up. In fact people who demonstrate high skill in both art and programming are so rare that even in these jobless times, they're almost guarenteed a job in the gaming industry if they want it.
Art is one of those things that you generally have to have a knack for or all the practice in the world isn't going to help. This is not to say the originally poster doesn't have the knack. I don't know. But if he doesn't, he should put his money toward hiring an artist instead of buying art books.
I went to college with quite a few guys that were ace programmers but couldn't sketch their way out of a wet paper bag. At the same time, a tremendous number of the greatest artists (even digital artists) barely have enough technical knowledge to turn on a pc let alone program one.
It's just that these two activities use very different parts of the brain and not many people have both parts sufficiently developed to be good at both.
Before someone jumps on the chance to correct me, I mean raid 15 not 50. It's a sunday, I think I'm going back to bed now ;-)
The issue here was that hardware raid is no less recoverable than software raid. Bad sectors tend to creap up a little at a time. If you have a drive showing bad sectors (that is the drive has run out of spare sectors to remap) you damn well should replace that drive ASAP. There is no real excuse for the partial failure of multiple drives that you discribe other than incompetence on the part of the admin. It doesn't matter if it's hardware or software raid.
As for total failure on two drives, it's conceivable that some files would not span across more than one failed disk. However, then there's no way to check the integrity of that file either. If you're looking at total failure on more than one disk, more often than not it's either a heat or power problem which likely means that other drives are suffering or not far behind and likewise cannot be trusted. This iswhere we start talking about things like raid 50 across seperate enclusres (in case of heat problems from ventilation failure and certainly disperate powersupplies), but the issue here is SW raid which is quite impracticle for any reasonable sized raid 50 array in software.
Likewise, files that small probably should not be on a raid 5 array anyway since they're going to cause all sorts of "small write" problems. Something that noone seems to understand is that raid 5 is not an all perfect solution. If you do not write a file the whole way across the array, that is hit every disk except parity, you have to read the other cluster that you didn't write to in order to calculate the new parity.
Of course, small writes are a significantly smaller problem on hardware raid than software since much of the redundancy from the reads and checksumming invovled with small writes is contained locally to the card and does no require the intervention of the cpu thereby keeping quite a bit of unecessary data off the system bus.
If you have quite a few small files and more than about 3 maybe 4 drives modern fast drives tops in your raid 5 array, you're going to kill your array performance on SW raid 5. Perhaps you should look at raid 10. as a more small file friendly solution as there are less disks to read and no checksumming to perform.
As for "forcing" the sytem to mount a damaged disk, total drive failure implies that the disk is well....totally failed. Let's stop and think carefully about this. That means it can't be read. I am not aware of any raid controllers that won't continue operating with a disk that has bad sectors in order for you to continue revovering what you can (virtually the whole array). In the case of total failure, most cards have some form of maintenance or forced mode in which they will continue to operate in an unrecoverable state so that you can salvage what little is there.
Now, lets discuss raid 0 and recovering files. Let's assume a striped set of just two disks since you'd be insane to stripe more than that. Too many points of failure, data integrity and all those problems. I just checked. A completely blank MS Word file (I know, I'm a heretic for not using Open Office) and that comes to 23.5 and an even 24k on disk. Now, golly, wonder what the cluster size is on my disk. (Counts on fingers) I reckon it must be 8K. Now I have a realatively small 60gig disk. Lets say that 16k clusters are more on the average for most people now. That means that even for a totally blank Word document, you need two full clusters. Hopefully I don't need to explain that the way striping (raid 0) works is that every other cluster goes on the other disk. Just how many many modern document types do you think won't span the striped set. Certainly no office documents and there aren't many source files either.
Raid 5 makes things a little more complicated since the most basic raid 5 set would be similar in situation with 2 data disks and 1 parity disk. One thing to point out, as the number of disks in a raid 5 set go up, so does the chance to recover small files. However, so does the overall volatility of the set as there a
If the machine is a dedicated file server this is conceivable. But in that case, you're wasting your time worrying about performance on the array anyway since your network will be a super bottleneck anyway. In this situation, the only purpose of raid 5 is for data integrity.
The real place where the performance side of raid 5 matters is for things like application (especially in the web sense) and database servers. That is, where the data consumed by the server is far more significant than the data dispatched to the clients. In cases like this, CPU cycles are generally of high value and any ability to offload work to other devoted hardware is a necessity.
I got a newsflash for you. If you're using raid 5 and you drop more than one drive at the same time, you're done period. It doesn't matter if you're running hardware or software raid. It's the way raid 5 works. RAID 5 can recreate the info from any one drive. But if it loses two drives, it can't determine both variables. There is no convicing about it. If two drives die, so does your array. Do not pass go. Do not collet $200.
It scares me that they let people like you play with the sort of computing resources that have 50TB of disk space.
I don't think it's going to make much difference. It seems new and novel now, but within a couple years, all the schools will have competing programs. And it will be like trying to shop for a cell phone now. All the schools will have different plans that are nearly impossible to compare to one another to really be objective.
Besides, some schools have been making things like laptops part of the tuition for years. This really isn't that new when you look at it this way. It's just the next evolution in perks one gets from going to college. It used to be access to a great library. Then it was great discounts when you flashed your student ID. The last few years, it's been very high speed internet access. Now it's free music downloads.
Move along. Nothing to see here.
Ummmm....is this where we should discuss the piss poor dacs that virtually all computer sound cards whether on board or add-on use? There are exceptions, but those are few and far between and are certainly not the rule. Should we also discuss the problems with rf/emi inside a computer case and what that will do to analog signals?
Anyone that claims to be serious about pro audio production on their computer better be running digital outside their box into an external dac, especially if you think you really need mackies!
Thinking parent poster must be deaf to not hear any problems with that setup.
Find a responsible roommate ASAP. Locking the door when you leave won't do crap if your roommate doesn't do the same. This may not help for your first semester as most universities insist that you suffer with your randomly selected roommate for at least one semester, but should improve there after.
Also, find a roommate you can trust. I've heard plenty of horror stories about the roommate being the ultimate thief when everything was settled. Again, this may not help your first sememster/year, but make it a priority. At least then you only have a small peice of your college life to suffer with minimal effects in your room for fear of theivery.
You don't get out much do you?
http://developer.novell.com/ndk/mono.htm
The first thing to realize is that a web service no longer indicates internet anymore. It's a shift in paradigm. It's time for a lot of you guys to throw out your blue polyester shirts and suspenders. Web simply means it's delivered from a central location and typically works through a browser or derivative of a browser. This doesn't mean 2 billion users will be streaming office from microsoft.com. It doesn't mean that when MS's servers go down that the entire world will be without Office.
.NET reaches full maturity and is available for other platforms. Realistically, Linux might be running MS office in the near future.
This is simply the realization of the thin client paradigm. As corporate environments go, it's about time.
And before anyone panics about all those stand alone machines out there (like us developers are all so fond of), there are a bunch of appies out there that are essentially written this way already. VS.NET is web driven. That front end is all xml/html driven. We see it with the MS management console and MSC snappins for it too. This is the sort of thing we're looking at with the future of office.
The front end will be web based. The back end will likely have a few different options and standalone on the local machine I would wager will still be one of them. But at the same time, the back end could be centralized greatly simplifying mangement. I wouldn't be suprised if the next incarnation of Visual Studio can be set up to compile on a central server.
This should in theory simplify development of the Office software and reduce all versions of Office to a single codebase once
Glad to see you keep up and base your opinions on current versions of software. Clippy was killed in Office 2k3. Try living in the present. Or at least within the last year.
Ok, I'm not too much older than you if you're still in college, but I'm going to play old curmudgeon anyway.
you can put together a decent solution for computing now for around a grand. Kick in another $250 for needing a good workstation board to get the right slots and say $600 ($300x2) for the two cards and you're still just under $2000. THIS IS CHEAP. I'm sorry. I know how many lawns I had to mow as a youngin to buy my first pentium 60. That was $2k for JUST the computer and monitor. That included a baseline 1 meg video card, no cdrom and no sound. The cdrom and sound card cost me another $400 a couple months later.
So cry me a freaking river. Get a weekend job. Stop spending so much money on booz. If this is a priority for you, then you'll find the money. If it's not a priority, then quit your pissing and moaning.
Don't confuse the CPU a system uses for the entire performance value of the system. There are different bus and memory architectures that can do a lot to differentiate the performance of a "pricey" Sun with an AMD and the "value" machine you'd assemble from commodity parts
SGI did this with Pentiums (II's or III's if I remember correctly), though a lot depends on marketing which has not beeb SGI's strong point as of late so don't site SGI as an anecdote to predict Suns failure also.
Universities set these arbitrary requirements simply so there computer centers aren't supporting decade old hardware cause you know there would be some student show up next year with his family's old 486 running windows 3.0 and demanding they get him connected to the network in his dorm room.
I promise you, I can take the original parent post's notebook (the 500) and put it on Tech's dorm network. I can surf the web, I can take it to class and the library, write papers and code on it and Tech would never know the difference.
If anyone is silly enough to actually beleive they need a particular clock speed (especially in the multiple ghz range), hard drive capacity, etc to simply have their computer at college and presumably connected to the campus' network, then quite frankly they shouldn't be attending Virginia Tech(at least not for computer science or a related feild).
lol! OK, fair enough. At least I'm not claiming Unix from logging into my redhat box or something. Jeez your picky ;-)
Solaris is an implimentation of UNIX System Release V. Or are you gonna try and get picky of who owns the UNIX trademark?
It was a joke. The last time I used UNIX (Solaris to be specific) is when I logged on to my Sparc to check my mail this morning.
It's the same old debate. Unix, apple and BSD are all dead. Have been for years....right? Oh wait, nevermind.
Yeah, I thought the same thing. "they don't sound like Bose". Isn't that a good thing? going to listen to my Mirages now....
Well, perhaps you are the visionary here, but I see that convergence as very far off.If ever. We all know how slow standards move. HDTV anyone?
I look at my Sony DB930 Reciever (not super high end by anyone's definition, was $700 new) and it's pretty chock full at component width and some of the actuall highend models from B&K, Denon, Bob Carver, etc are even taller (at the same component width) and just as full or fuller. If I must have one component in my system that big, I'd rather waste a little space in the other components so they stack nice and neat.
But not all the other components are wasting space. My 400 disc CD carasel is pretty chock full (once it has CD's in it) as would be one of those spiffy DVD carasels. Yeah, my single disc tray loading DVD player could be thinner. I wish it was actually, but I don't want it any narrower.
These types of devices will never make it into my home entertainment system until they at east fit in my AV cabinet with my other components. 17 inches people. It's not that hard.
Check out Amazon for a little ebook called Buzz Your MP3 by a fellow named Brian Freeman It's a nice little reference for exactly this.