It will arrive in a lovely wooden crate and sometime after morning coffee he will unpack it, walk over the the Z Series, open the door, slide it into place, connect the cooling hoses and close the door. He will then walk to the maintenance terminal, type in the secret code, and your Z Series now has 64 more processors.
So far. What are you going to do when a motherboard or CPU failure requires a multi-hour outage to replace ?
HA cluster? In many ways the HA cluster can be more resilient than a single "mainframe". There are situations where hardware failures can take down all the partitions.
Or filesystem corruption requires a multi-hour restore ?
Mainframes etc, aren't invulnerable to filesystem corruption. I think that's an OS stability statement, and frankly I haven't seen a NTFS failure (excepting raid controller failures, etc) in a long time.
Or a botched patching session requires an hour to roll back ?
There are solutions to this too, but they require planning for the contingency.
Lots of these problems are also solved by running vmware on a SAN attached disk array and taking snapshots on a on a regular basis to a secondary disk array (this still isn't a replacement for proper backups).
As the parent said, there is a difference between a reference and a pointer, what you are describing is a pointer, and your syntax continues to work just fine in C++.
Doctors typically proscribe name brand prescriptions regardless of the existence of a generic. You simply tell your the pharmacy you want the generic. The doctor will specify if generics are not allowed on the prescription and they pharmacy will tell you if they can't do it, but 99 times out of a hundred, you'll get a generic.
The problem is that the doctors are also prescribing the latest and greatest drug, rather than the 20 year old one. The perception is the new drug must be better. This brings up a bunch of problems, the new drug costs a fortune, it doesn't have a generic replacement because its still patented, and it may not even be more effective than the old one (no one knows, hence this article). The pharmacy won't replace a drug with a similar generic one that does the same thing, you have to get doctor approval for that. One thing that desperately needs to be done, is comparative studies between drugs. It should be a requirement that when you release a new drug it comes with fairly easy to understand documentation which says where its better or worse than the existing drugs. Right now, that information is intentionally hidden because the drug companies want to be able to sell a lot of the new drug at a high price. If the drug is a dud that is hard to do even with a lot of advertising.
BTW: I agree about the US politics... Its more a game to keep people distracted while the problems continue, and the ball keeps rolling in the same direction. For example, where is the party which actually believe in personal freedom? There isn't one, the republicans have proven they are as nanny state as the democrats. The health care problems in this country will continue to get worse until someone shows up with a plan which cleans the slate and starts over. At this point trying to patch up the current system is just going to result in more inefficiently. Having seen how bad our system really is, my personal thoughts are we do need government run health care to provide a basic level of service. Beyond that there is still a place for insurance and private health care. I just don't want to have to fight with the insurance/doctors all the time over billing codes, standard procedures, just to get my teeth cleaned.
Frankly, Linux isn't growing because it has fundamental problems, even in its "core competencies".
Last year, I was volunteering for for the local congress critter. They needed a number of PC's with web browsers to access their canvasing tools. Where I work, we have a parts room full of 2-7 year old desktops, so I borrowed a few with the intention of installing Linux and Firefox and loaning them out. I downloaded the most recent version of ubuntu, and openSuse and installed a mix. I would say in large number of the cases the install was _NOT_ smooth requiring some text file editing or worse to get the machine to a usable state. That was ok, I did that before donating them. The problem is when I loaned them the desktops they were plugging their own monitors in (donated by someone else). In probably about 50% of the cases the resolution/refresh rates could not be set by the GUI utilities even though in 100% of the cases i saw data was available with the DDC display utility. One guy requested I reset the refresh rate because it was hurting his eyes. The look I got when I ended up hand editing the xconfig file summed up the why Linux isn't anywhere near ready for general consumption.
This is acceptable if you have an "expert" pre-install the machine, and setup every piece of hardware and software that will be used. At that point the linux machines need big stickers that say "no user serviceable parts inside." I end up hand patching and compiling stuff all the time, I frankly cannot figure out how a normal user is suppose to figure this stuff out.
I see a lot of good suggestions, so I will mention a couple I don't see anyone else mentioning.
Get a shell extension viewer (ShellExView) Ive found things hooking themselves into some shell behavior which makes the machine run poorly.
Pay attention to ram usage in the process explorer. Often a bad/compromised application will show up as using a lot of ram (which can cause excessive paging, which may look like a slower disk). Another sign is the IO operations/sec fields.
Get a decent network monitoring tool and look for failed connections or share logins. Sometimes a document in the recent used list will be on a network share that isn't available. This will also help locate other strange behavior. Windows likes its network connections, try unplugging your network connection and see if it gets faster/slower. Sometimes things will immediately fail (causing everything to accelerate) or the timeouts have to actually expire causing everything to run slower.
Look at what applications have been recently installed. I had a problem with S3 sleep mode on my desktop recently that was caused by an evil application's licensing service. The problem with services is that they run under the svchost process and can be hard to track down, so as others have mentioned, stop all unnecessary services.
Maybe, you should pick a version and try sharing documents with a few different office users to see how well it works for you over 6 months or so. I predict when your done you might have your answer.
Our sysadmin has really been trying to use it on her own machine on and off for a few years. She has to exchange docs with lots of different people running lots of different versions of office (mostly vendors). Without fail, there is always some bug that forces her to use ms office. Recently she was trying to open a 20 page word document in OO, and it took something like 10 minutes to open. Turned out to be a known bug in OO. My boss is also trying to use it, but he leaves it open for days on end, and it misbehaves by locking up his X session to the point where the machine has to have X killed from a remote machine. Then there are the dozens and dozens of compatibility problems which aren't deal breakers but do cause problems. In the end it keeps turning out to be cheaper/easier just to pay the devil and use Office instead of dealing with all the OO problems.
Your experiences may be different, but I suspect that something like http://www.corel.com/servlet/Satellite/us/en/Product/1151523326841#tabview=tab4 may still be a better bet than OO if you really must use something other than office. It has compatibility issues too, but last time I tried it was a much better solution than OO.
Your making the assumption that it uses less energy than a conventional launch vehicle. It seems to me, that removing yourself from the universe (or whatever it takes) and then coming back might take more power than loading up a couple rockets with fuel and lighting a match.
Its better than vista when it came out. The problem is the vista "experience" has improved. There are better drivers, more optimization, most software packages have been updated to work with vista, etc. Frankly, I think win7 is mostly marketing fluff because its causing people to take a second look at what vista has become.
Ok, i'm running XP/2003 on my machines and I installed the win7 beta in a vmware session. It actually works pretty well compared to vista at about the same time development point. Thats to be expected though, win7 is pretty much just Vista SE from what I can tell. There aren't any big features over vista that scream gotta have, but as with every M$ release since 2000, there is at least one thing that annoys me. In this case its the removal of the classic mode. Call me an old Luddite, but frankly I'm pretty happy with the win95 start menu and theme with a title bar, menu bar and movable button bars that generally can be disabled to save screen space. Why remove a feature that a decent percentage of your users are using? They could add the little search box to the classic menu and it would be better than either one. Stupid, some junior programmer at M$ is probably driving the whole thing cause he can't figure out the code base.
The aesthetics still need work. They can't get a common icon theme thoughout the OS, for applications they own. For me it feels more cluttered and harder to use than XP. I don't feel the a UI expert was involved.
Yah, the "aesthetics" in W7 are a joke. I sent them "feedback" about it. I wasn't arguing for or against any particular look/feel. The problem is that there are about 5 entirely different ones. They range from windows 3.0 applications like notepad where all the options are in the menu bar, through some combination of menu/button bars to wordpad, paint and friends having a ribbon. Frankly, it screams disjoint little serfdoms each doing their own thing. There isn't any overarching control over the product saying what look/feel the product should have and enforcing that across the different development groups. Explorer has one look/feel, the applications another, and the system administration tools yet another. Plus, the removal of classic mode really makes it worse because in a strange sort of way, classic mode was more unified and consistent. Especially over different OS releases. The start menu/task bar keeps changing from windows release to release, but classic mode was mostly consistent. This was a huge advantage because rolling out the OS didn't require retraining a bunch of 45-60 year olds that can barely use windows to begin with. It took some of these people 5 years after win 95 before they were productive in that look and feel. M$ should have rolled some kind of customer feedback as part of windows update for XP/Vista to see what percentage of users were in classic mode before removing it. I suspect that its over double digit percentages because of the corporate desktop images which have it turned on by default.
Of course I could argue about the ribbon, but others have done that. What frustrates me the most is its inflexibility. The primary problem I have with the new UI's is how they waste screen real estate. I don't need huge icons everywhere when a 16x16 will get the point across. I have a massive amount of screen real estate, but wasting it on dead whitespace or buttons I never click still bothers me.
Thats about the same amount of time it takes my machine, but I can type my password in ~1 second and a second later be at the desktop. Ie generally will start in another second or two. This is because I have disabled most of the crap that puts itself into the system tray. Everything continues to work, except when I want to change my desktop resolution I actually have to right click on the desktop instead of the system tray... anyway most of that crap that rattles your disk after login is probably useless and you won't notice if you disable it. Bottom line, my XP machine goes from power on to useful work in ~15 seconds, but it generally is at S3. From S3 to useful work is really more like ~2 seconds while the monitors come back on and the HD spins back up. The machine consumes ~6W in S3, but all the USB, networking, monitors etc running in standby bump it to ~13W. With everything running but idle its ~210W (monitors included).
Well, the canned response part is pretty good now. As the systems add more canned responses they will become more seamless. I can imagine at some point, that to someone unfamiliar with a particular system the voice commands if sufficiently intuitive sounding might seem like natural speech interaction when they are listening to a trained operator.
My wife's car was in the shop recently, and they gave loaned us one of the new acura TL's with voice control. It works pretty well, you can control the ac units, the navigation system, radio, etc. It has a hundred or so phrases it understands, so its more like interacting with a phone system than a real human in that regard. The phrases can be things like "how long to the destination" or "Find nearest xxxx" where xxx can be "gas station", "ATM", "hospital" etc. There is a list at http://www.acuraworld.com/forums/f72/repost-voice-commands-new-owners-56829/
The moral being that some of the commands are intuitive, I figured a bunch of them out without looking at the manual and once I had memorized a number of them my wife was pretty impressed because it seemed like I was just talking to the car.
Anyway, Telling the computer to do something it hasn't been previously programmed to do will require some form of generic AI. Whether the interaction is voice, keyboard or something else is completely unrelated, that part is still a long way off.
As a side note, I got a chuckle because I tried the "fan speed" phrase, as that link hints, the voice recognition pretty much stops working as the fan speed on the ac unit increases. If you say "fan speed 7" you then have to reach over and turn the AC fan down in order to issue another command.
I work for a Disk->Disk->Tape vendor. There is definitely space for tape, for the following reasons.
Compression: Tape drives do in-band compression. This means that the LTO4 cart at 2:1 data can read/write data at over 300MB/sec. It also means that your getting 1.6T on that $50 cart. The lowest average compression ratio I've seen at a customers site is 1.5:1.
Encryption: The most recent versions of LTO/T10k/359x all support in-band encryption at full data rates. This means that when you backup 5T a night and send it offsite and it falls off the truck you don't have to worry about your data.
WORM: You can buy WORM media, which allows you to conform to various government regulations for track ability and accountability if you happen to be a health institution, financial institution etc.
Robust long term storage: Tape is rated and guaranteed for significantly longer than any disk subsystem on the market
Offsite bandwidth: Tape is suited to shipping, and its far easier/cheaper to send 10T of data on a half dozen tapes via fedex every night than trickle it over a high bandwidth high cost interdata center connection.
Power costs: With tape you can have exabytes of storage (near)online, in a library for just pennies a TB a year in power/cooling costs. MAID type disks help to fix this for disk, but the power costs are still significantly higher due to periodic spin-ups, keeping the electronics online, etc.
Raw bandwidth scalability/backup window: When configured properly, i've seen tape backup systems getting 30GBytes/sec+ the bottleneck is almost always how fast the source database disk, or whatever can feed the data at. The flip side is the price of the disk necessary to feed at these data rates. At those data rates the disks cost 1000x as much as the tape subsystem.
My internet connection died exactly one minute before new years and stayed that way for 8 hours.....
Later, while diagnosing the situation, I discovered it was probably my big feet kicking the switch in the closet with the party favors when I went to get the toy horns, the uplink port to the firewall was partially pulled out.
They would have to NAT it, because just changing the route isn't going to affect the source IP blocks. Of course then it would be pretty obvious if a huge chunk of traffic starts coming in from a small IP block, in that case it could get blocked too. Frankly, there probably isn't much TW can do other than upset their peering agreements if someone decides to block all their customers access.
Its about time, they are just getting some of their own medicine. The crap they have been pulling is frankly sickening. I canceled my cable a few years ago because they were charging to much for commercial laden crap. I could get by on PBS and HBO but that isn't a valid choice, so instead I am just a RR customer. I'm watching the AT&T u-verse creep toward my neighborhood. I hate AT&T as much as TW, but I will probably switch on principle. TW milks the crap out of the RR customers, how many other computer technologies are as stagnant as RR? When I signed up 10 years ago I got 1M up 3M down, now I have 512k up ~7M down, and I only have that because grande offers it. As soon as u-verse started showing up here TW magically starts offering turbo which they claim is 20/2, but you have to pay extra. I would have paid extra for that 10 years ago, but now there is enough bad blood they can fsck off. I'm just waiting to switch. Frankly after looking at the DOCSIS specs http://en.wikipedia.org/wiki/DOCSIS , there isn't any reason for the UP/DOWN disparity. Its just TW milking the market for everything they can get. I don't even consider them a full service ISP anymore, since they block ports that they originally didn't, canceled usenet service, etc..
Maybe more people would understand it, if it were available. I know I didn't understand the virge manual when I first received it, but after extensive study it made more sense. Now it all seems pretty obvious when I pick it up. Frankly, I find the hardware register documentation to often be the most concise method of understanding a piece of hardware....
I think you really need to reconsider your medical insurance. True, you don't currently need it. That is exactly why you should consider "insurance" rather than a "health program" mislabeled as insurance. Real health insurance covers unusual accidents, like getting kicked by the goat and ending up in the emergency room. Usually, this kind of health insurance is called major medical and has outrageous deductibles, but is quite cheap. Granted in most states you can just show up at the emergency room and get treated, but as soon as they find out you don't have insurance your going to face some ugly realities. Farm accidents are particularly nasty and without insurance the hospital is as likely to cut your legs off, rather than spend 300k attempting to rebuild them and send you through therapy so you can walk.
Depends, any software which requires a lot of clicking is probably still better with the mouse. I find it really hard to do really precise work with my pen. Anytime i go to click the button I end up inadvertently moving the cursor. The pen is great for sketching, shading etc, where the movements don't have to be hyper precise and the pressure sensitivity of the pen really adds to the function. In truth I end up using all three if I'm doing any drawing. One hand on the keyboard for some functions and the other hand alternates between the mouse and the pen, generally with the pen still in my fingers when I'm using the mouse.
It was "virtual memory" because the working set size of all the "processes" could exceed the physical memory size. Basically, memory was allocated and you were given a "handle" to it. When you wanted to use the memory, you made a call to start referencing the memory. This allowed the system to perform heap compaction, and swap memory regions that weren't currently active. So, these regions could be "virtual" because they didn't really exist in the address space until they were used. The point is, "virtual memory" didn't require virtual address spaces, unless you consider the implicit idea that the region handles were some kind of addressing range. This is similar to the way many other machines worked including the B5000 which had hardware assistance for this kind of model, and managed it in a more elegant way (because you didn't have to worry about swapping or moving active regions, the hardware would fault and allow fix-ups if a region that wasn't active got referenced).
That said, there are a lot of really smart people, who understand this model, and bemoan the current C/unix/windows model. With proper hardware assistance, there isn't any need for complex heap management and security is much easier to manage when the OS has fine grained control over individual data structures. I would also agree its a better model simply on the fact that heap fragmentation doesn't exist.
Well, then I guess the Mac OS people had it wrong along with a number of other systems. The mac had "virtual memory" in a single address space shared between processes equal to the size of ram (early on, without an MMU). Quite a number of other systems whose manufactures had virtual memory don't conform to the modern definition either.
I note a lot of people are insisting that "virtual memory" refers to the virtual address space given to a execution context, and what the author really means is "paging".
The funny thing is that these are traditionally poorly defined/understood terms which are gaining a hard consensus for the meanings due to some recent OS books, and poor comp-sci education which insists on a particular definition. Everyone is faulting M$ for using the term incorrectly, even though the original mac OS and other OS's used the term in the same way. Wikipedia defines it one way and then goes on to give historical systems which don't really adhere to the definition. For example the B5000 (considered the first commercial machine with virtual memory) didn't even have "contiguous" working memory as required by the wikipedia definition. It had what would be more specifically called multiple variable sized segments which could be individually swapped. Again, the mac OS evolved from a single process model to muliprocess, in the same address space (look up mac switcher) and implemented "virtual memory" using a system without a MMU by swaping the allocated pieces of memory to disk if they weren't currently locked (in use) and reallocating the memory. Aka they had "virtual memory" in single fragmented address space.
The other example is people use "paging" to describe the act of swaping portions of the memory to disk, misunderstanding that paging is more about splitting an address space or segment up into fixed pieces for address translation to physical, and that disk swapping of pages isn't required for paging. Aka, your system is still "paging" if you disable swapping.
Even the term swapping is unclear because the need to differentiate between swaping pages, and swapping whole processes (or even segments) resulted in people avoided the term swapping to describe systems which were swapping pages instead of segments/regions/processes. These systems were generally called "demand paged" or something similar to indicate that they didn't need to swap a complete process or dataset (see DOSSHELL).
So, give the guy a break, in may ways he is just as correct, if not more so.
Work for a while with C++ and its standard library's iterator concept, and things like for(p=begin; p!=end; ++p) feel like the natural way to do such things. It's easy to reason about, it's not more prone to off-by-one or fencepost errors, and I fail to see how it's harder to debug a core dump of code done this way.
I do work in C++ almost exclusively (I'm of the school, that C++ is a "better" C). I do use the STL and iterators somewhat frequently, their shortcomings are many, and for a different discussion.
When debugging something I find it much easier to see 100 in a register, than 0x1FF3E64010. The 100 tells me how far from the base I am, which in my code usually has another number floating around close by telling me how long the structure is, and what the structure base should be. I've gotten burnt far to many times because a pointer got set to the wrong location initially, and proceeded to wander off to nowhere and crash. When that happens, I don't know if the initial condition was wrong, some arithmetic error on the pointer was wrong, someone else whacked the data structure, etc. Basically all you know is that something bad happened. In my code I can usually pin it down pretty quickly (which is the same for everyone), but in other peoples code I have to completely read and understand the total picture before I can start to narrow down the causes. I can't tell you how many times some genius wrote some clever piece of pointer code that gets past code reviews and a few months of testing to fall on my lap for a few days of head banging that would have been instantly obvious or flat out impossible (due to reusing pointers) with array syntax and more rigorous style.
There are literally probably two dozen cases I could give against K&R style pointers, and not just on maintainability or readability, but on things like compiler optimization where pointer aliasing is a _MAJOR_ problem for optimizing and parallelizing compilers (there are probably hundreds of code optimization tricks published in papers over the last decade or so which just don't work in C due to the loose use of pointers, even with noalias switches), to code analysis tools that simply cannot find problems with pointer based code because they cannot figure out intent. These same tools have a 100% accuracy when you use an index out of bounds on a data structure.
I swore off using pointers a few years ago except when necessary (and whole other parts of the language) after seeing the mess that can be made at job after job. I f*ck up, so I need solutions to avoid it. It seems the majority of programmers think they are better than all the other programmers and never make mistakes so they don't try to create habits which force the mistakes out in the open. Its also why I started matching braces and using 10-25 character function/variable names. Making my variable names long forces me to pause and think about a good name. I've got a bunch of "habbits" which cause my code to fail what many people consider good taste, as a result I seem to actually be able to avoid some common problems. I write a lot of heavily multithreaded, real time, high performace code (my current project can litteral max out memory bandwith and process Gbytes of data a second) with what i consider is a pretty low bug count. Once in a while I even run into some poor basdard who ends up maintaing my code and they say that initally it drove them nuts, but now that they have been maintaing it for a while they like it.
Slashdot Mirror
It will arrive in a lovely wooden crate and sometime after morning coffee he will unpack it, walk over the the Z Series, open the door, slide it into place, connect the cooling hoses and close the door. He will then walk to the maintenance terminal, type in the secret code, and your Z Series now has 64 more processors.
More like the guy acts like he is messing with the hardware and when you turn around he types the secret key into the maintenance terminal. http://publib.boulder.ibm.com/infocenter/eserver/v1r2/index.jsp?topic=/eicaz/eicazzcod.htm
So far. What are you going to do when a motherboard or CPU failure requires a multi-hour outage to replace ?
HA cluster? In many ways the HA cluster can be more resilient than a single "mainframe". There are situations where hardware failures can take down all the partitions.
Or filesystem corruption requires a multi-hour restore ?
Mainframes etc, aren't invulnerable to filesystem corruption. I think that's an OS stability statement, and frankly I haven't seen a NTFS failure (excepting raid controller failures, etc) in a long time.
Or a botched patching session requires an hour to roll back ?
There are solutions to this too, but they require planning for the contingency.
Lots of these problems are also solved by running vmware on a SAN attached disk array and taking snapshots on a on a regular basis to a secondary disk array (this still isn't a replacement for proper backups).
As the parent said, there is a difference between a reference and a pointer, what you are describing is a pointer, and your syntax continues to work just fine in C++.
Doctors typically proscribe name brand prescriptions regardless of the existence of a generic. You simply tell your the pharmacy you want the generic. The doctor will specify if generics are not allowed on the prescription and they pharmacy will tell you if they can't do it, but 99 times out of a hundred, you'll get a generic.
The problem is that the doctors are also prescribing the latest and greatest drug, rather than the 20 year old one. The perception is the new drug must be better. This brings up a bunch of problems, the new drug costs a fortune, it doesn't have a generic replacement because its still patented, and it may not even be more effective than the old one (no one knows, hence this article). The pharmacy won't replace a drug with a similar generic one that does the same thing, you have to get doctor approval for that. One thing that desperately needs to be done, is comparative studies between drugs. It should be a requirement that when you release a new drug it comes with fairly easy to understand documentation which says where its better or worse than the existing drugs. Right now, that information is intentionally hidden because the drug companies want to be able to sell a lot of the new drug at a high price. If the drug is a dud that is hard to do even with a lot of advertising.
BTW: I agree about the US politics... Its more a game to keep people distracted while the problems continue, and the ball keeps rolling in the same direction. For example, where is the party which actually believe in personal freedom? There isn't one, the republicans have proven they are as nanny state as the democrats. The health care problems in this country will continue to get worse until someone shows up with a plan which cleans the slate and starts over. At this point trying to patch up the current system is just going to result in more inefficiently. Having seen how bad our system really is, my personal thoughts are we do need government run health care to provide a basic level of service. Beyond that there is still a place for insurance and private health care. I just don't want to have to fight with the insurance/doctors all the time over billing codes, standard procedures, just to get my teeth cleaned.
Here goes some karma...
Frankly, Linux isn't growing because it has fundamental problems, even in its "core competencies".
Last year, I was volunteering for for the local congress critter. They needed a number of PC's with web browsers to access their canvasing tools. Where I work, we have a parts room full of 2-7 year old desktops, so I borrowed a few with the intention of installing Linux and Firefox and loaning them out. I downloaded the most recent version of ubuntu, and openSuse and installed a mix. I would say in large number of the cases the install was _NOT_ smooth requiring some text file editing or worse to get the machine to a usable state. That was ok, I did that before donating them. The problem is when I loaned them the desktops they were plugging their own monitors in (donated by someone else). In probably about 50% of the cases the resolution/refresh rates could not be set by the GUI utilities even though in 100% of the cases i saw data was available with the DDC display utility. One guy requested I reset the refresh rate because it was hurting his eyes. The look I got when I ended up hand editing the xconfig file summed up the why Linux isn't anywhere near ready for general consumption.
This is acceptable if you have an "expert" pre-install the machine, and setup every piece of hardware and software that will be used. At that point the linux machines need big stickers that say "no user serviceable parts inside." I end up hand patching and compiling stuff all the time, I frankly cannot figure out how a normal user is suppose to figure this stuff out.
I see a lot of good suggestions, so I will mention a couple I don't see anyone else mentioning.
Get a shell extension viewer (ShellExView) Ive found things hooking themselves into some shell behavior which makes the machine run poorly.
Pay attention to ram usage in the process explorer. Often a bad/compromised application will show up as using a lot of ram (which can cause excessive paging, which may look like a slower disk). Another sign is the IO operations/sec fields.
Get a decent network monitoring tool and look for failed connections or share logins. Sometimes a document in the recent used list will be on a network share that isn't available. This will also help locate other strange behavior. Windows likes its network connections, try unplugging your network connection and see if it gets faster/slower. Sometimes things will immediately fail (causing everything to accelerate) or the timeouts have to actually expire causing everything to run slower.
Look at what applications have been recently installed. I had a problem with S3 sleep mode on my desktop recently that was caused by an evil application's licensing service. The problem with services is that they run under the svchost process and can be hard to track down, so as others have mentioned, stop all unnecessary services.
Maybe, you should pick a version and try sharing documents with a few different office users to see how well it works for you over 6 months or so. I predict when your done you might have your answer.
Our sysadmin has really been trying to use it on her own machine on and off for a few years. She has to exchange docs with lots of different people running lots of different versions of office (mostly vendors). Without fail, there is always some bug that forces her to use ms office. Recently she was trying to open a 20 page word document in OO, and it took something like 10 minutes to open. Turned out to be a known bug in OO. My boss is also trying to use it, but he leaves it open for days on end, and it misbehaves by locking up his X session to the point where the machine has to have X killed from a remote machine. Then there are the dozens and dozens of compatibility problems which aren't deal breakers but do cause problems. In the end it keeps turning out to be cheaper/easier just to pay the devil and use Office instead of dealing with all the OO problems.
Your experiences may be different, but I suspect that something like http://www.corel.com/servlet/Satellite/us/en/Product/1151523326841#tabview=tab4 may still be a better bet than OO if you really must use something other than office. It has compatibility issues too, but last time I tried it was a much better solution than OO.
I have minimized the ribbon. Its better that way, but I wasn't just talking about the ribbon. Besides minimizing it only partially helps.
Your making the assumption that it uses less energy than a conventional launch vehicle. It seems to me, that removing yourself from the universe (or whatever it takes) and then coming back might take more power than loading up a couple rockets with fuel and lighting a match.
Its better than vista when it came out. The problem is the vista "experience" has improved. There are better drivers, more optimization, most software packages have been updated to work with vista, etc. Frankly, I think win7 is mostly marketing fluff because its causing people to take a second look at what vista has become.
Ok, i'm running XP/2003 on my machines and I installed the win7 beta in a vmware session. It actually works pretty well compared to vista at about the same time development point. Thats to be expected though, win7 is pretty much just Vista SE from what I can tell. There aren't any big features over vista that scream gotta have, but as with every M$ release since 2000, there is at least one thing that annoys me. In this case its the removal of the classic mode. Call me an old Luddite, but frankly I'm pretty happy with the win95 start menu and theme with a title bar, menu bar and movable button bars that generally can be disabled to save screen space. Why remove a feature that a decent percentage of your users are using? They could add the little search box to the classic menu and it would be better than either one. Stupid, some junior programmer at M$ is probably driving the whole thing cause he can't figure out the code base.
The aesthetics still need work. They can't get a common icon theme thoughout the OS, for applications they own. For me it feels more cluttered and harder to use than XP. I don't feel the a UI expert was involved.
Yah, the "aesthetics" in W7 are a joke. I sent them "feedback" about it. I wasn't arguing for or against any particular look/feel. The problem is that there are about 5 entirely different ones. They range from windows 3.0 applications like notepad where all the options are in the menu bar, through some combination of menu/button bars to wordpad, paint and friends having a ribbon. Frankly, it screams disjoint little serfdoms each doing their own thing. There isn't any overarching control over the product saying what look/feel the product should have and enforcing that across the different development groups. Explorer has one look/feel, the applications another, and the system administration tools yet another. Plus, the removal of classic mode really makes it worse because in a strange sort of way, classic mode was more unified and consistent. Especially over different OS releases. The start menu/task bar keeps changing from windows release to release, but classic mode was mostly consistent. This was a huge advantage because rolling out the OS didn't require retraining a bunch of 45-60 year olds that can barely use windows to begin with. It took some of these people 5 years after win 95 before they were productive in that look and feel. M$ should have rolled some kind of customer feedback as part of windows update for XP/Vista to see what percentage of users were in classic mode before removing it. I suspect that its over double digit percentages because of the corporate desktop images which have it turned on by default.
Of course I could argue about the ribbon, but others have done that. What frustrates me the most is its inflexibility. The primary problem I have with the new UI's is how they waste screen real estate. I don't need huge icons everywhere when a 16x16 will get the point across. I have a massive amount of screen real estate, but wasting it on dead whitespace or buttons I never click still bothers me.
Thats about the same amount of time it takes my machine, but I can type my password in ~1 second and a second later be at the desktop. Ie generally will start in another second or two. This is because I have disabled most of the crap that puts itself into the system tray. Everything continues to work, except when I want to change my desktop resolution I actually have to right click on the desktop instead of the system tray... anyway most of that crap that rattles your disk after login is probably useless and you won't notice if you disable it. Bottom line, my XP machine goes from power on to useful work in ~15 seconds, but it generally is at S3. From S3 to useful work is really more like ~2 seconds while the monitors come back on and the HD spins back up. The machine consumes ~6W in S3, but all the USB, networking, monitors etc running in standby bump it to ~13W. With everything running but idle its ~210W (monitors included).
Well, the canned response part is pretty good now. As the systems add more canned responses they will become more seamless. I can imagine at some point, that to someone unfamiliar with a particular system the voice commands if sufficiently intuitive sounding might seem like natural speech interaction when they are listening to a trained operator.
My wife's car was in the shop recently, and they gave loaned us one of the new acura TL's with voice control. It works pretty well, you can control the ac units, the navigation system, radio, etc. It has a hundred or so phrases it understands, so its more like interacting with a phone system than a real human in that regard. The phrases can be things like "how long to the destination" or "Find nearest xxxx" where xxx can be "gas station", "ATM", "hospital" etc. There is a list at http://www.acuraworld.com/forums/f72/repost-voice-commands-new-owners-56829/
The moral being that some of the commands are intuitive, I figured a bunch of them out without looking at the manual and once I had memorized a number of them my wife was pretty impressed because it seemed like I was just talking to the car.
Anyway, Telling the computer to do something it hasn't been previously programmed to do will require some form of generic AI. Whether the interaction is voice, keyboard or something else is completely unrelated, that part is still a long way off.
As a side note, I got a chuckle because I tried the "fan speed" phrase, as that link hints, the voice recognition pretty much stops working as the fan speed on the ac unit increases. If you say "fan speed 7" you then have to reach over and turn the AC fan down in order to issue another command.
I work for a Disk->Disk->Tape vendor. There is definitely space for tape, for the following reasons.
My internet connection died exactly one minute before new years and stayed that way for 8 hours.....
Later, while diagnosing the situation, I discovered it was probably my big feet kicking the switch in the closet with the party favors when I went to get the toy horns, the uplink port to the firewall was partially pulled out.
They would have to NAT it, because just changing the route isn't going to affect the source IP blocks. Of course then it would be pretty obvious if a huge chunk of traffic starts coming in from a small IP block, in that case it could get blocked too. Frankly, there probably isn't much TW can do other than upset their peering agreements if someone decides to block all their customers access.
Its about time, they are just getting some of their own medicine. The crap they have been pulling is frankly sickening. I canceled my cable a few years ago because they were charging to much for commercial laden crap. I could get by on PBS and HBO but that isn't a valid choice, so instead I am just a RR customer. I'm watching the AT&T u-verse creep toward my neighborhood. I hate AT&T as much as TW, but I will probably switch on principle. TW milks the crap out of the RR customers, how many other computer technologies are as stagnant as RR? When I signed up 10 years ago I got 1M up 3M down, now I have 512k up ~7M down, and I only have that because grande offers it. As soon as u-verse started showing up here TW magically starts offering turbo which they claim is 20/2, but you have to pay extra. I would have paid extra for that 10 years ago, but now there is enough bad blood they can fsck off. I'm just waiting to switch. Frankly after looking at the DOCSIS specs http://en.wikipedia.org/wiki/DOCSIS , there isn't any reason for the UP/DOWN disparity. Its just TW milking the market for everything they can get. I don't even consider them a full service ISP anymore, since they block ports that they originally didn't, canceled usenet service, etc..
Maybe more people would understand it, if it were available. I know I didn't understand the virge manual when I first received it, but after extensive study it made more sense. Now it all seems pretty obvious when I pick it up. Frankly, I find the hardware register documentation to often be the most concise method of understanding a piece of hardware....
I think you really need to reconsider your medical insurance. True, you don't currently need it. That is exactly why you should consider "insurance" rather than a "health program" mislabeled as insurance. Real health insurance covers unusual accidents, like getting kicked by the goat and ending up in the emergency room. Usually, this kind of health insurance is called major medical and has outrageous deductibles, but is quite cheap. Granted in most states you can just show up at the emergency room and get treated, but as soon as they find out you don't have insurance your going to face some ugly realities. Farm accidents are particularly nasty and without insurance the hospital is as likely to cut your legs off, rather than spend 300k attempting to rebuild them and send you through therapy so you can walk.
Depends, any software which requires a lot of clicking is probably still better with the mouse. I find it really hard to do really precise work with my pen. Anytime i go to click the button I end up inadvertently moving the cursor. The pen is great for sketching, shading etc, where the movements don't have to be hyper precise and the pressure sensitivity of the pen really adds to the function. In truth I end up using all three if I'm doing any drawing. One hand on the keyboard for some functions and the other hand alternates between the mouse and the pen, generally with the pen still in my fingers when I'm using the mouse.
It was "virtual memory" because the working set size of all the "processes" could exceed the physical memory size. Basically, memory was allocated and you were given a "handle" to it. When you wanted to use the memory, you made a call to start referencing the memory. This allowed the system to perform heap compaction, and swap memory regions that weren't currently active. So, these regions could be "virtual" because they didn't really exist in the address space until they were used. The point is, "virtual memory" didn't require virtual address spaces, unless you consider the implicit idea that the region handles were some kind of addressing range. This is similar to the way many other machines worked including the B5000 which had hardware assistance for this kind of model, and managed it in a more elegant way (because you didn't have to worry about swapping or moving active regions, the hardware would fault and allow fix-ups if a region that wasn't active got referenced).
That said, there are a lot of really smart people, who understand this model, and bemoan the current C/unix/windows model. With proper hardware assistance, there isn't any need for complex heap management and security is much easier to manage when the OS has fine grained control over individual data structures. I would also agree its a better model simply on the fact that heap fragmentation doesn't exist.
Well, then I guess the Mac OS people had it wrong along with a number of other systems. The mac had "virtual memory" in a single address space shared between processes equal to the size of ram (early on, without an MMU). Quite a number of other systems whose manufactures had virtual memory don't conform to the modern definition either.
I note a lot of people are insisting that "virtual memory" refers to the virtual address space given to a execution context, and what the author really means is "paging".
The funny thing is that these are traditionally poorly defined/understood terms which are gaining a hard consensus for the meanings due to some recent OS books, and poor comp-sci education which insists on a particular definition. Everyone is faulting M$ for using the term incorrectly, even though the original mac OS and other OS's used the term in the same way. Wikipedia defines it one way and then goes on to give historical systems which don't really adhere to the definition. For example the B5000 (considered the first commercial machine with virtual memory) didn't even have "contiguous" working memory as required by the wikipedia definition. It had what would be more specifically called multiple variable sized segments which could be individually swapped. Again, the mac OS evolved from a single process model to muliprocess, in the same address space (look up mac switcher) and implemented "virtual memory" using a system without a MMU by swaping the allocated pieces of memory to disk if they weren't currently locked (in use) and reallocating the memory. Aka they had "virtual memory" in single fragmented address space.
The other example is people use "paging" to describe the act of swaping portions of the memory to disk, misunderstanding that paging is more about splitting an address space or segment up into fixed pieces for address translation to physical, and that disk swapping of pages isn't required for paging. Aka, your system is still "paging" if you disable swapping.
Even the term swapping is unclear because the need to differentiate between swaping pages, and swapping whole processes (or even segments) resulted in people avoided the term swapping to describe systems which were swapping pages instead of segments/regions/processes. These systems were generally called "demand paged" or something similar to indicate that they didn't need to swap a complete process or dataset (see DOSSHELL).
So, give the guy a break, in may ways he is just as correct, if not more so.
It's not the average where half are under and half over, the median is the point where half are over and half under.
Except that all the intelligence tests I'm aware of, follow a normal distribution, and therefor the mean and median are the same value.
Work for a while with C++ and its standard library's iterator concept, and things like for(p=begin; p!=end; ++p) feel like the natural way to do such things. It's easy to reason about, it's not more prone to off-by-one or fencepost errors, and I fail to see how it's harder to debug a core dump of code done this way.
I do work in C++ almost exclusively (I'm of the school, that C++ is a "better" C). I do use the STL and iterators somewhat frequently, their shortcomings are many, and for a different discussion.
When debugging something I find it much easier to see 100 in a register, than 0x1FF3E64010. The 100 tells me how far from the base I am, which in my code usually has another number floating around close by telling me how long the structure is, and what the structure base should be. I've gotten burnt far to many times because a pointer got set to the wrong location initially, and proceeded to wander off to nowhere and crash. When that happens, I don't know if the initial condition was wrong, some arithmetic error on the pointer was wrong, someone else whacked the data structure, etc. Basically all you know is that something bad happened. In my code I can usually pin it down pretty quickly (which is the same for everyone), but in other peoples code I have to completely read and understand the total picture before I can start to narrow down the causes. I can't tell you how many times some genius wrote some clever piece of pointer code that gets past code reviews and a few months of testing to fall on my lap for a few days of head banging that would have been instantly obvious or flat out impossible (due to reusing pointers) with array syntax and more rigorous style.
There are literally probably two dozen cases I could give against K&R style pointers, and not just on maintainability or readability, but on things like compiler optimization where pointer aliasing is a _MAJOR_ problem for optimizing and parallelizing compilers (there are probably hundreds of code optimization tricks published in papers over the last decade or so which just don't work in C due to the loose use of pointers, even with noalias switches), to code analysis tools that simply cannot find problems with pointer based code because they cannot figure out intent. These same tools have a 100% accuracy when you use an index out of bounds on a data structure.
I swore off using pointers a few years ago except when necessary (and whole other parts of the language) after seeing the mess that can be made at job after job. I f*ck up, so I need solutions to avoid it. It seems the majority of programmers think they are better than all the other programmers and never make mistakes so they don't try to create habits which force the mistakes out in the open. Its also why I started matching braces and using 10-25 character function/variable names. Making my variable names long forces me to pause and think about a good name. I've got a bunch of "habbits" which cause my code to fail what many people consider good taste, as a result I seem to actually be able to avoid some common problems. I write a lot of heavily multithreaded, real time, high performace code (my current project can litteral max out memory bandwith and process Gbytes of data a second) with what i consider is a pretty low bug count. Once in a while I even run into some poor basdard who ends up maintaing my code and they say that initally it drove them nuts, but now that they have been maintaing it for a while they like it.