I never wore a helmet on a bicycle growing up either; but my brother (4 years younger) was required to. Many states were changing that in the late 1980's and early 1990's, and its very typical that a child under the age of 13 is required to wear a helmet by law now.
That said, I would not make such an exception even with a bike lane. The exception should only be for professional cyclists. Non-professional cyclists should follow regular pedestrian laws - moving against traffic, etc.
Drivers who hit and kill cyclists never saw them in the first place, helmet or not. I had this debate with an anti-helmet buddy of mine who two weeks later was hit by a car and suffered a major concussion. The driver thought she hit a squirrel. He still has problems concentrating.
I don't agree with helmet laws forcing people to not be stupid, but I know as many bikers who've been hit as I do that haven't.
And let me guess, he blamed the driver?
Reality is most cycling accidents are at minimum the fault of both parties. Drivers for not paying quite enough attention, and cyclists for doing something they shouldn't have been (e.g. blowing through a stop sign, traffic light, etc.).
Reality is drivers are already required to look both ways - for both vehicular traffic and non-vehicular traffic - when backing out of a driveway or entering an intersection. So saying "drivers won't look" is a misnomer when it comes to cyclists moving against traffic. (And, btw, people on skateboards/rollerskates/rollerblades can go just as fast as cyclists, are required to go against traffic, and have a far far less capability to stop.)
No, the real problem is cyclists are small and drivers aren't given enough experience when learning to drive to identify small targets; They learn that pedestrian-sized obstacles are on pavements.
Cyclists should wear helmets because it can save their life if hit by a car, not to stop a bruise when they fall over at traffic lights because their fancy shoes didn't unclip.
No. The problem is that there are really two types of cyclists: (i) professional, and (ii) non-professional. Both types need to be treated differently.
Professional cyclists need to, and should, be able to move with traffic for purposes of training, safety, etc. They are also more likely to have all the proper equipment, training, and physique for doing so. In this respect they are not vehicles, but they are not really pedestrians either.
Non-professional cyclists - e.g. the average citizen, children, etc. - have no place moving with traffic. They are pedestrians in every sense. The only time they should be moving with traffic is when they are in sufficiently large groups - e.g. a tour group.
The problem is that the cycling organizations that support the Professional Cyclists try to push their views on all cyclists. They try to be pedestrians when they want to be, and vehicles when they want to be, and in both cases ignoring the laws. As a result, they really do put a lot more people in jeopardy as they are not capable of doing what the professional cyclists can.
Until we really recognize that there are two distinct groups of cyclists, and we differentiate between them the problem will persist.
Reality is that anything that is part of what leads up to the scientific finding out to be available when asked for. Sure, they'll only publish what they want to, but in investigating it, all the materials do need to be available; and when investigating related matters (e.g. judicial matters) everything must be on the table.
If scientists do work for the government, then in the US FOIA (Freedom Of Information Act) requests should be sufficient to compel the information so outside observers can investigate the work and findings for whatever reason they so choose. The scientists are, of course, free to challenge any results by said requesting party - e.g. challenge the credentials of the person(s) reviewing the work so as to show that they are or are not capable of understanding what they are looking at. (For instance, a chemist reviewing biology notes may have some insights but would not necessariy be fully qualified to comment on them in whole.)
If the scientists do work for private organizations without any money from government, then standard work practices ought to apply - if requested by subpeona they should comply. Again, the organization and its lawyers can challenge the credentials of the person(s) reviewing it.
The only thing to be gained by hiding the work is to hide the biases, intents, and motives behind it, and to hide any fraudulent results; and to prohibit others from making further findings based on the work (e.g. someone noticing a special attribute that was overlooked).
Yes, peer reviewing has its place; but so does the abilities of others to review scientific work. The other option, of course, is they can do the work, publish, but be prohibited from having it taught about anywhere - without full access its just as useless.
They're just arguing for a data-only network where VOIP provides all Voice related services, and SMS is done via a SIP/Jabber-like protocol...yep, that's where cellular is going - that's really what 4G is suppose to be (only the carriers in the US have relabled 3.5G to 4G so they could claim coverage for something that doesn't exist yet - it'll all about the marketing).
And, when they finally do that - when cellular is finally data-only+VOIP - then, and only then, will I actually pay for a data plan.
Go India and make it happen! Show us how cheap cellular really can be!
I most certainly agree that a dedicated server room and (separate if possible) dedicated telco room is a must. Why separate? So you don't have to let the telco technicians near your servers any more than you have to. Less hands in the pot, and less temptation for strangers.
Second, put in a series of switch rooms near each office area, and run fibre between them and the server room. As noted, allow for extra space so more fiber can be run later; then put CAT-6/7 patch panels in to the switch rooms, and run CAT-6/7 to each desk from there. Again, think about having extra space and set it up so you can easily replace the ethernet cable with fiber at some point in the future.
Third, get a good network-centric PBX, think Asterisk or something from Cisco (if you can afford it), and then put in VOIP phones at each desk instead of putting in separate RJ-11 cabling for analog phones.
Fourth, any where that needs a network connection on the floor by equipment should get its own fiber optic link, with a local switch, and short ethernet lines (CAT6/7). Why? You minimize any interference issues with the equipment on the floor. It might not be much, but it'll be worth it.
Also, if you are running between buildings or between sections of a building that are on (for whatever reason) different power grids, save yourself the time and put in fiber comms to isolate them. We have two buildings where I work now that are on separate power grids, and for a long time they were running ethernet directly between them, and kept burning out ports when a t-storm came through - a simple fiber optic isolator solved the problem.
Now, you might think doing the fiber thing - or at the very least planning for it - is over kill. However, think of it as an investment. If not for the companies use, then as a value-add should the company need to sell the building - the next tenant could be a Facebook/Amazon/Walmart requiring massive data-througput so don't cut yourself short by not making it easy to put in while you have the opportunity to do so.
Wow, you make conclusions based on having no specs and assuming the piloted planes are FA-18 or F-22 or something else semi-modern and the drones are biplanes. You got spec sheets on these drones that don't yet exist? Who's to say the drones won't be as fast as a normal jet, just more maneuverable (due to no pilot to keep alive while turning)? Drones would rule the sky so long as an idiot doesn't program them.
I don't have to assume that to go with my analysis above.
Consider that the drones need to be able to be controlled by human as AI technology is not good enough for C2 to be done on-board - it has to be remote controlled, and that in itself makes some limits - sure the drone may have a lot of functionality and do things in a very limited way, but when it comes to fighting a human will still have to be in control, which limits reaction time (as communications have a certain latency to them by nature), and also introduces a major issue - jamming of the comms.
Consider also that without a global satellite based communications network, the drones would be limited in range by how far the comms could work. Thus far, only the US, China, and perhaps Russia have such capabilities - something that is a far, far cry from any middle east country.
Remember too that a US drone was taken by (i) jamming its comms, and (ii) providing an alternate GPS signal to make it go where they wanted it to safely and allow further study. Drone technology just isn't good enough (even at the cutting edge) for anything else yet - and we're probably taking decades away for it to get there without a major AI break-through.
So yes, the drone could be just as fast or faster than a normal jet, but they'd still have some very big limitations that would neutralize that issue when it comes to combat, especially given the capabilities of an F16/18/22, as well as other aircraft available to the US and its allies - even to China, North Korea, and Russia.
What you're proposing wouldn't make much difference in our air superiority as:
- the jets would be able to return/reload/resume faster than more drones could be brought up
- ammunition could be used far outside the drones range by the jets to bring the drones down
- the jets tactical abilities are far superior to the drones, making close range combat useless
- the jets would be able to take down the drones before the drones would be able to detect the jets
- the jets would be able to buzz the drones, creating deadly turbulence for the drones in ways that drone operators would (and software) would not be able to easily account.
- the list goes on
Now if you put up 10,000 drones it would become difficult to take them all down before some reached their targets. But more likely than not, you would take one down and have a chain taken out with it - either by the fire that took it down, or by it running into (and cascading into) numerous drones, etc. And gain, 1 F-18/16/22 would be able to easily take out several hundred of those drones by itself beyond the range of the drone. (The heat signature would not even half to come into play.) And that's not even going up to the carpet bombing techniques that could be used as well.
How well Windows supported it was largely up to how well the video card drivers supported it. Some systems required rebooting in order to get it to recognize the additional monitor; others would work without a problem. It was typically consistent for any given driver, but very hit-and-miss between video cards/driver versions.
Before Windows 2000, yes. Since then, what you describe would be extremely unusual.
I had that problem on a number of WinXP systems, so no it's still very much the norm. Now may be the Video driver rewrite for Vista corrected some of that for Vista and Win7, and now Win8. Microsoft has been slowly building in proper multi-monitor support, but even with what they provide there still a lot of details left to the video card drivers.
When I disconnect the external monitor, it automagically moves APP windows and panes to laptop monitor, that is configured as secondary screen. When I replug the monitor, APP windows that previously where on it, return to it, as do the panes...
I'm running 12.04 Kubuntu and no, it doesn't work properly when I undock my laptop; and yes, I keep the external monitor (DVI) as the primary and "to the left of" the internal monitor (LVDS).
According to the OP, it has multi-monitor support, but things like actually remembering the configuration you apply is inexplicably beyond its capabilities.
It remembers the configuration until you change the configuration.
That is, with a desktop where you have several monitors connected all the time it won't be an issue. But with a laptop where you may have an external monitor part of the time then it is an issue whenever you switch between laptop only mode, and laptop plus external monitor mode. What's most annoying is when you have the external display as the primary; when you disconnect it the multi-monitor dialog prompting on to reconfigure shows up on the external monitor, not the only remaining monitor - so you're kind of screwed. Currently I make it a point to reconfigure to clone mode before undocking my laptop.
Windows XP did multi-monitors fairly well. Windows 7 handles it excellently. I have five monitors and when replacing one of the video cards, I changed which monitors were plugged into which card. As soon as Windows 7 booted, it automatically corrected for switching the cables around so that the monitors were all exactly as they were when I powered down the system in spite of every monitor being plugged into a different card and port.
Not to say Win7 isn't lacking some features, but nothing free or cheap software like Ultramon doesn't fix (IE: fine-tuning relative positions, multi-monitor wallpaper, taskbar across all monitors), but the essential parts of multi-monitors are handled very well.
How well Windows supported it was largely up to how well the video card drivers supported it. Some systems required rebooting in order to get it to recognize the additional monitor; others would work without a problem. It was typically consistent for any given driver, but very hit-and-miss between video cards/driver versions.
Because unofficial public betas are a great way to evaluate, test, and fix your product. See Siri and iOS maps.
That only works it works well enough that people are willing to put up with the remaining deficiencies. Works fine for search; but not so much for the general OS.
You can have perfectly written code in the program, but if your compiler has bugs (read: some part of its code that doesn't work right with the OS or processor instructions) you're screwed. NASA still uses 40 yr old equipment because they'd have to re-do/re-confirm all of their proofs if they upgrade to a new OS or hardware with a different instruction set.
GP is mostly referring to the custom assembly programming that is done, no compiler necessary - just a straight assembler that can be proven to generate the correct output for the given input. Any program in a higher level language than assembly cannot be guaranteed in such a manner.
However, NASA also uses a lot of Java and other stuff, so not everything is to that level
And what exactly is he doing over ethernet that needs that much speed? I'm only just now looking at upgrading our small business network to gigabit. A couple of years ago the cost of a 48 port gigabit switch was pretty high, but now it's very reasonable
Well, I got introduced to Linux via Slackware on several LS-120's; but never installed it as it was too daunting at the time; but I quickly purchased SuSe 6 (back in 1999) and later installed that. Not being quite the fan of SuSe (even then) I later tried RedHat and Mandrake. Dual booting never really worked too well for me, so I didn't truely commit to Linux until I got another computer to use for a server - initially a Sun IPC (circa 1991) that I got NetBSD on and switched to Red Hat (the only distro to still support it); and later got an Intel P90 that I loaded up with Slackware - only, I did Slackware for the initial install and then upgraded everything bit by bit using vanilla sources. My desktop eventually converted to Slackware as well. I had a laptop for a while running Debian. I've since moved to Gentoo as it manages the sources for me; and my work computers are all Kubuntu. I usually try to keep servers to Debian, and I've lately tried out Arch under a VM (I like rolling distros). My home server will likely move to Debian, and my personal laptops might switch to Kubuntu - mostly due to lack of time to keep running Gentoo (which I really do prefer).
Your account is just an MS schill account, judging by your comment history. Vista was a steaming pile of crap when it was released. After four years of work on it, MS scrapped anything innovative and rushed for a year just to get *something* out the door. SP1 fixed a lot of the issues but I still have problems with my one remaining Vista box that have nothing to do with driver support. It is just...buggy. Win7 was great, but it sounds like they are once again rushing a Win8 release to get something out the door. I predict a moderate disaster, followed in about 12-18 months by a really solid new release for the desktop.
Well, I don't use Windows any more myself. We got Vista Ultimate for my wife's computer and it's pretty nice as far as Windows goes (I much prefer KDE 4 myself).
Vista's big issues were: (i) UAC, (ii) they rewrote the Audio and Video drivers as user-space drivers with a minimal kernnel presence, and (iii) they made some major changes to Vista between the last RC and RTM releases.
The UAC was a big thing, but it was mainly pointing out what Microsoft had been telling developers for years - yes, they originally advocated using the Admin account; but they did change that stance long before Vista came out; but it wasn't until Vista came out that developers actually paid attention; Win7 was a lot better in that respect simply because many had already fixed the issues.
The driver rewrite was big, so drivers couldn't be quickly ported from XP to Vista. But that's not that big a deal in the end. Oddly it made Windows more like the Linux/Unix/Mac model with Xorg/X-Windows than anything.
The real issue was the changing of the driver interfaces between the last RC and RTM releases - this basically caught many off-guard so drivers just were not as available in RTM as they were with the last RC release. If they hadn't done that, then the rep Vista got would not have been so bad.
For many trees, a Git repository with full history, and a checked out working tree, can be smaller than the equivalent SVN checkout.
(SVN may have changed this recently, I've stopped using it) ; SVN keeps a second pristine working copy of the files in hidden folders for comparison purposes, Git keeps everything packed in compressed files.
SVN changed their working copy information in 1.7 - now you have a central.svn folder for the entire working copy instead of in all the sub-folders. It's a little more painful as before you could just copy a subfolder somewhere else to copy it; now you can't - you have to go through a checkout. It gains them some efficiencies though. I don't think they do any compression on the files though.
As to git, it's more efficient in that the whole repository is there with only a single version of each object. However, it doesn't track folders that don't yet have any contents (something I find really annoying).
Difference between git on a server and on the desktop is that on the desktop you use a normal git repository - which has all the changes and a working copy; while on the server you only publish the bare-bones git repository - which is only the changes.
True. However, for some things that is actually a detraction. For example, companies do a lot of paperwork, much of which needs to be controlled as to who has access to it. A centralized system such as Subversion or CVS is really good for those kinds of things, while a decentralized system like git is not - too easy to get those sensitive things into the wrong hands, and no ability to control permissions.
And yes, I do use SVN and git for various things. Git is great for developers as it lets you take the repo on the road very easily and still do all your tracking, while SVN your have to dial-back home to really do your tracking while on the road - which can be a pain at times.
SVN/TortoiseSVN will do a 3-way merge (for conflict resolution) on a MS word document, at least for the newer releases of MS Word (2010+). I am not sure if the document has to be a.docx or not though.
TortoiseSVN does it via a custom diff handler for the doc/docx types. And no, it doesn't have to be docx, and yes it will work with MS Word 2003 (for which the script was originally written).
The problem is that you can't do a diff between Word documents. Source control and diffing tools work great on source code because it's all just plain text. But for things like Word documents or Powerpoint presentations things get a little more complicated. Sure with a version control service you won't lose any previous versions, but you don't know what changed between versions either. To me this is the major thing missing for MS Office, and I can't believe they haven't done it yet. Sure they have track changes, but you have to remember to turn that on. There's no reason you should have to track changes. You should be able to take two versions of the document and MS Word should be able to tell you what has changed between the two versions. I think that if OpenOffice developed this feature it would be a killer feature that might get people to actually start using it. Because as you pointed out, when you have 4 different versions of the same document from 4 different people, it's nice to be able to figure out which changes those 4 people actually made without going through all 50 pages of the document.
TortoiseSVN has a script that uses Word's features to diff Word files in Subversion. I think it can done for Excel and other parts of office too. However, few know about those features, and fewer still use them.
Slashdot Mirror
I never wore a helmet on a bicycle growing up either; but my brother (4 years younger) was required to. Many states were changing that in the late 1980's and early 1990's, and its very typical that a child under the age of 13 is required to wear a helmet by law now.
That said, I would not make such an exception even with a bike lane. The exception should only be for professional cyclists. Non-professional cyclists should follow regular pedestrian laws - moving against traffic, etc.
Drivers who hit and kill cyclists never saw them in the first place, helmet or not. I had this debate with an anti-helmet buddy of mine who two weeks later was hit by a car and suffered a major concussion. The driver thought she hit a squirrel. He still has problems concentrating. I don't agree with helmet laws forcing people to not be stupid, but I know as many bikers who've been hit as I do that haven't.
And let me guess, he blamed the driver?
Reality is most cycling accidents are at minimum the fault of both parties. Drivers for not paying quite enough attention, and cyclists for doing something they shouldn't have been (e.g. blowing through a stop sign, traffic light, etc.).
Reality is drivers are already required to look both ways - for both vehicular traffic and non-vehicular traffic - when backing out of a driveway or entering an intersection. So saying "drivers won't look" is a misnomer when it comes to cyclists moving against traffic. (And, btw, people on skateboards/rollerskates/rollerblades can go just as fast as cyclists, are required to go against traffic, and have a far far less capability to stop.)
No, the real problem is cyclists are small and drivers aren't given enough experience when learning to drive to identify small targets; They learn that pedestrian-sized obstacles are on pavements. Cyclists should wear helmets because it can save their life if hit by a car, not to stop a bruise when they fall over at traffic lights because their fancy shoes didn't unclip.
No. The problem is that there are really two types of cyclists: (i) professional, and (ii) non-professional. Both types need to be treated differently.
Professional cyclists need to, and should, be able to move with traffic for purposes of training, safety, etc. They are also more likely to have all the proper equipment, training, and physique for doing so. In this respect they are not vehicles, but they are not really pedestrians either.
Non-professional cyclists - e.g. the average citizen, children, etc. - have no place moving with traffic. They are pedestrians in every sense. The only time they should be moving with traffic is when they are in sufficiently large groups - e.g. a tour group.
The problem is that the cycling organizations that support the Professional Cyclists try to push their views on all cyclists. They try to be pedestrians when they want to be, and vehicles when they want to be, and in both cases ignoring the laws. As a result, they really do put a lot more people in jeopardy as they are not capable of doing what the professional cyclists can.
Until we really recognize that there are two distinct groups of cyclists, and we differentiate between them the problem will persist.
Reality is that anything that is part of what leads up to the scientific finding out to be available when asked for. Sure, they'll only publish what they want to, but in investigating it, all the materials do need to be available; and when investigating related matters (e.g. judicial matters) everything must be on the table.
If scientists do work for the government, then in the US FOIA (Freedom Of Information Act) requests should be sufficient to compel the information so outside observers can investigate the work and findings for whatever reason they so choose. The scientists are, of course, free to challenge any results by said requesting party - e.g. challenge the credentials of the person(s) reviewing the work so as to show that they are or are not capable of understanding what they are looking at. (For instance, a chemist reviewing biology notes may have some insights but would not necessariy be fully qualified to comment on them in whole.)
If the scientists do work for private organizations without any money from government, then standard work practices ought to apply - if requested by subpeona they should comply. Again, the organization and its lawyers can challenge the credentials of the person(s) reviewing it.
The only thing to be gained by hiding the work is to hide the biases, intents, and motives behind it, and to hide any fraudulent results; and to prohibit others from making further findings based on the work (e.g. someone noticing a special attribute that was overlooked).
Yes, peer reviewing has its place; but so does the abilities of others to review scientific work. The other option, of course, is they can do the work, publish, but be prohibited from having it taught about anywhere - without full access its just as useless.
They're just arguing for a data-only network where VOIP provides all Voice related services, and SMS is done via a SIP/Jabber-like protocol...yep, that's where cellular is going - that's really what 4G is suppose to be (only the carriers in the US have relabled 3.5G to 4G so they could claim coverage for something that doesn't exist yet - it'll all about the marketing).
And, when they finally do that - when cellular is finally data-only+VOIP - then, and only then, will I actually pay for a data plan.
Go India and make it happen! Show us how cheap cellular really can be!
I most certainly agree that a dedicated server room and (separate if possible) dedicated telco room is a must. Why separate? So you don't have to let the telco technicians near your servers any more than you have to. Less hands in the pot, and less temptation for strangers.
Second, put in a series of switch rooms near each office area, and run fibre between them and the server room. As noted, allow for extra space so more fiber can be run later; then put CAT-6/7 patch panels in to the switch rooms, and run CAT-6/7 to each desk from there. Again, think about having extra space and set it up so you can easily replace the ethernet cable with fiber at some point in the future.
Third, get a good network-centric PBX, think Asterisk or something from Cisco (if you can afford it), and then put in VOIP phones at each desk instead of putting in separate RJ-11 cabling for analog phones.
Fourth, any where that needs a network connection on the floor by equipment should get its own fiber optic link, with a local switch, and short ethernet lines (CAT6/7). Why? You minimize any interference issues with the equipment on the floor. It might not be much, but it'll be worth it.
Also, if you are running between buildings or between sections of a building that are on (for whatever reason) different power grids, save yourself the time and put in fiber comms to isolate them. We have two buildings where I work now that are on separate power grids, and for a long time they were running ethernet directly between them, and kept burning out ports when a t-storm came through - a simple fiber optic isolator solved the problem.
Now, you might think doing the fiber thing - or at the very least planning for it - is over kill. However, think of it as an investment. If not for the companies use, then as a value-add should the company need to sell the building - the next tenant could be a Facebook/Amazon/Walmart requiring massive data-througput so don't cut yourself short by not making it easy to put in while you have the opportunity to do so.
Wow, you make conclusions based on having no specs and assuming the piloted planes are FA-18 or F-22 or something else semi-modern and the drones are biplanes. You got spec sheets on these drones that don't yet exist? Who's to say the drones won't be as fast as a normal jet, just more maneuverable (due to no pilot to keep alive while turning)? Drones would rule the sky so long as an idiot doesn't program them.
I don't have to assume that to go with my analysis above.
Consider that the drones need to be able to be controlled by human as AI technology is not good enough for C2 to be done on-board - it has to be remote controlled, and that in itself makes some limits - sure the drone may have a lot of functionality and do things in a very limited way, but when it comes to fighting a human will still have to be in control, which limits reaction time (as communications have a certain latency to them by nature), and also introduces a major issue - jamming of the comms.
Consider also that without a global satellite based communications network, the drones would be limited in range by how far the comms could work. Thus far, only the US, China, and perhaps Russia have such capabilities - something that is a far, far cry from any middle east country.
Remember too that a US drone was taken by (i) jamming its comms, and (ii) providing an alternate GPS signal to make it go where they wanted it to safely and allow further study. Drone technology just isn't good enough (even at the cutting edge) for anything else yet - and we're probably taking decades away for it to get there without a major AI break-through.
So yes, the drone could be just as fast or faster than a normal jet, but they'd still have some very big limitations that would neutralize that issue when it comes to combat, especially given the capabilities of an F16/18/22, as well as other aircraft available to the US and its allies - even to China, North Korea, and Russia.
What you're proposing wouldn't make much difference in our air superiority as:
- the jets would be able to return/reload/resume faster than more drones could be brought up
- ammunition could be used far outside the drones range by the jets to bring the drones down
- the jets tactical abilities are far superior to the drones, making close range combat useless
- the jets would be able to take down the drones before the drones would be able to detect the jets
- the jets would be able to buzz the drones, creating deadly turbulence for the drones in ways that drone operators would (and software) would not be able to easily account.
- the list goes on
Now if you put up 10,000 drones it would become difficult to take them all down before some reached their targets. But more likely than not, you would take one down and have a chain taken out with it - either by the fire that took it down, or by it running into (and cascading into) numerous drones, etc. And gain, 1 F-18/16/22 would be able to easily take out several hundred of those drones by itself beyond the range of the drone. (The heat signature would not even half to come into play.) And that's not even going up to the carpet bombing techniques that could be used as well.
Before Windows 2000, yes. Since then, what you describe would be extremely unusual.
I had that problem on a number of WinXP systems, so no it's still very much the norm. Now may be the Video driver rewrite for Vista corrected some of that for Vista and Win7, and now Win8. Microsoft has been slowly building in proper multi-monitor support, but even with what they provide there still a lot of details left to the video card drivers.
When I disconnect the external monitor, it automagically moves APP windows and panes to laptop monitor, that is configured as secondary screen. When I replug the monitor, APP windows that previously where on it, return to it, as do the panes...
I'm running 12.04 Kubuntu and no, it doesn't work properly when I undock my laptop; and yes, I keep the external monitor (DVI) as the primary and "to the left of" the internal monitor (LVDS).
According to the OP, it has multi-monitor support, but things like actually remembering the configuration you apply is inexplicably beyond its capabilities.
It remembers the configuration until you change the configuration. That is, with a desktop where you have several monitors connected all the time it won't be an issue. But with a laptop where you may have an external monitor part of the time then it is an issue whenever you switch between laptop only mode, and laptop plus external monitor mode. What's most annoying is when you have the external display as the primary; when you disconnect it the multi-monitor dialog prompting on to reconfigure shows up on the external monitor, not the only remaining monitor - so you're kind of screwed. Currently I make it a point to reconfigure to clone mode before undocking my laptop.
Windows XP did multi-monitors fairly well. Windows 7 handles it excellently. I have five monitors and when replacing one of the video cards, I changed which monitors were plugged into which card. As soon as Windows 7 booted, it automatically corrected for switching the cables around so that the monitors were all exactly as they were when I powered down the system in spite of every monitor being plugged into a different card and port.
Not to say Win7 isn't lacking some features, but nothing free or cheap software like Ultramon doesn't fix (IE: fine-tuning relative positions, multi-monitor wallpaper, taskbar across all monitors), but the essential parts of multi-monitors are handled very well.
How well Windows supported it was largely up to how well the video card drivers supported it. Some systems required rebooting in order to get it to recognize the additional monitor; others would work without a problem. It was typically consistent for any given driver, but very hit-and-miss between video cards/driver versions.
Because unofficial public betas are a great way to evaluate, test, and fix your product. See Siri and iOS maps.
That only works it works well enough that people are willing to put up with the remaining deficiencies. Works fine for search; but not so much for the general OS.
Vista was a disaster
Mainly because drivers weren't updated correctly to the new architecture.
And that was mainly due to a drastic change in the driver interfaces between the last RC and RTM.
You can have perfectly written code in the program, but if your compiler has bugs (read: some part of its code that doesn't work right with the OS or processor instructions) you're screwed. NASA still uses 40 yr old equipment because they'd have to re-do/re-confirm all of their proofs if they upgrade to a new OS or hardware with a different instruction set.
GP is mostly referring to the custom assembly programming that is done, no compiler necessary - just a straight assembler that can be proven to generate the correct output for the given input. Any program in a higher level language than assembly cannot be guaranteed in such a manner.
However, NASA also uses a lot of Java and other stuff, so not everything is to that level
And what exactly is he doing over ethernet that needs that much speed? I'm only just now looking at upgrading our small business network to gigabit. A couple of years ago the cost of a 48 port gigabit switch was pretty high, but now it's very reasonable
You did see this article, no?
Well, I got introduced to Linux via Slackware on several LS-120's; but never installed it as it was too daunting at the time; but I quickly purchased SuSe 6 (back in 1999) and later installed that. Not being quite the fan of SuSe (even then) I later tried RedHat and Mandrake. Dual booting never really worked too well for me, so I didn't truely commit to Linux until I got another computer to use for a server - initially a Sun IPC (circa 1991) that I got NetBSD on and switched to Red Hat (the only distro to still support it); and later got an Intel P90 that I loaded up with Slackware - only, I did Slackware for the initial install and then upgraded everything bit by bit using vanilla sources. My desktop eventually converted to Slackware as well. I had a laptop for a while running Debian. I've since moved to Gentoo as it manages the sources for me; and my work computers are all Kubuntu. I usually try to keep servers to Debian, and I've lately tried out Arch under a VM (I like rolling distros). My home server will likely move to Debian, and my personal laptops might switch to Kubuntu - mostly due to lack of time to keep running Gentoo (which I really do prefer).
Your account is just an MS schill account, judging by your comment history. Vista was a steaming pile of crap when it was released. After four years of work on it, MS scrapped anything innovative and rushed for a year just to get *something* out the door. SP1 fixed a lot of the issues but I still have problems with my one remaining Vista box that have nothing to do with driver support. It is just...buggy. Win7 was great, but it sounds like they are once again rushing a Win8 release to get something out the door. I predict a moderate disaster, followed in about 12-18 months by a really solid new release for the desktop.
Well, I don't use Windows any more myself. We got Vista Ultimate for my wife's computer and it's pretty nice as far as Windows goes (I much prefer KDE 4 myself). Vista's big issues were: (i) UAC, (ii) they rewrote the Audio and Video drivers as user-space drivers with a minimal kernnel presence, and (iii) they made some major changes to Vista between the last RC and RTM releases.
The UAC was a big thing, but it was mainly pointing out what Microsoft had been telling developers for years - yes, they originally advocated using the Admin account; but they did change that stance long before Vista came out; but it wasn't until Vista came out that developers actually paid attention; Win7 was a lot better in that respect simply because many had already fixed the issues.
The driver rewrite was big, so drivers couldn't be quickly ported from XP to Vista. But that's not that big a deal in the end. Oddly it made Windows more like the Linux/Unix/Mac model with Xorg/X-Windows than anything.
The real issue was the changing of the driver interfaces between the last RC and RTM releases - this basically caught many off-guard so drivers just were not as available in RTM as they were with the last RC release. If they hadn't done that, then the rep Vista got would not have been so bad.
3rd Rule of Vista: If a driver says "stop", times out, or blue screens the up-time is over.
But then you wouldn't have listened to the first two rules.
For many trees, a Git repository with full history, and a checked out working tree, can be smaller than the equivalent SVN checkout.
(SVN may have changed this recently, I've stopped using it) ; SVN keeps a second pristine working copy of the files in hidden folders for comparison purposes, Git keeps everything packed in compressed files.
SVN changed their working copy information in 1.7 - now you have a central .svn folder for the entire working copy instead of in all the sub-folders. It's a little more painful as before you could just copy a subfolder somewhere else to copy it; now you can't - you have to go through a checkout. It gains them some efficiencies though. I don't think they do any compression on the files though.
As to git, it's more efficient in that the whole repository is there with only a single version of each object. However, it doesn't track folders that don't yet have any contents (something I find really annoying).
Difference between git on a server and on the desktop is that on the desktop you use a normal git repository - which has all the changes and a working copy; while on the server you only publish the bare-bones git repository - which is only the changes.
git doesn't need a central server.
True. However, for some things that is actually a detraction. For example, companies do a lot of paperwork, much of which needs to be controlled as to who has access to it. A centralized system such as Subversion or CVS is really good for those kinds of things, while a decentralized system like git is not - too easy to get those sensitive things into the wrong hands, and no ability to control permissions.
And yes, I do use SVN and git for various things. Git is great for developers as it lets you take the repo on the road very easily and still do all your tracking, while SVN your have to dial-back home to really do your tracking while on the road - which can be a pain at times.
As always, best tool for the job.
SVN/TortoiseSVN will do a 3-way merge (for conflict resolution) on a MS word document, at least for the newer releases of MS Word (2010+). I am not sure if the document has to be a .docx or not though.
TortoiseSVN does it via a custom diff handler for the doc/docx types. And no, it doesn't have to be docx, and yes it will work with MS Word 2003 (for which the script was originally written).
The problem is that you can't do a diff between Word documents. Source control and diffing tools work great on source code because it's all just plain text. But for things like Word documents or Powerpoint presentations things get a little more complicated. Sure with a version control service you won't lose any previous versions, but you don't know what changed between versions either. To me this is the major thing missing for MS Office, and I can't believe they haven't done it yet. Sure they have track changes, but you have to remember to turn that on. There's no reason you should have to track changes. You should be able to take two versions of the document and MS Word should be able to tell you what has changed between the two versions. I think that if OpenOffice developed this feature it would be a killer feature that might get people to actually start using it. Because as you pointed out, when you have 4 different versions of the same document from 4 different people, it's nice to be able to figure out which changes those 4 people actually made without going through all 50 pages of the document.
TortoiseSVN has a script that uses Word's features to diff Word files in Subversion. I think it can done for Excel and other parts of office too. However, few know about those features, and fewer still use them.
+infiniti
one of the few times I really wish I had some mod points...