I agree with you, there is no need to email.doc or.xls files, and I make it a point to email either plaintext or PDF files. However, most people do not understand that, and OpenOffice.org will open.doc files with little difficulty.
Most computer users are "computer illiterate." Case in point: how often does someone attach a.doc file that contains nothing more than an embedded image to an email they send you? It is ridiculous, but if you ask them why they did that, they just shrug. It is not just a MS Office user issue: I have had people send me OpenDocument files with nothing more than an embedded image as well, and they shrug too. People really know very little about their computers, or the software that runs on those computers.
It is my opinion that they should take the time to learn more about computing, but most people either do not have the time or simply do not care, and really, what reason do they have for caring? If their boss demanded higher productivity out of them, they might have a reason to care, but how many people would even think of their computer as a way to accelerate their work? How many people would even think that their computer already has enough software for productivity enhancements? In my entire life, I have met only one person who tried to script Windows, and only a handful of people who could write anything more than the most basic batch-style shell scripts in *nix. Nobody else I know has even given consideration to that as a possibility, and several people I know have even expressed doubt that it is even possible when I mention it to them.
"The people who maintain the tools you are using need to be Free (eg need to be able to modify it)."
No, they need legal access to the source code. The Windows team at Microsoft has legal access to the Windows codebase, and they can and do maintain it for all Windows users. Now, you and I probably both disagree with Microsoft's tactics with Windows, but the fact of the matter is that most people, especially individuals, having Microsoft maintain the Windows codebase is completely acceptable. In fact, outside of the world of servers and business computing, I cannot think of many examples of third parties being paid to maintain some code. This is one of the reasons that RMS is out of touch with the general public: he still views computing from the angle of the 1980s.
"The bottom line is that its not all about YOU and YOUR needs, its about a communities, if you avoid being part of that community then expect your needs to be ignored."
Agreed, but there is no need for free software zealotry. If people do not understand the decision they have, then educate them. If they still choose proprietary software...fine, that is their decision, and they had the right to make it.
"they had finally come around to the point of view that it was OK for third-party software houses to sell applications that would run on their OS. How many people are as far ahead of their time as Stallman was in 1984?"
Hm...a big company, usually referred to with a three letter acronym, comes to mind...
Sure, there are, and you are probably already using such hardware without even realizing it (except for the "Penguin on the box" part). It does, however, increase the cost of the hardware, and requires the hardware to be redesigned with all sorts of other functionality, like reflashing for firmware updates, loading the firmware from that flash cell (if that is to happen automatically, then the hardware must be designed to do it automatically), etc. There is also a more complex testing requirement, to ensure that a firmware update wouldn't suddenly brick the hardware. For some devices, though, this makes sense, such as USB connected devices or internal hard drives. For others, it just increases the complexity and cost, including development costs, so why would a company want to do it? Especially a company that releases new hardware frequently, like nVidia?
"--Software that runs hardware. You are buying the hardware, the software is just something that helps make it go. Thus it isn't a problem to have anyone able to copy and redistribute it. Heck, might even be to your benefit as maybe they make it better. Your money is in the device, so the software needn't be restricted. Embedded devices would be an example."
This is an interesting issue in the free software world, that is difficult to approach fairly. It is becoming increasingly common to use programmable logic (FPGAs, CPLDs, etc.) instead of specialty ASIC microchips in hardware these days. That means that for the hardware to run, there must be firmware available, either on a flash cell (which increases the cost of the hardware) or in the driver (which now has a free software complication).
Take nVidia as an example. Many people are confused when nVidia says that they cannot release an open source driver because of licensing agreements with other companies, but this is probably true. nVidia likely subcontracted some of its logic designs, and implemented those designs in firmware for an FPGA of some sort. The firm they contracted with probably demanded royalties for the logic designs, and why shouldn't they? Even RMS does not demand to know the internal designs of his hardware, and the contractor that demanded the royalties probably had no idea (and no need to know) how the logic would be implemented. The same is likely true of Broadcom wireless cards, various video capture cards that have no free software drivers, and all that specialty hardware out there.
This is why, although I am a big supporter of free software, I draw the line with firmware. Binary blobs do not encumber free software, because really, FPGA firmware is not really software, it is a hardware design.
The "freedom" that Stallman refers to has nothing to do with developer freedom, it has to do with the freedom of the users of software. You, as a user, can do anything you want with the software, but you, as a developer or distributor, must grant other users the same freedom. The GPL is about protecting the freedoms of users; your assertion about the BSD license is correct in that the BSD license "protects" the freedom of developers.
"For example, how many of us can say with hand on heart that we don't use an MP3 decoder? A nVidia graphics card? Firmware for the Intel wireless cards? In RMS's eyes we've tainted our freedom, but in reality these compromises allow us MORE freedom of choice, not less."
One of the issues here is that most people need to work with others, and as free software remains in the minority, the need to work with others forces us to use proprietary or entangled software. If you asked for a sound sample from someone, what is the likelihood that you would receive an OGG? What is the likelihood that you would even know what they will send you? I have had professors record lectures and post Real encoded files, and forced me to paw through Livna just to find codecs. If you are given equipment for a lab, how often do you get to specify what sort of graphics or network cards will be inserted into your computer? More often than not, you'll be given whatever is on hand, and you'll just have to deal with that. Most of us are not in a position to demand free-software-friendly hardware or media, nor are we in a position to refuse non-free hardware or media. It is not that we really WANT to use MP3 (it is a pain in the ass to get MP3 support in Fedora), but sometimes, you just have to.
Which is why we prefer the term "free-libre," which at least forces the uninitiated to ask, "What do you mean?" rather than jumping to the conclusion that we are talking about the price of software.
The problem with Stallman's approach is the assumption that most people want the free software ideal. The reality is that most people are not even knowledgeable enough about their computers to even understand what free software is all about, why it matters, and why they should care. All they see is Windows with driver support in one corner, Mac OS X working out of the box on bundled hardware in the other corner, and Linux/BSD/etc. in the last corner with poor (but slowly improving) driver support that may or may not work out of the box.
What Stallman needs to do is catch up with the biggest development in the computing world of the past 25 years: the growth of computer users who do not know anything about their computers, and do not care to know. Most people do not care about the legal or technical issues surrounding their software, they just want to get online and do stuff. Stallman insists that when somebody sends you a.doc file, you should refuse to open it and insist that they send you a PDF or ODT file instead. Great when you are dealing with engineers and programmers, but not so great when you are dealing with people who think you need to create a.doc file in order to attach an image to an email.
Disclaimer: I am a big supporter of free software, and I do wish that more people would learn more about their computers so they could at least understand that they have a choice.
Sadly, much as I like KDE, this is basically the case. Nobody really knows what Plasmoids are useful for, other than simplistic applets that we already had with other systems. It is kind of interesting to having something like Folder View...but not really useful. Unfortunately, the KDE team spent so much time worrying about Plasma and plasmoids and getting it all working that they neglected things like KHotKeys, KSSL, etc.
Maybe this is fixed in the nightly builds or in 4.2? I am using 4.1.3 right now, and yes, that option *exists* but it does not work, it does not propagate the public keys globally, and it does not retain those settings after hitting "OK." Claiming that is the solution to my problem is kind of like calling your alpha release "version 4.0 stable." Hmm...
"the "OLE framework" in KDE is called KParts, and the infrastructure hasn't changed one bit between KDE3 and KDE4."
Except that the use of KParts has changed. In KDE3, all the KParts components played well with each other (except for the Kontact KParts, which only embedded in Kontact), which is exactly what OLE is supposed to do. In KDE4, a few components still embed in one another, but nothing on the level of KDE 3. The OP was claiming that KDE 4.1 was approaching 3.5 in terms of functionality; where are the useful, play-nice-with-others KParts?
"it doesn't embed into Dolphin, no, because that's not Dolphin's design goal. i don't have 4.1 nearby to test this on, but in 4.2 you can navigate directly into tarballs seamlessly in Konqueror."
Then Dolphin was poorly designed. I do not need a file manager if all it does is browse normal, already mounted file systems. Dolphin certain supports some level of OLE, the fact that it cannot embed an ArK component is, once again, an oversight, and an embarrassing one. Maybe this will be fixed in 4.2.x? 4.3?
"currently Ark relies on KParts for previewing files without extracting. an "open with" that would extract to a temporary location and launch the app would be nice, though."
And once again, they do not play well with others. Why not have an Okular component embed into ArK? Why force me to extract a PDF file just to view it? If the file manager does not embed an ArK component, and ArK cannot embed an Okular component, then why would I use Dolphin/ArK when I could just use a terminal? In 3.5, there was no question: KPDF embedded in ArK, ArK embedded in Konqueror, and the software stack was more useful than trying to navigate using just a terminal.
"which shortcuts would those be? actually, better yet, go to bugs.kde.org and report it there so it can be handled."
Notice that they did not even PLAN to fix it in 4.1. Maybe it will be fixed in 4.2? I can only hope so, because it is clumsy, annoying, and frankly stupid for shortcuts to fail. What is very odd, though, is the kxkb shortcuts work in Fluxbox; oh wait, that is confirmed too:
Again, NO PLANS to fix. Why are there no plans to fix this? Keyboard shortcuts are the only way to keep a large GUI like KDE from being too clumsy to use, but they are sitting around scratching their heads and not even trying to get this working. Again, one is forced to ask just what design methodology they are adhering to, if any. Another commenter noted that there are other shortcut daemons; is that really what we are stuck with?
I take it you are not very demanding with features or stability? Things that are completely broken in 4.1:
SSL -- There is no SSL configuration tool, poor documentation on where SSL certificates are stored or how they are stored, and bugfixes are barely even on the horizon right now.
Keyboard shortcuts -- not only are global shortcuts still not working, but KDE4 seems to kill shortcuts set by other applications, even when those shortcuts are working when I run the application not in KDE.
OLE -- 3.5 had solid OLE system that worked exceptionally well. 4.1 has an OLE system that is flaky, poorly unified, and poorly used. Maybe 4.2 will fix it? Maybe we won't see a fix until 5.0.
Bluetooth -- I should NOT be using Nautilus for browsing Bluetooth filesystems.
ArK -- I should not have to extract files from an archive to view them. Assuming that ArK will even get me that far, which it sometimes will not.
Samba -- Samba support should be integrated with Dolphin, or supported by embedding smb4k using the OLE system; see above.
Configuration -- I should be able to rely on my configuration settings remaining set. Over and over, I see my settings being forgotten when I hit "Apply," even for things that should be a no brainer: setting the default application to open a text file.
You can check the KDE bugzilla if you are curious about just how many things need to be fixed. KDE 4 is a complete mess, and was completely mishandled. It is getting to the point where, embarrassing as it would be, they should probably scrap it and start over by porting KDE3 to Qt4.
Except that you do not need to write KDE apps or Qt apps for your program to run properly in KDE. You can run GTK applications in KDE, I do it all the time, and it is not a problem at all.
Plasmoids can be embedded in the taskbar, and could be useful for something like a little weather applet. Since Plasmoids use SVG, they fit well no matter what size you choose for your taskbar. There is potential there for something useful, but really, potential will only get us so far.
Uhm, 4.1 is only marginally better than alpha quality. Perfect example: yesterday, I needed to import a CA public key for use in all my KDE apps. There is no tool for this, and I actually had to use 'cat' to append the certificate to the system certificates file. That is an embarrassing oversight, and forces one to question just what sort of design practices, if any, were adhered to by the KDE 4 team.
You say that 4.0 was a temporary step backward from 3.5? 4.1 is still a step backward, just slightly less of one. 3.5 derived a lot of its power from a very solid, well refined OLE framework, and 4.1 has yet to even approach that. In 3.5, it was seamless to browse a tarball, because the ArK component would embed right into Konqueror. ArK does not embed into Dolphin or Konqueror in 4.1, and in standalone ArK, you cannot open most files without extracting, which is annoying and basically defeats the purpose of a tool like ArK. Many users, myself included, use (or used to use) keyboard shortcuts for various actions -- yet that is still completely broken in KDE 4.1, and worse yet, some application shortcuts are broken if you run the application with KDE as the WM, but work just fine if you use something else.
If the KDE team does not get their act together fast, and give people some sort of hope with the 4.2 release, KDE is going to die.
A malware attack on a hospital may very well affect life-or-death systems, if it is carried into the hospital on a USB key. Some doctor is reviewing patient records, brings it all home on his USB key for some reason, and brings that key back into the office -- now the malware is inside, on that internal LAN. It may not infect the life critical systems, but it may clog the network and prevent those hosts from communicating with whatever other systems they need to communicate with. My hope would be that those systems are on a separate LAN altogether from every other system, but I would doubt that the budget allowed for doubling up on LAN cabling like that.
That may not be as simple as it sounds. Sure, it is technologically feasible to lock down a computer system, but there are matters of money and politics to consider. Consider the expense of hiring a full time security team that can tune ACLs and security policies and monitor the hospital network for intrusions. Here in America, hospitals, especially public hospitals, often have to fight for every dollar just to afford medical equipment, and there is constant political wrangling about paying for healthcare. Investing millions of dollars per hospital to create a secure IT infrastructure is a difficult move to justify when you are engaged in a battle for money for other equipment, and a lot of people either do not understand or do not care about the risks patients face from IT failures.
There is also the matter of commercialization of healthcare software. Gone are the days when a hospital's IT staff would roll their own middle tier and front end systems -- healthcare software systems are now purchased from companies that "specialize" in such products. Those companies often market proprietary software, compile it for the world's most popular desktop OS, and send shrink-wrapped copies to hospitals. That software can force choices upon the hospital, like requiring a certain database that only runs on a certain server OS or preventing certain ACLs from being in place because of the manner in which the software utilizes system resources. It is neither malice nor incompetence, it is just a byproduct of the system we have in place for managing our healthcare centers.
Personally, I have never understood how utilities might wind up in a situation where their systems may be vulnerable to a malware attack. I would think that the critical systems in utilities would be offline and running some sort of highly application-specific software, but I could be wrong.
People often forget that many real world, physical system have internet connections, and therefore many people cannot even fathom the idea of a cyberattack resulting in human death. I read about a hack a few years ago that nearly resulting in a man being shot and killed by a SWAT team: somebody had spoofed the phone system and issued a call to 911 indicating that he was holding a family hostage, and the SWAT team arrived and very nearly shot the father in that house. The kid who executed the hack never even considered the possibility that a SWAT team with automatic weapons might actually fire their guns during the confusion (or so he said when he was arrested by the FBI). TFA indicates that a malware attack hit a UK hospital and shut down the computer systems, forcing doctors and nurses to search for paper records.
Easy: have the car automatically accelerate if something is coming toward the rear end of the car. Sounds just as safe as having a car that automatically breaks.
Worse is that people tend to attribute useful innovations to the wrong source. How many people do you hear attributing the mouse to Apple? How many other people do you hear trying to correct the first group by telling them that Xerox invented the mouse? When an innovation goes poorly, the people who came up with it become a joke and are remembered because of that joke; but when it goes well, it is usually some company like Apple or Microsoft that popularizes it, and nobody remembers the original innovators.
Slashdot Mirror
Linux users don't have sex! I have documentary evidence of this:
http://xkcd.com/196/
I agree with you, there is no need to email .doc or .xls files, and I make it a point to email either plaintext or PDF files. However, most people do not understand that, and OpenOffice.org will open .doc files with little difficulty.
.doc file that contains nothing more than an embedded image to an email they send you? It is ridiculous, but if you ask them why they did that, they just shrug. It is not just a MS Office user issue: I have had people send me OpenDocument files with nothing more than an embedded image as well, and they shrug too. People really know very little about their computers, or the software that runs on those computers.
Most computer users are "computer illiterate." Case in point: how often does someone attach a
It is my opinion that they should take the time to learn more about computing, but most people either do not have the time or simply do not care, and really, what reason do they have for caring? If their boss demanded higher productivity out of them, they might have a reason to care, but how many people would even think of their computer as a way to accelerate their work? How many people would even think that their computer already has enough software for productivity enhancements? In my entire life, I have met only one person who tried to script Windows, and only a handful of people who could write anything more than the most basic batch-style shell scripts in *nix. Nobody else I know has even given consideration to that as a possibility, and several people I know have even expressed doubt that it is even possible when I mention it to them.
"The people who maintain the tools you are using need to be Free (eg need to be able to modify it)."
No, they need legal access to the source code. The Windows team at Microsoft has legal access to the Windows codebase, and they can and do maintain it for all Windows users. Now, you and I probably both disagree with Microsoft's tactics with Windows, but the fact of the matter is that most people, especially individuals, having Microsoft maintain the Windows codebase is completely acceptable. In fact, outside of the world of servers and business computing, I cannot think of many examples of third parties being paid to maintain some code. This is one of the reasons that RMS is out of touch with the general public: he still views computing from the angle of the 1980s.
"The bottom line is that its not all about YOU and YOUR needs, its about a communities, if you avoid being part of that community then expect your needs to be ignored."
Agreed, but there is no need for free software zealotry. If people do not understand the decision they have, then educate them. If they still choose proprietary software...fine, that is their decision, and they had the right to make it.
"they had finally come around to the point of view that it was OK for third-party software houses to sell applications that would run on their OS. How many people are as far ahead of their time as Stallman was in 1984?"
Hm...a big company, usually referred to with a three letter acronym, comes to mind...
Sure, there are, and you are probably already using such hardware without even realizing it (except for the "Penguin on the box" part). It does, however, increase the cost of the hardware, and requires the hardware to be redesigned with all sorts of other functionality, like reflashing for firmware updates, loading the firmware from that flash cell (if that is to happen automatically, then the hardware must be designed to do it automatically), etc. There is also a more complex testing requirement, to ensure that a firmware update wouldn't suddenly brick the hardware. For some devices, though, this makes sense, such as USB connected devices or internal hard drives. For others, it just increases the complexity and cost, including development costs, so why would a company want to do it? Especially a company that releases new hardware frequently, like nVidia?
"--Software that runs hardware. You are buying the hardware, the software is just something that helps make it go. Thus it isn't a problem to have anyone able to copy and redistribute it. Heck, might even be to your benefit as maybe they make it better. Your money is in the device, so the software needn't be restricted. Embedded devices would be an example."
This is an interesting issue in the free software world, that is difficult to approach fairly. It is becoming increasingly common to use programmable logic (FPGAs, CPLDs, etc.) instead of specialty ASIC microchips in hardware these days. That means that for the hardware to run, there must be firmware available, either on a flash cell (which increases the cost of the hardware) or in the driver (which now has a free software complication).
Take nVidia as an example. Many people are confused when nVidia says that they cannot release an open source driver because of licensing agreements with other companies, but this is probably true. nVidia likely subcontracted some of its logic designs, and implemented those designs in firmware for an FPGA of some sort. The firm they contracted with probably demanded royalties for the logic designs, and why shouldn't they? Even RMS does not demand to know the internal designs of his hardware, and the contractor that demanded the royalties probably had no idea (and no need to know) how the logic would be implemented. The same is likely true of Broadcom wireless cards, various video capture cards that have no free software drivers, and all that specialty hardware out there.
This is why, although I am a big supporter of free software, I draw the line with firmware. Binary blobs do not encumber free software, because really, FPGA firmware is not really software, it is a hardware design.
"Perhaps the parent's POV on what Stallman thinks of doc files might be a bit extreme to some people but he has a point."
http://www.gnu.org/philosophy/no-word-attachments.html
I really was not kidding: Stallman does believe that you should demand free media if you are sent entangled or proprietary media.
The "freedom" that Stallman refers to has nothing to do with developer freedom, it has to do with the freedom of the users of software. You, as a user, can do anything you want with the software, but you, as a developer or distributor, must grant other users the same freedom. The GPL is about protecting the freedoms of users; your assertion about the BSD license is correct in that the BSD license "protects" the freedom of developers.
"For example, how many of us can say with hand on heart that we don't use an MP3 decoder? A nVidia graphics card? Firmware for the Intel wireless cards? In RMS's eyes we've tainted our freedom, but in reality these compromises allow us MORE freedom of choice, not less."
One of the issues here is that most people need to work with others, and as free software remains in the minority, the need to work with others forces us to use proprietary or entangled software. If you asked for a sound sample from someone, what is the likelihood that you would receive an OGG? What is the likelihood that you would even know what they will send you? I have had professors record lectures and post Real encoded files, and forced me to paw through Livna just to find codecs. If you are given equipment for a lab, how often do you get to specify what sort of graphics or network cards will be inserted into your computer? More often than not, you'll be given whatever is on hand, and you'll just have to deal with that. Most of us are not in a position to demand free-software-friendly hardware or media, nor are we in a position to refuse non-free hardware or media. It is not that we really WANT to use MP3 (it is a pain in the ass to get MP3 support in Fedora), but sometimes, you just have to.
Which is why we prefer the term "free-libre," which at least forces the uninitiated to ask, "What do you mean?" rather than jumping to the conclusion that we are talking about the price of software.
The problem with Stallman's approach is the assumption that most people want the free software ideal. The reality is that most people are not even knowledgeable enough about their computers to even understand what free software is all about, why it matters, and why they should care. All they see is Windows with driver support in one corner, Mac OS X working out of the box on bundled hardware in the other corner, and Linux/BSD/etc. in the last corner with poor (but slowly improving) driver support that may or may not work out of the box.
.doc file, you should refuse to open it and insist that they send you a PDF or ODT file instead. Great when you are dealing with engineers and programmers, but not so great when you are dealing with people who think you need to create a .doc file in order to attach an image to an email.
What Stallman needs to do is catch up with the biggest development in the computing world of the past 25 years: the growth of computer users who do not know anything about their computers, and do not care to know. Most people do not care about the legal or technical issues surrounding their software, they just want to get online and do stuff. Stallman insists that when somebody sends you a
Disclaimer: I am a big supporter of free software, and I do wish that more people would learn more about their computers so they could at least understand that they have a choice.
Sadly, much as I like KDE, this is basically the case. Nobody really knows what Plasmoids are useful for, other than simplistic applets that we already had with other systems. It is kind of interesting to having something like Folder View...but not really useful. Unfortunately, the KDE team spent so much time worrying about Plasma and plasmoids and getting it all working that they neglected things like KHotKeys, KSSL, etc.
"Konqueror -> Settings -> Configure Konqueror -> Crypto -> SSL Signers -> Import."
Maybe this is fixed in the nightly builds or in 4.2? I am using 4.1.3 right now, and yes, that option *exists* but it does not work, it does not propagate the public keys globally, and it does not retain those settings after hitting "OK." Claiming that is the solution to my problem is kind of like calling your alpha release "version 4.0 stable." Hmm...
"the "OLE framework" in KDE is called KParts, and the infrastructure hasn't changed one bit between KDE3 and KDE4."
Except that the use of KParts has changed. In KDE3, all the KParts components played well with each other (except for the Kontact KParts, which only embedded in Kontact), which is exactly what OLE is supposed to do. In KDE4, a few components still embed in one another, but nothing on the level of KDE 3. The OP was claiming that KDE 4.1 was approaching 3.5 in terms of functionality; where are the useful, play-nice-with-others KParts?
"it doesn't embed into Dolphin, no, because that's not Dolphin's design goal. i don't have 4.1 nearby to test this on, but in 4.2 you can navigate directly into tarballs seamlessly in Konqueror."
Then Dolphin was poorly designed. I do not need a file manager if all it does is browse normal, already mounted file systems. Dolphin certain supports some level of OLE, the fact that it cannot embed an ArK component is, once again, an oversight, and an embarrassing one. Maybe this will be fixed in 4.2.x? 4.3?
"currently Ark relies on KParts for previewing files without extracting. an "open with" that would extract to a temporary location and launch the app would be nice, though."
And once again, they do not play well with others. Why not have an Okular component embed into ArK? Why force me to extract a PDF file just to view it? If the file manager does not embed an ArK component, and ArK cannot embed an Okular component, then why would I use Dolphin/ArK when I could just use a terminal? In 3.5, there was no question: KPDF embedded in ArK, ArK embedded in Konqueror, and the software stack was more useful than trying to navigate using just a terminal.
"which shortcuts would those be? actually, better yet, go to bugs.kde.org and report it there so it can be handled."
http://bugs.kde.org/show_bug.cgi?id=165441#c2
Notice that they did not even PLAN to fix it in 4.1. Maybe it will be fixed in 4.2? I can only hope so, because it is clumsy, annoying, and frankly stupid for shortcuts to fail. What is very odd, though, is the kxkb shortcuts work in Fluxbox; oh wait, that is confirmed too:
http://markmail.org/message/dxz6fntbrp73cljl
Again, NO PLANS to fix. Why are there no plans to fix this? Keyboard shortcuts are the only way to keep a large GUI like KDE from being too clumsy to use, but they are sitting around scratching their heads and not even trying to get this working. Again, one is forced to ask just what design methodology they are adhering to, if any. Another commenter noted that there are other shortcut daemons; is that really what we are stuck with?
Which university are you at? Mine basically holds all rights to any research done by students or faculty using university equipment.
You can check the KDE bugzilla if you are curious about just how many things need to be fixed. KDE 4 is a complete mess, and was completely mishandled. It is getting to the point where, embarrassing as it would be, they should probably scrap it and start over by porting KDE3 to Qt4.
Except that you do not need to write KDE apps or Qt apps for your program to run properly in KDE. You can run GTK applications in KDE, I do it all the time, and it is not a problem at all.
Plasmoids can be embedded in the taskbar, and could be useful for something like a little weather applet. Since Plasmoids use SVG, they fit well no matter what size you choose for your taskbar. There is potential there for something useful, but really, potential will only get us so far.
Uhm, 4.1 is only marginally better than alpha quality. Perfect example: yesterday, I needed to import a CA public key for use in all my KDE apps. There is no tool for this, and I actually had to use 'cat' to append the certificate to the system certificates file. That is an embarrassing oversight, and forces one to question just what sort of design practices, if any, were adhered to by the KDE 4 team.
You say that 4.0 was a temporary step backward from 3.5? 4.1 is still a step backward, just slightly less of one. 3.5 derived a lot of its power from a very solid, well refined OLE framework, and 4.1 has yet to even approach that. In 3.5, it was seamless to browse a tarball, because the ArK component would embed right into Konqueror. ArK does not embed into Dolphin or Konqueror in 4.1, and in standalone ArK, you cannot open most files without extracting, which is annoying and basically defeats the purpose of a tool like ArK. Many users, myself included, use (or used to use) keyboard shortcuts for various actions -- yet that is still completely broken in KDE 4.1, and worse yet, some application shortcuts are broken if you run the application with KDE as the WM, but work just fine if you use something else.
If the KDE team does not get their act together fast, and give people some sort of hope with the 4.2 release, KDE is going to die.
When the need for 64 bit OSes starts to become significant. How many people have a significant need for 64 bit computing right now?
A malware attack on a hospital may very well affect life-or-death systems, if it is carried into the hospital on a USB key. Some doctor is reviewing patient records, brings it all home on his USB key for some reason, and brings that key back into the office -- now the malware is inside, on that internal LAN. It may not infect the life critical systems, but it may clog the network and prevent those hosts from communicating with whatever other systems they need to communicate with. My hope would be that those systems are on a separate LAN altogether from every other system, but I would doubt that the budget allowed for doubling up on LAN cabling like that.
That may not be as simple as it sounds. Sure, it is technologically feasible to lock down a computer system, but there are matters of money and politics to consider. Consider the expense of hiring a full time security team that can tune ACLs and security policies and monitor the hospital network for intrusions. Here in America, hospitals, especially public hospitals, often have to fight for every dollar just to afford medical equipment, and there is constant political wrangling about paying for healthcare. Investing millions of dollars per hospital to create a secure IT infrastructure is a difficult move to justify when you are engaged in a battle for money for other equipment, and a lot of people either do not understand or do not care about the risks patients face from IT failures.
There is also the matter of commercialization of healthcare software. Gone are the days when a hospital's IT staff would roll their own middle tier and front end systems -- healthcare software systems are now purchased from companies that "specialize" in such products. Those companies often market proprietary software, compile it for the world's most popular desktop OS, and send shrink-wrapped copies to hospitals. That software can force choices upon the hospital, like requiring a certain database that only runs on a certain server OS or preventing certain ACLs from being in place because of the manner in which the software utilizes system resources. It is neither malice nor incompetence, it is just a byproduct of the system we have in place for managing our healthcare centers.
Personally, I have never understood how utilities might wind up in a situation where their systems may be vulnerable to a malware attack. I would think that the critical systems in utilities would be offline and running some sort of highly application-specific software, but I could be wrong.
People often forget that many real world, physical system have internet connections, and therefore many people cannot even fathom the idea of a cyberattack resulting in human death. I read about a hack a few years ago that nearly resulting in a man being shot and killed by a SWAT team: somebody had spoofed the phone system and issued a call to 911 indicating that he was holding a family hostage, and the SWAT team arrived and very nearly shot the father in that house. The kid who executed the hack never even considered the possibility that a SWAT team with automatic weapons might actually fire their guns during the confusion (or so he said when he was arrested by the FBI). TFA indicates that a malware attack hit a UK hospital and shut down the computer systems, forcing doctors and nurses to search for paper records.
Easy: have the car automatically accelerate if something is coming toward the rear end of the car. Sounds just as safe as having a car that automatically breaks.
Ah yes, XHTML, a very important step. I understand the Clippy was part of an early AJAX experiment...
</sarcasm>
Worse is that people tend to attribute useful innovations to the wrong source. How many people do you hear attributing the mouse to Apple? How many other people do you hear trying to correct the first group by telling them that Xerox invented the mouse? When an innovation goes poorly, the people who came up with it become a joke and are remembered because of that joke; but when it goes well, it is usually some company like Apple or Microsoft that popularizes it, and nobody remembers the original innovators.