I have no idea but this topic fits perfectly in the ever degenerating "ask slashdot" series. The type of questions asked under this topic can mostly be answered with "ask google" (or sometimes a simple "Duh!" is appropriate).
However, this particular "question" may form an interesting start for a debate about why micropayments don't work.
IMHO it is for the same reason that we had the dot com implosion: in order to work, micropayments require lots of transactions in order to generate a significant amount of revenue. Typically the critical amount of transactions never materializes.
Well they have a point that the combination HTML/javascript is cumbersome. At least they bother to address the issue whereas the rest of the world still struggles to acknowledge the issue.
I use dreamweaver occasionally and despite some annoying bugs this is a very nice development environment. People dismissing it as just another text editor are really not doing this product justice. I've spent quite a bit of time evaluating open source alternatives to dreamweaver but nothing I've seen so far even comes close. The key feature of dreamweave that I absolutely depend on is nested templates (define site layout/menus/etc. separately from the pages on the site). If anyone knows of an easy to use, oss alternative for that I'd be grateful.
Kazaa has one huge problem for me: large downloads invariably corrupt. That doesn't really matter for movies since it just means you have a few seconds of distorted sound/picture. However, if you're downloading things that need to be 100% intact like (cough) a linux iso (cough) it is very convenient if the p2p client bothers to verify that what it is downloading is the same as that should be downloaded.
Shareaza solves this problem, kazaa doesn't. However, the shareaza network is so small currently that there isn't much to download. With 3 million users at any time, the Kazaa network is unbeatable in that respect.
Well it is easy to point the finger at MS when even this open source safe haven called slashdot does not meet any w3c standard. If standards are so important, why can't this site adhere to them?
The reason GUI tools never work in linux is that they are an afterthought. Configuration always boils down to editing text file XYZ in some location which typically varies from distribution to distribution (something that often is worked areound using symbolic links!). There's a few problems with this approach: 1) each configuration file has its own syntax therefore automated editing (e.g. through a gui) needs to be custom designed (and consequently lacking/severely limited). 2) sysadmins typically have to learn how to configure a tool in addition to learning to understand what it does (consequently sysadmins tend to be expensive). 3) sysadmins have to be aware of distribution specifics in order to succesfully configure a system.
Most attempts at making linux user friendly work around these issues rather than fixing them. As a result, you can't go to the shop and buy software for linux. You can only buy software bundled with linux because that is the only way to get it working in a somewhat reliable fashion. It is also the reason many software developers currently do not have a linux version of their software. Getting software running on the zillion variations of linux is a support nightmare.
I was referring to the fact that generics in Java is nothing new. The proposal for inclusion of generics in Java has been around for quite some time. At the time the 1.4 spec was being finalized it was nearly ready. I've been aware for more than a year now that the intention was to include generics in Java 1.5.
I agree that probably Java's generics are a bit of a compromise between the Java object model, generics security and type safety. Arguably, C++ is more powerfull but less safe in that respect. Probably Beta does a better job at implementing generics. However, generics are not the solution to all problems. Essentially separation of concerns is not possible for all problem dimensions (which is what both OO and generics try to achieve).
Sounds like you don't get the point. The reason templates are useful is because it reduces the amount of code you need to write for common situations in a typesafe way. This makes your code more compact (duplicate code) and less error-prone (it's all type safe and you only write it once).
BTW. generics are old news. The prototype compiler and the proposal have been available for at least a year. On the javalobby there's a link to some additional language extensions: http://www.javalobby.org/thread.jsp?forum=61&threa d=6009 (lets hope/. gets this right). Lets hope all these goodies end up in 1.5.
Actually I would recommend against insisting on a particular format. Not everybody is a tex fan and I don't know anyone who ever worked with docbook. It's content that matters, not the format. You don't want to scare away interested authors by putting all sorts of constraints on the wordprocessor. IEEE and ACM for instance no longer do this, instead they have guidelines for formatting your document and templates for most popular wordprocessors are available.
I mostly work with framemaker and sometimes with ms word. It always pisses me off if TeX is required. It's not my problem that the editor likes TeX. He/she should be glad I'm willing to invest the time to (freely) provide content.
As for the functionality, there are two aspects: handling the subscriptions and handling the review process. Good solutions for both aspects exist. There's basically no need to reinvent the wheel.
Why can't kernel compiling linux users figure out how to use the control panel? It's not that hard. It was designed to be useable by idiots. It's not completely idiot proof, but then what software is?
Of course you can save the more essential stuff like bookmarks, cookies, mailfolders etc. by making a backup copy. This is how I upgraded from 1.0 RC1 all the way to 1.2.1 (via 1.0 RC2 1.0, 1.1a, 1.1b 1.1, 1.2b, 1.2). It would be nice if the mozilla setup would automatically migrate stuff that can be migrated rather than just dumping everything ontop of the existing setup and hoping for the best.
I installed 1.2 as soon as it was released. I appreciate the quick update but must say I didn't notice anything wrong with the pages I visited. Apparently, this bug mostly affected some advertisements on some sites (cnn amongst others). You'd almost consider it a feature because of this.
I used a gnu licensed set of python files called cyberchair (http://www.cyberchair.org/) to manage the submission and review process for a small conference last year. If you know python, it is relatively easy to set up. The version distributed by the author is somewhat out of date and has a number of undocumented bugs that you need to work around. However, I found that it was reasonably functional after some tinkering. Basically cyberchair saved me a lot of time since it takes care of collecting abstracts, author information etc. Also you can set it up to request reviewers to use a web form to submit their reviews, generate overviews for conference chair and reviewers, etc.
As for scalability, it is all pretty light weight and basically only requires apache and a recent version of python. It should be relatively easy to adapt this for a journal. After all the review process for journals is similar to that of conferences.
The guy who developed cyberchair also runs a small service company that can arrange the whole thing for you (including webhosting if needed) for a reasonable fee. This is actually the route I recommend unless you have too much spare time. For budgetary reasons, this was not feasible in my case unfortunately.
There's a few similar opensource packages for managing conferences that I can't remember the name of right now but I'm sure google will help you with that.
The benchmark becomes even more interesting once you realize that the java benchmark includes the overhead of starting up java and jit compiling the class file. It still is faster.
Rather than getting xenophobic about foreigners, you could also demand that they are treated the same as 'native' americans (not just the indians though, pun intended). That way foreigners compete on quality and not on salary.
Of course the whole reason foreigners are so popular is that it is a modern form of slavery. They work long hours for low wages. And you always have the whip of the immigration authorities to slap them with if they complain (being fired usually boils down to being put on the plane back home). Take away that 'advantage' and it suddenly becomes worthwhile to hire locals (provided there are well educated natives available, it's mostly the uneducated ones that get fired). Of course in the long term, IT shops will emigrate towards India. Labor cost is the driving cost in IT and its about 1/10th in e.g. India.
Silver bullets still do not exist. New technologies and methodologies are often hyped as such and naturally fail to live up to the hype. However, that does not mean they are useless.
These technologies and methodologies have allowed us to keep pace with Moore's law. The amount of software we develop today would simply be impossible using the state of the art in 1970. We routinely poor out millions of lines of code in mammoth projects that take sever hundreds or thousands of man-years to complete. The only reason we are able to do so is because of improvements in development methodologies and technology.
The (huge) research field that studies these technologies and approaches is called software engineering.
Here in the Netherlands, the left-wing green party yesterday also proposed embracing open source. In a quite extensive report on their website (http://www.groenlinks.nl/nieuws/4001428.html, in Dutch). They motivate the proposal quite well. There are a few minor details that they got wrong (most notably, Linus' last name is misspelled and the fact that a closed source format is used for the actual report) but overall the message is that closed source is bad and open source can be beneficial for both economical reasons and other reasons such as security, reliability and openness. Considering the report is written by a non technical person for a non technical audience, the effort should be applauded.
With the upcoming election in January, I hope this will be one of the election themes.
There's a lot of GUI environments on linux. Most are a work in progress and generally they don't work together very well.
Most of the applications I use on a daily basis have alternatives on linux. Sadly, most of the alternatives are of lesser quality in terms of usability, stability, features, etc. Actually there are a few apps I'd like to spend some time with. For example evolution seems like a decent mail client that has some interesting features.
Probably these are issues I could work around. I'm quite comfortable tweaking linux to suit my needs. Currently I'm fooling around with Debian in vmware.
What I can't work around is that most of the games I enjoy are windows only. In addition I have several pieces of hardware which are either not supported at all or with degraded performance/functionality/stability under linux. I didn't invest in an audigy + geforce ti 4200 just to listen to stereo sound and run a couple of outdated games. Not to mention my scanner, camera, wheel, usb mouse. Probably, with a lot of work, I can make most of these things work to some extent. However, the fact remains that they were designed for windows and not for linux.
I browse with mozilla and I mail with outlook xp. Outlook is currently one of the best mail clients (in terms of features) and the mozilla mail client still needs a lot of work to even come close. Security is an issue with outlook only if you can't find your way to the preferences dialog to adjust security to the appropriate settings.
Spamnet is catching most of the spam on my machine now. I installed it two months ago and it has caught up of 95 percent of the spam I receive. More importantly, it hasn't miscategorized a single message.
The problem is that you need all these high level abstractions to reduce the workload of creating large systems. There's just no way you could have those VB monkeys be productive in C and there's just no way you are going to replace them with competent C programmers. Besides, competent programmers are more productive using high abstraction level tools.
BSD is just a kernel + a small toolset. As soon as you start running all the regular stuff on top of it performance is comparable to a full blown linux/mac os X/windows desktop. Proof: mac os X, remove the none BSD stuff and see what's left: no ui, no friendly tools, no easy access to all connected devices.
There's two ways to adopt new technology: you can throw away everything you have and move to new technology in a revolutionary fashion. This is highly disruptive to an organization. Excessive training cost, numerous problems, etc. are more than likely. In short, you don't want revolutions in your IT infrastructure. The second option is to evolve your system. Replace things that would otherwise require new investments in licenses, things that you are not happy with etc.
If your exchange server is working fine, don't replace it unless you would otherwise upgrade to a new version. If your IIS webserver is adequate for serving the handful of word documents in your organization, don't fix it unless you need certain functionality offered by apache. Linux has a lower TCO for stuff like filesharing and printersharing. However, you have already invested in your existing infrastructure and if its working fine there's no reason to make additional investments in a functional equivalent.
Peercast also supports firewalled nodes. Streams can both be pushed and pulled. The only problem is that two firewalled nodes can't connect to each other directly. A firewalled node can push streams to none firewalled nodes and it can pull streams from none firewalled nodes.
The protocol is the same as is used by shoutcast. All that peercast does is relay the stream. Typically clients such as winamp do some buffering.
Essentially, when streaming, you tell shoutcast/oddcast or whatever encoder you are using to stream the music to your peercast client rather than a regular shoutcast server. Peercast then inserts some meta information into the peercast network so that other clients can find your stream. Other peercast clients can connect to your client to receive the stream. Peercast treats received streams exactly the same as streams that come from oddcast so anyone listening to your stream can also relay it further on. Theoretically, this allows a modem user to stream to a nearly endless amount of users by uploading just 1 stream. In practice it is better to allow for two or three streams so you have somewhat better reliability. Of course you need bandwidth in order to be able to relay. Low bandwidth ogg streams propagate better through the network than 128kbit mp3 streams. Ogg streams of around 45 kbit seem to be popular among peercast users.
Recently the ability to relay regular shoutcast streams was added so if you have a regular shoutcast server, you may save some bandwidth by encouraging your users to connect using peercast rather than directly to your server. Alternatively you can set up your own peercast node next to your regular server.
It's funny that this debate rages in a country that has seen severe problems with a severely outdated and erroneous voting infrastructure. Nobody has problems there with (proprietary) punch card machines but as soon as computers are involved everybody gets worried. Arguably it wasn't technology that failed during the last elections but the process after the election during which both parties spent several millions on campaigns trying to prove that they won rather than just recounting the votes (which was an option all along) or holding a state wide reelection (which even in third world countries is common practice in case of doubt).
I'm sure there is room for an open source voting system next to the many excellent commercial products available (which outside the US are widely being used and which tested in practice). Let the market decide. Let the government focus on certification rather than specific products. Voting machines (electronic and mechanical) should meet certain standards with respect to reliability, ease of use, accessibility, acceptable margin of error etc. Any standard in this area is better than none (which currently seems to be the case).
People trust their life to certified proprietary medical software, nasa launches billions worth of equipment using certified proprietary software, if you travel by car, you are using tons of certified proprietary embedded software. The keyword is certification. We trust this software because independent third parties have assessed that the software does what it advertises to do in a sufficiently reliable fashion.
Certification is currently uncommon in commercial software engineering. Not in the last place because most so called software engineers are not even qualified to tie their shoelaces properly. Any idiot who has read VB for dummies can claim to be a software engineer.
I'm specifically referring to overzealous linux sysadmins who tend to overconfigure their systems or to sysadmins who just go with the defaults of the redhat/mandrake firewall.
Real UNIX sysadmins are expensive and rare. You should see the monkeys that maintain our local network. If I want a virtual server added to our apache server it is more efficient for me to look up the documentation and take our sysadmin by the hand to guide him through the process (this actually happened). Most organizations have shitty sysadmins. Luckily ours never tried their hands on a firewall (that means security sucks by default and there are no restrictions on network usage:-).
Slashdot Mirror
I have no idea but this topic fits perfectly in the ever degenerating "ask slashdot" series. The type of questions asked under this topic can mostly be answered with "ask google" (or sometimes a simple "Duh!" is appropriate).
However, this particular "question" may form an interesting start for a debate about why micropayments don't work.
IMHO it is for the same reason that we had the dot com implosion: in order to work, micropayments require lots of transactions in order to generate a significant amount of revenue. Typically the critical amount of transactions never materializes.
Well they have a point that the combination HTML/javascript is cumbersome. At least they bother to address the issue whereas the rest of the world still struggles to acknowledge the issue.
I use dreamweaver occasionally and despite some annoying bugs this is a very nice development environment. People dismissing it as just another text editor are really not doing this product justice. I've spent quite a bit of time evaluating open source alternatives to dreamweaver but nothing I've seen so far even comes close. The key feature of dreamweave that I absolutely depend on is nested templates (define site layout/menus/etc. separately from the pages on the site). If anyone knows of an easy to use, oss alternative for that I'd be grateful.
Kazaa has one huge problem for me: large downloads invariably corrupt. That doesn't really matter for movies since it just means you have a few seconds of distorted sound/picture. However, if you're downloading things that need to be 100% intact like (cough) a linux iso (cough) it is very convenient if the p2p client bothers to verify that what it is downloading is the same as that should be downloaded.
Shareaza solves this problem, kazaa doesn't. However, the shareaza network is so small currently that there isn't much to download. With 3 million users at any time, the Kazaa network is unbeatable in that respect.
Well it is easy to point the finger at MS when even this open source safe haven called slashdot does not meet any w3c standard. If standards are so important, why can't this site adhere to them?
The reason GUI tools never work in linux is that they are an afterthought. Configuration always boils down to editing text file XYZ in some location which typically varies from distribution to distribution (something that often is worked areound using symbolic links!). There's a few problems with this approach: 1) each configuration file has its own syntax therefore automated editing (e.g. through a gui) needs to be custom designed (and consequently lacking/severely limited). 2) sysadmins typically have to learn how to configure a tool in addition to learning to understand what it does (consequently sysadmins tend to be expensive). 3) sysadmins have to be aware of distribution specifics in order to succesfully configure a system.
Most attempts at making linux user friendly work around these issues rather than fixing them. As a result, you can't go to the shop and buy software for linux. You can only buy software bundled with linux because that is the only way to get it working in a somewhat reliable fashion. It is also the reason many software developers currently do not have a linux version of their software. Getting software running on the zillion variations of linux is a support nightmare.
I was referring to the fact that generics in Java is nothing new. The proposal for inclusion of generics in Java has been around for quite some time. At the time the 1.4 spec was being finalized it was nearly ready. I've been aware for more than a year now that the intention was to include generics in Java 1.5.
I agree that probably Java's generics are a bit of a compromise between the Java object model, generics security and type safety. Arguably, C++ is more powerfull but less safe in that respect. Probably Beta does a better job at implementing generics. However, generics are not the solution to all problems. Essentially separation of concerns is not possible for all problem dimensions (which is what both OO and generics try to achieve).
Sounds like you don't get the point. The reason templates are useful is because it reduces the amount of code you need to write for common situations in a typesafe way. This makes your code more compact (duplicate code) and less error-prone (it's all type safe and you only write it once).
a d=6009 (lets hope /. gets this right). Lets hope all these goodies end up in 1.5.
BTW. generics are old news. The prototype compiler and the proposal have been available for at least a year. On the javalobby there's a link to some additional language extensions: http://www.javalobby.org/thread.jsp?forum=61&thre
Actually I would recommend against insisting on a particular format. Not everybody is a tex fan and I don't know anyone who ever worked with docbook. It's content that matters, not the format. You don't want to scare away interested authors by putting all sorts of constraints on the wordprocessor. IEEE and ACM for instance no longer do this, instead they have guidelines for formatting your document and templates for most popular wordprocessors are available.
I mostly work with framemaker and sometimes with ms word. It always pisses me off if TeX is required. It's not my problem that the editor likes TeX. He/she should be glad I'm willing to invest the time to (freely) provide content.
As for the functionality, there are two aspects: handling the subscriptions and handling the review process. Good solutions for both aspects exist. There's basically no need to reinvent the wheel.
Why can't kernel compiling linux users figure out how to use the control panel? It's not that hard. It was designed to be useable by idiots. It's not completely idiot proof, but then what software is?
Of course you can save the more essential stuff like bookmarks, cookies, mailfolders etc. by making a backup copy. This is how I upgraded from 1.0 RC1 all the way to 1.2.1 (via 1.0 RC2 1.0, 1.1a, 1.1b 1.1, 1.2b, 1.2). It would be nice if the mozilla setup would automatically migrate stuff that can be migrated rather than just dumping everything ontop of the existing setup and hoping for the best.
I installed 1.2 as soon as it was released. I appreciate the quick update but must say I didn't notice anything wrong with the pages I visited. Apparently, this bug mostly affected some advertisements on some sites (cnn amongst others). You'd almost consider it a feature because of this.
I used a gnu licensed set of python files called cyberchair (http://www.cyberchair.org/) to manage the submission and review process for a small conference last year. If you know python, it is relatively easy to set up. The version distributed by the author is somewhat out of date and has a number of undocumented bugs that you need to work around. However, I found that it was reasonably functional after some tinkering. Basically cyberchair saved me a lot of time since it takes care of collecting abstracts, author information etc. Also you can set it up to request reviewers to use a web form to submit their reviews, generate overviews for conference chair and reviewers, etc.
As for scalability, it is all pretty light weight and basically only requires apache and a recent version of python. It should be relatively easy to adapt this for a journal. After all the review process for journals is similar to that of conferences.
The guy who developed cyberchair also runs a small service company that can arrange the whole thing for you (including webhosting if needed) for a reasonable fee. This is actually the route I recommend unless you have too much spare time. For budgetary reasons, this was not feasible in my case unfortunately.
There's a few similar opensource packages for managing conferences that I can't remember the name of right now but I'm sure google will help you with that.
The benchmark becomes even more interesting once you realize that the java benchmark includes the overhead of starting up java and jit compiling the class file. It still is faster.
Rather than getting xenophobic about foreigners, you could also demand that they are treated the same as 'native' americans (not just the indians though, pun intended). That way foreigners compete on quality and not on salary.
Of course the whole reason foreigners are so popular is that it is a modern form of slavery. They work long hours for low wages. And you always have the whip of the immigration authorities to slap them with if they complain (being fired usually boils down to being put on the plane back home). Take away that 'advantage' and it suddenly becomes worthwhile to hire locals (provided there are well educated natives available, it's mostly the uneducated ones that get fired). Of course in the long term, IT shops will emigrate towards India. Labor cost is the driving cost in IT and its about 1/10th in e.g. India.
At least your question here suggests that they might be. Nothing worse than an overzealous, clueless sysadmin.
Silver bullets still do not exist. New technologies and methodologies are often hyped as such and naturally fail to live up to the hype. However, that does not mean they are useless.
These technologies and methodologies have allowed us to keep pace with Moore's law. The amount of software we develop today would simply be impossible using the state of the art in 1970. We routinely poor out millions of lines of code in mammoth projects that take sever hundreds or thousands of man-years to complete. The only reason we are able to do so is because of improvements in development methodologies and technology.
The (huge) research field that studies these technologies and approaches is called software engineering.
Here in the Netherlands, the left-wing green party yesterday also proposed embracing open source. In a quite extensive report on their website (http://www.groenlinks.nl/nieuws/4001428.html, in Dutch). They motivate the proposal quite well. There are a few minor details that they got wrong (most notably, Linus' last name is misspelled and the fact that a closed source format is used for the actual report) but overall the message is that closed source is bad and open source can be beneficial for both economical reasons and other reasons such as security, reliability and openness. Considering the report is written by a non technical person for a non technical audience, the effort should be applauded.
With the upcoming election in January, I hope this will be one of the election themes.
There's a lot of GUI environments on linux. Most are a work in progress and generally they don't work together very well.
Most of the applications I use on a daily basis have alternatives on linux. Sadly, most of the alternatives are of lesser quality in terms of usability, stability, features, etc. Actually there are a few apps I'd like to spend some time with. For example evolution seems like a decent mail client that has some interesting features.
Probably these are issues I could work around. I'm quite comfortable tweaking linux to suit my needs. Currently I'm fooling around with Debian in vmware.
What I can't work around is that most of the games I enjoy are windows only. In addition I have several pieces of hardware which are either not supported at all or with degraded performance/functionality/stability under linux. I didn't invest in an audigy + geforce ti 4200 just to listen to stereo sound and run a couple of outdated games. Not to mention my scanner, camera, wheel, usb mouse. Probably, with a lot of work, I can make most of these things work to some extent. However, the fact remains that they were designed for windows and not for linux.
I browse with mozilla and I mail with outlook xp. Outlook is currently one of the best mail clients (in terms of features) and the mozilla mail client still needs a lot of work to even come close. Security is an issue with outlook only if you can't find your way to the preferences dialog to adjust security to the appropriate settings.
Spamnet is catching most of the spam on my machine now. I installed it two months ago and it has caught up of 95 percent of the spam I receive. More importantly, it hasn't miscategorized a single message.
The problem is that you need all these high level abstractions to reduce the workload of creating large systems. There's just no way you could have those VB monkeys be productive in C and there's just no way you are going to replace them with competent C programmers. Besides, competent programmers are more productive using high abstraction level tools.
BSD is just a kernel + a small toolset. As soon as you start running all the regular stuff on top of it performance is comparable to a full blown linux/mac os X/windows desktop. Proof: mac os X, remove the none BSD stuff and see what's left: no ui, no friendly tools, no easy access to all connected devices.
There's two ways to adopt new technology: you can throw away everything you have and move to new technology in a revolutionary fashion. This is highly disruptive to an organization. Excessive training cost, numerous problems, etc. are more than likely. In short, you don't want revolutions in your IT infrastructure. The second option is to evolve your system. Replace things that would otherwise require new investments in licenses, things that you are not happy with etc.
If your exchange server is working fine, don't replace it unless you would otherwise upgrade to a new version. If your IIS webserver is adequate for serving the handful of word documents in your organization, don't fix it unless you need certain functionality offered by apache. Linux has a lower TCO for stuff like filesharing and printersharing. However, you have already invested in your existing infrastructure and if its working fine there's no reason to make additional investments in a functional equivalent.
Peercast also supports firewalled nodes. Streams can both be pushed and pulled. The only problem is that two firewalled nodes can't connect to each other directly. A firewalled node can push streams to none firewalled nodes and it can pull streams from none firewalled nodes.
The protocol is the same as is used by shoutcast. All that peercast does is relay the stream. Typically clients such as winamp do some buffering.
Essentially, when streaming, you tell shoutcast/oddcast or whatever encoder you are using to stream the music to your peercast client rather than a regular shoutcast server. Peercast then inserts some meta information into the peercast network so that other clients can find your stream. Other peercast clients can connect to your client to receive the stream. Peercast treats received streams exactly the same as streams that come from oddcast so anyone listening to your stream can also relay it further on. Theoretically, this allows a modem user to stream to a nearly endless amount of users by uploading just 1 stream. In practice it is better to allow for two or three streams so you have somewhat better reliability. Of course you need bandwidth in order to be able to relay. Low bandwidth ogg streams propagate better through the network than 128kbit mp3 streams. Ogg streams of around 45 kbit seem to be popular among peercast users.
Recently the ability to relay regular shoutcast streams was added so if you have a regular shoutcast server, you may save some bandwidth by encouraging your users to connect using peercast rather than directly to your server. Alternatively you can set up your own peercast node next to your regular server.
It's funny that this debate rages in a country that has seen severe problems with a severely outdated and erroneous voting infrastructure. Nobody has problems there with (proprietary) punch card machines but as soon as computers are involved everybody gets worried. Arguably it wasn't technology that failed during the last elections but the process after the election during which both parties spent several millions on campaigns trying to prove that they won rather than just recounting the votes (which was an option all along) or holding a state wide reelection (which even in third world countries is common practice in case of doubt).
I'm sure there is room for an open source voting system next to the many excellent commercial products available (which outside the US are widely being used and which tested in practice). Let the market decide. Let the government focus on certification rather than specific products. Voting machines (electronic and mechanical) should meet certain standards with respect to reliability, ease of use, accessibility, acceptable margin of error etc. Any standard in this area is better than none (which currently seems to be the case).
People trust their life to certified proprietary medical software, nasa launches billions worth of equipment using certified proprietary software, if you travel by car, you are using tons of certified proprietary embedded software. The keyword is certification. We trust this software because independent third parties have assessed that the software does what it advertises to do in a sufficiently reliable fashion.
Certification is currently uncommon in commercial software engineering. Not in the last place because most so called software engineers are not even qualified to tie their shoelaces properly. Any idiot who has read VB for dummies can claim to be a software engineer.
I'm specifically referring to overzealous linux sysadmins who tend to overconfigure their systems or to sysadmins who just go with the defaults of the redhat/mandrake firewall.
:-).
Real UNIX sysadmins are expensive and rare. You should see the monkeys that maintain our local network. If I want a virtual server added to our apache server it is more efficient for me to look up the documentation and take our sysadmin by the hand to guide him through the process (this actually happened). Most organizations have shitty sysadmins. Luckily ours never tried their hands on a firewall (that means security sucks by default and there are no restrictions on network usage