I can't stand misguided attempts at making an "interactive web experience." Why should I have to install a plugin in my browser just to visit a website? Why should I have to deal with javascript errors because someone decided that having a list of 12 links on the left side of a web page was too "ugly," and tried to make a mouseover effect? Why should I be required to use a bloated browser? I have seen very few websites that had legitimate uses for applets, and even there, the applets are discrete, small, and designed to do one thing well.
Youtube most certainly does have codecs. They are just easier to install, and generally invisible to the user, because Youtube is designed better. There is no reason that you couldn't do what Youtube does without using Flash.
I would argue that adding more complexity to a system, without actually adding more functionality to it, adds more points of failure and reduces the overall quality of the system. Sometimes, people manage to use Flash to accomplish something useful, but when a website is done completely with Flash, it just adds complexity without adding much real value.
Most websites that use Flash have no reason to use it. Why encourage people to rewrite perfectly functional, HTML/CGI/Javascript websites with Flash, wasting their time, my time, their money, and my bandwidth, by bending over and taking it?
If you are fortunate enough to live in a big city, your public library will probably grant you access to a lot of scientific literature for free. The New York City public library carries various journals, although there is one specific branch where those may be accessed.
Really though, I agree with you. The general public, beyond big cities, should have unfettered access to scientific literature.
I have disabled flash, because the benefits of having it enabled were near zero and flash ads consume a lot of CPU time.
I have blacklisted most of the common advertising services (doubleclick.net, etc.), and so if slashdot is getting its ads delivered by one or more of those services, the ads are not delivered to me.
In all seriousness, LaTeX/troff really do work better for writing HUGE papers with complex layouts and formulas. I've seen OOo and MSO consume tremendous amounts of memory trying to deal with lengthy engineering papers, whereas LaTeX/troff take a few minutes to CPU time to compile the paper and spit out DVI or PS.
Admittedly, typesetting software is annoying when you are first learning it, and you are sitting there wondering why you are inserting markup all over the place when all you wanted was a table, but after the initial hurdle it really isn't so bad. I'll admit, it really isn't for everyone, especially not for people who aren't comfortable entering commands into a shell, but for people who are, I'd recommend at least checking out LaTeX or troff.
I've used OpenOffice.org throughout college, for various papers, including papers with reasonably complex layouts and levels of embedded objects, and I only started having problems when my documents became excessively long. MS Office didn't handle those documents any better, and so I've just migrated away from WYSIWYG editors all together, which I've discovered is pretty common in my field.
Frankly, any generalization is inherently misguided. OOo is extremely useful for certain tasks, and I have seen people do things with OOo that they couldn't see any way to do with MSO. OOo has some problems, yes, but to claim that there is absolutely no use for it is just as bad as claiming that there are no problems with it at all.
The main problem with end users running Linux is that we have been encouraging them to just insert an installer disk, install it, and "everything will work." This is a sentiment we have been promoting among people who wouldn't be able to do that with a Windows installer disk. Worse than that, we've been acting like setting up a multiboot system is a simple task, when it really isn't. Partition management? Swap? These are terms that are completely confusing to most people, and yet they encounter them as soon as they try installing a Linux distro. Configuring drivers to work is a nightmare, especially on distros committed to shipping only free software; worse is when someone discovers that they don't have MP3 support, and don't really care about the legal issues surrounding it.
That being said, it is also fair to point out that most people have no clue about the various features a typing X11 UI has over Windows, or the various ways the system can be accessed (remotely, uucp, etc.). There is no perceived advantage, so why would they want to have it installed, or worse, go through the process of installing it?
Your comment is typical of people from my generation: we grew up with webmail, and so we can't imagine anything else:
"Oh yes, it's a fine idea in theory, but in Gmail, if I need to find serial numbes (something I've need to do several times in the last year) I can just search for the game name voila!...To replicate that functionality I would have to backup my entire email pretty much every time I got any, which is completely impractical. Not to mention tedious."
Allow me to introduce you to the brave old world of email clients; I personally use Kontact/KMail, but as you play video games I will infer that you are a Windows user and therefore will want Thunderbird. Email clients started as programs that read your local mail file, but now they also can use POP3 or IMAP4 to check your email on a mail server. I have KMail set up to automatically check my email every 5 minutes for 4 different accounts; searching through messages is built right in. Many people I know use Thunderbird, and they have similar setups through that. Searching is built right in, and in KMail I can even search using a regular expression, which last I checked Google's web interface lacked.
Email clients are usually integrated on some level with calendar/todo programs, such as Sunbird. This is useful if you want to schedule an event with other people and have them automatically emailed of changes to that schedule, or if you receive an email and want to create a todo from it. I believe Google also has this functionality, although it seems more developed in Kontact or Evolution (I guess that one really depends on your perspective).
Really, this is symptomatic of a larger problem: most people were never aware of what their PCs were capable of, especially people who joined the Internet after 1995, and so most people think that they must use a website to accomplish certain tasks. Another example is my university's VPN -- most people had no idea that they could configure their computer to connect to our Cisco VPN concentrator, let alone the details of how to actually do it, and so it was underused. Then the university paid for web based software that used a Java applet to set up a VPN tunnel, and suddenly everyone was using it, and nobody knew or cared that our Cisco VPN (which was a lot more robust) was offline. Now most people walk around thinking that the only way to accomplish such a task is through a website of some sort.
RMS is on the ball with this one, and he is not the first person to say any of it. Web apps themselves are inherently not-free (as in freedom), and this issue came up at the last FUDCon. Sure, one can release the source code of their web app, and allow people to run it on their servers, but unless your willing to allow those third party servers to exchange data with yours, part of the point of using a web app at all will be lost. Free software (as in freedom) is important for all the reasons Stallman pointed out, and with web apps, that becomes very clear: websites routinely make unpopular changes and force their users to deal with it. I'm sure anyone here can think of an example or two.
True, although it is fair to point out that OOo can also connect to shared databases or database files for its bibliography sources. Someone else pointed out that EndNote can automatically import citations from other papers, including PDFs of those papers, which is a pretty useful feature, although I have to wonder about someone who wouldn't have already looked up the cited sources and gotten the bibliography entries from those. Ultimately, I guess it depends on how your write your papers...
It is more a question of whether or not citizens will be able to access government data in a meaningful way. If the government wants to standardize its data, it can, assuming it contracts with a company that actually knows what it is doing (this is the real hitch). Government employees need to be able to continue doing what they normally do, and have the standardization happen automatically -- such as a MS.doc to ODF converter that silently makes the conversion whenever a file is saved, or another tool that automatically indexes files as they are saved. Such things already exist, it is just a matter of implementing on the scale of the government.
We are not talking about the government sharing data on individual citizens or on military secrets. We are talking about things involving government spending, contracts, loans, grants, etc. Things that citizens should have access to, but have trouble organizing.
I must wonder, though, why bother with anything more than the citation database? OpenOffice.org will format your bibliography directly from the citation database, in the same manner that it can build other tables and indexes from database sources. Why bother with a "citation manager" for this seemingly simple task?
Virtually all citation management software is difficult to deal with. To be honest, I find it more intuitive to just enter the citations manually into the OpenOffice.org bibliography database than to use any of the tools currently out there. Some people swear by EndNote, but I can't deal with it.
The point is to confuse a jury. Think about their case: they are claiming that an EULA which prevents reverse engineering the software was violated, but the case centers around reverse engineering their file format. The only possible way to win such a case is to confuse the people deciding it, and confusing a judge isn't as easy as confusing a jury (a judge has a certain level of education, whereas a jury isn't required to have any education at all). I can imagine their case already: keep shouting present the EULA clause about reverse engineering the software, then keep shouting reverse engineering whenever the issue of the file format comes up; also, get the GMU people to say that they reverse engineered the file format (which is legal but a jury probably would have trouble making a distinction).
Fedora 9 has gone a long way toward that. Out of the box, with the default GNOME environment (not my thing, but whatever), it is a pretty clean, one browser, one mail client, etc.
What's really bad is when someone with that mindset is told, "Yes, it is free to acquire, but the long-term costs are higher, here are pretty charts and graphs to look at," and they swallow it hook, line and sinker.
Because if the primary reason for using open source software is to save money, it is easy for a proprietary vendor to try and gain an edge in a particular market by lowering or eliminating their acquisition price and focusing on service contracts. There is nothing inherent to open source that guarantees that it will cost less to buy than proprietary code, nor is there anything inherent to open source that guarantees that the long term costs will be less than with proprietary code. Some of the most stable, reliable software platforms on Earth are proprietary -- z/VM, VMS, etc.
When the primary reason is to remain free from vendor lock-in, or to have the freedom to modify the code as needed, or the freedom to redistribute the code as needed, then it becomes much harder for proprietary vendors to compete.
Slashdot Mirror
Most websites that use Flash have no reason to use it. Why encourage people to rewrite perfectly functional, HTML/CGI/Javascript websites with Flash, wasting their time, my time, their money, and my bandwidth, by bending over and taking it?
When did Ubuntu become a certified Unix?
If you are fortunate enough to live in a big city, your public library will probably grant you access to a lot of scientific literature for free. The New York City public library carries various journals, although there is one specific branch where those may be accessed.
Really though, I agree with you. The general public, beyond big cities, should have unfettered access to scientific literature.
What's an ad?
Real authors use troff?
In all seriousness, LaTeX/troff really do work better for writing HUGE papers with complex layouts and formulas. I've seen OOo and MSO consume tremendous amounts of memory trying to deal with lengthy engineering papers, whereas LaTeX/troff take a few minutes to CPU time to compile the paper and spit out DVI or PS.
Admittedly, typesetting software is annoying when you are first learning it, and you are sitting there wondering why you are inserting markup all over the place when all you wanted was a table, but after the initial hurdle it really isn't so bad. I'll admit, it really isn't for everyone, especially not for people who aren't comfortable entering commands into a shell, but for people who are, I'd recommend at least checking out LaTeX or troff.
I've used OpenOffice.org throughout college, for various papers, including papers with reasonably complex layouts and levels of embedded objects, and I only started having problems when my documents became excessively long. MS Office didn't handle those documents any better, and so I've just migrated away from WYSIWYG editors all together, which I've discovered is pretty common in my field.
Frankly, any generalization is inherently misguided. OOo is extremely useful for certain tasks, and I have seen people do things with OOo that they couldn't see any way to do with MSO. OOo has some problems, yes, but to claim that there is absolutely no use for it is just as bad as claiming that there are no problems with it at all.
Are you kidding? My first serious girlfriend did math with me (cue up "in bed" joke), far beyond what we were being taught in school.
Except that Apple will undoubtedly try to enforce this patent. Oh well, I'm a Fluxbox user, so I suppose I'm not really affected.
The main problem with end users running Linux is that we have been encouraging them to just insert an installer disk, install it, and "everything will work." This is a sentiment we have been promoting among people who wouldn't be able to do that with a Windows installer disk. Worse than that, we've been acting like setting up a multiboot system is a simple task, when it really isn't. Partition management? Swap? These are terms that are completely confusing to most people, and yet they encounter them as soon as they try installing a Linux distro. Configuring drivers to work is a nightmare, especially on distros committed to shipping only free software; worse is when someone discovers that they don't have MP3 support, and don't really care about the legal issues surrounding it.
That being said, it is also fair to point out that most people have no clue about the various features a typing X11 UI has over Windows, or the various ways the system can be accessed (remotely, uucp, etc.). There is no perceived advantage, so why would they want to have it installed, or worse, go through the process of installing it?
Please, DOS 5 is where it's at. No PC can be without it.
Your comment is typical of people from my generation: we grew up with webmail, and so we can't imagine anything else:
"Oh yes, it's a fine idea in theory, but in Gmail, if I need to find serial numbes (something I've need to do several times in the last year) I can just search for the game name voila!...To replicate that functionality I would have to backup my entire email pretty much every time I got any, which is completely impractical. Not to mention tedious."
Allow me to introduce you to the brave old world of email clients; I personally use Kontact/KMail, but as you play video games I will infer that you are a Windows user and therefore will want Thunderbird. Email clients started as programs that read your local mail file, but now they also can use POP3 or IMAP4 to check your email on a mail server. I have KMail set up to automatically check my email every 5 minutes for 4 different accounts; searching through messages is built right in. Many people I know use Thunderbird, and they have similar setups through that. Searching is built right in, and in KMail I can even search using a regular expression, which last I checked Google's web interface lacked.
Email clients are usually integrated on some level with calendar/todo programs, such as Sunbird. This is useful if you want to schedule an event with other people and have them automatically emailed of changes to that schedule, or if you receive an email and want to create a todo from it. I believe Google also has this functionality, although it seems more developed in Kontact or Evolution (I guess that one really depends on your perspective).
Really, this is symptomatic of a larger problem: most people were never aware of what their PCs were capable of, especially people who joined the Internet after 1995, and so most people think that they must use a website to accomplish certain tasks. Another example is my university's VPN -- most people had no idea that they could configure their computer to connect to our Cisco VPN concentrator, let alone the details of how to actually do it, and so it was underused. Then the university paid for web based software that used a Java applet to set up a VPN tunnel, and suddenly everyone was using it, and nobody knew or cared that our Cisco VPN (which was a lot more robust) was offline. Now most people walk around thinking that the only way to accomplish such a task is through a website of some sort.
RMS is on the ball with this one, and he is not the first person to say any of it. Web apps themselves are inherently not-free (as in freedom), and this issue came up at the last FUDCon. Sure, one can release the source code of their web app, and allow people to run it on their servers, but unless your willing to allow those third party servers to exchange data with yours, part of the point of using a web app at all will be lost. Free software (as in freedom) is important for all the reasons Stallman pointed out, and with web apps, that becomes very clear: websites routinely make unpopular changes and force their users to deal with it. I'm sure anyone here can think of an example or two.
True, although it is fair to point out that OOo can also connect to shared databases or database files for its bibliography sources. Someone else pointed out that EndNote can automatically import citations from other papers, including PDFs of those papers, which is a pretty useful feature, although I have to wonder about someone who wouldn't have already looked up the cited sources and gotten the bibliography entries from those. Ultimately, I guess it depends on how your write your papers...
It is more a question of whether or not citizens will be able to access government data in a meaningful way. If the government wants to standardize its data, it can, assuming it contracts with a company that actually knows what it is doing (this is the real hitch). Government employees need to be able to continue doing what they normally do, and have the standardization happen automatically -- such as a MS .doc to ODF converter that silently makes the conversion whenever a file is saved, or another tool that automatically indexes files as they are saved. Such things already exist, it is just a matter of implementing on the scale of the government.
We are not talking about the government sharing data on individual citizens or on military secrets. We are talking about things involving government spending, contracts, loans, grants, etc. Things that citizens should have access to, but have trouble organizing.
Seriously, it took a team of researchers to figure this out?
"You can easily give your bibliography to others who don't use BibTeX"
Uhm...assuming they are using Zotero/EndNote/whatever? How is that better?
I must wonder, though, why bother with anything more than the citation database? OpenOffice.org will format your bibliography directly from the citation database, in the same manner that it can build other tables and indexes from database sources. Why bother with a "citation manager" for this seemingly simple task?
Virtually all citation management software is difficult to deal with. To be honest, I find it more intuitive to just enter the citations manually into the OpenOffice.org bibliography database than to use any of the tools currently out there. Some people swear by EndNote, but I can't deal with it.
The point is to confuse a jury. Think about their case: they are claiming that an EULA which prevents reverse engineering the software was violated, but the case centers around reverse engineering their file format. The only possible way to win such a case is to confuse the people deciding it, and confusing a judge isn't as easy as confusing a jury (a judge has a certain level of education, whereas a jury isn't required to have any education at all). I can imagine their case already: keep shouting present the EULA clause about reverse engineering the software, then keep shouting reverse engineering whenever the issue of the file format comes up; also, get the GMU people to say that they reverse engineered the file format (which is legal but a jury probably would have trouble making a distinction).
Fedora 9 has gone a long way toward that. Out of the box, with the default GNOME environment (not my thing, but whatever), it is a pretty clean, one browser, one mail client, etc.
Vista SP2 is going to remove my email (Thunderbird) and photo apps (Picasa)?
What's really bad is when someone with that mindset is told, "Yes, it is free to acquire, but the long-term costs are higher, here are pretty charts and graphs to look at," and they swallow it hook, line and sinker.
Because if the primary reason for using open source software is to save money, it is easy for a proprietary vendor to try and gain an edge in a particular market by lowering or eliminating their acquisition price and focusing on service contracts. There is nothing inherent to open source that guarantees that it will cost less to buy than proprietary code, nor is there anything inherent to open source that guarantees that the long term costs will be less than with proprietary code. Some of the most stable, reliable software platforms on Earth are proprietary -- z/VM, VMS, etc.
When the primary reason is to remain free from vendor lock-in, or to have the freedom to modify the code as needed, or the freedom to redistribute the code as needed, then it becomes much harder for proprietary vendors to compete.