That still leaves you vulnerable to viruses that exploit holes in the service implementation on that port. Most of the (non-email) viruses I've seen recently were spread from the inside of a company. One system -- a laptop that someone had hooked up at home against company policy -- was infected, then spread to all the other internal systems. We don't have (many) firewalls segregating our internal network, so the virus was free to spread.
There are 3 ways to prevent this. In order of preference: turning off unused ports; "personal" firewalls running on every system; internal segmentation with ACLs/firewalls. Ideally, all 3. None of these seem to be common practice on Windows systems.:(
I haven't seen much mention of Lotus SmartSuite in this conversation. IBM owns Lotus. Lotus SmartSuite was a pretty decent competitor to Microsoft Office up until a couple years ago. (Millenium Edition was OK, and 97 was pretty good.) The code should still be pretty decent. At least there should be some good stuff we could rip out of it and use in OpenOffice, KOffice, Gnome Office. So why doesn't IBM open the code for SmartSuite? Perhaps there are parts that they can't free, but there's got to be a lot of good code there even after removing encumbered code.
So my big question is, if IBM is talking about Linux office suites, why haven't they opened or ported the one they own?
I've given some presentations on Open Source. More to a lower-level management and techie crowd. I do include a full slide on why / when NOT to use Open Source:
If Open Source solutions do not exist
Or they are not mature enough
If maintenance costs exceed license costs
Need someone to blame for problems
It's scary to take full responsibility for things
It costs money to change
Don't replace things just to replace them
Fear, Uncertainty, Doubt (FUD)
Summary: You need to figure out if it's right for your situation
You wouldn't believe how much more credibility that gives me. They're more willing to trust me, because I was willing to admit to the down-sides, presenting both sides of the argument. Another advantage I have is that I have experience with a variety of systems, including Microsoft, Novell, and Linux. I've always maintained that pushing Linux / Open Source into places where it isn't ready does more harm than good for the Free Software movement.
They can already easily make your professional life a nightmare. First, they could randomly decide to sue you for some arbitrary reason. With as much money as they have, they could find lawyers who would bog you down in a legal quagmire, taking years and hundreds of thousands of dollars in legals expenses. Second, they could create a collection of APIs called Win32, and require you to use it if you want a programming job. Third, they could create a collection of APIs called.NET Framework, give it the same name as a whole bunch of their other products, deprecate their previous Win32 APIs, and charge you a bunch of money for the luxury. Fourth, they could ceritify you as an MCSE and make you work on Windows servers all day.
Rivendell aims to be a complete radio broadcast automation solution, with facilities for the acquisition, management, scheduling and playout of audio content.
We actually had someone come talk to our UNIX Users Group about that very subject this past December. The speaker was Ben Oberkfell of KWUR. He's also president of the Washington University ACM chapter. He spoke about how they've set up automated DJing, play-lists, streaming, and just about everything else using Linux and some Open Source tools he built. It doesn't appear that we've posted the presentation on-line, but if you email him, I'm sure he'd be happy to send you the presentation and help you get your station going using his software. Finding his email from the above links is left as an exercise for the reader.
If the majority of your applications are for Linux, but you need Windows for a few, you should use a solution such as Citrix. Other similar solutions include Windows RTP, VNC, and VMWare. Clients for all of these will work on Linux.
Another point I'd make is that you can run many Open Source applications on Windows. Open Office will definitely run on Windows, and saves you from the Microsoft Office licensing beast.
The meta-object compiler (moc) used to turn me off as well. However, since you just add that as an additional step in the Makefile, it really isn't that big of a deal. There are still a few issues with the non-standard syntax (syntax-highlighting, lint, refactoring tools, etc.) but they aren't all that terrible.
The bigger issue to me is language choice. GTK has bindings for almost every commonly used language. Qt doesn't support much besides C++. I prefer the extra flexibility of GTK for that reason.
I do note that Open Source projects using KDE/Qt seem to get to a usable stage much quicker than any other toolkit. My guess is that that means that the Qt framework makes writing programs faster. Now, if only it had a libglade-type RAD system to separate GUI design from implementation, it'd kick ass. And no, Qt Designer doesn't cut it, since you have to lay out your widgets more statically, and then recompile.
I'm wondering when the CSS 3 standard will be completed. It started in 1999, I think. Four years is pretty long, considering how quickly the first 2 versions came out.
I seem to recall Fortran 90 finally coming out in maybe 1993 or so. I think they were originally going to call it Fortran 88.
I guess there are several problems: 1) standards are designed by committee; 2) unlike the first revision of a standard, more people are involved, each having their own agenda; 3) they want to test out the features in actual implementations before ratifying the standard; 4) if the features of the standard are overly difficult to implement (see #2) it'll take longer to work out the kinks.
Scientist 1: We can mine this Helium-3, and it'll produce almost infinite power. Scientist 2: Really? Scientist 1: Yeah, it'll give us 1000 times the amount of power of coal, per pound. Scientist 2: So what's the down-side? Scientist 1: It costs 1 million times as much to produce. Scientist 2: Wow, that's a great idea!
Significant interplanetary spaceflight? You've got to be kidding! Sending out a few small probes a year is not terribly significant. Especially since in this context, they'd have to stop at the moon to pick up the fuel.
The problem is not that it's too hard to explain. The problem is that it's too simple: "we'd like to lease computers and/or compute time to you".
There are several difficulties though: 1) companies remember how expensive mainframes are/were, and they're leased in the same pay-per-use manner; 2) companies are comfortable owning computers, even with short amortization schedules; 3) companies may not be (and should not be) comfortable with their information being stored on someone else's computer; 4) it's more profitable for the computer companies to sell computer services this way, which makes companies think that it's going to be more costly in the end.
So the computer companies are trying to sell it as something else. They're failing miserably. Maybe if they tried selling it as what it really is, and showing the actual savings by sharing resource, the idea would be more profitable.
Yes, but I don't know how easy it would be to gather that info from within a netfilter module. I know we can get PID and executable name, since netstat can do that from user-space. I suppose we could get the other info out of the kernel process table. To do the hash, we'd have to get the executable's path (or maybe process table includes the inode/handle) and run the hash on that file.
Seems like a lot more work than I had initially envisioned. As a first cut, I think just filtering by name and/or PID would be a good proof of concept.
I've often thought of implementing per-process packet filtering myself. I'm pretty sure it could be implemented in netfilter (iptables), but it doesn't appear that anyone has written such a netfilter module. Basically, for packets originating from the local host, the module would take a look at the local port of the connection and run 'netstat -nap' to see what program is connected to that port. One of the big problems is that you'd just be tying it to a PID or program name, not a particular executable. In other words, if you allowed "mozilla", an attacker could write his own executable called "mozilla" and it'd still be allowed through. I'm not sure if there's any good way around this.
I think are is typically just a short-hand for most are. I suppose that there's a danger in using such a short-cut, allowing you to make the mistake of starting to think that the generalization really does apply to all. But when I say something like "black men can dance better than white men" I mean that black men tend to be better at dancing. To me, thinking that I meant that every black man can dance better than any white man would be a ludicrous interpretation. Linking in some genetic cause for such is even more ridiculous.
Hah! When I Googled, everything pointed back to people quoting you! The small "I" after the period was the give-away. Thanks for pointing me in the right direction. I found plenty of references to the Margaret Mead versions. Hard to tell exactly which one is the exact wording she used though; there are plenty of variants out there.
BTW, I added you as a Friend, since I liked a lot of the posts I read of yours. (I didn't even realize they were all yours at first.)
Hmm, that's a good point about religious beliefs changing, but they definitely change at a much slower pace.
I still have to disagree about scientific "laws" being assumed to be true. While we act under the assumption that they are "true enough" a good scientist should always remember that all "laws" can be proven (usually only slightly) untrue.
First of all, "race" in the modern US is more of a self-selection than most of us are aware. Most "black" people are probably 30-70% genetically European. Any rational classification system would place anyone with over 50% white genes as a "white" person, yet most of these people decide (possibly without realizing, and most likely their parents choose for them; I suppose that you could say that society often dictates as well) which sub-culture to "belong" to. You are right in stating that social pressures have a stronger influence on racial differences than anything biological or genetic.
But there are some biological differences. In addition to those mentioned in another reply to yours, there are different frequencies of diseases among the "races". For instance, blacks have a much higher incidence rate of sickle-cell anemia. These are of course mostly due to genetic heridity, although there are also instances of disease spread among members of the sub-culture.
So admittedly, most racial differences are due to the sub-culture in which the person was raised. But that does not make those differences any less real. Black men tend to be better at basketball than white men, because they grew up playing it more. Saying that they're better doesn't mean that they're genetically pre-disposed to be better though.
I think you may misunderstand science. (I'm of the belief that Science is our national religion.) Science never proves things to be true. It can only prove things to be wrong. Any scientific theory is based on assumptions, and we "believe" those theories until they are proven to be untrue. Our belief in the theories is based on a proponderance of evidence, but there's no way to prove anything 100%, so scientists are always refining the current theories.
I think the main difference between science and (traditional) religion is that scientists are willing to eventually give up their current beliefs if the evidence against it becomes overwhelming.
I agree -- this is more about outright theft of the CSS and look and feel copyrighted material than freedom of speech or parody. Besides, one test of parody is whether the parody can be confused with the original. To tell you the truth, I didn't find the "parody" site to be at funny. It's more of a counter-point than parody. A political opinion is protected, but not when it is a copy of another (opposing) political opinion.
Slashdot Mirror
Morphon is a good WYSIWYG editor for DocBook and other XML dialects. It comes with DocBook stylesheets built in. Morphon is now free (as in beer).
That still leaves you vulnerable to viruses that exploit holes in the service implementation on that port. Most of the (non-email) viruses I've seen recently were spread from the inside of a company. One system -- a laptop that someone had hooked up at home against company policy -- was infected, then spread to all the other internal systems. We don't have (many) firewalls segregating our internal network, so the virus was free to spread.
:(
There are 3 ways to prevent this. In order of preference: turning off unused ports; "personal" firewalls running on every system; internal segmentation with ACLs/firewalls. Ideally, all 3. None of these seem to be common practice on Windows systems.
And of course, keep your patches up to date.
Famous Hot Weiner? Sounds like a Freudian subliminal message!
I haven't seen much mention of Lotus SmartSuite in this conversation. IBM owns Lotus. Lotus SmartSuite was a pretty decent competitor to Microsoft Office up until a couple years ago. (Millenium Edition was OK, and 97 was pretty good.) The code should still be pretty decent. At least there should be some good stuff we could rip out of it and use in OpenOffice, KOffice, Gnome Office. So why doesn't IBM open the code for SmartSuite? Perhaps there are parts that they can't free, but there's got to be a lot of good code there even after removing encumbered code.
So my big question is, if IBM is talking about Linux office suites, why haven't they opened or ported the one they own?
- If Open Source solutions do not exist
- Or they are not mature enough
- If maintenance costs exceed license costs
- Need someone to blame for problems
- It's scary to take full responsibility for things
- It costs money to change
- Don't replace things just to replace them
- Fear, Uncertainty, Doubt (FUD)
- Summary: You need to figure out if it's right for your situation
You wouldn't believe how much more credibility that gives me. They're more willing to trust me, because I was willing to admit to the down-sides, presenting both sides of the argument. Another advantage I have is that I have experience with a variety of systems, including Microsoft, Novell, and Linux. I've always maintained that pushing Linux / Open Source into places where it isn't ready does more harm than good for the Free Software movement.I would tend to agree with that.
They can already easily make your professional life a nightmare. First, they could randomly decide to sue you for some arbitrary reason. With as much money as they have, they could find lawyers who would bog you down in a legal quagmire, taking years and hundreds of thousands of dollars in legals expenses. Second, they could create a collection of APIs called Win32, and require you to use it if you want a programming job. Third, they could create a collection of APIs called .NET Framework, give it the same name as a whole bunch of their other products, deprecate their previous Win32 APIs, and charge you a bunch of money for the luxury. Fourth, they could ceritify you as an MCSE and make you work on Windows servers all day.
We actually had someone come talk to our UNIX Users Group about that very subject this past December. The speaker was Ben Oberkfell of KWUR. He's also president of the Washington University ACM chapter. He spoke about how they've set up automated DJing, play-lists, streaming, and just about everything else using Linux and some Open Source tools he built. It doesn't appear that we've posted the presentation on-line, but if you email him, I'm sure he'd be happy to send you the presentation and help you get your station going using his software. Finding his email from the above links is left as an exercise for the reader.
If the majority of your applications are for Linux, but you need Windows for a few, you should use a solution such as Citrix. Other similar solutions include Windows RTP, VNC, and VMWare. Clients for all of these will work on Linux.
Another point I'd make is that you can run many Open Source applications on Windows. Open Office will definitely run on Windows, and saves you from the Microsoft Office licensing beast.
Interesting -- they used C to create a thunk layer into C++ code instead of the other way around. Are there any studies of how well that works?
The meta-object compiler (moc) used to turn me off as well. However, since you just add that as an additional step in the Makefile, it really isn't that big of a deal. There are still a few issues with the non-standard syntax (syntax-highlighting, lint, refactoring tools, etc.) but they aren't all that terrible.
The bigger issue to me is language choice. GTK has bindings for almost every commonly used language. Qt doesn't support much besides C++. I prefer the extra flexibility of GTK for that reason.
I do note that Open Source projects using KDE/Qt seem to get to a usable stage much quicker than any other toolkit. My guess is that that means that the Qt framework makes writing programs faster. Now, if only it had a libglade-type RAD system to separate GUI design from implementation, it'd kick ass. And no, Qt Designer doesn't cut it, since you have to lay out your widgets more statically, and then recompile.
I'm wondering when the CSS 3 standard will be completed. It started in 1999, I think. Four years is pretty long, considering how quickly the first 2 versions came out.
I seem to recall Fortran 90 finally coming out in maybe 1993 or so. I think they were originally going to call it Fortran 88.
I guess there are several problems: 1) standards are designed by committee; 2) unlike the first revision of a standard, more people are involved, each having their own agenda; 3) they want to test out the features in actual implementations before ratifying the standard; 4) if the features of the standard are overly difficult to implement (see #2) it'll take longer to work out the kinks.
Scientist 1: We can mine this Helium-3, and it'll produce almost infinite power.
Scientist 2: Really?
Scientist 1: Yeah, it'll give us 1000 times the amount of power of coal, per pound.
Scientist 2: So what's the down-side?
Scientist 1: It costs 1 million times as much to produce.
Scientist 2: Wow, that's a great idea!
Significant interplanetary spaceflight? You've got to be kidding! Sending out a few small probes a year is not terribly significant. Especially since in this context, they'd have to stop at the moon to pick up the fuel.
The problem is not that it's too hard to explain. The problem is that it's too simple: "we'd like to lease computers and/or compute time to you".
There are several difficulties though: 1) companies remember how expensive mainframes are/were, and they're leased in the same pay-per-use manner; 2) companies are comfortable owning computers, even with short amortization schedules; 3) companies may not be (and should not be) comfortable with their information being stored on someone else's computer; 4) it's more profitable for the computer companies to sell computer services this way, which makes companies think that it's going to be more costly in the end.
So the computer companies are trying to sell it as something else. They're failing miserably. Maybe if they tried selling it as what it really is, and showing the actual savings by sharing resource, the idea would be more profitable.
Yes, but I don't know how easy it would be to gather that info from within a netfilter module. I know we can get PID and executable name, since netstat can do that from user-space. I suppose we could get the other info out of the kernel process table. To do the hash, we'd have to get the executable's path (or maybe process table includes the inode/handle) and run the hash on that file.
Seems like a lot more work than I had initially envisioned. As a first cut, I think just filtering by name and/or PID would be a good proof of concept.
I've often thought of implementing per-process packet filtering myself. I'm pretty sure it could be implemented in netfilter (iptables), but it doesn't appear that anyone has written such a netfilter module. Basically, for packets originating from the local host, the module would take a look at the local port of the connection and run 'netstat -nap' to see what program is connected to that port. One of the big problems is that you'd just be tying it to a PID or program name, not a particular executable. In other words, if you allowed "mozilla", an attacker could write his own executable called "mozilla" and it'd still be allowed through. I'm not sure if there's any good way around this.
I think are is typically just a short-hand for most are. I suppose that there's a danger in using such a short-cut, allowing you to make the mistake of starting to think that the generalization really does apply to all. But when I say something like "black men can dance better than white men" I mean that black men tend to be better at dancing. To me, thinking that I meant that every black man can dance better than any white man would be a ludicrous interpretation. Linking in some genetic cause for such is even more ridiculous.
Hah! When I Googled, everything pointed back to people quoting you! The small "I" after the period was the give-away. Thanks for pointing me in the right direction. I found plenty of references to the Margaret Mead versions. Hard to tell exactly which one is the exact wording she used though; there are plenty of variants out there.
BTW, I added you as a Friend, since I liked a lot of the posts I read of yours. (I didn't even realize they were all yours at first.)
Hmm, that's a good point about religious beliefs changing, but they definitely change at a much slower pace.
I still have to disagree about scientific "laws" being assumed to be true. While we act under the assumption that they are "true enough" a good scientist should always remember that all "laws" can be proven (usually only slightly) untrue.
First of all, "race" in the modern US is more of a self-selection than most of us are aware. Most "black" people are probably 30-70% genetically European. Any rational classification system would place anyone with over 50% white genes as a "white" person, yet most of these people decide (possibly without realizing, and most likely their parents choose for them; I suppose that you could say that society often dictates as well) which sub-culture to "belong" to. You are right in stating that social pressures have a stronger influence on racial differences than anything biological or genetic.
But there are some biological differences. In addition to those mentioned in another reply to yours, there are different frequencies of diseases among the "races". For instance, blacks have a much higher incidence rate of sickle-cell anemia. These are of course mostly due to genetic heridity, although there are also instances of disease spread among members of the sub-culture.
So admittedly, most racial differences are due to the sub-culture in which the person was raised. But that does not make those differences any less real. Black men tend to be better at basketball than white men, because they grew up playing it more. Saying that they're better doesn't mean that they're genetically pre-disposed to be better though.
I think you may misunderstand science. (I'm of the belief that Science is our national religion.) Science never proves things to be true. It can only prove things to be wrong. Any scientific theory is based on assumptions, and we "believe" those theories until they are proven to be untrue. Our belief in the theories is based on a proponderance of evidence, but there's no way to prove anything 100%, so scientists are always refining the current theories.
I think the main difference between science and (traditional) religion is that scientists are willing to eventually give up their current beliefs if the evidence against it becomes overwhelming.
Excellent post.
I'm curious about your sig. Is it original, or can you attribute it?
I agree -- this is more about outright theft of the CSS and look and feel copyrighted material than freedom of speech or parody. Besides, one test of parody is whether the parody can be confused with the original. To tell you the truth, I didn't find the "parody" site to be at funny. It's more of a counter-point than parody. A political opinion is protected, but not when it is a copy of another (opposing) political opinion.