Does it bother anyone else that the judges are gaining power all the time? What do we need congress for anyway, since any judge can just change the rules in the middle of the game?
Some of these laws open up the door for way too much interpretation. How is it proven that a file someone deleted was indeed copyrighted by the person who deleted it? Who gets the benefit of the doubt? Was a line crossed "a little bit"? What if the damages are unforseen by the copyright holder, or due to a bug in any of the software down the line (filesystem bug in the latest kernel shows up the time they delete that file)? There are no means to actually determine what happened, what was used to make it happen, or what people knew at the time it happened.
Guess who gets to decide? Your judge, that's who. Who needs those other branches of government, anyway?
One of the things that makes market economies work is that people know where they stand in the law, most of the time. If lawsuits are flying all over the place, and nobody knows where the money's gonna land, who would invest? George Will wrote an interesting column about businesses getting as far away from Mississippi as they can because the judges work so hard against those evil "Big Businesses". And, by coincidence I'm sure, Mississippi is the poorest state in the nation.
We need to put power back in the hands of Congress. I actually don't like that idea much either -- I'd prefer that we put power back in the hands of the states -- but it's better than the whims of appointed judges.
And another note, how is this law "necessary and proper" (the clause of the Constitution under which Congress presumably passes the law)? Why can't it be reserved to the states? Why does this need to be a nation-wide law? Think about how many laws on the federal books could be left to the states. Sure, interstate commerce, treaties, federal taxes: all those are "necessary and proper" at the federal level, but this?
With so much work (including DARPA funding) going into reiserfs version 4, it seems the *BSDs might be missing out.
I like FreeBSD from what I've done, and I see a lot of good technical merits when compared with GNU/linux. How feasible is it to port reisrefs to FreeBSD, assuming that it proves its worth?
Can you give some examples of unix utility differences, and possibly some reasons for the differences? What might a GNU user be surprised about as he tries out FreeBSD?
Yeah, good point. It's always more of a "red flag" (pun intended) to citizens when the government starts banning books.
I don't much agree with some of those liberal places on campus, but they certainly have a right to express their opinions, especially since they aren't getting in anyone's way at all in this situation. In some situations people try to force their opinions onto an unwilling audience or queit someone else's opinion, and I don't think that behavior is always protected by the first amendment. For example, I've heard rumors that at some colleges, certain speakers are unable to speak to a willing audience because protesters make the speech impractical (shouting or whatnot), and I really don't think that kind of behavior is protected.
One thing I have to say is that liberals are a lot better about saying "no" to government. When conservative groups or capitalists really dislike something (i.e. overtaxation or new government agencies or the like), they seem to mostly just use their vote and not otherwise protest. I would be interested to read about some counterexamples, as this is just an observation of mine.
Because the school certainly took it upon themselves to interpret the law in an unprecedented way. The school isn't really sure that anyone violated a law, but UCSD is trying to enforce it. Who gave UCSD the power to interpret the law like this?
I suppose the target of the enforcement can always appeal, so I expect the system to work in this situation. Most likely, a little pushing-back by the che cafe will get the whole thing thrown out and forgotten.
Incidently, I attend UCSD. I was there about an hour ago.
Where can I find such a company? I wouldn't mind overbidding a few projects to help a manager spend his budget. I'm just a helpful guy, what can I say?
Fake a transaction with LOCKs? What about the "I" in "ACID": Isolation? Or were you thinking to do an exclusive LOCK to block writing *and* reading?
Not only that, aren't there a few performance issues with taking down tables at a time to do a few simple operations? Doesn't that mean you can have no more than 1 concurrent (not that 1 is concurrent, mind you) connection from the app that always LOCKs the table?
You're right, it is a kind of subsidy. So is any kind of insurance. If some genius *knows* he's going to invent something worth $700M, he can go about the venture himself. However, he isn't quite so confident, so he takes the safe route and signs away his brain for the sake of a regular income. Some people then come up with a $700M invention, other people don't. To the company it's just risk analysis. However people, unlike companies, need a regular income to feed their families.
I agree with you last statement. PostgreSQL and MySQL came about in different ways and became popular for different reasons.
Since that time, both have been improving a lot. I don't know much about mysql, but I hear it's filling in some of those features I like so much. I will probably continue to use postgres for a long time because I really do like it. But it's rediculous to act like it's a competition. You put it best: "They're both kicking serious ass." It's good if MySQL gets another feature, not a setback to the postgres guys.
Deleting a column will be added in 7.3, which is already beta (beta 2 next week, I think). I am running the beta on my development workstation, and it works nicely (including drop column).
The one thing that bugs me is that you can't change the type of an attribute. MySQL changes the type and casts all the data. PostgreSQL you can drop/add, but you end up with a lot of empty attributes. Sure, it can't work for things that don't cast, but usually you don't change you mind between the "polygon" type and the "datetime" type or something weird like that. It's normally like int->float or maybe int->text or something that's castable within reason.
I still like postgres more though:) For those interested, I highly recommend the beta for 7.3. Lots of great new stuff.
Well, you didn't leave an email, so I'll type out some answers in here. You can always email me at jdaNOSPAMvis at empNOSPAMires dot org (some decoding required), or you can join the mailing lists.
pl/pgsql isn't quite as bad as it looks (although I wouldn't write in it outside of a DB:)
Here's a simple function:
CREATE FUNCTION addone(INT) RETURNS INT AS ' BEGIN
RETURN $1 + 1; END; ' LANGUAGE 'plpgsql';
Then do
SELECT addone(7);
and you should get 8.
Postgres has awesome support for user-defined finctions. There are many languages you can code them in (or you can write a new pl handler if you want your own). You can write functions that return result sets and then place them just as you would tables. You can wirte in C for speed. You can write aggregate functions. In other words, I don't know what more you could ask for, and I'd be interested to know if oracle's support for user defined functions was as good.
Packages however: nope. There was talk about implementing them. If you want to help push the developers a little they'd go for it I think. I asked them about it before and they said that basically nobody came forward with either a detailed proposal or a patch, so it kind of fell through. As long as it doesn't get in the way of other stuff, or cause lots of complexity in other pieces of code, they seem to be OK with new features.
For populating a database do you mean filling it in with auto-generated values for testing? I suppose you could do that, although personally I would use a python script. Tell me what you need to do more specifically, and I'll try to point you in the right direction.
I really like PostgreSQL, and it's nice to see that it's gaining some acceptence. In particular, I like the way the developers develop. I really find it interesting reading the mailing lists (pgsql-general and pgsql-hackers).
It's interesting to note how much MySQL seems to be improving as well. I don't follow that very closely, but appearently they're filling in those "missing features" at a rapid pace (though it's not like an RDBMS is about feature count). I still like PostgreSQL more, but it's nice to see that both are improving so much.
I think this guy puts too much blame on the vendors. Vendors supply what people want, and people, in general, want bad software fast rather than good software a little slower. I don't buy into that, I get software based primarily on it's technical merit (which includes security, of course).
Sometimes vendors imply that "everyone writes crappy software". I think that's bad because the consumers might not understand the idea of an application or OS that works consistantly.
However, I don't see any clear way around that kind of marketing. And I certainly don't see any productive (as opposed to counterproductive) way of *forcing* people to write good software without public demand. After all, liability would decrease the number of free software developers. Not only that, what about software developed outside the US? Import laws? That just sounds like a bad idea.
If you look at proprietary software and intellectual property protection, you see a model that makes some sense and certainly fits within an economic system. Developers are paid in accordance with demand -- demand is the best indicator for the benefit that your individual product provides.
Open source software is a development model that never accounts for real benefits provided; that is, nobody even knows how much a given piece of free/os software helps anyone. If I were to donate $100 to "free software", in what proportion would I donate to the individual products, or people? Is $100 enough? Too much? I don't know where to start, because I don't know how much help Linus is vs. Alan Cox, nor do I know how valuable Linux is when I could just grab FreeBSD, or any other BSD, or Hurd. I use linux, but if it never existed, how much would I be affected?
The point is that nobody knows. The developers are not compensated as accurately as in more standard business models. In my opinion, this is not economic justice.
I am not proposing a solution to this problem. I believe that the best software is created in this way. I also believe that no amount of socialism would correct this problem... the government has no better way of deciding a product's value than the people do (if you ask me, it isn't usually as good at such decisions... but that's a different topic).
As far as a credible source, you might try looking into RMS's writings, as well as the interviews/statements by MS. Both address the problem that developers don't always get compensated correctly (remember "unamerican"?).
Facts:
1) A Russian researched and made a discovery in Russia that is critical of a current practice (Adobe's software design).
2) He tries to share the idea with people in the U.S.
3) The U.S. government is trying to stop the discovery from being spread to the public.
Who made the iron curtain again?
Other thoughts:
You can't copyright an idea, just expression.
Patents are void if prior use is proven (Adobe's weak encryption is nothing new AFAIK).
Patents are to protect new methods from use by competition; they don't stop people from learning about them or sharing information.
It doesn't seem the DMCA fits well into copyright or any form of intellectual property...
Slashdot Mirror
Does it bother anyone else that the judges are gaining power all the time? What do we need congress for anyway, since any judge can just change the rules in the middle of the game?
Some of these laws open up the door for way too much interpretation. How is it proven that a file someone deleted was indeed copyrighted by the person who deleted it? Who gets the benefit of the doubt? Was a line crossed "a little bit"? What if the damages are unforseen by the copyright holder, or due to a bug in any of the software down the line (filesystem bug in the latest kernel shows up the time they delete that file)? There are no means to actually determine what happened, what was used to make it happen, or what people knew at the time it happened.
Guess who gets to decide? Your judge, that's who. Who needs those other branches of government, anyway?
One of the things that makes market economies work is that people know where they stand in the law, most of the time. If lawsuits are flying all over the place, and nobody knows where the money's gonna land, who would invest? George Will wrote an interesting column about businesses getting as far away from Mississippi as they can because the judges work so hard against those evil "Big Businesses". And, by coincidence I'm sure, Mississippi is the poorest state in the nation.
We need to put power back in the hands of Congress. I actually don't like that idea much either -- I'd prefer that we put power back in the hands of the states -- but it's better than the whims of appointed judges.
And another note, how is this law "necessary and proper" (the clause of the Constitution under which Congress presumably passes the law)? Why can't it be reserved to the states? Why does this need to be a nation-wide law? Think about how many laws on the federal books could be left to the states. Sure, interstate commerce, treaties, federal taxes: all those are "necessary and proper" at the federal level, but this ?
With so much work (including DARPA funding) going into reiserfs version 4, it seems the *BSDs might be missing out. I like FreeBSD from what I've done, and I see a lot of good technical merits when compared with GNU/linux. How feasible is it to port reisrefs to FreeBSD, assuming that it proves its worth?
Can you give some examples of unix utility differences, and possibly some reasons for the differences? What might a GNU user be surprised about as he tries out FreeBSD?
Yeah, good point. It's always more of a "red flag" (pun intended) to citizens when the government starts banning books.
I don't much agree with some of those liberal places on campus, but they certainly have a right to express their opinions, especially since they aren't getting in anyone's way at all in this situation. In some situations people try to force their opinions onto an unwilling audience or queit someone else's opinion, and I don't think that behavior is always protected by the first amendment. For example, I've heard rumors that at some colleges, certain speakers are unable to speak to a willing audience because protesters make the speech impractical (shouting or whatnot), and I really don't think that kind of behavior is protected.
One thing I have to say is that liberals are a lot better about saying "no" to government. When conservative groups or capitalists really dislike something (i.e. overtaxation or new government agencies or the like), they seem to mostly just use their vote and not otherwise protest. I would be interested to read about some counterexamples, as this is just an observation of mine.
Because the school certainly took it upon themselves to interpret the law in an unprecedented way. The school isn't really sure that anyone violated a law, but UCSD is trying to enforce it. Who gave UCSD the power to interpret the law like this?
I suppose the target of the enforcement can always appeal, so I expect the system to work in this situation. Most likely, a little pushing-back by the che cafe will get the whole thing thrown out and forgotten.
Incidently, I attend UCSD. I was there about an hour ago.
Not only that, but as I understand it they don't really give you any complete body of code, just chunks.
That means that you can't really be sure it's even the code that they say it is, because you can't compile it all together and test it.
Although I really don't know what I'm talking about.
"we must spend our entire budget"
Where can I find such a company? I wouldn't mind overbidding a few projects to help a manager spend his budget. I'm just a helpful guy, what can I say?
Fake a transaction with LOCKs? What about the "I" in "ACID": Isolation? Or were you thinking to do an exclusive LOCK to block writing *and* reading?
Not only that, aren't there a few performance issues with taking down tables at a time to do a few simple operations? Doesn't that mean you can have no more than 1 concurrent (not that 1 is concurrent, mind you) connection from the app that always LOCKs the table?
I don't think that either company has ever suggested anything other than self-interest wrt open source.
Now they don't have any interest to back MySQL, so they try to encourage the use of DB2 or whatever else.
You're right, it is a kind of subsidy. So is any kind of insurance. If some genius *knows* he's going to invent something worth $700M, he can go about the venture himself. However, he isn't quite so confident, so he takes the safe route and signs away his brain for the sake of a regular income. Some people then come up with a $700M invention, other people don't. To the company it's just risk analysis. However people, unlike companies, need a regular income to feed their families.
I agree with you last statement. PostgreSQL and MySQL came about in different ways and became popular for different reasons.
Since that time, both have been improving a lot. I don't know much about mysql, but I hear it's filling in some of those features I like so much. I will probably continue to use postgres for a long time because I really do like it. But it's rediculous to act like it's a competition. You put it best: "They're both kicking serious ass." It's good if MySQL gets another feature, not a setback to the postgres guys.
Jeff Davis
Deleting a column will be added in 7.3, which is already beta (beta 2 next week, I think). I am running the beta on my development workstation, and it works nicely (including drop column).
:) For those interested, I highly recommend the beta for 7.3. Lots of great new stuff.
The one thing that bugs me is that you can't change the type of an attribute. MySQL changes the type and casts all the data. PostgreSQL you can drop/add, but you end up with a lot of empty attributes. Sure, it can't work for things that don't cast, but usually you don't change you mind between the "polygon" type and the "datetime" type or something weird like that. It's normally like int->float or maybe int->text or something that's castable within reason.
I still like postgres more though
Jeff Davis
Well, you didn't leave an email, so I'll type out some answers in here. You can always email me at jdaNOSPAMvis at empNOSPAMires dot org (some decoding required), or you can join the mailing lists.
:)
pl/pgsql isn't quite as bad as it looks (although I wouldn't write in it outside of a DB
Here's a simple function:
CREATE FUNCTION addone(INT) RETURNS INT AS '
BEGIN
RETURN $1 + 1;
END;
' LANGUAGE 'plpgsql';
Then do
SELECT addone(7);
and you should get 8.
Postgres has awesome support for user-defined finctions. There are many languages you can code them in (or you can write a new pl handler if you want your own). You can write functions that return result sets and then place them just as you would tables. You can wirte in C for speed. You can write aggregate functions. In other words, I don't know what more you could ask for, and I'd be interested to know if oracle's support for user defined functions was as good.
Packages however: nope. There was talk about implementing them. If you want to help push the developers a little they'd go for it I think. I asked them about it before and they said that basically nobody came forward with either a detailed proposal or a patch, so it kind of fell through. As long as it doesn't get in the way of other stuff, or cause lots of complexity in other pieces of code, they seem to be OK with new features.
For populating a database do you mean filling it in with auto-generated values for testing? I suppose you could do that, although personally I would use a python script. Tell me what you need to do more specifically, and I'll try to point you in the right direction.
Regards,
Jeff Davis
I really like PostgreSQL, and it's nice to see that it's gaining some acceptence. In particular, I like the way the developers develop. I really find it interesting reading the mailing lists (pgsql-general and pgsql-hackers).
It's interesting to note how much MySQL seems to be improving as well. I don't follow that very closely, but appearently they're filling in those "missing features" at a rapid pace (though it's not like an RDBMS is about feature count). I still like PostgreSQL more, but it's nice to see that both are improving so much.
Jeff
Far from "ignoring" the situation, the developers are developing a native win32 port as we speak.
I think this guy puts too much blame on the vendors. Vendors supply what people want, and people, in general, want bad software fast rather than good software a little slower. I don't buy into that, I get software based primarily on it's technical merit (which includes security, of course).
Sometimes vendors imply that "everyone writes crappy software". I think that's bad because the consumers might not understand the idea of an application or OS that works consistantly.
However, I don't see any clear way around that kind of marketing. And I certainly don't see any productive (as opposed to counterproductive) way of *forcing* people to write good software without public demand. After all, liability would decrease the number of free software developers. Not only that, what about software developed outside the US? Import laws? That just sounds like a bad idea.
JeffHmmm... reminded me of an article about code rewriting that slashdot posted a while back.
Slashdot Story: How To Make Software Projects Fail
Relevant interview link
If you look at proprietary software and intellectual property protection, you see a model that makes some sense and certainly fits within an economic system. Developers are paid in accordance with demand -- demand is the best indicator for the benefit that your individual product provides.
Open source software is a development model that never accounts for real benefits provided; that is, nobody even knows how much a given piece of free/os software helps anyone. If I were to donate $100 to "free software", in what proportion would I donate to the individual products, or people? Is $100 enough? Too much? I don't know where to start, because I don't know how much help Linus is vs. Alan Cox, nor do I know how valuable Linux is when I could just grab FreeBSD, or any other BSD, or Hurd. I use linux, but if it never existed, how much would I be affected?
The point is that nobody knows. The developers are not compensated as accurately as in more standard business models. In my opinion, this is not economic justice.
I am not proposing a solution to this problem. I believe that the best software is created in this way. I also believe that no amount of socialism would correct this problem... the government has no better way of deciding a product's value than the people do (if you ask me, it isn't usually as good at such decisions... but that's a different topic).
As far as a credible source, you might try looking into RMS's writings, as well as the interviews/statements by MS. Both address the problem that developers don't always get compensated correctly (remember "unamerican"?).
Jeff Davis
Facts:
1) A Russian researched and made a discovery in Russia that is critical of a current practice (Adobe's software design).
2) He tries to share the idea with people in the U.S.
3) The U.S. government is trying to stop the discovery from being spread to the public.
Who made the iron curtain again?
Other thoughts:
You can't copyright an idea, just expression. Patents are void if prior use is proven (Adobe's weak encryption is nothing new AFAIK). Patents are to protect new methods from use by competition; they don't stop people from learning about them or sharing information. It doesn't seem the DMCA fits well into copyright or any form of intellectual property...
Jeff Davis