Here are some basic observations on the Iraq situation and what we should do to correct them:
There is no singular insurgency in Iraq. Instead there are at least 4 major Sunni groups. According to the International Crisis Group, we have seen a trend of consolidation in insurgent groups, several hundred of which sprung up after the war. Another ICG report details how Saddam used inter-clan strife as a way of ensuring that nobody could ever challenge his rule, so I doubt that these insurgent groups were new with the war, but probably were there before the war (but directed at different targets). According to a BBC report just before the war, Saddam was distributing various weapons to the general public including RPG's, assault rifles, etc. and I expect that the plan was to create the sort of chaos that we are seeing now.
Secondly, I think it is a mistake to see the insurgents as offering an alternative form of governance for Iraq. The Insurgent occupation of Fallujah and their propaganda suggests that they are Safalist in their rhetoric and actions, and see the idea form of government as some sort of mob rule.
A second major factor is the continuing use by the Iraqi government of sectarian militias in both official and unofficial capacities. THis is completely unacceptable because it means that Iraq is engulfed in a viscious civil war, and one side is using our troops as human shields while not playing by any reasonable rules. Bush has shown great hesitation to do what is necessary, which is to tell the Iraqi government in no uncertain terms that if they continue this sort of thing, we will be withdrawing our support for them, and providing their leadership with absolutely no protection even if this means temporarily redeploying back to Saudi Arabia until we have an opportunity to start the nation building process again, perhaps with additional partners from the region.
A second thing which needs to be done is that we need to state clearly that we are currently guests of the Iraqi government and will be more than happy to leave when asked to do so, but that if the current Iraqi government should fall and not be replaced with a representative government, we reserve the right to come back.
A third thing which needs to be done is that we need to scale back on the use of security contractors. These groups, such as Blackwater, have a fundamental conflict of interest in their involvement (i.e their contracts end when things become stable), and so they don't have the incentive to take the big picture into account.
A fourth thing we need to do is start talking to Iran and Sytia. We need to tell them in no uncertain terms that we are willing to given them conditional security guarantees provided that they are willing to respect certain bright lines in the area (no funding attacks on Israeli civilians inside the Green Line, no full nuclear fuel cycle without extremely intrusive inspections, and we want them to start working with us to build a stable Iraq. This is necessary because without those guarantees, the only rational response of Iran and Syria is to meddle in Iraq, fund both sides in the civil war, and more or less ensure that we cannot actually invade them without abandoning our other allies in the world.
Ok, how about no employment in government institutions, institutions, or vendors without putting three years into either of these organizations (the military of the conscientious objector corps)? It doesn't make it involuntary, but your life will be a *lot* easier if you do (for example, Microsoft, Intel, etc. could only hire people below a certain age if they had done these sorts of service).
I don't know about the desirability of storing OLAP data in non-relational ways. The really high-end (Terradata) solutions tend to actually process live relational data and get around the issue of large data sets through parallelism (i.e. instead of running one query on one monster 100TB database, you run small piecess of the query on 1000 ~100GB databases). The data is then aggregated back into the form for the OLAP query. BizgressMPP uses a similar solution. With appropriate star schemas, you can effectively run the query only on the interesting pieces of the database.
The major advantage this has over summary OLAP solutions is that you have a great deal more flexibility, so your system isn't prejudiced against new questions (i.e. if someone asks a new question you don't have a long wait to get the answer). Sure there is a tradeoff-- relational OLAP solutions take more time and effort to initially set up and maintain, but they are more powerful.
The user has different permissions? Or you want to limit connections by a specific user to specific computers or want to require some other credentials depending where they log in from?
I have never seen a situation where different permissions are required so I very much doubt that this is common. In PostgreSQL, you just set it up to specify that cetain user and/or host combinations require a different form of authentication. For example, you could use Kerberos for intranet connections, and PAM with a certificate for connections from home.
I would just add that textual searching in many RDBMS's is possible. PostgreSQL has good full text searching capabilities, and query ranking capabilities. But you are right-- basic SQL does not handle the sorts of information required, so extensions are required speciically for this purpose.
Note that XML doesn't provide a perfect fit for this either. What one really needs is a good full text querying standard (which doesn't exist).
Is that different for mom-and-pop shops depending on 2-3 employees than it is for 40000 employee corporations?
BTW, I do agree that employers tend to have a little more power than employees in this case, but I don;t think you can generalize too much.
BTW, the worst employment contracts I have seen are from small to midsize businesses. Often this is because the people making the decisions don't really understand the consequences of them. I get the same thing too with service contracts customers ask me to sign. Lawyers get second opinions and then add strange fluff into them.
Heck, I was once asked to sign their standard contract for contractors... Evidently they didn;t read the contract too closely because here I was building software and asked to ensure that all building codes were respected (wrong kind of contractor. Ooops).
If you are fluent in Perl and/or have good database design/stored procedure coding experience, please don't hesitate to send me a resume (mailto:chris@metatrontechcom). Since my company does pretty much exclusively FOSS work, we don't really rely on non-compete and IP-assignment clauses the way IP-centered firms do.
It doesn't matter if "people" think more carefully, as long as our elected officials can effectively veto any decision we make. Practically no one wanted to be in Iraq in the first place. GWB put us there, regardless, after playing public and world opinion like a Stradivarius. In that case, you would have thought it would have cost him the re-election, right? I am not sure I would agree with your characterization as to popular support for the war at the beginning. My own experience suggests that the public was pretty divided with a slight majority supporting it.
I don;t know. Every OLAP system I have ever worked with has been relational at least in the broadest sense.
Even things like CUBE are relational operations because they take relational data, synthesize it, and so forth.
Of course you may also be talking about parallelism in distributed databases or column-centered databases, but even there you are still doing nothing but manipulating keyed, ordered tuples to create new ordered tuples (which may or may not be keyed).
How many Americans think of war as something that someone else will fight in?
My main point is that making sure that the military service is really spread around would help to personalize the costs. It is easy to support a war when it is strangers fighting in your name. Would people support the same war if it meant that they would personally know people who were fighting in the conflict?
BTW, my own views of Iraq are that we never should have been there (as I said before the war), but that this is a unique war which has no easy answers about what to do now. Ultimately actions are less important than intents now, and when we did make the mistake of deposing Saddam, we became responsible for the country. Although this might make our job harder in the short run, I think we need to be willing to redeploy if the Iraqi government doesn't step up to the plate regarding various offices using sectarian militas for their own internal security. We may need to leave Iraq long enough to let the current government fall apart, put together a newer, larger coalition, and move in again. Iraq is harder to leave than Vietnam because we have no centralized opposition. At least when we left Vietnam, the VC could fill the power vacuum quickly but in Iraq, we will see more bloodshed, violence, and the like, and it will be our falt whether we are there or not.
BTW, I really think that the Iraqi government needs to sue Blackwater, and I believe that the key factor is that Blackwater can only be sued in an American court for crimes committed in Iraq, and only under US legal principles including common law (CPA-27 specifies that contractors are under the sole jurisdiction of sending states). We should not be using mercinaries in Iraq any more than the Iraqi government should be using sectarian militias, and Blackwater's actions certainly suggest that wrongful death suits in both the Al-Iraqya and Nasoor Square incidents might be worth pursuing. IANAL, though.
That is exactly what I am talking about. Every politician's son and every politician's daughter. Every young American, whether rich or poor, influential or not, etc. Those that are conscientious objectors would be obligated to form a special unit of community service workers who would be shipped to foreign countries to undertake infrastructure development projects as part of US foreign aid, or maybe used for projects back at home (though a lot of that should be left to the private sector).
Great examples of work for conscientious objector corps might include going to Israel and Lebannon after border wars to help survey and mitigate the cost of unexploded munitions, providing training for incident response for cities relating to chemical weapons attacks, and other life-saving programs (but still connected with military concerns).
When the WTC was hit, one of my first comments about it was to note the parallels between that attack and the start of WWI.
Of course, the world isnt the same place it was then. Instead of massive wars, you are more likely to see a sort of perpetual war as we see in US involvement in the Persian Gulf right now. Our unwillingness to consider leaving Iraq for the right reasons more or less ensures that the status quo will continue. (The right reasons are that certain branches of the Iraqi civil war are hiding behind our troops, committing attrocities, and fueling the violence, and we should not allow this to continue, even if it means pulling out until after this current government falls.)
A more likely scenario is that no nukes are used, and instead we will see cruise missiles with conventional warheads be used against Iran in retaliation. A nuclear strike against Iran by anyone except as a response to a similar strike would be massively stupid. I have said before that Israel would not be able to survive the repercussions* of launching such an attack, and even the US, I believe, would be sorely wounded from the resulting international backlash.**
* The single largest factor in the Israeli economy is trade relations with the EU. Last time the EU nations and companies started cancelling orders (during Sharon's Operation Defensive Shield), the Israeli economy temporarily collapsed. If such were to happen again but last for more than a few weeks, Iran wouldn't have to militarily retaliate in order to cripple Israel's ability to defend itself. ** The single largest factor in the US economy is the international price of oil which is largely controlled by OPEC. Most OPEC nations are in the Arab League as well.
Also note that Iran's major game at the moment is quite simple. They know that the current administration wants to take action against them and that this may include military action. So they are doing something incredibly rational, which is to sink as many of our troops into quagmires on their borders as possible. As long as Iraq and Afghanistan are unstable, it is not clear that the US can afford to spend the effort on a third war. Yes, we have the troops, but they are all deployed on various other fronts on our quest to promote our influence, and committing them elsewhere could well mean abandoning Taiwan, South Korea, and other places (i.e at best, we have enough forces to fight one more front. If we use that carelessly, then we lose our ability to defend important strategic allies).
Compulsatory military service would help prevent that. At least in theory, the impact of the war would be spread more broadly, thus making people think beforehand a bit more carefully.
In Vietnam, we had been there for a long time before the draft came up. If the draft had to happen in advance, do you think we would have been there at all?
I actually think we ought to have mandatory military service for everyone along with the option of other forms of community service for conscientious objectors (maybe infrastructure development).
The major reason is that foriegn policy regarding war is rarely made with a real understanding of personal cost in lives, lost loved ones, etc. War has become something that other people fight for us so we can be cavalier about it.
I think it would change the dynamics fundamentally if every community had people serving in the armed forces at all times. We pay the price at any rate, but at least this way, we as a society know the price when we pay it-- it isn't something that just happens to someone else.
BTW, I have lost a friend to the Iraq war. It wasn't an IED, an insurgent bullet, or anything like that which ended her life. It was the PTSD (she committed suicide about a year after getting back). How many others have lost friends?
Just to let you know where I stand on Iraq. This is a war we never should have fought. We shouldn't have gotten involved, and I said that at the beginning. However, we are there now, and for better or worse, we are now responsible for Iraq's well-being. Therefore IMO, the only acceptable discussion now is how we help the Iraqis to emerge from their civil war. While we as a country made a mistake, it would be an even bigger one to leave simply because the cost of staying in American lives is too high.
I actually think we need to be prepared to leave, and we need to use that as a threat we can use to get the current Iraqi government to stop using militias in official or unofficial capacities (we should probably scale down our use of private paramilitary groups such as Blackwater as well).
However, back to the China question: we will be involved in military confrontations with them at some point there is no doubt. However, as a democracy, we owe it to the people to spread the burden of the decision making, the politics, and the price fairly. Volunteers are great. But mandatory military service would put this country back into the hands of the people rather than the global planners.
In LDAP you can do both. In fact, you can even combine them,
The basic idea is that an object is a set of named attributes (names may be duplicate) attached to a path specification. The dn or distinguished name is actually just the path to the object.
For hierarchical querying you can do things like:
"Give me all InetOrgPersons with the zip of 12345" "Give me all OU's which are children of dc=example,dc=com whose name begins with s" "Give me all child objects of OU=sales,O=Example,dc=example,dc=com" "Give me all objects which are in the subtree starting with OU=sales,O=Example,dc=example,dc=com"
It is actually a little more complicated than this because what is actually happening is you are doing a hierarchical query first, and applying a filter to it. In LDAP, you have basically 3 modes of hierarchical queries: base, children, and subtree. The filters are then applied to the results before they are handed back. Personally I think that LDAP is a horrible monster and needs to be disposed of, but it is hierarchical and is a good example of where you can go with that model.
What you can;t readily do is "Give me any employees matching the following criteria, and add the name and address of the manager to each." (THis would require pulling the employees, and then pulling the managers separately, then aggregating the data structures).
When looking at the flaws of a hierarchical model, I think that LDAP is the best example of a system which does what it needs to do but not much more.
The basic issue is that a relational database management system is basically a big math engine which allows you to express any sort of complex operation mathematically, and get a usable result out of it. SQL is basically an imperfect representation of relational algebra (quite frankly, I think that Codd's 12 rules have one subtle issue as well, but that is beside the point). LDAP and similar systems, on the other hand, are designed to do nothing more than object storage and lookup.
Now, despite the fact that search issues are largely solved in LDAP implementations (i.e one can search on any field), the fact is, it is only programmed to give you data back in a set structure. Synthetic structures (for example, the result of SQL JOINs) are not possible in LDAP, and adding them would result in a number of inherent math issues.
Hence if you want to do reporting you can do that. You just retrieve all the objects, iterate through them, process the data in your app, and generate the report. Needless to say, this is pretty painful and not portable at all, however, it is clearly doable because I have seen people do this with middleware and SQL.... (SQL-Ledger for example does this in a large number of places.)
Also I would note that MySQL's authentication system is a real pain to manage for a large user-base, and the fact that usernames are not guaranteed to be unique (since the host is part of the auth credential) strikes me as a foot-gun.
It would be much nicer to have a nice system like PostgreSQL has, where you can assign various hosts to auth methods, but usernames are unique. Also external auth methods like Kerberos and PAM would be nice (PostgreSQL already has these, though I think that the Kerberos support could be improved).
Actually once you start getting into stored procedures, race conditions and the like could play havoc with statement-level replication. I would not use that unless I had no other choice and then only for a few narrowly defined sets of databases.
I don't care about default or administrator-defined values. If the option was available for administrators to *require* Strict mode (no turning it off by arbitrary clients), and if administrators could tell the RDBMS not to start if, say, innodb wasn't found, and if client applications could easily test for innodb support, then I would complain a lot less.
The problem is a fundamental one relating to an attitude that data integrity checking should be an optional feature available to the client. As long as this is the case, it cannot be relied on.
Actually the real problam is more fundamental. As long as the application can say "don't validate my data" then you can;t assume that the data in the db has been validated. Some clients may use strict mode while other clients may allow for data trunati
1) TotalRekall (Python-based) 2) PgAccess (a TCL front-end and form builder for PostgreSQL) 3) Once:Radix (a web-based front-end builder for PostgreSQL).
OnceRadix is quite new and I think it is well thought out in a lot of ways.
Besides, RDBMS should probably not be your primary validation tool anyhow; for their error messages are too cryptic and too divorced from the app's perspective. And there are ways to write nightly batch reports that point out accounts with suspicious info. I actually disagree with you here. The database should be your primary validation tool, and the application only does secondary checking. The reason is that this is the best way to remove invalid data bugs from an application later on. Think about it this way, DQL DDL gives you the ability to define data storage and presentation formats in ways which allow for powerful checking. Applications should trap the errors and at least inform users that some data was not of an acceptable format. Of course, at the same time, applications also need to verify input when it comes to issues like buffer overruns, an application having access to only a subset of valid data, or where additional error handling needs to be in place.
Note of course that all software bugs can be categorized as either data validity bugs (i.e. doing the right things on bad data), data flow bugs (doing the wrong things on good data), or both (doing the wrong things on bad data). Separating flow and data validation makes a lot of sense.
Why not? It is not as if MySQL safeguards your data anyway. Having these dangerous features might actually make things safer because people would be faced with these sorts of issues frequently.
Actually both sorts of databases exist today. Think of LDAP vs. an RDBMS. Hierarchical and object databases allow for easy storage and retrieval of data objects, while relational databases excel at mathematical manipulations of the data.
Tabular databases have one edge that hierarchical databases will never be able to match: mathematically powerful manipulations can be described mathematically. For example, aggregates and joins don't make sense in a hierarchical database (you would probably fetch all related objects and do your own data manipulation). Hence where you have a single application or a single agreed-upon data structure, and where mathematical transformations of the data is not necessary (again think of LDAP, which is mostly about looking up specific records), hierarchical organization works pretty well.
When you need to be able to store and manage complex data in complex ways, however, the relational system really shines. Now you can create synthetic relations by mathematically manipulating other relations, and thus you can do things with the data you can only dream about doing in a hierarchical model.
Consider the following scenario. You store all financial data in a database. At the end of the year, someone asks you, "How much revenue did we have in March?" Assume you did not store specific summary data for that. In a relational model, we just create a query which gives us the answer. In a hierarchical model, we have to retrieve all invoices and do our own aggragates on them ourselves.
Slashdot Mirror
Here are some basic observations on the Iraq situation and what we should do to correct them:
There is no singular insurgency in Iraq. Instead there are at least 4 major Sunni groups. According to the International Crisis Group, we have seen a trend of consolidation in insurgent groups, several hundred of which sprung up after the war. Another ICG report details how Saddam used inter-clan strife as a way of ensuring that nobody could ever challenge his rule, so I doubt that these insurgent groups were new with the war, but probably were there before the war (but directed at different targets). According to a BBC report just before the war, Saddam was distributing various weapons to the general public including RPG's, assault rifles, etc. and I expect that the plan was to create the sort of chaos that we are seeing now.
Secondly, I think it is a mistake to see the insurgents as offering an alternative form of governance for Iraq. The Insurgent occupation of Fallujah and their propaganda suggests that they are Safalist in their rhetoric and actions, and see the idea form of government as some sort of mob rule.
A second major factor is the continuing use by the Iraqi government of sectarian militias in both official and unofficial capacities. THis is completely unacceptable because it means that Iraq is engulfed in a viscious civil war, and one side is using our troops as human shields while not playing by any reasonable rules. Bush has shown great hesitation to do what is necessary, which is to tell the Iraqi government in no uncertain terms that if they continue this sort of thing, we will be withdrawing our support for them, and providing their leadership with absolutely no protection even if this means temporarily redeploying back to Saudi Arabia until we have an opportunity to start the nation building process again, perhaps with additional partners from the region.
A second thing which needs to be done is that we need to state clearly that we are currently guests of the Iraqi government and will be more than happy to leave when asked to do so, but that if the current Iraqi government should fall and not be replaced with a representative government, we reserve the right to come back.
A third thing which needs to be done is that we need to scale back on the use of security contractors. These groups, such as Blackwater, have a fundamental conflict of interest in their involvement (i.e their contracts end when things become stable), and so they don't have the incentive to take the big picture into account.
A fourth thing we need to do is start talking to Iran and Sytia. We need to tell them in no uncertain terms that we are willing to given them conditional security guarantees provided that they are willing to respect certain bright lines in the area (no funding attacks on Israeli civilians inside the Green Line, no full nuclear fuel cycle without extremely intrusive inspections, and we want them to start working with us to build a stable Iraq. This is necessary because without those guarantees, the only rational response of Iran and Syria is to meddle in Iraq, fund both sides in the civil war, and more or less ensure that we cannot actually invade them without abandoning our other allies in the world.
Ok, how about no employment in government institutions, institutions, or vendors without putting three years into either of these organizations (the military of the conscientious objector corps)? It doesn't make it involuntary, but your life will be a *lot* easier if you do (for example, Microsoft, Intel, etc. could only hire people below a certain age if they had done these sorts of service).
I don't know about the desirability of storing OLAP data in non-relational ways. The really high-end (Terradata) solutions tend to actually process live relational data and get around the issue of large data sets through parallelism (i.e. instead of running one query on one monster 100TB database, you run small piecess of the query on 1000 ~100GB databases). The data is then aggregated back into the form for the OLAP query. BizgressMPP uses a similar solution. With appropriate star schemas, you can effectively run the query only on the interesting pieces of the database.
The major advantage this has over summary OLAP solutions is that you have a great deal more flexibility, so your system isn't prejudiced against new questions (i.e. if someone asks a new question you don't have a long wait to get the answer). Sure there is a tradeoff-- relational OLAP solutions take more time and effort to initially set up and maintain, but they are more powerful.
Define "different security?"
The user has different permissions?
Or you want to limit connections by a specific user to specific computers or want to require some other credentials depending where they log in from?
I have never seen a situation where different permissions are required so I very much doubt that this is common. In PostgreSQL, you just set it up to specify that cetain user and/or host combinations require a different form of authentication. For example, you could use Kerberos for intranet connections, and PAM with a certificate for connections from home.
I would just add that textual searching in many RDBMS's is possible. PostgreSQL has good full text searching capabilities, and query ranking capabilities. But you are right-- basic SQL does not handle the sorts of information required, so extensions are required speciically for this purpose.
Note that XML doesn't provide a perfect fit for this either. What one really needs is a good full text querying standard (which doesn't exist).
So.....
Is that different for mom-and-pop shops depending on 2-3 employees than it is for 40000 employee corporations?
BTW, I do agree that employers tend to have a little more power than employees in this case, but I don;t think you can generalize too much.
BTW, the worst employment contracts I have seen are from small to midsize businesses. Often this is because the people making the decisions don't really understand the consequences of them. I get the same thing too with service contracts customers ask me to sign. Lawyers get second opinions and then add strange fluff into them.
Heck, I was once asked to sign their standard contract for contractors... Evidently they didn;t read the contract too closely because here I was building software and asked to ensure that all building codes were respected (wrong kind of contractor. Ooops).
If you are fluent in Perl and/or have good database design/stored procedure coding experience, please don't hesitate to send me a resume (mailto:chris@metatrontechcom). Since my company does pretty much exclusively FOSS work, we don't really rely on non-compete and IP-assignment clauses the way IP-centered firms do.
I don;t know. Every OLAP system I have ever worked with has been relational at least in the broadest sense.
Even things like CUBE are relational operations because they take relational data, synthesize it, and so forth.
Of course you may also be talking about parallelism in distributed databases or column-centered databases, but even there you are still doing nothing but manipulating keyed, ordered tuples to create new ordered tuples (which may or may not be keyed).
How many Americans think of war as something that someone else will fight in?
My main point is that making sure that the military service is really spread around would help to personalize the costs. It is easy to support a war when it is strangers fighting in your name. Would people support the same war if it meant that they would personally know people who were fighting in the conflict?
BTW, my own views of Iraq are that we never should have been there (as I said before the war), but that this is a unique war which has no easy answers about what to do now. Ultimately actions are less important than intents now, and when we did make the mistake of deposing Saddam, we became responsible for the country. Although this might make our job harder in the short run, I think we need to be willing to redeploy if the Iraqi government doesn't step up to the plate regarding various offices using sectarian militas for their own internal security. We may need to leave Iraq long enough to let the current government fall apart, put together a newer, larger coalition, and move in again. Iraq is harder to leave than Vietnam because we have no centralized opposition. At least when we left Vietnam, the VC could fill the power vacuum quickly but in Iraq, we will see more bloodshed, violence, and the like, and it will be our falt whether we are there or not.
BTW, I really think that the Iraqi government needs to sue Blackwater, and I believe that the key factor is that Blackwater can only be sued in an American court for crimes committed in Iraq, and only under US legal principles including common law (CPA-27 specifies that contractors are under the sole jurisdiction of sending states). We should not be using mercinaries in Iraq any more than the Iraqi government should be using sectarian militias, and Blackwater's actions certainly suggest that wrongful death suits in both the Al-Iraqya and Nasoor Square incidents might be worth pursuing. IANAL, though.
That is exactly what I am talking about. Every politician's son and every politician's daughter. Every young American, whether rich or poor, influential or not, etc. Those that are conscientious objectors would be obligated to form a special unit of community service workers who would be shipped to foreign countries to undertake infrastructure development projects as part of US foreign aid, or maybe used for projects back at home (though a lot of that should be left to the private sector).
Great examples of work for conscientious objector corps might include going to Israel and Lebannon after border wars to help survey and mitigate the cost of unexploded munitions, providing training for incident response for cities relating to chemical weapons attacks, and other life-saving programs (but still connected with military concerns).
You know, it is darn ironic.
When the WTC was hit, one of my first comments about it was to note the parallels between that attack and the start of WWI.
Of course, the world isnt the same place it was then. Instead of massive wars, you are more likely to see a sort of perpetual war as we see in US involvement in the Persian Gulf right now. Our unwillingness to consider leaving Iraq for the right reasons more or less ensures that the status quo will continue. (The right reasons are that certain branches of the Iraqi civil war are hiding behind our troops, committing attrocities, and fueling the violence, and we should not allow this to continue, even if it means pulling out until after this current government falls.)
A more likely scenario is that no nukes are used, and instead we will see cruise missiles with conventional warheads be used against Iran in retaliation. A nuclear strike against Iran by anyone except as a response to a similar strike would be massively stupid. I have said before that Israel would not be able to survive the repercussions* of launching such an attack, and even the US, I believe, would be sorely wounded from the resulting international backlash.**
* The single largest factor in the Israeli economy is trade relations with the EU. Last time the EU nations and companies started cancelling orders (during Sharon's Operation Defensive Shield), the Israeli economy temporarily collapsed. If such were to happen again but last for more than a few weeks, Iran wouldn't have to militarily retaliate in order to cripple Israel's ability to defend itself.
** The single largest factor in the US economy is the international price of oil which is largely controlled by OPEC. Most OPEC nations are in the Arab League as well.
Also note that Iran's major game at the moment is quite simple. They know that the current administration wants to take action against them and that this may include military action. So they are doing something incredibly rational, which is to sink as many of our troops into quagmires on their borders as possible. As long as Iraq and Afghanistan are unstable, it is not clear that the US can afford to spend the effort on a third war. Yes, we have the troops, but they are all deployed on various other fronts on our quest to promote our influence, and committing them elsewhere could well mean abandoning Taiwan, South Korea, and other places (i.e at best, we have enough forces to fight one more front. If we use that carelessly, then we lose our ability to defend important strategic allies).
Compulsatory military service would help prevent that. At least in theory, the impact of the war would be spread more broadly, thus making people think beforehand a bit more carefully.
In Vietnam, we had been there for a long time before the draft came up. If the draft had to happen in advance, do you think we would have been there at all?
I actually think we ought to have mandatory military service for everyone along with the option of other forms of community service for conscientious objectors (maybe infrastructure development).
The major reason is that foriegn policy regarding war is rarely made with a real understanding of personal cost in lives, lost loved ones, etc. War has become something that other people fight for us so we can be cavalier about it.
I think it would change the dynamics fundamentally if every community had people serving in the armed forces at all times. We pay the price at any rate, but at least this way, we as a society know the price when we pay it-- it isn't something that just happens to someone else.
BTW, I have lost a friend to the Iraq war. It wasn't an IED, an insurgent bullet, or anything like that which ended her life. It was the PTSD (she committed suicide about a year after getting back). How many others have lost friends?
Just to let you know where I stand on Iraq. This is a war we never should have fought. We shouldn't have gotten involved, and I said that at the beginning. However, we are there now, and for better or worse, we are now responsible for Iraq's well-being. Therefore IMO, the only acceptable discussion now is how we help the Iraqis to emerge from their civil war. While we as a country made a mistake, it would be an even bigger one to leave simply because the cost of staying in American lives is too high.
I actually think we need to be prepared to leave, and we need to use that as a threat we can use to get the current Iraqi government to stop using militias in official or unofficial capacities (we should probably scale down our use of private paramilitary groups such as Blackwater as well).
However, back to the China question: we will be involved in military confrontations with them at some point there is no doubt. However, as a democracy, we owe it to the people to spread the burden of the decision making, the politics, and the price fairly. Volunteers are great. But mandatory military service would put this country back into the hands of the people rather than the global planners.
In LDAP you can do both. In fact, you can even combine them,
The basic idea is that an object is a set of named attributes (names may be duplicate) attached to a path specification. The dn or distinguished name is actually just the path to the object.
For hierarchical querying you can do things like:
"Give me all InetOrgPersons with the zip of 12345"
"Give me all OU's which are children of dc=example,dc=com whose name begins with s"
"Give me all child objects of OU=sales,O=Example,dc=example,dc=com"
"Give me all objects which are in the subtree starting with OU=sales,O=Example,dc=example,dc=com"
It is actually a little more complicated than this because what is actually happening is you are doing a hierarchical query first, and applying a filter to it. In LDAP, you have basically 3 modes of hierarchical queries: base, children, and subtree. The filters are then applied to the results before they are handed back. Personally I think that LDAP is a horrible monster and needs to be disposed of, but it is hierarchical and is a good example of where you can go with that model.
What you can;t readily do is "Give me any employees matching the following criteria, and add the name and address of the manager to each." (THis would require pulling the employees, and then pulling the managers separately, then aggregating the data structures).
When looking at the flaws of a hierarchical model, I think that LDAP is the best example of a system which does what it needs to do but not much more.
The basic issue is that a relational database management system is basically a big math engine which allows you to express any sort of complex operation mathematically, and get a usable result out of it. SQL is basically an imperfect representation of relational algebra (quite frankly, I think that Codd's 12 rules have one subtle issue as well, but that is beside the point). LDAP and similar systems, on the other hand, are designed to do nothing more than object storage and lookup.
Now, despite the fact that search issues are largely solved in LDAP implementations (i.e one can search on any field), the fact is, it is only programmed to give you data back in a set structure. Synthetic structures (for example, the result of SQL JOINs) are not possible in LDAP, and adding them would result in a number of inherent math issues.
Hence if you want to do reporting you can do that. You just retrieve all the objects, iterate through them, process the data in your app, and generate the report. Needless to say, this is pretty painful and not portable at all, however, it is clearly doable because I have seen people do this with middleware and SQL.... (SQL-Ledger for example does this in a large number of places.)
Also I would note that MySQL's authentication system is a real pain to manage for a large user-base, and the fact that usernames are not guaranteed to be unique (since the host is part of the auth credential) strikes me as a foot-gun.
It would be much nicer to have a nice system like PostgreSQL has, where you can assign various hosts to auth methods, but usernames are unique. Also external auth methods like Kerberos and PAM would be nice (PostgreSQL already has these, though I think that the Kerberos support could be improved).
Because MySQL handles something like:
Insert into mytable (myvalue) values (random());
so well.....
Actually once you start getting into stored procedures, race conditions and the like could play havoc with statement-level replication. I would not use that unless I had no other choice and then only for a few narrowly defined sets of databases.
I don't care about default or administrator-defined values. If the option was available for administrators to *require* Strict mode (no turning it off by arbitrary clients), and if administrators could tell the RDBMS not to start if, say, innodb wasn't found, and if client applications could easily test for innodb support, then I would complain a lot less.
The problem is a fundamental one relating to an attitude that data integrity checking should be an optional feature available to the client. As long as this is the case, it cannot be relied on.
Actually the real problam is more fundamental. As long as the application can say "don't validate my data" then you can;t assume that the data in the db has been validated. Some clients may use strict mode while other clients may allow for data trunati
1) TotalRekall (Python-based)
2) PgAccess (a TCL front-end and form builder for PostgreSQL)
3) Once:Radix (a web-based front-end builder for PostgreSQL).
OnceRadix is quite new and I think it is well thought out in a lot of ways.
Note of course that all software bugs can be categorized as either data validity bugs (i.e. doing the right things on bad data), data flow bugs (doing the wrong things on good data), or both (doing the wrong things on bad data). Separating flow and data validation makes a lot of sense.
You can solve query issues in modern hierarchical databases. For example:
"Give me all inetOrgUsers who's state attribute is 'Kentucky' and whose OU is 'Sales'"
The larger problem is that hierarchical databases have representational issues which prevents the flexibility that you get in a relational model.
Why not? It is not as if MySQL safeguards your data anyway. Having these dangerous features might actually make things safer because people would be faced with these sorts of issues frequently.
Actually both sorts of databases exist today. Think of LDAP vs. an RDBMS. Hierarchical and object databases allow for easy storage and retrieval of data objects, while relational databases excel at mathematical manipulations of the data.
Tabular databases have one edge that hierarchical databases will never be able to match: mathematically powerful manipulations can be described mathematically. For example, aggregates and joins don't make sense in a hierarchical database (you would probably fetch all related objects and do your own data manipulation). Hence where you have a single application or a single agreed-upon data structure, and where mathematical transformations of the data is not necessary (again think of LDAP, which is mostly about looking up specific records), hierarchical organization works pretty well.
When you need to be able to store and manage complex data in complex ways, however, the relational system really shines. Now you can create synthetic relations by mathematically manipulating other relations, and thus you can do things with the data you can only dream about doing in a hierarchical model.
Consider the following scenario. You store all financial data in a database. At the end of the year, someone asks you, "How much revenue did we have in March?" Assume you did not store specific summary data for that. In a relational model, we just create a query which gives us the answer. In a hierarchical model, we have to retrieve all invoices and do our own aggragates on them ourselves.