Or just set the thing down next to a speaker and listen for the characteristic click-click-buzz when it's transmitting. Some phones are even powerful enough that my car stereo picks it up, or leaks into the signal of a nearby land-line (the PBX phones are work are real sensitive to it).
On occasion I've put my phone down next to an LED flashlight, and when the phone rings the flashlight turns on!
Hahahahahahaha. Oracle has to be one of the least secure products (and companies for that better) I've ever seen. Far worse than anything MS ever put out. Just do a search on Bugtraq for people complaining about security bugs they've filed years (!) ago finally being patched, only to discover that the "fix" doesn't really fix the underlying issue but just papers over one particular vector of exploiting it.
Despite whatever marketing rhetoric Larry may spout, Oracle's attitude toward security is still what MS's was 7 years ago (la-la-la-I'm-not-listening!).
I don't think I could consider a database where "not losing data" is one of it's weaker points. I haven't lost data on a SQL Server since version 6.5. NEVER.
In all seriousness, I think what the OP was referring to is MySQL's tendency to silently change input data rather than aborting the transaction with an error. Examples include things like invalid dates, or out-of-range numbers being truncated. It's been a major MySQL gripe for a long time. The latest version (finally) has a "strict" mode that enables checks for things like that, but I don't believe it's on by default. Not 100% sure what the status of it is as I'm a PostgreSQL junkie.
Oh, and of course for complex queries Postgres kicks SQL Servers ass, performance-wise.
Has anyone done any tests lately to see if that is still the case? EULA be damned, post as an AC if you have to:P
I hear SQL Server 2005 finally got MVCC and did away with row locking. Hard to believe even 2003 still locked rows for every write, but there it is. Even without that edge I'd like to believe PostgreSQL will still kick ass on complex queries, but haven't done any empirical testing against the new MSSQL engine.
I see you've never worked with big databases. Clustering? Redo logs?
With MySQL, forget it.
PostgreSQL has "redo logs" (it calls them Write Ahead Logging). You can do log archiving as a form of incremental backups, or you can copy the WAL segments to another server for hot standby -- we have an HA cluster at work that uses this method; more reliable than a SAN-based cluster as there's no single point of failure. You can do point-in-time recovery and switch between different timelines if necessary.
There's also asynchronous replication via slony (good for on-line read-only cache servers), or synchronous replication with several other methods.
Oracle has an advantage in the multimaster clustering arena. It's not until your databases get really big that you need that, but the that can be a very tricky subject as most databases aren't designed with the idea of not having everything always consistent. PostgreSQL can't do that yet but is actually gaining ground here -- you could in theory do something similar with two-phase commit over separate servers.
Last I checked Oracle was still slightly better with GIS and spatial data. PostGIS has historically had some indexing problems, but I admit it's been a while since I've looked at it.
One thing you can't yet get with PostgreSQL is the privilege of paying oodles of money to Larry & co for unhelpful support staff that usually have trouble understanding what the problem is:-P
Maybe it's more difficult to set up if you don't like to edit the pg_hba.conf and postgresql.conf files, but other than that...
Editing pg_hba.conf is nothing compared to having to configure a TNS Listener instance and names files on all the clients, even with (or perhaps in spite of) the "help" of the wizard.
PostgreSQL is an absolute breeze to install and configure.
UT99 is still played more on the LAN at every college in the area than any other game
I assume by UT99 you mean the original Unreal Tournament? I can see why it's still played, that game was fun.
In UT, the characters you control are fast, agile, and reasonably tough. And man oh man you could jump. By contrast, the controls of UT2003/04 (and even Q3A to a lesser extent) may be more realistic, but seem slow and boggy.
Combine with wide, varying maps and creative weapon design, and you've got a hit. I used to run a UT server for some friends back in the day. It was always fun to play with the Relics mod, and we had a few custom mods (Last Team Standing gametype was so much fun).
Hey, I didn't even notice that I was replying to BadAnalogyGuy! Well, to make a bad analogy, small scale fusion reactions are nothing like the Sun, despite what Spider-man 2 would have you believe.
It's actually very difficult to create a fusion chain reaction. Pretty much impossible for an uncontained reaction to last longer than a few atoms (which wouldn't result in all that much heat). Just getting the fuel really hot isn't enough -- the atoms have to be close enough together. It works in the sun because that has its own containment field -- large amounts of gravity keep the pressure extremely high at the center.
Here on Earth gravity isn't really an option, in order to get fusion working you have to pressurize the fuel somehow. Farnsworth fusors like this one use a strong electromagnetic shell to repel ionized gas and accelerate it into the center of a sphere. The effect is similar to aiming the electron gun of several CRTs at precisely the same point (actually Farnsworth worked mostly on television electronics, so the similarity is no coincidence).
The worst that can happen if it overheats is that the equipment generating the EM field is damaged, causing the individual atoms of the fuel to repel from each other and the reaction to halt. Because fusion is so hard to keep going, it's a lot more fool-resistant than trying to build a fission reactor. It's also much much harder to extract a useful amount of energy -- it may not even be possible from this type of reactor because it takes so much to keep it running.
An H-bomb achieves the necessary pressure and temperature by setting off a conventional fission warhead, fusing all the available fuel in one shot. You'd never be able to achieve that kind of uncontrolled reaction in a jar without an equally violent trigger.
(We sort of already knew that, because as a major petroleum supplier they don't _need_ nuclear power, but the deuterium makes it even more obvious.)
Why does everybody assume that Iran's claim about civilian nuclear power generation is completely false?
As a major petroleum supplier they don't _need_ nuclear power, but they might _want_ it for a variety of reasons.
1. Less petroleum burned locally = More to sell 2. Pollution issues 3. If the reserves are getting low, they'd be the first to know (and probably wouldn't advertise it to they buyers who might panic and start looking at alternate energy sources)
Now, I think it would be naive to think that Iran isn't interested in developing nuclear weapons. I'd say the truth is probably somewhere in the middle -- they want nuclear power for civilian applications, but if it can help produce nuclear weapons it kills two birds with one stone.
Unfortunately, the most efficient type of reactor is also the best kind for producing weapons-grade materiel. Conventional designs use up uranium way too fast to be long-term viable.
Correct, Farnsworth Fusors are "hot" fusion devices. They use intense electromagnetic fields to accelerate particles rather than lasers or other more fanciful methods, but the end result is the same.
The problem with America is stupidity. I'm not saying there should be a capital punishment for stupidity, but why don't we just take the safety labels off of everything and let the problem solve itself?
I agree with the sentiment, unfortunately the fatality rate would probably be too low. Most stupid people would just end up injuring themselves, and as healthcare is becoming more and more socialized the rest of us would end up footing the bill...
I'm hoping that maybe...just maybe with the many Dem.s that won seats....they seem to be more centrist that the liberal Dems of the recent past...maybe they will have a fiscal concious, that will look to cut pork, rather than dig deeper into our coI'm hoping that maybe...
There's also no chance in hell that the Democrats will do anythingto cut government spending and everyone knows it. The Republicans have been as bad as the Democrats at that in the last several years but the Dems will never do it. You can bet they're already working on which taxes to increase and by how much and what gun control laws they can try to push through.
Oh, absolutely. Dems have always been about spending as much as possible on whatever social program catches their fancy.
What makes it so frustrating is that the Republican party used to be fiscally conservative. Now both of the major parties want to keep spending and spending even though the budget is already hemorrhaging. Just another reason why the two-party system sucks and never should have been allowed to happen (it's exactly what several of the founding fathers were afraid of happening).
I've looked at autopatcher before (and even used it on a spare machine just to see what the fuss was about). My only concern about it is that there's no guarantee about where the patches are coming from -- how can you be sure they haven't been tampered with?
Yes, I've done customized unattended installs before (even over netboot), but it's a pain because you have to build a new image every time you want to change something. Much more flexible to just do a clean install and run a script off a network location that sets everything up.
I don't think you can do it with the "offical" slipstreaming process. I've seen guides to doing that before, but basically they involve copying all of the setup files for the individual patches onto the CD and using a custom install script to apply the patches near the end of the install. Seems really inefficient plus you have to use an answer file which means your setup process is no longer interactive in a few places (notably disk partitioning IIRC).
It might also be possible to manually replace individual files on the CD with patched versions, but I'm leery of doing that in case there's also some registry changes or other issues that would get missed.
Or better yet slipstream the SP into your install CD and only use that. Nothing to have to remember to apply. An "XP with SP2" CD is a much more secure way to install than an original XP CD and applying SP2 post install.
The main reason is because service packs can do what they call "slipstreaming". You apply the service pack against the original install CD, create an image from that, and burn yourself a new CD. When you install from your new CD you already have everything that was part of the service pack, so you have a lot less that needs to be downloaded after the install is complete. Since the files on the CD were updated, it doesn't take any longer to install the OS than normal (versus waiting 15 minutes for an SP to apply even when it's local). Great for admins who frequently build machines.
Don't tell me to just use ghost or dd or some other disk imaging solution. I've found that a fresh install is a lot cleaner (no filesystem resizing / conversion, no SID issues, no cruft in your image from when you logged in to set stuff up). Combined with a script to tweak some default settings and group policy to automatically install the appropriate software, it's just as automated as an image and doesn't take much longer. Waiting for 70 security updates to install (even from a local WSUS mirror) is probably the longest part of the whole process.
Yes, but the problem is that even with a slipstreamed SP2 (which is all I use anymore), there's still 60-70 updates that it has to download and install. That's a number which will only get bigger over the next year.
That's why we want SP3, so that it can be slipstreamed in and vastly reduce that number.
Slashdot Mirror
Or just set the thing down next to a speaker and listen for the characteristic click-click-buzz when it's transmitting. Some phones are even powerful enough that my car stereo picks it up, or leaks into the signal of a nearby land-line (the PBX phones are work are real sensitive to it).
On occasion I've put my phone down next to an LED flashlight, and when the phone rings the flashlight turns on!
Hahahahahahaha. Oracle has to be one of the least secure products (and companies for that better) I've ever seen. Far worse than anything MS ever put out. Just do a search on Bugtraq for people complaining about security bugs they've filed years (!) ago finally being patched, only to discover that the "fix" doesn't really fix the underlying issue but just papers over one particular vector of exploiting it.
Despite whatever marketing rhetoric Larry may spout, Oracle's attitude toward security is still what MS's was 7 years ago (la-la-la-I'm-not-listening!).
If you're paying a million for a 16-way you're an idiot. Partition your data and cluster cheaper machines. That's what Google does.
I don't think I could consider a database where "not losing data" is one of it's weaker points. I haven't lost data on a SQL Server since version 6.5. NEVER.
Don't tell that to the Russians...
In all seriousness, I think what the OP was referring to is MySQL's tendency to silently change input data rather than aborting the transaction with an error. Examples include things like invalid dates, or out-of-range numbers being truncated. It's been a major MySQL gripe for a long time. The latest version (finally) has a "strict" mode that enables checks for things like that, but I don't believe it's on by default. Not 100% sure what the status of it is as I'm a PostgreSQL junkie.
Oh, and of course for complex queries Postgres kicks SQL Servers ass, performance-wise.
:P
Has anyone done any tests lately to see if that is still the case? EULA be damned, post as an AC if you have to
I hear SQL Server 2005 finally got MVCC and did away with row locking. Hard to believe even 2003 still locked rows for every write, but there it is. Even without that edge I'd like to believe PostgreSQL will still kick ass on complex queries, but haven't done any empirical testing against the new MSSQL engine.
I see you've never worked with big databases. Clustering? Redo logs?
:-P
With MySQL, forget it.
PostgreSQL has "redo logs" (it calls them Write Ahead Logging). You can do log archiving as a form of incremental backups, or you can copy the WAL segments to another server for hot standby -- we have an HA cluster at work that uses this method; more reliable than a SAN-based cluster as there's no single point of failure. You can do point-in-time recovery and switch between different timelines if necessary.
There's also asynchronous replication via slony (good for on-line read-only cache servers), or synchronous replication with several other methods.
Oracle has an advantage in the multimaster clustering arena. It's not until your databases get really big that you need that, but the that can be a very tricky subject as most databases aren't designed with the idea of not having everything always consistent. PostgreSQL can't do that yet but is actually gaining ground here -- you could in theory do something similar with two-phase commit over separate servers.
Last I checked Oracle was still slightly better with GIS and spatial data. PostGIS has historically had some indexing problems, but I admit it's been a while since I've looked at it.
One thing you can't yet get with PostgreSQL is the privilege of paying oodles of money to Larry & co for unhelpful support staff that usually have trouble understanding what the problem is
Maybe it's more difficult to set up if you don't like to edit the pg_hba.conf and postgresql.conf files, but other than that...
Editing pg_hba.conf is nothing compared to having to configure a TNS Listener instance and names files on all the clients, even with (or perhaps in spite of) the "help" of the wizard.
PostgreSQL is an absolute breeze to install and configure.
Or connected a USB hard drive to the Xbox with a spliced up cable ;)
UT99 is still played more on the LAN at every college in the area than any other game
I assume by UT99 you mean the original Unreal Tournament? I can see why it's still played, that game was fun.
In UT, the characters you control are fast, agile, and reasonably tough. And man oh man you could jump. By contrast, the controls of UT2003/04 (and even Q3A to a lesser extent) may be more realistic, but seem slow and boggy.
Combine with wide, varying maps and creative weapon design, and you've got a hit. I used to run a UT server for some friends back in the day. It was always fun to play with the Relics mod, and we had a few custom mods (Last Team Standing gametype was so much fun).
Hey, I didn't even notice that I was replying to BadAnalogyGuy! Well, to make a bad analogy, small scale fusion reactions are nothing like the Sun, despite what Spider-man 2 would have you believe.
It's actually very difficult to create a fusion chain reaction. Pretty much impossible for an uncontained reaction to last longer than a few atoms (which wouldn't result in all that much heat). Just getting the fuel really hot isn't enough -- the atoms have to be close enough together. It works in the sun because that has its own containment field -- large amounts of gravity keep the pressure extremely high at the center.
Here on Earth gravity isn't really an option, in order to get fusion working you have to pressurize the fuel somehow. Farnsworth fusors like this one use a strong electromagnetic shell to repel ionized gas and accelerate it into the center of a sphere. The effect is similar to aiming the electron gun of several CRTs at precisely the same point (actually Farnsworth worked mostly on television electronics, so the similarity is no coincidence).
The worst that can happen if it overheats is that the equipment generating the EM field is damaged, causing the individual atoms of the fuel to repel from each other and the reaction to halt. Because fusion is so hard to keep going, it's a lot more fool-resistant than trying to build a fission reactor. It's also much much harder to extract a useful amount of energy -- it may not even be possible from this type of reactor because it takes so much to keep it running.
An H-bomb achieves the necessary pressure and temperature by setting off a conventional fission warhead, fusing all the available fuel in one shot. You'd never be able to achieve that kind of uncontrolled reaction in a jar without an equally violent trigger.
(We sort of already knew that, because as a major petroleum supplier they don't _need_ nuclear power, but the deuterium makes it even more obvious.)
Why does everybody assume that Iran's claim about civilian nuclear power generation is completely false?
As a major petroleum supplier they don't _need_ nuclear power, but they might _want_ it for a variety of reasons.
1. Less petroleum burned locally = More to sell
2. Pollution issues
3. If the reserves are getting low, they'd be the first to know (and probably wouldn't advertise it to they buyers who might panic and start looking at alternate energy sources)
Now, I think it would be naive to think that Iran isn't interested in developing nuclear weapons. I'd say the truth is probably somewhere in the middle -- they want nuclear power for civilian applications, but if it can help produce nuclear weapons it kills two birds with one stone.
Unfortunately, the most efficient type of reactor is also the best kind for producing weapons-grade materiel. Conventional designs use up uranium way too fast to be long-term viable.
Correct, Farnsworth Fusors are "hot" fusion devices. They use intense electromagnetic fields to accelerate particles rather than lasers or other more fanciful methods, but the end result is the same.
The problem with America is stupidity. I'm not saying there should be a capital punishment for stupidity, but why don't we just take the safety labels off of everything and let the problem solve itself?
I agree with the sentiment, unfortunately the fatality rate would probably be too low. Most stupid people would just end up injuring themselves, and as healthcare is becoming more and more socialized the rest of us would end up footing the bill...
If you immediately know the candlelight is fire, the meal was cooked a long time ago.
I'm hoping that maybe...just maybe with the many Dem.s that won seats....they seem to be more centrist that the liberal Dems of the recent past...maybe they will have a fiscal concious, that will look to cut pork, rather than dig deeper into our coI'm hoping that maybe...
Wait, so our only hope is conservative Democrats?
/my head a splode
:)
There's also no chance in hell that the Democrats will do anythingto cut government spending and everyone knows it. The Republicans have been as bad as the Democrats at that in the last several years but the Dems will never do it. You can bet they're already working on which taxes to increase and by how much and what gun control laws they can try to push through.
Oh, absolutely. Dems have always been about spending as much as possible on whatever social program catches their fancy.
What makes it so frustrating is that the Republican party used to be fiscally conservative. Now both of the major parties want to keep spending and spending even though the budget is already hemorrhaging. Just another reason why the two-party system sucks and never should have been allowed to happen (it's exactly what several of the founding fathers were afraid of happening).
I like lower taxes as much as anyone, but when the alternative is to have the economy crumble and all my cash become worthless, I'll vote for the tax.
Granted, the correct answer is to cut government spending, but that's something that will take a lot more political muscle to pull off.
I've looked at autopatcher before (and even used it on a spare machine just to see what the fuss was about). My only concern about it is that there's no guarantee about where the patches are coming from -- how can you be sure they haven't been tampered with?
Yes, I've done customized unattended installs before (even over netboot), but it's a pain because you have to build a new image every time you want to change something. Much more flexible to just do a clean install and run a script off a network location that sets everything up.
I don't think you can do it with the "offical" slipstreaming process. I've seen guides to doing that before, but basically they involve copying all of the setup files for the individual patches onto the CD and using a custom install script to apply the patches near the end of the install. Seems really inefficient plus you have to use an answer file which means your setup process is no longer interactive in a few places (notably disk partitioning IIRC).
It might also be possible to manually replace individual files on the CD with patched versions, but I'm leery of doing that in case there's also some registry changes or other issues that would get missed.
4our
I like it. I like it a lot.
Or better yet slipstream the SP into your install CD and only use that. Nothing to have to remember to apply. An "XP with SP2" CD is a much more secure way to install than an original XP CD and applying SP2 post install.
The main reason is because service packs can do what they call "slipstreaming". You apply the service pack against the original install CD, create an image from that, and burn yourself a new CD. When you install from your new CD you already have everything that was part of the service pack, so you have a lot less that needs to be downloaded after the install is complete. Since the files on the CD were updated, it doesn't take any longer to install the OS than normal (versus waiting 15 minutes for an SP to apply even when it's local). Great for admins who frequently build machines.
Don't tell me to just use ghost or dd or some other disk imaging solution. I've found that a fresh install is a lot cleaner (no filesystem resizing / conversion, no SID issues, no cruft in your image from when you logged in to set stuff up). Combined with a script to tweak some default settings and group policy to automatically install the appropriate software, it's just as automated as an image and doesn't take much longer. Waiting for 70 security updates to install (even from a local WSUS mirror) is probably the longest part of the whole process.
Yes, but the problem is that even with a slipstreamed SP2 (which is all I use anymore), there's still 60-70 updates that it has to download and install. That's a number which will only get bigger over the next year.
That's why we want SP3, so that it can be slipstreamed in and vastly reduce that number.