> Ever been on trial? Or served on a jury? Here's how it works: if you, at any point, invoke the 5th amendment, the judge instructs the jury to disregard that invocation. The judge instructs the jury not to consider the invocation of the 5th amendment to be incriminating in any way.
I've done neither, but being at trial is not the issue here. Your right to remain silent based on the 5th amandment exists outside the courtroom as well, for example when questioned by police, and obviously there is no jury and judge present in that situation, and it is also the kind of situation that this ruling applies to.
At any rate, I do agree with your reading of the 5th, btu I disagree with the extremely limited scope you seem to give to it. As with most things in law and esp. in a constitution, it is about the purpose and not the letter.
> Nope. By definition, invoking the 5th amendment is not an admission. That's how the system works.
That is how the system IS SUPPOSED to work, yes. Think about this a bit..
What I said is that when you only allow people to pleed the 5th in case the information they are demanded to give would incriminate them, then you have created a situation where pleading the 5th can only be interpreted as withholding incriminating information, and thus as proof that such information exists. Even more so when things go into the direction of you actually having to show that you have a reason to plead the 5th.
> In all honesty though, I am puzzled by it. I mean, Slashdot must generate some good ad revenue, so why can't they afford some decent editors? There are mods at forums working for free that do a better job.
> In other words, if the defendant HAD had a real and appreciable fear that his name would be used to incriminate him, he could have simply refused to answer on those grounds.
The problem with that reasoning is that in that case not givign your name is incriminating in itself.
The 5th amandment becomes a useless piece of paper that way.
> If a cop stops me on the street for no good reason and hassles me, I'll go along with it, as long as we're on the street and it's mano-a-mano. Once we're no longer on his playing field, the game changes. There's a time to assert your "rights," and on the street where the officer is on his home turf is not the best time...
The problem with that is that by going along with it you already lost the right to be free from unwarranted meddling by yolour government.
The uncalled for hassling is exactly the thign to fight.
They forgot the 'online' part a long time ago judging from whats being discussed on yro.. or it must mean: our ONLINE ranting about rights or such.. Still have to find out why 'you' is in the title tho.
Yes, but at this moment that is limited to 4.8 (and up for as far as I know) but doesn't run properly on 5.2.1. It also doesn't take advantage of SMP wen available for multithreading.
That is kinda sad because binaries built on 5.2.1 do work and do properly support smp.. and as things are, I am far more bothered by restrictions on distribution of such binaries then by not having Suns blessing really.
At any rate, I guess we'll have to wait till at least 5.3 before getting 5.x binaries and proper SMP and such... or till Sun opensources JAVA.. My bet is on 5.3 for now.
> if a virus can spread that requires users to unzip a password protected zip attachment, then run the executable do you have any hope whatsoever for operating system or anti-virus companies to solve the social engineering problem of email viruses
Good point, and against that there is little help from technical means.
> nothing short of education can fix these problems.
I completely agree that education is the key.
That said, I believe that the whole way that Windows works (and for that matter, KDE has a bit of this as well) is not helpfull because it teaches a user to not think and take whatever the computer does for granted. Empowering the user? fine. Making complicated things easy? fine. Acting as if complex things are easy? Completely wrong. Worst of all is giving the impression that you don't have to think about what you are doing tho.
Education also has to come in other forms, and you rightfully point at schools having a task in that.. I'd say that you as parent also have a task in it, even more so when you are a knowledgable user yourself. (sorry, just getting a bit annoyed with parents who don't think they have a responsibility in teaching their kids how to behave, not saying that you are one of them)
If you'd have clicked on the link n the header of my post and read up a bit, you would have found the question to your answer.
I still use Amigas, and have been using them since the late 80s.
I currently have two working machines, a modified Amiga 500 (68030, 8mb, scsi) and a modified Amiga 2000 (68040, 16mb, scsi)
I have seen a lot of viruses and trojans over the years, but no spyware and adware as we know it now. Of course, there was little point to those since most Amigas didn't have a net connection.
So I guess it depends on your definition of malware, but for me every bit of software that on purpose tries to mess up your machine while either hiding itself, or giving the impression of doing something else, qualifies as malware.
> Statistics? How about trying each for yourself? A fresh install of Windows XP requires 49, as stated by going to the Windows update site, security updates
Well, besides the fact that the post I replied to was comparing it to windows 2000.. are that 49 updates for remotely exploitable things or things that can be exploited without requiring user actions? In other words, can I expect to be able to install those updates without getting hacked/infected before having done so?
> What does a default install of RedHat require for the number of security vulnerabilies?
I don't use redhat myself, and I believe their distro sucks. Nonetheless, taking a peek at their info tells me that update 1 for the current enterprise server contains some 50 packages, most of which are not security updates but other bugfixes.
I use FreeBSD myself, and for a workstation install, there have been exactly 2 security issues in 4.9 that would be of importance, and both relate to openssh which is enabled by default. For a server install there are some 2 dozen security updates.
Of course that is excluding all kinds of additional packages that one might install. But that is just like not counting IIS vulnerabilities on the list of XP vulnerabilities.
> It would have to be signficantly less than 49.
Not really. the original poster was claiming it would be significantly more, and was comparing with Windows 2000 (which is used a lot more then XP for business use at the moment) It seems that at best it is similar.
Nescape lost in part due to IE getting bundled with WIndows, in part because NS made a crappy product themselves.
Same applies to Real.
That said, the problem is not so much that they bundle their products as such but that they close off the market for others.
What could be asked from them is to provide OEMs with a bare bones version of the OS and optional components. What they do now is easier, but it costs in the long term due to it closing off any competition on top of the windows platform.
TO me it sounds a lot better to simply force usage of open standards. Not by forcing Microsoft to use them so much but by the government demanding that any format used for government purposes is freely implementable.
This is both in the interest of the public and the government, and guarantees that documentation will be available independent of Microsoft or any other producer of software.
Due to Microsofts current position as by far the largest player for desktop operating system and productivity software, they should be forced by law to provide everyone equal access to the internals of their system. If they don't want that, they must allow a significant competitor to exist. I'd say a 2/3 market coverage would be a nice line for such a thing.
Asking them to stop integrating features however opposes their possibilities to develop. They may not be great at technological development maybe, but they are obviously good at developing things that appeal to the average person using a computer more then others do.
If their possibilities for lock-in are taken from them, then they have to maintain this position by means of quality and providing what their users need. As it is, they maintain it by creating incompatibility and lock-in and high cost of migration.
Such things are for as far as I am concerned what the law calls barriers to competition entering the market and should be dealt with.
> Sure there's a problem with it due to the fact that Symantec and McAfee and all of these companies sell AV software already. What if they didn't sell it yet? and MS beat them to the punch.
So.. instead of having MS pay fines and the like, they should be required to use open specifications and open and unencumbered standards.
That way they can inegrate what they like, but so can everyone else.
I agree btw with regards to it being a good thing if MS would provide anti-virus software with windows. How deeply it should be integrated into the system is another thing. The less it depends on the system, the less vulnerable it will be to being fooled by a compromised system into ignoring the compromise.
> Not quite correct. In recent versions of Outlook, executable attachments cannot be opened or saved without messing with the registry
Thats helpfull, tho for what I can see, this only works for new installations. At any rate...
> Microsoft does patch their OS quickly. The only problem is that many many people don't install the patches they provide.
While we can argue about if they do patch fast enough, you are definitely right about users not installing their patches.
When comparing Microsoft today to Microsoft 5 years ago, they have made a giant leap when it comes to security. That said, none of their systems was designed to be used by multiple users simultaneously, and the results of that are still deeply embedded in their designs.
> Linux is no more secure than Windows. I'm sure if you added up all the vulnerabilities in Windows 2000 and compared them to a list of vulnerabilities in all the software on a standard Linux distribution of the same age, Linux would have at least as many as Windows. The only reason Linux doesn't have the same bad reputation as Windows in terms of security is because there are many less Internet-facing Linux machines around,
First of all, I'd like to see some statistics on that because I strongly doubt there are more vulnerabilities.
But regardless, your statement is not true. The first reason for Linux being more secure is a stricter seperation between what is considered kernel and what is not.
This doesn't mean Linux or any Unix variation is flawless, they have their own problems, and one of the big ones is still privilege escalation due to setuid binaries/scripts.
Such bugs being exposed to remote attackers however happens a lot less often.
Because Linux and Windows mostly get used in different ways, its kinda pointless to really compare numbers anyway.
If you'd want to look at a situation where things compare a lot better, I'd look at IIS and Apache. While Apache's marketshare is bigger, IIS does have a substantial market, and in many cases they are in direct competition with eachother.
I'd really suggest looking at actually compromised machines over time for those two.
What I do know is that despite IIS having a smaller marketshare, the majority of exploit probes that I get in the logs of my webserver are IIS related.
> As Linux becomes a more accepted desktop OS, there will be worms attacking Linux machines, and its "secure" reputation will dissolve. Make sure you're ready for it, because it's not going to be pretty.
While often brought up, the marketshare argument doesn't match reality at all.
Besides the Apache/IIS example above, I suggest lookign at for example the Amiga platform.
While it has a fanatical group of followers still, and had a much larger group of followers in the late 80s and early 90s, it has never had a marketshare of any significance outside some niche markets.
Yet, viruses and malware are a substantial problem on this platform, and both had reached a maturity level that the PC equivalents took quite a few years to catch up with.
The Amiga platform also contains a few features and was surrounded by a culture that make it extremely vulnerable for particular kinds of malware, esp. bootsector infecting viruses. Disk images and disks being the primary way of exchanging software being a large factor in that.
At any rate, a platform needs to have enough users to allow any kind of succesfull virus or trojan, but beyond that popularity seems to be a minor factor, and ease of infecion seems to be a much larger factor.
> I STILL cannot understand why it is wrong for Microsoft to release products the compete with existing software.
In itself there is nothing wrong with that.
What is wrong is:
1. Using their monopoly in operating systems to give themself a technical advantage.
2. Using their operating system monopoly to give themselves a market advantage (by for example bundling it with their OS)
Why?
Because both result in it being impossible to compete with them, and as a result prevent competition. It is called anti-competitive behavior, and it is illegal if you have a monopoly already. Not having that illegal would mean allowing mega-corporations that determien every aspect of life and that are unchallangable.
So, while they may enter other markets, they may only do so without using their OS monopoly.
Outlook still allows running attachments, it is just an extra click to confirm you want to do it really.
The problem with this is that people are too used to clicking yes when asked and will do so here as well.
The only solution is to not allow it at all and to have people take very conscious and specific actuions (which preferably also demand knowing what they are doing before even being possible)
Inserting another click is not a solution. Requiring the user to think does go a long way to solving this.
Thunderbird at least requires you to save it to disk and run it outside thunderbird if it doesn't have a handler defined for a file. It wont allow you to run a random program with the file or run the file itself.
> Especially since Sourceforge points out to anyone starting a new project, very clearly, that they are responsible for backing up all of their project's data.
Which is obviously good advise. You should do so with your data at home (who does that..) but even more so with data you host somewhere else.
When the hosting is free as well as seemed to be the case here, thats yet more reason to keep your own backups..
-You can set the phone to go into speakerphone mode if you need to take notes or utilize the PDA portion while you're on the phone. It works very well, provided the ambient noise isn't too loud. You can also plug-in a headset or use a bluetooth headset to talk on the phone while using the PDA (I use that a lot).
I don't feel like letting the entire train I am in hear whatever message or number it is I am writing down really. A headset solves a large part of the issue btw, but doesn't solve my other issue, screensize.
-Most (if not all) PDA/Phone combos offer a "flight mode" that disables the transmitter/receiver when you're on a plane and turns the unit into a PDA.
So does mine, and yet I'm still told to turn it off in-flight, while they have no issue with me using my PDA at all.
Come back when IE realizes that fixed positioning is relative to the viewport and not to the canvas (and as a result, is not the same as absolute positioning)
That btw is just one of the many CSS things that are broken in IE, and one that you can't just turn on/off by using the correct doctype.
Regardless of which doctype, ie does nto comply with the w3c standards regardign size, or the latest sp for ie6 must have fixed that..
Slashdot Mirror
> Ever been on trial? Or served on a jury? Here's how it works: if you, at any point, invoke the 5th amendment, the judge instructs the jury to disregard that invocation. The judge instructs the jury not to consider the invocation of the 5th amendment to be incriminating in any way.
I've done neither, but being at trial is not the issue here. Your right to remain silent based on the 5th amandment exists outside the courtroom as well, for example when questioned by police, and obviously there is no jury and judge present in that situation, and it is also the kind of situation that this ruling applies to.
At any rate, I do agree with your reading of the 5th, btu I disagree with the extremely limited scope you seem to give to it. As with most things in law and esp. in a constitution, it is about the purpose and not the letter.
> Nope. By definition, invoking the 5th amendment is not an admission. That's how the system works.
That is how the system IS SUPPOSED to work, yes.
Think about this a bit..
What I said is that when you only allow people to pleed the 5th in case the information they are demanded to give would incriminate them, then you have created a situation where pleading the 5th can only be interpreted as withholding incriminating information, and thus as proof that such information exists.
Even more so when things go into the direction of you actually having to show that you have a reason to plead the 5th.
> In all honesty though, I am puzzled by it. I mean, Slashdot must generate some good ad revenue, so why can't they afford some decent editors? There are mods at forums working for free that do a better job.
I think you answered your own question...
> Fear is just another method to gain absolute power.
I don't like quoting the man in general, but Hitler said somethign rather relevant about that..
It translates to:
'Make the people feel so they won't think'
It does a good hob again and again... when you are interested in absolute power that is.
> In other words, if the defendant HAD had a real and appreciable fear that his name would be used to incriminate him, he could have simply refused to answer on those grounds.
The problem with that reasoning is that in that case not givign your name is incriminating in itself.
The 5th amandment becomes a useless piece of paper that way.
> If a cop stops me on the street for no good reason and hassles me, I'll go along with it, as long as we're on the street and it's mano-a-mano. Once we're no longer on his playing field, the game changes. There's a time to assert your "rights," and on the street where the officer is on his home turf is not the best time...
The problem with that is that by going along with it you already lost the right to be free from unwarranted meddling by yolour government.
The uncalled for hassling is exactly the thign to fight.
> With the AC you replied to modded down to -1, for someone viewing at 0, it would look like you are calling the younger-than-dos-guy an idiot...
And the lesson is, never jump to a conclusion without havign looked at all the information at your disposal (ie, readign at -1 in this case)
They forgot the 'online' part a long time ago judging from whats being discussed on yro.. or it must mean: our ONLINE ranting about rights or such..
Still have to find out why 'you' is in the title tho.
Yes, but at this moment that is limited to 4.8 (and up for as far as I know) but doesn't run properly on 5.2.1. It also doesn't take advantage of SMP wen available for multithreading.
That is kinda sad because binaries built on 5.2.1 do work and do properly support smp.. and as things are, I am far more bothered by restrictions on distribution of such binaries then by not having Suns blessing really.
At any rate, I guess we'll have to wait till at least 5.3 before getting 5.x binaries and proper SMP and such... or till Sun opensources JAVA.. My bet is on 5.3 for now.
All that said, this is of course a good step.
> if a virus can spread that requires users to unzip a password protected zip attachment, then run the executable do you have any hope whatsoever for operating system or anti-virus companies to solve the social engineering problem of email viruses
Good point, and against that there is little help from technical means.
> nothing short of education can fix these problems.
I completely agree that education is the key.
That said, I believe that the whole way that Windows works (and for that matter, KDE has a bit of this as well) is not helpfull because it teaches a user to not think and take whatever the computer does for granted. Empowering the user? fine. Making complicated things easy? fine. Acting as if complex things are easy? Completely wrong. Worst of all is giving the impression that you don't have to think about what you are doing tho.
Education also has to come in other forms, and you rightfully point at schools having a task in that.. I'd say that you as parent also have a task in it, even more so when you are a knowledgable user yourself. (sorry, just getting a bit annoyed with parents who don't think they have a responsibility in teaching their kids how to behave, not saying that you are one of them)
I should wake up before posting ;P
Of course.. you would have found the answer to your question instead of the question to your answer. ah well..
(goes for another cup of coffee)
If you'd have clicked on the link n the header of my post and read up a bit, you would have found the question to your answer.
I still use Amigas, and have been using them since the late 80s.
I currently have two working machines, a modified Amiga 500 (68030, 8mb, scsi) and a modified Amiga 2000 (68040, 16mb, scsi)
I have seen a lot of viruses and trojans over the years, but no spyware and adware as we know it now. Of course, there was little point to those since most Amigas didn't have a net connection.
So I guess it depends on your definition of malware, but for me every bit of software that on purpose tries to mess up your machine while either hiding itself, or giving the impression of doing something else, qualifies as malware.
> Statistics? How about trying each for yourself? A fresh install of Windows XP requires 49, as stated by going to the Windows update site, security updates
Well, besides the fact that the post I replied to was comparing it to windows 2000.. are that 49 updates for remotely exploitable things or things that can be exploited without requiring user actions? In other words, can I expect to be able to install those updates without getting hacked/infected before having done so?
> What does a default install of RedHat require for the number of security vulnerabilies?
I don't use redhat myself, and I believe their distro sucks. Nonetheless, taking a peek at their info tells me that update 1 for the current enterprise server contains some 50 packages, most of which are not security updates but other bugfixes.
I use FreeBSD myself, and for a workstation install, there have been exactly 2 security issues in 4.9 that would be of importance, and both relate to openssh which is enabled by default. For a server install there are some 2 dozen security updates.
Of course that is excluding all kinds of additional packages that one might install. But that is just like not counting IIS vulnerabilities on the list of XP vulnerabilities.
> It would have to be signficantly less than 49.
Not really. the original poster was claiming it would be significantly more, and was comparing with Windows 2000 (which is used a lot more then XP for business use at the moment)
It seems that at best it is similar.
Yes, people will.. but a lot less often then when its as easy as it is now.
You can forget fixing this problem completely, but it doesn't have to be as buig an issue as it is right now.
Nescape lost in part due to IE getting bundled with WIndows, in part because NS made a crappy product themselves.
Same applies to Real.
That said, the problem is not so much that they bundle their products as such but that they close off the market for others.
What could be asked from them is to provide OEMs with a bare bones version of the OS and optional components. What they do now is easier, but it costs in the long term due to it closing off any competition on top of the windows platform.
TO me it sounds a lot better to simply force usage of open standards. Not by forcing Microsoft to use them so much but by the government demanding that any format used for government purposes is freely implementable.
This is both in the interest of the public and the government, and guarantees that documentation will be available independent of Microsoft or any other producer of software.
Due to Microsofts current position as by far the largest player for desktop operating system and productivity software, they should be forced by law to provide everyone equal access to the internals of their system. If they don't want that, they must allow a significant competitor to exist. I'd say a 2/3 market coverage would be a nice line for such a thing.
Asking them to stop integrating features however opposes their possibilities to develop. They may not be great at technological development maybe, but they are obviously good at developing things that appeal to the average person using a computer more then others do.
If their possibilities for lock-in are taken from them, then they have to maintain this position by means of quality and providing what their users need. As it is, they maintain it by creating incompatibility and lock-in and high cost of migration.
Such things are for as far as I am concerned what the law calls barriers to competition entering the market and should be dealt with.
> And this is different from our present situation how?
That there are still some limits to what they can do, at least in some places on this planet.
> Sure there's a problem with it due to the fact that Symantec and McAfee and all of these companies sell AV software already. What if they didn't sell it yet? and MS beat them to the punch.
So.. instead of having MS pay fines and the like, they should be required to use open specifications and open and unencumbered standards.
That way they can inegrate what they like, but so can everyone else.
I agree btw with regards to it being a good thing if MS would provide anti-virus software with windows. How deeply it should be integrated into the system is another thing. The less it depends on the system, the less vulnerable it will be to being fooled by a compromised system into ignoring the compromise.
> Not quite correct. In recent versions of Outlook, executable attachments cannot be opened or saved without messing with the registry
Thats helpfull, tho for what I can see, this only works for new installations. At any rate...
> Microsoft does patch their OS quickly. The only problem is that many many people don't install the patches they provide.
While we can argue about if they do patch fast enough, you are definitely right about users not installing their patches.
When comparing Microsoft today to Microsoft 5 years ago, they have made a giant leap when it comes to security. That said, none of their systems was designed to be used by multiple users simultaneously, and the results of that are still deeply embedded in their designs.
> Linux is no more secure than Windows. I'm sure if you added up all the vulnerabilities in Windows 2000 and compared them to a list of vulnerabilities in all the software on a standard Linux distribution of the same age, Linux would have at least as many as Windows. The only reason Linux doesn't have the same bad reputation as Windows in terms of security is because there are many less Internet-facing Linux machines around,
First of all, I'd like to see some statistics on that because I strongly doubt there are more vulnerabilities.
But regardless, your statement is not true. The first reason for Linux being more secure is a stricter seperation between what is considered kernel and what is not.
This doesn't mean Linux or any Unix variation is flawless, they have their own problems, and one of the big ones is still privilege escalation due to setuid binaries/scripts.
Such bugs being exposed to remote attackers however happens a lot less often.
Because Linux and Windows mostly get used in different ways, its kinda pointless to really compare numbers anyway.
If you'd want to look at a situation where things compare a lot better, I'd look at IIS and Apache. While Apache's marketshare is bigger, IIS does have a substantial market, and in many cases they are in direct competition with eachother.
I'd really suggest looking at actually compromised machines over time for those two.
What I do know is that despite IIS having a smaller marketshare, the majority of exploit probes that I get in the logs of my webserver are IIS related.
> As Linux becomes a more accepted desktop OS, there will be worms attacking Linux machines, and its "secure" reputation will dissolve. Make sure you're ready for it, because it's not going to be pretty.
While often brought up, the marketshare argument doesn't match reality at all.
Besides the Apache/IIS example above, I suggest lookign at for example the Amiga platform.
While it has a fanatical group of followers still, and had a much larger group of followers in the late 80s and early 90s, it has never had a marketshare of any significance outside some niche markets.
Yet, viruses and malware are a substantial problem on this platform, and both had reached a maturity level that the PC equivalents took quite a few years to catch up with.
The Amiga platform also contains a few features and was surrounded by a culture that make it extremely vulnerable for particular kinds of malware, esp. bootsector infecting viruses. Disk images and disks being the primary way of exchanging software being a large factor in that.
At any rate, a platform needs to have enough users to allow any kind of succesfull virus or trojan, but beyond that popularity seems to be a minor factor, and ease of infecion seems to be a much larger factor.
> I STILL cannot understand why it is wrong for Microsoft to release products the compete with existing software.
In itself there is nothing wrong with that.
What is wrong is:
1. Using their monopoly in operating systems to give themself a technical advantage.
2. Using their operating system monopoly to give themselves a market advantage (by for example bundling it with their OS)
Why?
Because both result in it being impossible to compete with them, and as a result prevent competition. It is called anti-competitive behavior, and it is illegal if you have a monopoly already. Not having that illegal would mean allowing mega-corporations that determien every aspect of life and that are unchallangable.
So, while they may enter other markets, they may only do so without using their OS monopoly.
Outlook still allows running attachments, it is just an extra click to confirm you want to do it really.
The problem with this is that people are too used to clicking yes when asked and will do so here as well.
The only solution is to not allow it at all and to have people take very conscious and specific actuions (which preferably also demand knowing what they are doing before even being possible)
Inserting another click is not a solution. Requiring the user to think does go a long way to solving this.
Thunderbird at least requires you to save it to disk and run it outside thunderbird if it doesn't have a handler defined for a file. It wont allow you to run a random program with the file or run the file itself.
And so I host it myself ;P
> Especially since Sourceforge points out to anyone starting a new project, very clearly, that they are responsible for backing up all of their project's data.
Which is obviously good advise. You should do so with your data at home (who does that..) but even more so with data you host somewhere else.
When the hosting is free as well as seemed to be the case here, thats yet more reason to keep your own backups..
-You can set the phone to go into speakerphone mode if you need to take notes or utilize the PDA portion while you're on the phone. It works very well, provided the ambient noise isn't too loud. You can also plug-in a headset or use a bluetooth headset to talk on the phone while using the PDA (I use that a lot).
I don't feel like letting the entire train I am in hear whatever message or number it is I am writing down really. A headset solves a large part of the issue btw, but doesn't solve my other issue, screensize.
-Most (if not all) PDA/Phone combos offer a "flight mode" that disables the transmitter/receiver when you're on a plane and turns the unit into a PDA.
So does mine, and yet I'm still told to turn it off in-flight, while they have no issue with me using my PDA at all.
Uhm, install the linux-threads port? Mind pointing at the discussion you are refering to in the archives btw? because they don't seem to show it..
You could take a peek here for some info on how to get that to run.
Now.. could you come with a source that is a bit more verifiable then some AC claim?
Come back when IE realizes that fixed positioning is relative to the viewport and not to the canvas (and as a result, is not the same as absolute positioning)
That btw is just one of the many CSS things that are broken in IE, and one that you can't just turn on/off by using the correct doctype.
Regardless of which doctype, ie does nto comply with the w3c standards regardign size, or the latest sp for ie6 must have fixed that..