If I remember correctly, SEDA uses a few thread pools to handle clients at different stages of their work.
From his website: "We have built a number of applications to demonstrate the SEDA framework. Haboob is a a high-performance Web server including support for both static and dynamic pages that outperforms both Apache and Flash (which are implemented in C) on a SPECWeb99-like benchmark. Other applications include a Gnutella packet router and Arashi, a Web-based email service similar to Yahoo! Mail."
I remember seeing a sign that said "eye-dentiscan" or something like that, with the "eye" in a different font or color (don't remember exactly), which might make it seem like just "dentiscan". I took it to be a brandname for the specific retinal identification system that was being used. Might that be it?
There definitely was an MS-DOS 4:-). I have disks for it somewhere (albeit 5 1/4). The reason you don't hear much about it is because it sucked, hehe. For the life of me I can't remember _why_ it sucked, but I remember everyone sticking with 3.3 before skipping directly to 5 for some pretty good reason...
Seriously, though. Yes, Intel is unlikely to commit their own engineers to work on implementations for other chipsets. However, the JIT is fully isolated from the rest of the code, all of which is under a modified BSD license. Translation? Anyone can spend the time required to write the implementation, and I seriously doubt that Intel will pass up the _already written_ code (and even if they did, it should be simple enough for anyone to plug into the rest of the codebase as not to bee a major deterrent).
So, let's see here:
w/Intel
Intel Chips: Intel
Other Chips: Other (Intel?)
No Intel
Intel Chips: Other
Other Chips: Other
So if you work with Intel, you get Intel's support on their Chips, and the strong possibility that you will have to go elsewhere for other platforms. If you don't work with Intel, you don't get Intel's support at all.
ORP is designed as a runtime platform that has a very clean separation between the core, the JIT, and the garbage collector. The idea is that, right now, it is difficult to compare various algorithms as to implement them often requires a different core runtime (which could ruin the accuracy of the data). Initially, it operates over Java byte codes, but it was designed to be much more than that. In every press release MRL has released, they have never called it a JVM.
Recently, at Java Grande 2001, some people from MRL gave a talk about their work on ORP and put their slides online at their SourceForge project. In this presentation, it is mentioned that they have are "considering adding support for CLI". CLI stands for "Comman Language Infrastructure", the core parts of.NET, for those people who aren't keeping up with the flurry of 3-letter acronyms that begin with "CL".
Having someone like Intel behind a project to create an open-source runtime would lend it credibility to both end-users and corporate decision makers. It would also give powerful support to the development of JITs for Intel platforms.
A while back I attempted to get some support for a project that would involve talking to Intel (that Miguel de Icaza from Ximian can even be seen on the mailing list for) but was unable to gather enough people to make approaching Intel sound useful (this was also before Java Grande). As of right now, I've all but given up on that project; especially seeing that having a large project like Mono around (which is under GPL, instead of the modified BSD license of ORP, which gathers more of the pure "free software"/GNU advocates) would discourage individual contributors and organizations (such as MRL) from commiting much of their efforts.
I have been thinking about approaching Apache or FreeBSD about trying to put together a project with Intel to work on this. If anyone is interested in this, has opinions, comments, questions, concerns, arguments, just plain wants to send me death threats (hey, I like e-mail), feel free to send them to saurik@saurik.com.
It's important to note that Rhys' project is concentrating on building a light-weight interpreter for embedded systems, and has no real goal of constructing a full featured runtime for Unix. Really, he is most interested in selling compiler technology.
In a message to his mailing list yesterday he once again stressed his dedication to his compiler technology and even mentioned that, in order to diversify his company's offerings, he is likely to focus more on re-compiling.NET to work on the JVM.
I was working on an alternative project with the goal of using Intel's ORP to build a fully featured runtime, but due to lack of support I've all but dropped the project. One of my earlier project statements (which was posted to Rhys' mailing list) can be found at:
This was being worked on a few years ago by some people at The University of Queensland. Unfortunately, they got tired of the project (and, if I remember correctly, that they weren't getting much popular support).
"UQBT - A Resourceable and Retargetable Binary Translator"
To note, they mention that they got some funding from Sun for a few years. (Likely either causing or due to their work on writing a gcc compiler back-end that emits Java byte-codes.)
While not a good explanation, you may be interested in one of the main uses of said information: lowering taxes. If your house has been appraised to be XXX dollars, but you feel that is too much, when re-appraisals roll around you can argue that your house costs too much and site "comparables", other houses in your neighborhood which are similar to yours (same neighborhood, same township, similar size) that have been appraised to be a lower amount, or have even sold for lower amounts (to prove that it may not have been worth much to begin with). This is really the best argument you have during this time.
If you want to see some worse data, check out Cook county (includes Chicago). From http://www.cookcountyassessor.com/ (direct to search: http://www.cookcountyassessor.co m/startsearch.html) you can not only find out the appraisal information, you can see the number of bathrooms, bedrooms, state of the attic, basement, size of garage, really a full break down of the entire state of the house.
I don't see much wrong with it myself (most of this was all in listings and such when I bought the house in the first place), but the shear amount of information that can be compiled does hit me back sometimes:).
But the system as a whole is worthless. The desktop can be repaired in the matter of a couple hours, totally rebuilt from scratch in no more than a day of installations that require only minimal thought. The only thing... the ONLY thing of worth on any of my computers are the data files that I work on. Financial data... e-mail... business documents... source code... the very things that are usually owned by _regular users_ who read e-mail. As I said in my comment:
<snip> The only thing that is hopefully protected are the libraries, programs, and misc. config files (which I, at least, don't value very greatly seeing that they are rather replacable parts with nowhere near the importance attached to them as my user data... destroy all the dll's and exe's on a Windows machine and its at most a day long frustration). </snip>
If the reason you mention this is because of multiple users (where keeping the system running might also be considered important, although everywhere I have worked we would have rather lost the server than even one document of data...), then here we see a "weakness" on the part of UNIX systems. With a Windows NT file server, noone would ever use the computer for reading e-mail, and therefor never have the opportunity to delete any OS files anyway. Much rather shift the danger over to disposable workstations. In this case, if they fry their computer, I could just unplug the computer from their stuff, drop in a replacement, and with the file server they won't notice the difference.
Oh, and your last comment:
> Outlook executing scripts automatically? > Not OS dependant?
Outlook doesn't execute anything automatically. You have to double click the attachment (which any e-mail client sufficiently easy for working with attachments does (such as Netscape)). I believe the best comment I have seen about this was from "YU Nicks NE Way" (responding to "Microsoft Patents Package Management"):
<reprint> Sorry, you lose, but thanks for playing. That's a good/. urban myth, but it isn't true. Neither Outlook nor Outlook Express will execute WSH code when an e-mail is read or previewed. All the "experts" who were claiming outherwise have retracted their claims. Turns out that the only mailer anyone can think of that will run dangerous script when an e-mail is opened is Gnu Emacs. </reprint>
If what you want is a difficult route to executing a file (which would probably get the people who know not to execute files even madder), by all means you can configure your system to do this. A configuration issue, which UNIX people don't want held against them on their systems ("but this isn't secure out of the box!" IS a rather annoying argument, hehe), so why hold it against us:).
If you have the option of attaching files to e-mail, then you have this problem. People target Windows only because so many people use it.
If I attach a shell script to an e-mail and send it to someone running Linux/SunOS/etc., and he is stupid enough to run it, there is no inherent protection given by those OSs to keep his files from being wiped out. The script can delete all the data files that person has been working on, all the rest of his e-mail, SEND the exploit to other users of the system (and even your contacts if the script was designed to try the address lists of various mail clients the user might be using), modify the user's login script to make any of the above happen every login, etc.. The only thing that is hopefully protected are the libraries, programs, and misc. config files (which I, at least, don't value very greatly seeing that they are rather replacable parts with nowhere near the importance attached to them as my user data... destroy all the dll's and exe's on a Windows machine and its at most a day long frustration).
The script could even use the login script to add an alias for su to some ~/.su file that it writes in the hope that this is a computer that only this one user works on.
The problem here has nothing at all to do with Windows or any inherent "lack of security" on its part, and has _only_ to do with users who don't believe that running programs or scripts can cause that much damage (as you put, "Windows Dummies"). But what is your solution to that? Just giving another OS to people doesn't make them smarter. It might make them much more frustrated when they have to think about everything they want to do, that's true... but some people just aren't going to get to get any better. I know a president of a bank (which shall remain nameless) who has his secretary handle all of his e-mail, converting it along the way to paper compatible with his in/out box interface to the world.
Then there is the problem that even if you did change everyone over to a more complex operating system, even got them to the point where they could write their own programs for it, they would continue sending each other little joke e-mails and executable files, and eventually stumble on a virus. The people who worked for a company I worked for last year did this all the time. They all had access to all the data files on the network from their accounts (legacy flat file database), all the documents, etc.; and when the company started scanning for viruses regularly it was found they ended up infecting over a thousand files. Not because they didn't know what the risks were, but because they enjoyed opening the executables and watching the animated jokes that unfolded from them (most of which likely didn't even contain a virus). They would get the thing forwarded from one of their friends, then they would forward it to all their friends, etc..
Everyone who is making comments like about how Microsoft is patenting package management or automation ("Microsoft Patented CRON" was my favorite example of this one) needs to read the referenced patents.
If you actually look, this patent is a modification to existing patents by other companies, such as Compaq's 1996 "US5586304: Automatic computer upgrading" ( http://www.patents.ibm.com/details?&pn10=US5586304), which begins with "A method for use in upgrading a resource of a computer from an existing version of the resource to a later version of the resource.", which seems to conflict with just about every helpful installation program ever created.
How about IBM's 1995 "US5473772: Automatic update of static and dynamic files at a remote network node in response to calls issued by or for application programs" ( http://www.patents.ibm.com/details?&pn10=US0547377 2) which seems to cover applications like rsync and maybe even CVS.
All that Microsoft's patent is is a small extension of a few other patents (Apple already got automating program upgrade in 1998, IBM got automatic maintenance, an earlier Microsoft patent covering remote software discovery (probably for Windows Update like behavior), and Samsung's system of letting the app developer update a server and have it distribute down to clients) making them more focused for what they are trying to do. BUT, everyone chooses to attack this patent:).
Xerox's >>1985 http://www.patents.ibm.com/details?&pn10=US0455841 3) actually seems to have the source code based package management systems covered (although it might be just different enough to be talking more about source code management than of binary applications).
There were a bunch of other goodies there but I am already way past when I was supposed to leave this morning to point them out:).
Well, one more I just hit on... IBM's 1987 "US4649473: Flexible data transmission for message based protocols" ( http://www.patents.ibm.com/details?&pn10=US4649473). This one seems pretty close to Microsoft's Message Queue system... Maybe someone should point that out to someone, hehe.
One thing that has been bothering me is the software RAID support patches from ftp.kernel.org. Does anyone know whether the stock kernels are going to be supporting the 0.90 raid tools in future versions? (For all I know, they already are...) What I am talking about are the patches from ftp://ftp.us.kernel.org/pub/linux/daemons/raid/alp ha/ . Sometimes I wonder whether or not I should be installing some of the RAID patches on a kernel since it almost seems as if the change might have been merged in already, but not quite...
Red Hat does have an FTP installation option, as well as an SMB one. FezBox simply chose not to support those (along with the installation from local harddrive option). This is probably because those three methods of installation require extra drivers/applications above and beyond what is built into the kernel (ok, I don't know 100% why the harddrive option was like this), and Red Hat used to require (as of 5.2) a "supplemental diskette" for these installations (now, in 6.0 and RawHide, there are actually two diskettes, one for network installations (bootnet) such as FTP, SMB, NFS, etc., and one for non-network installations (boot)).
Hehe, I started removing the doubles with the graphical X's, which then removed both copies instead of one, and then tried to put them all back from the preferences page, and now I have NO slashboxes, and I can't add any either:(.
I think he meant that he was the first person to post a comment to this item, but since two other people posted during the same minute (5:45-5:46), he failed in his quest:).
Of course Outlook supports POP3, but that's just for getting mail. You couldn't store your mail in addition to your calendar, todo list, etc. on the server. Using IMAP instead will at least give you the ability to store your mail on the server, but Microsoft has chosen not to support storing anothing other than mail on an IMAP mail store. OpenMail gives all these abilities back to you.
It isn't designed for the individual who justs wants mail, and shouldn't be read as such. Its for the corporation who wants to do scheduling and the such off their e-mail system, and don't want to use NT to do it.
I can't speak for whether or not the product itself works well, or if it crashes all the time as people have claimed, but I do know that the arguments being placed against it as far as Oulook connectivity and its pointlessness are far from on target.
Slashdot Mirror
Another model is SEDA ("Staged Event-Driven Architecture"). This was mainly examined by the guy who wrote what became Java's new java.nio package.
Here's a link: http://www.cs.berkeley.edu/~mdw/proj/sandstorm/
If I remember correctly, SEDA uses a few thread pools to handle clients at different stages of their work.
From his website:
"We have built a number of applications to demonstrate the SEDA framework. Haboob is a a high-performance Web server including support for both static and dynamic pages that outperforms both Apache and Flash (which are implemented in C) on a SPECWeb99-like benchmark. Other applications include a Gnutella packet router and Arashi, a Web-based email service similar to Yahoo! Mail."
I remember seeing a sign that said "eye-dentiscan" or something like that, with the "eye" in a different font or color (don't remember exactly), which might make it seem like just "dentiscan". I took it to be a brandname for the specific retinal identification system that was being used. Might that be it?
There definitely was an MS-DOS 4 :-). I have disks for it somewhere (albeit 5 1/4). The reason you don't hear much about it is because it sucked, hehe. For the life of me I can't remember _why_ it sucked, but I remember everyone sticking with 3.3 before skipping directly to 5 for some pretty good reason...
Seriously, though. Yes, Intel is unlikely to commit their own engineers to work on implementations for other chipsets. However, the JIT is fully isolated from the rest of the code, all of which is under a modified BSD license. Translation? Anyone can spend the time required to write the implementation, and I seriously doubt that Intel will pass up the _already written_ code (and even if they did, it should be simple enough for anyone to plug into the rest of the codebase as not to bee a major deterrent).
So, let's see here:
w/Intel
Intel Chips: Intel
Other Chips: Other (Intel?)
No Intel
Intel Chips: Other
Other Chips: Other
So if you work with Intel, you get Intel's support on their Chips, and the strong possibility that you will have to go elsewhere for other platforms. If you don't work with Intel, you don't get Intel's support at all.
How about the total absence of Mono implementations on any platform?
The thing that bugs me about this is that there is already a quite good, open-source runtime out there. Intel has charged their Microprocessor Research Laboratory (MRL) with the creation of ORP, the Open Research Platform.
.NET, for those people who aren't keeping up with the flurry of 3-letter acronyms that begin with "CL".
ORP is designed as a runtime platform that has a very clean separation between the core, the JIT, and the garbage collector. The idea is that, right now, it is difficult to compare various algorithms as to implement them often requires a different core runtime (which could ruin the accuracy of the data). Initially, it operates over Java byte codes, but it was designed to be much more than that. In every press release MRL has released, they have never called it a JVM.
Recently, at Java Grande 2001, some people from MRL gave a talk about their work on ORP and put their slides online at their SourceForge project. In this presentation, it is mentioned that they have are "considering adding support for CLI". CLI stands for "Comman Language Infrastructure", the core parts of
Having someone like Intel behind a project to create an open-source runtime would lend it credibility to both end-users and corporate decision makers. It would also give powerful support to the development of JITs for Intel platforms.
A while back I attempted to get some support for a project that would involve talking to Intel (that Miguel de Icaza from Ximian can even be seen on the mailing list for) but was unable to gather enough people to make approaching Intel sound useful (this was also before Java Grande). As of right now, I've all but given up on that project; especially seeing that having a large project like Mono around (which is under GPL, instead of the modified BSD license of ORP, which gathers more of the pure "free software"/GNU advocates) would discourage individual contributors and organizations (such as MRL) from commiting much of their efforts.
I have been thinking about approaching Apache or FreeBSD about trying to put together a project with Intel to work on this. If anyone is interested in this, has opinions, comments, questions, concerns, arguments, just plain wants to send me death threats (hey, I like e-mail), feel free to send them to saurik@saurik.com.
It's important to note that Rhys' project is concentrating on building a light-weight interpreter for embedded systems, and has no real goal of constructing a full featured runtime for Unix. Really, he is most interested in selling compiler technology.
.NET to work on the JVM.
In a message to his mailing list yesterday he once again stressed his dedication to his compiler technology and even mentioned that, in order to diversify his company's offerings, he is likely to focus more on re-compiling
I was working on an alternative project with the goal of using Intel's ORP to build a fully featured runtime, but due to lack of support I've all but dropped the project. One of my earlier project statements (which was posted to Rhys' mailing list) can be found at:
What I'm Planning: Alternative Project w/ Slightly Different Goals (was: Suggestions) [ http://lists.saurik.net/message.xsp?id=62 ]
This was being worked on a few years ago by some people at The University of Queensland. Unfortunately, they got tired of the project (and, if I remember correctly, that they weren't getting much popular support).
Their website is at :
http://www.csee.uq.edu.au/~csmweb/uqbt.html
"UQBT - A Resourceable and Retargetable Binary Translator"
To note, they mention that they got some funding from Sun for a few years. (Likely either causing or due to their work on writing a gcc compiler back-end that emits Java byte-codes.)
While not a good explanation, you may be interested in one of the main uses of said information: lowering taxes. If your house has been appraised to be XXX dollars, but you feel that is too much, when re-appraisals roll around you can argue that your house costs too much and site "comparables", other houses in your neighborhood which are similar to yours (same neighborhood, same township, similar size) that have been appraised to be a lower amount, or have even sold for lower amounts (to prove that it may not have been worth much to begin with). This is really the best argument you have during this time.
:).
If you want to see some worse data, check out Cook county (includes Chicago). From http://www.cookcountyassessor.com/ (direct to search: http://www.cookcountyassessor.co m/startsearch.html) you can not only find out the appraisal information, you can see the number of bathrooms, bedrooms, state of the attic, basement, size of garage, really a full break down of the entire state of the house.
I don't see much wrong with it myself (most of this was all in listings and such when I bought the house in the first place), but the shear amount of information that can be compiled does hit me back sometimes
But the system as a whole is worthless. The desktop can be repaired in the matter of a couple hours, totally rebuilt from scratch in no more than a day of installations that require only minimal thought. The only thing... the ONLY thing of worth on any of my computers are the data files that I work on. Financial data... e-mail... business documents... source code... the very things that are usually owned by _regular users_ who read e-mail. As I said in my comment:
/. urban myth, but it isn't true. Neither Outlook nor Outlook Express will execute WSH code when an e-mail is read or previewed. All the "experts" who were claiming outherwise have retracted their claims. Turns out that the only mailer anyone can think of that will run dangerous script when an e-mail is opened is Gnu Emacs.
:).
<snip>
The only thing that is hopefully protected are the libraries, programs, and misc. config files (which I, at least, don't value very greatly seeing that they are rather replacable parts with nowhere near the importance attached to them as my user data... destroy all the dll's and exe's on a Windows machine and its at most a day long frustration).
</snip>
If the reason you mention this is because of multiple users (where keeping the system running might also be considered important, although everywhere I have worked we would have rather lost the server than even one document of data...), then here we see a "weakness" on the part of UNIX systems. With a Windows NT file server, noone would ever use the computer for reading e-mail, and therefor never have the opportunity to delete any OS files anyway. Much rather shift the danger over to disposable workstations. In this case, if they fry their computer, I could just unplug the computer from their stuff, drop in a replacement, and with the file server they won't notice the difference.
Oh, and your last comment:
> Outlook executing scripts automatically?
> Not OS dependant?
Outlook doesn't execute anything automatically. You have to double click the attachment (which any e-mail client sufficiently easy for working with attachments does (such as Netscape)). I believe the best comment I have seen about this was from "YU Nicks NE Way" (responding to "Microsoft Patents Package Management"):
<reprint>
Sorry, you lose, but thanks for playing. That's a good
</reprint>
If what you want is a difficult route to executing a file (which would probably get the people who know not to execute files even madder), by all means you can configure your system to do this. A configuration issue, which UNIX people don't want held against them on their systems ("but this isn't secure out of the box!" IS a rather annoying argument, hehe), so why hold it against us
If you have the option of attaching files to e-mail, then you have this problem. People target Windows only because so many people use it.
:).
If I attach a shell script to an e-mail and send it to someone running Linux/SunOS/etc., and he is stupid enough to run it, there is no inherent protection given by those OSs to keep his files from being wiped out. The script can delete all the data files that person has been working on, all the rest of his e-mail, SEND the exploit to other users of the system (and even your contacts if the script was designed to try the address lists of various mail clients the user might be using), modify the user's login script to make any of the above happen every login, etc.. The only thing that is hopefully protected are the libraries, programs, and misc. config files (which I, at least, don't value very greatly seeing that they are rather replacable parts with nowhere near the importance attached to them as my user data... destroy all the dll's and exe's on a Windows machine and its at most a day long frustration).
The script could even use the login script to add an alias for su to some ~/.su file that it writes in the hope that this is a computer that only this one user works on.
The problem here has nothing at all to do with Windows or any inherent "lack of security" on its part, and has _only_ to do with users who don't believe that running programs or scripts can cause that much damage (as you put, "Windows Dummies"). But what is your solution to that? Just giving another OS to people doesn't make them smarter. It might make them much more frustrated when they have to think about everything they want to do, that's true... but some people just aren't going to get to get any better. I know a president of a bank (which shall remain nameless) who has his secretary handle all of his e-mail, converting it along the way to paper compatible with his in/out box interface to the world.
Then there is the problem that even if you did change everyone over to a more complex operating system, even got them to the point where they could write their own programs for it, they would continue sending each other little joke e-mails and executable files, and eventually stumble on a virus. The people who worked for a company I worked for last year did this all the time. They all had access to all the data files on the network from their accounts (legacy flat file database), all the documents, etc.; and when the company started scanning for viruses regularly it was found they ended up infecting over a thousand files. Not because they didn't know what the risks were, but because they enjoyed opening the executables and watching the animated jokes that unfolded from them (most of which likely didn't even contain a virus). They would get the thing forwarded from one of their friends, then they would forward it to all their friends, etc..
God, I'm rambling...
Everyone who is making comments like about how Microsoft is patenting package management or automation ("Microsoft Patented CRON" was my favorite example of this one) needs to read the referenced patents.
4 ), which begins with "A method for use in upgrading a resource of a computer from an existing version of the resource to a later version of the resource.", which seems to conflict with just about every helpful installation program ever created.
7 2) which seems to cover applications like rsync and maybe even CVS.
:).
1 3) actually seems to have the source code based package management systems covered (although it might be just different enough to be talking more about source code management than of binary applications).
:).
3 ). This one seems pretty close to Microsoft's Message Queue system... Maybe someone should point that out to someone, hehe.
If you actually look, this patent is a modification to existing patents by other companies, such as Compaq's 1996 "US5586304: Automatic computer upgrading" ( http://www.patents.ibm.com/details?&pn10=US558630
How about IBM's 1995 "US5473772: Automatic update of static and dynamic files at a remote network node in response to calls issued by or for application programs" ( http://www.patents.ibm.com/details?&pn10=US054737
All that Microsoft's patent is is a small extension of a few other patents (Apple already got automating program upgrade in 1998, IBM got automatic maintenance, an earlier Microsoft patent covering remote software discovery (probably for Windows Update like behavior), and Samsung's system of letting the app developer update a server and have it distribute down to clients) making them more focused for what they are trying to do. BUT, everyone chooses to attack this patent
Xerox's >>1985 http://www.patents.ibm.com/details?&pn10=US045584
There were a bunch of other goodies there but I am already way past when I was supposed to leave this morning to point them out
Well, one more I just hit on... IBM's 1987 "US4649473: Flexible data transmission for message based protocols" ( http://www.patents.ibm.com/details?&pn10=US464947
One thing that has been bothering me is the software RAID support patches from ftp.kernel.org. Does anyone know whether the stock kernels are going to be supporting the 0.90 raid tools in future versions? (For all I know, they already are...) What I am talking about are the patches from ftp://ftp.us.kernel.org/pub/linux/daemons/raid/alp ha/ . Sometimes I wonder whether or not I should be installing some of the RAID patches on a kernel since it almost seems as if the change might have been merged in already, but not quite...
Red Hat does have an FTP installation option, as well as an SMB one. FezBox simply chose not to support those (along with the installation from local harddrive option). This is probably because those three methods of installation require extra drivers/applications above and beyond what is built into the kernel (ok, I don't know 100% why the harddrive option was like this), and Red Hat used to require (as of 5.2) a "supplemental diskette" for these installations (now, in 6.0 and RawHide, there are actually two diskettes, one for network installations (bootnet) such as FTP, SMB, NFS, etc., and one for non-network installations (boot)).
Hehe, I started removing the doubles with the graphical X's, which then removed both copies instead of one, and then tried to put them all back from the preferences page, and now I have NO slashboxes, and I can't add any either :(.
I think he meant that he was the first person to post a comment to this item, but since two other people posted during the same minute (5:45-5:46), he failed in his quest :).
Of course Outlook supports POP3, but that's just for getting mail. You couldn't store your mail in addition to your calendar, todo list, etc. on the server. Using IMAP instead will at least give you the ability to store your mail on the server, but Microsoft has chosen not to support storing anothing other than mail on an IMAP mail store. OpenMail gives all these abilities back to you.
It isn't designed for the individual who justs wants mail, and shouldn't be read as such. Its for the corporation who wants to do scheduling and the such off their e-mail system, and don't want to use NT to do it.
I can't speak for whether or not the product itself works well, or if it crashes all the time as people have claimed, but I do know that the arguments being placed against it as far as Oulook connectivity and its pointlessness are far from on target.