Reverse Polish notation, also called "postfix notation", has the arguments first, followed by the operator. For example, "3 2 +". Lisp programs are typically written directly in the language's abstract syntax, using the prefix notation and parens with which many are familiar. For example, "(+ 3 2)".
Common Lisp's reader (aka parser) is, however, fully under program control. Through CL's readtables, one can extend the basic prefix notation, e.g. CL's sharpsign-S macro to read in structure objects is "#S(typename:field1 val1...)" as opposed to calling MAKE-typename the usual way "(MAKE-typename:field1 val1...)". One particulary perverse use of Common Lisp's readtables is a package that implements an infix syntax for basic arithmetic operators, e.g. "#I(2 + 3)".
Unlike static languages like C or Java, almost everything about the Common Lisp run-time environment is under program control. Even the object system can be modified (through something called the Meta Object Protocol). And idiomatic Common Lisp code is often times within a factor of two of idiomatic C code in terms of performance.
There are some serious problems with biometric identification systems, including:
privacy concerns (who stores and distributes biometric data?)
religious concerns (Christian groups have often associated bio IDs with "the mark of the beast")
environmental concerns (e.g. hygiene)
personal concerns (the stigmata of criminal association, as with fingerprints; or: some people have fingers cut off, or are mute, or have cataracts, etc.)
biometric IDs are subject to all the same compromises as passwords (e.g. replay), but you can't change a biometric trait as easily as you can a password
Gartner wrote several great studies of various identification systems. Take a look at "Are Smart Cards Doomed to Die" by Andrew Phillips (a market analysis published 26 Apr 200) and "Security Applications of Biometrics: Perspective" by Ant Allan (a technology overview published 12 Feb 2001).
How difficult is it to configure one's web browser so that it rejects most of the scripting junk out there? If you are using IE, check out the security zones feature that allows you to toggle scripting, cookies, and so forth depending on to which of four security zones a particular site belongs. I'm sure the free browsers have something much more sophisticated. Use it!
I'm not as much concerned with speed as with feature set and consistency. Reliable and consistent 416-Kbps symmetric DSL and routable IP address space---like a/27 or/26---would entice me away from my reliable and semi-consistent 2-Mbps CATV connection with only four (freaking!) IPs. But my connectivity needs aren't typical of your average Unreal freak, and unfortunately, the DSL market sucks and the cable providers aren't really in the biz market at the low end.
It's all rather frustrating.
Er, I had a point. Oh well.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
My setup isn't as sophisticated as I'd like, but it gets the wires off the floor. I have two enclosures, an old AT&T DataPhone cabinet (about 1.5m tall by.65m wide by.5m deep) and a free-standing "ladder"-style rack (3m tall, standard width). The DataPhone cabinet holds two DDS-2 tape drives, a keyboard/monitor switch, and four computers: two "pizza box"-style Intergraph TD-40 workstations, a 486 in a standard tower case, and a Packard Bell desktop. The rack holds most of the networking kit in the apartment (both firewalls, the cable modem, an Ethernet hub, the Synoptics chassis, and the Token Ring MAU). The disk array chassis and the UPS remain freestanding.
The most important reason to throw stuff in a rack is cable management. Common cabling is bundled up neatly and hung off the screw holes on the enclosures. Bundles of CAT-5 cable are snugly wrapped in masking tape every half meter, with Ethernet in a bundle separate from the Token Ring. Power cabling runs up the opposite side of the rack from the data cabling, and I have taken care to cross power and data cables at right angles to minimize cross-talk. Almost everything is labeled neatly. If you want to spend the money, you can color code your cabling (I usually do this when doing wiring jobs for money).
Be careful when using tie-wraps: They can be cinched too tight and cause internal breaks in delicate wiring (e.g. SCSI or UTP cabling). I usually use wire ties like you get with boxes of trash bags to truss everything in place.
Make friends with other sysadmins. A friend of mine gave me all sorts of nice telecomms hardware (including the Dataphone cabinet and brackets for running wiring) in trade for helping him clean out his machine room. I regularly go through the trash at work looking for useful items, e.g. an HP SCSI enclosure or Sun monitor cables. I once found two 4-GB SCSI disks in the trash and was able to make do with them for several years until I bought enough disks for a software RAID array. A consultant friend of mine was able to get four dual-processor Intergraph workstations for less than USD 300---all with 128 MB RAM, on-board Ethernet and SCSI. Even though they're Pentium-133s, Windows, Linux, and Solaris fully support the hardware and run great as domain controllers, file servers, and Unix workstations. Another guy traded me a monitor for a working SGI Indy that was loaded with memory. Make new friends and keep the old, my aunt always said.
I have something like 20 computers in my apartment, from an old IMSAI 8080 on up to a DEC Alpha. While I'd be really hurting if I had to replace it all, I think I've purchased maybe one complete system out of the whole lot. If you're having a hard time finding racks and such, nose around the local swap meets or hamfests, contact wholesalers and salvagers, and hang out around a college data center. Somebody's bound to get rid of something useful, and most times they'll just give it away to be rid of it.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
And they breathe...how?
on
Eco-Terrorism
·
· Score: 1
You're both asking a much deeper question than just "was John Brown right". The real question is, does the end ever justify the means? And to be honest, I don't really have a good answer for or against. Love your neighbor, love the Increate---that's pretty much the only ethical foundation most of us have, and it's not enough to answer this question.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
What if you can't boot to a working system? Thank God NetBSD would desplay interrupt and Ethernet MAC ID, because it would crash hard soon after and that was the info I needed to make things work.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Please contact me via email. I would gladly take the machine off your hands for the price of shipping. It would have a good home, right next to the Alpha, also running OpenVMS (thank god for the hobbyist program).
P.S. I am serious.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
You're wrong. Even the regular XF86_SVGA server must directly access the IO space, and it does so through the kernel. On my Alpha, when I still ran Linux on it, I had nothing but problems with X. A perverse combination of user operations (e.g. a fill or scroll) would completely lock up the machine.
Yes, there are times when some poorly-written X app kills the server, and there are times when the X server bugs out without killing the OS. But your statement "X and the Kernel are clearly different" is a gross oversimplification, ignoring, iovec(), memory-mapped IO, and the like. Even the most lowly frame buffer drivers must directly access the hardware.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
A friend of mine was getting an error from the 2000 installer that sounds similar to yours. Basically, just before it finished formatting the 30 GB IDE disk (either NTFS or FAT), it would crash with an error message. Of course, it worked fine under Windows 98 and DOS (dunno if he tried Linux). It turned out that he had defective memory. Upon replacement of the DIMM, 2000 installed without complain.
Maybe you should have your memory checked out?
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
If you're poor, I would assume that meant you didn't have $52.28 to spend buying a piece of paper. As a college student supporting myself and making less than $9K a year, even $10 was a lot of money and not to be squandered.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
That's the other "his" point. Both Kerberos and NTLM require unique account names within an administrative domain and can't (easily) use DNs in the place of the User Principal Name or the SAM Account Name.
And yes, the old (NT4) domain system was basically a flat file (the SAM database), but this isn't the reason behind the requirement for unique UPNs in Active Directory. Kerberos too uses a flat database, and there has traditionally been no way to partition a realm (that is to say, the entire realm is propagated among masters and slaves in Kerberos). The replication and partitioning limitations in Active Directory are derived from AD's roots in Kerberos, not NT4.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
I'm in the process of building a system that integrates Windows and
Unix logon data. My environment currently includes Debian GNU/Linux,
Microsoft Windows, NetBSD, Silicon Graphics IRIX, and Sun Solaris, and
I hope to bring Compaq OpenVMS and Apple MacOS (pre-OSX) clients into
the mix within the next six to twelve months. What makes a difficult
project even more challenging is that I'm attempting to integrate file
sharing as well as domain logon. The environment I'm building is in
it's third generation.
The biggest problem with Active Directory has nothing to do with
Kerberos, and everything to do with getting directory information out
to clients in the realm. For example, NetBSD doesn't currently use
the PAM mechanism, and Solaris has a difficult time understanding
Microsoft's somewhat peculiar schema. To this end, I've decided to
use NIS to publish certain directory data to my Unix clients (user
name, UID, GID, GECOS, home, and shell; group name, GID, and members), and
revisit client directory access issues as the technology matures. And
since I can't get SFU to install on my DC (the DC's DNS domain name
and the AD domain name are different, and this causes problems for
SFU's installer), I've resorted to creating the necessary auxiliary
classes by hand and using some glue code to dump user and group
information from LDAP into the relevant NIS databases.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
I mean, a web server named after a punctuation mark? How do you pronounce "!"? I'm going to get some weird looks from my hacker buddies if I say I'm running Microsoft's GPL-ed IPv6 web server, "bang".
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
I finally understand why planet-wide exploration and colonization didn't really happen until Columbus. It must have been as fantastically expensive to cross an ocean then as it is to go to the moon or Mars now. Modern society is facing similar problems to Columbus: Why bother with the expense and danger of exploration & colonization when there is plenty to explore, research, mine, trade, etc. locally? What is really funny is that Columbus was able to get money for his expeditions in the same way NASA has: National ego. Spain had the English and Portugese to one-up, and the American's had the Soviets and the Chinese. I guess it just goes to show that not much changes, even in 500 years.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Religions aren't as stable as you claim. Just look at the past five hundred years or so of
Christian history. But I guess that's one of the differences between a religion and science: The
best you're going to get from a scientist is the claim that a particular model fits all existing
evidence. There are no secrets. There are no revelations. If new evidence invalidates the
model, the model is either revised or discarded. I'd be lying to you if I told you that there wasn't
dogmatic belief in the scientific world, but the idea is to keep it to a minimum, and that keeps
everyone intellectually honest (at least).
For sure, science doesn't have all the answers. But at least they tell you that up front.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Slashdot Mirror
Nice troll. Unfortunately, you are wrong.
Reverse Polish notation, also called "postfix notation", has the arguments first, followed by the operator. For example, "3 2 +". Lisp programs are typically written directly in the language's abstract syntax, using the prefix notation and parens with which many are familiar. For example, "(+ 3 2)".
Common Lisp's reader (aka parser) is, however, fully under program control. Through CL's readtables, one can extend the basic prefix notation, e.g. CL's sharpsign-S macro to read in structure objects is "#S(typename :field1 val1 ...)" as opposed to calling MAKE-typename the usual way "(MAKE-typename :field1 val1 ...)". One particulary perverse use of Common Lisp's readtables is a package that implements an infix syntax for basic arithmetic operators, e.g. "#I(2 + 3)".
Unlike static languages like C or Java, almost everything about the Common Lisp run-time environment is under program control. Even the object system can be modified (through something called the Meta Object Protocol). And idiomatic Common Lisp code is often times within a factor of two of idiomatic C code in terms of performance.
There are some serious problems with biometric identification systems, including:
Gartner wrote several great studies of various identification systems. Take a look at "Are Smart Cards Doomed to Die" by Andrew Phillips (a market analysis published 26 Apr 200) and "Security Applications of Biometrics: Perspective" by Ant Allan (a technology overview published 12 Feb 2001).
That's an entirely acceptable distraction. :)
It is never moral to kill another human being. It is, however, sometimes expedient.
How difficult is it to configure one's web browser so that it rejects most of the scripting junk out there? If you are using IE, check out the security zones feature that allows you to toggle scripting, cookies, and so forth depending on to which of four security zones a particular site belongs. I'm sure the free browsers have something much more sophisticated. Use it!
I'm not as much concerned with speed as with feature set and consistency. Reliable and consistent 416-Kbps symmetric DSL and routable IP address space---like a /27 or /26---would entice me away from my reliable and semi-consistent 2-Mbps CATV connection with only four (freaking!) IPs. But my connectivity needs aren't typical of your average Unreal freak, and unfortunately, the DSL market sucks and the cable providers aren't really in the biz market at the low end.
It's all rather frustrating.
Er, I had a point. Oh well.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Heh. Fortunately for me, my girlfriend has her own place. I doubt I'd get away with all the crap I have if I was married. :)
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
My setup isn't as sophisticated as I'd like, but it gets the wires off the floor. I have two enclosures, an old AT&T DataPhone cabinet (about 1.5m tall by .65m wide by .5m deep) and a free-standing "ladder"-style rack (3m tall, standard width). The DataPhone cabinet holds two DDS-2 tape drives, a keyboard/monitor switch, and four computers: two "pizza box"-style Intergraph TD-40 workstations, a 486 in a standard tower case, and a Packard Bell desktop. The rack holds most of the networking kit in the apartment (both firewalls, the cable modem, an Ethernet hub, the Synoptics chassis, and the Token Ring MAU). The disk array chassis and the UPS remain freestanding.
The most important reason to throw stuff in a rack is cable management. Common cabling is bundled up neatly and hung off the screw holes on the enclosures. Bundles of CAT-5 cable are snugly wrapped in masking tape every half meter, with Ethernet in a bundle separate from the Token Ring. Power cabling runs up the opposite side of the rack from the data cabling, and I have taken care to cross power and data cables at right angles to minimize cross-talk. Almost everything is labeled neatly. If you want to spend the money, you can color code your cabling (I usually do this when doing wiring jobs for money).
Be careful when using tie-wraps: They can be cinched too tight and cause internal breaks in delicate wiring (e.g. SCSI or UTP cabling). I usually use wire ties like you get with boxes of trash bags to truss everything in place.
Make friends with other sysadmins. A friend of mine gave me all sorts of nice telecomms hardware (including the Dataphone cabinet and brackets for running wiring) in trade for helping him clean out his machine room. I regularly go through the trash at work looking for useful items, e.g. an HP SCSI enclosure or Sun monitor cables. I once found two 4-GB SCSI disks in the trash and was able to make do with them for several years until I bought enough disks for a software RAID array. A consultant friend of mine was able to get four dual-processor Intergraph workstations for less than USD 300---all with 128 MB RAM, on-board Ethernet and SCSI. Even though they're Pentium-133s, Windows, Linux, and Solaris fully support the hardware and run great as domain controllers, file servers, and Unix workstations. Another guy traded me a monitor for a working SGI Indy that was loaded with memory. Make new friends and keep the old, my aunt always said.
I have something like 20 computers in my apartment, from an old IMSAI 8080 on up to a DEC Alpha. While I'd be really hurting if I had to replace it all, I think I've purchased maybe one complete system out of the whole lot. If you're having a hard time finding racks and such, nose around the local swap meets or hamfests, contact wholesalers and salvagers, and hang out around a college data center. Somebody's bound to get rid of something useful, and most times they'll just give it away to be rid of it.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
You're both asking a much deeper question than just "was John Brown right". The real question is, does the end ever justify the means? And to be honest, I don't really have a good answer for or against. Love your neighbor, love the Increate---that's pretty much the only ethical foundation most of us have, and it's not enough to answer this question.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
What if you can't boot to a working system? Thank God NetBSD would desplay interrupt and Ethernet MAC ID, because it would crash hard soon after and that was the info I needed to make things work.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Seals and porpoises are mammals. What do they do, hold their breath for a couple of miles?
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Please contact me via email. I would gladly take the machine off your hands for the price of shipping. It would have a good home, right next to the Alpha, also running OpenVMS (thank god for the hobbyist program).
P.S. I am serious.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
The rest of the series wasn't THAT good.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
You're wrong. Even the regular XF86_SVGA server must directly access the IO space, and it does so through the kernel. On my Alpha, when I still ran Linux on it, I had nothing but problems with X. A perverse combination of user operations (e.g. a fill or scroll) would completely lock up the machine.
Yes, there are times when some poorly-written X app kills the server, and there are times when the X server bugs out without killing the OS. But your statement "X and the Kernel are clearly different" is a gross oversimplification, ignoring, iovec(), memory-mapped IO, and the like. Even the most lowly frame buffer drivers must directly access the hardware.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Cloning sure beats having to wait for somebody to die (or "disappear"). Plus, you wouldn't have to worry about all that tissue rejection nastyness.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
A friend of mine was getting an error from the 2000 installer that sounds similar to yours. Basically, just before it finished formatting the 30 GB IDE disk (either NTFS or FAT), it would crash with an error message. Of course, it worked fine under Windows 98 and DOS (dunno if he tried Linux). It turned out that he had defective memory. Upon replacement of the DIMM, 2000 installed without complain.
Maybe you should have your memory checked out?
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
If you're poor, I would assume that meant you didn't have $52.28 to spend buying a piece of paper. As a college student supporting myself and making less than $9K a year, even $10 was a lot of money and not to be squandered.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
That's the other "his" point. Both Kerberos and NTLM require unique account names within an administrative domain and can't (easily) use DNs in the place of the User Principal Name or the SAM Account Name.
And yes, the old (NT4) domain system was basically a flat file (the SAM database), but this isn't the reason behind the requirement for unique UPNs in Active Directory. Kerberos too uses a flat database, and there has traditionally been no way to partition a realm (that is to say, the entire realm is propagated among masters and slaves in Kerberos). The replication and partitioning limitations in Active Directory are derived from AD's roots in Kerberos, not NT4.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
I'm in the process of building a system that integrates Windows and Unix logon data. My environment currently includes Debian GNU/Linux, Microsoft Windows, NetBSD, Silicon Graphics IRIX, and Sun Solaris, and I hope to bring Compaq OpenVMS and Apple MacOS (pre-OSX) clients into the mix within the next six to twelve months. What makes a difficult project even more challenging is that I'm attempting to integrate file sharing as well as domain logon. The environment I'm building is in it's third generation.
Getting Kerberos 5 clients to authenticate against AD is pretty easy and (for MIT Kerberos 5 and SEAM clients) well documented. See Windows 2000 Kerberos Interoperability, Step-by-Step Guide to MS Kerberos 5 Interoperability, and SEAM 1.0 Interoperability Documentation. You should be able to adapt the SEAM instructions to pam_krb5 on Linux, or you could use login.krb5 (which comes with MIT Kerberos 5). Note that Heimdal clients don't support DES-CBC-MD5.
The biggest problem with Active Directory has nothing to do with Kerberos, and everything to do with getting directory information out to clients in the realm. For example, NetBSD doesn't currently use the PAM mechanism, and Solaris has a difficult time understanding Microsoft's somewhat peculiar schema. To this end, I've decided to use NIS to publish certain directory data to my Unix clients (user name, UID, GID, GECOS, home, and shell; group name, GID, and members), and revisit client directory access issues as the technology matures. And since I can't get SFU to install on my DC (the DC's DNS domain name and the AD domain name are different, and this causes problems for SFU's installer), I've resorted to creating the necessary auxiliary classes by hand and using some glue code to dump user and group information from LDAP into the relevant NIS databases.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
I mean, a web server named after a punctuation mark? How do you pronounce "!"? I'm going to get some weird looks from my hacker buddies if I say I'm running Microsoft's GPL-ed IPv6 web server, "bang".
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Ever heard of UUCP? As far as I know, it works over IP as well as dialup.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
I finally understand why planet-wide exploration and colonization didn't really happen until Columbus. It must have been as fantastically expensive to cross an ocean then as it is to go to the moon or Mars now. Modern society is facing similar problems to Columbus: Why bother with the expense and danger of exploration & colonization when there is plenty to explore, research, mine, trade, etc. locally? What is really funny is that Columbus was able to get money for his expeditions in the same way NASA has: National ego. Spain had the English and Portugese to one-up, and the American's had the Soviets and the Chinese. I guess it just goes to show that not much changes, even in 500 years.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
That's amazing! I've got the same combination on my luggage.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
True, rules should apply to everyone. However, judgements should fit the crime and the criminal.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16
Religions aren't as stable as you claim. Just look at the past five hundred years or so of Christian history. But I guess that's one of the differences between a religion and science: The best you're going to get from a scientist is the claim that a particular model fits all existing evidence. There are no secrets. There are no revelations. If new evidence invalidates the model, the model is either revised or discarded. I'd be lying to you if I told you that there wasn't dogmatic belief in the scientific world, but the idea is to keep it to a minimum, and that keeps everyone intellectually honest (at least).
For sure, science doesn't have all the answers. But at least they tell you that up front.
Rev. Dr. Xenophon Fenderson, the Carbon(d)ated, KSC, DEATH, SubGenius, mhm21x16