Slashback: Errata, Futurity, Portality

Slashed back tonight: The (slight) return of the Y2K behemoth, good news for those locked out of port 80 by the recent unpleasantness, one interested party's response to Stephen Hawking's genetic-engineering ideas, and even an update on the Scarfo key-logging story.

Better than world-wide anarchy and privation. kejoki writes: "I came into work today and nobody had voicemail. We use an ancient AT&T system 25 (Merlin) with the Audix automated attendant/voice mail system ... not my bailiwick but the boss was going nuts trying to figure it out.

He finally called his System 25 guy and found out that quite a few people were having the same problem. Inspiration hit, and he set the system date back before 31 Dec 1999 ... whammo! The voice mail returneth.

AT&T->Lucent->Avaya, of course, no longer supports the system...as a matter of fact the boss seems to recall getting a letter from AT&T saying that they'd be taking care of the Y2K problems which might be in their equipment; but another soon after saying that support for the System 25 would be dropped as of 31 Dec 1999 ... hmmm.

Oddly enough, he's had a problem with the system giving a database I/O error for a while, but since he reset the date that has also vanished.

All very interesting. At any rate, if you have a System 25 and you can't get your voice mail, set back the date!"

And in related news, Che Fox writes :"The OpenLDAP project is one of the first to be hit by a major bug due to the S1G (one billion seconds) Unix time rollover. The slurpd replication daemon, which pushes changes from the master LDAP server to the slaves, no longer works now that time has rolled over to 1 billion seconds. This means that all LDAP-using networks in the world that use OpenLDAP and slave servers to replicate the data (very common) are now broken. There is a fix available against both the 1.2 and 2.x OpenLDAP releases in the OpenLDAP CVS repository."

You may assume your former activities for the moment. Agent Green writes: "I was checking out my firewall logs this morning and noticed an unusual amount of port 80 traffic and come to find out...it seems that AT&T Broadband has lifted their port 80 restrictions on its residential network. Let's see how long this lasts ..."

Probably until the next worm that takes over everyone's port 80, whatever OS it runs under.

So what did one giant say to the other? jshep writes: "Inventor Ray Kurzweil recently responded to physicist Stephen Hawking's concerns regarding the progression of AI (previous Slashdot story can be viewed here). Kurzweil takes aim at Hawking's suggestion that we use genetic engineering to augment the power of the human brain."

The man behind the curtain is ... uh, vital to national security! camusflage writes: "Reuters has a story (courtesy of Yahoo) that says the judge in the Nicodemo Scarfo believes the "national security" gambit about as much as the /. community does regarding the use of keyloggers. The most choice quote is "I don't know what it means. It's gobbledygook. More gobbledygook," referring to the argument put forth that the keylogger is a sensitive piece of national security. An assistant U.S. Attorney indicated he would provide "classified and unclassified summaries of the system's operation and more affidavits detailing the national security aspects at stake," next Friday."

Comments

Wow, 2 things to ba happy about!

[Enigma2175]

1) I think it is fantastic that the judge in the Scarfo case isn't dazzled by the FBI's "National Security" defense. This case has absolutely nothing to do with national security, the FBI is trying to establish precedent above the law. This time it is the keylogging technique, next time it is Carnivore v.2.0 that they try to hide behind the "national security" shield.

2) Being a subscriber, I am extremely happy that AT&T has lifted the ban in HTTP servers (I know I may assume too much given the anecdotal source). Most of the servers that run on the @home network are small, low traffic servers that don't cause much of a problem(unless they are infected). They must be worried about losing even the small percentage of customers that run web servers. Economic hard times are hitting everywhere...

national security...

[nettdata]

I could see how divulging how the keylogger works could be a national security issue... once it's been released how it works, people could start looking for the tell-tales, and then once word gets out about how many people are actually being logged, all hell breaks loose... both in and outside the US.

Re:national security...

[WhiteWolf666]

This is insane. How do we know that its just a keylogger? How do we know that its not a virus? How do we know that the government is only using it on people it has obtained a warrant for? How do we know that the government does not install it first, and then gets the warrant if necessary?

National security can go to hell. Worst case scenario (and this is unlikely)---the FBI looses its ability to log keys, and has to go back to doing real detective work. Big Deal. The U.S. will go on, and we will know that our rights are not being infringed.

If the FBI thinks that said documents really contain information pertaining to the case, subpeona the password.

I think if the government has the probably cause(is that the standard? whatever it is) necessary to get a warrant, the person targeted should know about it, and be able to challenge it, in the interets of privacy.

You have a right to know if you have been charged, so why don't you have a right to know if you privacy has been violated? Similarly, don't you have a right to not incriminate yourself? If the fifth amendment prevents the government from using a subpeona to get a password, than it is at least idealogical consistant that is should protect you from unknowingly giving the government your password.
And if it doesn't, and they say no, its a crime in and of itself, and the government than has every right to put the wiretap on whatever computers said person uses. This still provides a reasonable chance to catch the crook, or stop his illegal behavior, and still protects privacy rights. Yeah, this may reduce the probablity of catching them, but hey, it is better for 1000 guilty men to go free, than for 1 innocent man to be punished.

And, of course, I need to say this.

They that give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -- Benjamin Franklin, 1759.

Yet Another.....

[[amorphis]]

The Jive Forum BBS software was hit by the bug as well, for the same reason as everybody else: the sort order changes when the values are stored in a character field.
Here's a comment from one of the developers regarding the design decision:

Hey all,

Thought I would respond since I'm a Jive developer. There were quite a few reasons for the date to be stored as it is:

1) Java uses the millesecond values since 1970 as its native date format. However, unlike Unix, this value is stored as a 64 bit long instead of a 32 bit integer. Effectively, this means there will never be date overflow. In any case, using the millesecond value is very easy and fast in Java.

2) Database support for dates is horrible. Most db's have a DATETIME or TIMESTAMP column type. However, all databases seem to implement them differently. Further, support for 64 bit numbers is also poorly supported across many databases. Therefore, we were forced to go with our own encoding (millesecond values), and to use character columns instead of numeric ones. This lets Jive work with over 10 different databases instead of 1 or 2.

3) Yep, we never thought about the date rollover bug until about a month and a half ago. Adding a few padding 0's was a simple fix and was released on Aug 8th as Jive 2.0.

-Matt

Re:genetic engineering the brain

[Cuthalion]

I don't think anybody would consider treatments for sickle-cell anemia (which occurs primarily in Africans and African-Americans) immoral genetic engineering, for instance.

Not really on topic, but the reason that sickle-cell anemia is more prevalant in Africa (and other tropical regions) is as follows:

Sickle-cell anemia is a recessive trait. If you have only one of them, you are not anemic. However, for whatever reason, you ARE more resistant to malaria.

Microsoft & code theft

[Anonymous Coward]

Recently, I've witnessed an increase in posts claiming that microsoft has stolen linux code for use in their Windows 2000 operating system. These posts never give concrete examples, but instead rely on strawman arguments and wild claims.

Well folks, I've got access to the Windows 2000 code (I'm a CS graduate student at Rutgers University, which has access to it for educational purposes), and I've discovered the truth, which I'll get to in a moment.

First, a brief history.

Windows 2000 and NT are based on VMS design principles. VMS (Virtual Memory System) was DEC's operating system for for their VAX (Virtual Address eXtender) and later Alpha computers. It was a contemporary to Unix, but unlike AT&T, they didn't allow it to be ported to other architectures, which explains why it never gained Unix's popularity, despite being more secure and more powerful.

In the early 80s (while future "Open Source" icon Linux Torvalds was still wearing underoos), Microsoft did experiment with their own Unix operating system, and even planned on phasing out MS-DOS in favor of it, but scrapped that plan due to problems with Unix's ease of use, inefficient use of system resources, and archaic design. They sold MS-Unix to SCO, and signed a non-compete agreement.

This explains why Microsoft chose to use VMS as a basis for their next generation operating system.

Returning back to the Windows 2000 source code: I've examined it thoroughly, and cross compared it in design, style, and composition.

Microsoft stealing Linux source code for windows 2000 is as likely as Spam being the next theme ingredient on Iron Chef.

Re:Microsoft & code theft

[Zeinfeld]

This explains why Microsoft chose to use VMS as a basis for their next generation operating system.

There is another reason. DEC screwed Dave Culter, the chief architect of VMS at just the time Microsoft got serious about operating systems again. Cutler left DEC with most of the core VMS kernel team for Microsoft. Shortly after DEC signed a co-development agreement with Microsoft under which Windows-NT would become the successor to VMS.

A few years later Rashid, the chief architect of CMU's MACH microkernel, which is generaly considered one of the bes UNIX kernels arround also joined Microsoft.

Bill works by offering top engineers seven figure golden hellos. That is how he got the team who designed Word to switch from Xerox Parc.

Re:Nobody but nobody...

[Winged Cat]

The problem with arguments like that, is that the only people who will pay them any heed are exactly the people who would care about the morality. If you ask them to stop, they may...but the unethical will continue blithely on. And then [PICK_RESEARCH_FIELD] will be dominated by the unethical: zealots (religious or otherwise) out to make life miserable for those who do not believe as they do, corporations who care for little save their current quarter's bottom line (even at the expense of their own future), and the just plain uninformed (who lack any basic education about the field with which to make moral decisions).

The result, then, is that it is the duty of those with any shred of ethics to aggressively pursue their research at the fastest possible speed, so that [PICK_RESEARCH_FIELD] may be tapped for the benefit of all humanity and countermeasures developed against its negative applications. This ensures that the ethical decisions that matter will be made by those who have some ethics, whether or not they have the fullest possible advantage of foresight. Perhaps the debate can rage while the research goes on, to give them some foresight...but to hold back the research until the ethics have been settled is to ensure catastrophe. (Besides, for many of these fields, the ethics may never be settled. See, for example, the continued debate over abortion, decades after - at least in the US - the issue was settled by the Supreme Court.)

Re:Nobody but nobody...

[eric17]

It is, in fact, why I believe that neither Communism nor Objectivism will ever actually work: both assume a priori that all involved in the system will behave like perfect little robots with no free will. Face it: just as there will always be overly-motivated people who will do anything to succeed including screwing over their mother, there will be slackers, and every option inbetween.


I offer a second opinion:
Objectivism won't work because people will never figure out the difference between a principled person acting in rational self-interest and a emotionless automaton with no ethics, principles or values.

Re:Port 80 blocking

[Anonymous Coward]

I had the same experience; this works in New Brunswick too. Here is the patch I made:

--- tcp_input.c Sun Dec 10 18:49:44 2000
+++ /tmp/tcp_input.c Mon Sep 3 12:23:45 2001
@@ -2074,7 +2074,7 @@
/* These use the socket TOS..
* might want to be the received TOS
*/
- if(th->ack)
+ if(th->ack && !(th->source == 80 && th->fin))
return 1;

if(th->syn) {