Thanks to email@example.com for pinning down my half-remembered quotation in the preceding segment (RISKS 9.60): > The actual quote is "Those who cannot remember the past are condemned > to repeat it." from George Santayana's "The Life of Reason". The grandfather of all command-control-communication (C3) systems was an air defense system called SAGE, a rather tortured acronym for Semi- Automatic Ground Environment. As I reported earlier in RISKS 8.74, some of the missles that operated under SAGE had a serious social problem: they tended to have inadvertent erections at inappropriate times. A more serious problem was that SAGE, as it was built, would have worked only in peacetime. That seemed to suit the Air Force just fine. SAGE was designed in the mid to late 1950s, primary by MIT Lincoln Lab, with follow-up development by IBM and by nonprofits System Development Corp. and Mitre Corp. The latter two were spun off from RAND and MIT, respectively, primarily for this task. SAGE was clearly a technological marvel for its time, employing digitized radar data, long distance data communications via land lines and ground-air radio links, the largest computer (physically) built before or since, a special-purpose nonstop timesharing system, and a large collection of interactive display terminals. SAGE was necessarily designed top-down because there had been nothing like it before -- it was about 10 years ahead of general purpose timesharing systems and 20 years ahead of personal computers and workstations. While the designers did an outstanding job of solving a number of technical problems, SAGE would have been relatively useless as a defense system if a manned bomber attack had occurred for the following reasons. 1. COUNTERMEASURES. Each SAGE system was designed to automatically track aircraft within a certain geographic area based on data from several large radars. While the system worked well under peacetime conditions, an actual manned bomber attack would likely have employed active radar jamming, radar decoys, and other countermeasures. The jamming would have effectively eliminated radar range information and would even have made azimuth data imprecise, which meant that the aircraft tracking programs would not have worked. In other words, this was a air defense system that was designed to work only in peacetime! (Some "Band-aids" were later applied to the countermeasures vulnerability problem, but a much simpler system would have worked better under expected attack conditions.) 2. HARDENING. Whereas MIT had strongly recommended that the SAGE computers and command centers be put in hardened, underground facilities so that they could at least survive near misses, the "bean counters" in the Pentagon decided that this would be too expensive. Instead, they specified above-ground concrete buildings without windows. This was, of course, well suited to peacetime use. 3. PLACEMENT. While the vulnerabilities designed into SAGE by MIT and the Pentagon made it relatively ineffective as a defense system, the Air Defense Command added a finishing blunder by siting most of the SAGE computer facilities in such a way that they would be bonus targets! This was an odd side effect of military politics and sociology, as discussed below. In the 1950s, General Curtis Lemays's Strategic Air Command consistently had first draw on the financial resources of the Defense Department. This was due to the ongoing national paranoia regarding Soviet aggression and some astute politicking by Lemay and his supporters. One thing that Lemay insisted on for his elite SAC bases was that they have the best Officers Clubs around. MIT had recommended that the SAGE computer facilities be located remotely, away from both cities and military bases, so that they would not be bonus targets in the event of an attack. When the Air Defense Command was called upon to select SAGE sites, however, they realized that their people would not enjoy being assigned to the boondocks, so they decided to put the SAGE centers at military bases instead. Following up on that choice, the Air Defense Command looked for military bases with the best facilities, especially good O-clubs. Sure enough, SAC had the best facilities around, so they put many of the SAGE sites on SAC bases. Given that SAC bases would be prime targets in any manned bomber attack, the SAGE centers thus became bonus targets that would be destroyed without extra effort. Thus the peacetime lifestyle interests of the military were put ahead of their defense responsibilities. SAGE might be regarded as successful in the sense that no manned bomber attack occurred during its life and that it might have served as a deterrant to those considering an attack. There were reports that the Soviet Union undertook a similar experimental development in the same time period, though that story may have been fabricated by Air Force intelligence units to help justify investment in SAGE. In any case, the Russians didn't deploy such a system, either because they lacked the capability to build a computerized, centralized "air defense" system such as SAGE or had the good sense not to expend their resources on such a vulnerable kludge. (Next segment: command-control catches on.) -Les Earnest (Les@Sail.Stanford.edu)
The sendmail problem to which our moderator frequently refers is actually a design problem in the SMTP protocol (that sendmail implements) and so is well within the domain of RISKS discussions. Especially since the design problem has been recognized since at least 1983, but SMTP is so entrenched that the problem has never been fixed. The way in which SMTP sends a message is, roughly: 1. Send message headers to server, etc. & get responses 2. Send DATA keyword to server & get affirmative response 3. Send entire message to server with no response 4. Send "." to Server 5. WAIT UNTIL SERVER HAS DELIVERED MESSAGE AND RESPONDS.  is the problem. At any given time during , there are two possibilities. Either the server has crashed and will never send an answer, or the server is still in the process of transmitting the mail. There is NO WAY for the sender to tell the difference between these two possibilities. This is because: a. SMTP presumes that TCP is reliable, so it doesn't allow the sender to retransmit anything. Hence, the sender can't send any data during the wait in  to find out if the server is still alive. b. TCP, to be efficient for long lived connections, does not guarantee continual monitoring of the connection. If TCP has no data to send, it never sends any data (except for the Berkeley-derived keep-alive option, which are not part of the TCP spec, but see below). c. EVEN IF TCP did monitor the connection, the SMTP process could die or go into an infinite loop without resetting the TCP connection. (You can't catch this case: Halting Problem). I can vaguely remember hearing of this happening in 1983. I truly believe this problem is unsolvable, without changing SMTP. The Programming Marines, who build and sell systems must solve it anyway. Or alleviate it, anyway. The ad-hoc solution is to set a timeout; assume the message takes no more than time T to process by the server, so if the sender waits for 2T or so, the server must have crashed. This helps, but it clearly doesn't solve the problem, since the sender has no metric by which to judge T except the length of the message. For example, a message transmitted to a Large, Risky Mailing List might take an order of magnitude longer to process than a simple user-to-user message. Hence, it is likely that messages to large mailing lists will tend to fail with the sender timing out and resetting the connection; this condition is noticed by the server AFTER transmitting at least part of the message. The sender later retries, so you get multiple messages. Very many messages, if you are unlucky, or its a Monday or you moderate a very large and risky mailing list. (BTW, it was decided in SMTP's design that it was better to have multiple messages than to have messages get lost, so it is not considered acceptable for the SMTP server to queue the message but not deliver it during the pause in ). The general problem: building functionality on top of a "reliable system" does not GUARANTEE that the result is reliable. The reliability must, in some way, be verified by the topmost level. Hence, you balance your bank statement to make sure your checks (cheques.uk) were really cashed shortly after you sent them off, and you look for cancel checks if you get a red notice from a bill you sent off. Reference: J. Saltzer, D. Reed, D. Clark, "End-to-End Arguments in System Design", Proc. IEEE, April 1981. - Geof [Well, on RISKS-9.64 I removed all of the offending addresses and I did not get any reports of multiples. So, here goes RISKS-9.65. PGN]
It is now possible for you to file your income tax return electronically. Yes, you. Of course the IRS will not accept it directly from you; they only deal with "tax preparers' like H&R Block, the owners of CompuServe. Not to worry. There are value-added re-sellers who will accept your return from you, check it and forward it to the IRS. I have seen ads now from two different such re-sellers: "Nelco" and "Nexus Direct." PGN's misgivings to the contrary notwithstanding, it is not likely to open up the IRS's computers to tampering by taxpayers. From their point of view it is as safe as having you send it to them by snailmail and a lot cheaper than key-entering it themselves. From the taxpayers' point of view, it is a great deal safer than putting it in the snailmail and having it transcribed by a temp at the IRS. Of course, a copy of your return can end up in the computers of the services. On the other hand, that happens whenever you deal with a tax preparer. That has only marginally impacted the success of H&RB. Incidentally, when you file with H&RB, or any number of other preparers, they offer to give you your refund on the spot. It is really in the form of a loan, but the loan is re-paid directly to them by the IRS. The IRS likes this arrangement because they can pay the refund by EFT/EDI. Not only do they not handle the paper return, but they do not have to prepare a check for the refund. It is almost as if the IRS has opened thousands of offices across the country where they will prepare your return for you and give you your refund on the spot. Of course, since it is for-profit, it may be a little more efficient. Who pays? Well who always pays? At least there are some choices. How about it H&RB? How long will it be before I can "GO 1040" on CompuServe? William Hugh Murray, Fellow, Information System Security, Ernst & Young 2000 National City Center Cleveland, Ohio 44114 21 Locust Avenue, Suite 2D, New Canaan, Connecticut 06840
Bob Munck <firstname.lastname@example.org> writes: <> While the software had been rigorously tested in laboratory <> environments before it was introduced, the unique combination of <> events that led to this problem couldn't be predicted. > > Don't they mean "wasn't"? The rest of the report seems (to me) to > be reasonably detailed, well explained, and apparently honest, but > this one little dissemblance ruins the whole thing. Is there any > justification for the assertion that the prediction was (and is) > _impossible_ in these circumstances? I noticed this too. On reflection, though, I'm sort of glad that they said it this way. Of course this error could have been predicted. Any particular error could be predicted, if the right person looked at it at the right time. But what *is* impossible is to predict all errors, all of the time, with probability 1. So, if you interpret the above statement above as, "The event occurred randomly in a situation where the probability of such errors could not be reduced to zero," then I think it is a reasonable description of the truth. Certainly it is a better reaction than the other common thing that one sees in these press statements, which is something like, "Steps are being taken to ensure that no similar outage can ever occur again." This is an unrealistic and false reaction which would cause me, at least, far more concern than the AT&T reaction. Of course they will try to prevent such problems. But it is equally obvious that the probability of such events will remain positive. -- David desJardins
Coming back from London about a year ago I was seated next to a pilot from a Munich based company who flew 757s. He'd overshot his flying time and was heading back to Munich from Kenya over London. We talked about the Airbus crash and it seems that in pilot circles it was attributed to the fact that the older Airbus machines had Rolls-Royce engines. It seems that when given full power they reacted about 3 seconds faster that those on the A320. In his opinion these three seconds were vital. Needless to say, the A320 pilot was familiar with the older model(s) and this may have played a role in his judgement. On a side note the pilot also mentioned that the 757, which also has the fly by wire system, sometimes "hangs". The only way to get the electronics to function again is to power cycle the lot.
We reported earlier on the diskettes containing a destructive Trojan horse program. Something like 20,000 disks were mailed from London around 11 December 1989. When executed, it trashed PCs in at least Britain, Norway, Sweden, Belgium, Denmark and California. Joseph W. Popp, 39, a zoologist from the Cleveland area, has been arrested in Ohio on a federal fugitive warrant, and charged by Scotland Yard in London with blackmail and extortion. In court in Cleveland, Popp told U.S. Magistrate Joseph W. Bartunek he is under a psychiatrist's care and must take drugs for a mental condition. The judge has scheduled psychiatric evaluation prior to any further hearing. Popp apparently (has/had?) worked for the World Health Organization.
Please report problems with the web pages to the maintainer