VIRUS-L Digest Tuesday, 15 Oct 1991 Volume 4 : Issue 191 Today's Topics: Response re DIET (PC) Re: Hardware Help needed in testing software security combination, please. Measures Experiences with hardware protection (Thunderbyte) Re: is vshield working? (PC) Origins of "Worm" More hardware! BGS9 (Amiga) Rejects unformatted disk (Mac II) New files on risc (PC) safembr 1.3 available (PC) Appenders and prependers VIRUS-L is a moderated, digested mail forum for discussing computer virus issues; comp.virus is a non-digested Usenet counterpart. Discussions are not limited to any one hardware/software platform - diversity is welcomed. Contributions should be relevant, concise, polite, etc. Please sign submissions with your real name. Send contributions to VIRUS-L@IBM1.CC.LEHIGH.EDU (that's equivalent to VIRUS-L at LEHIIBM1 for you BITNET folks). Information on accessing anti-virus, documentation, and back-issue archives is distributed periodically on the list. Administrative mail (comments, suggestions, and so forth) should be sent to me at: krvw@CERT.SEI.CMU.EDU. Ken van Wyk ---------------------------------------------------------------------- Date: Sun, 13 Oct 91 13:55:03 >From: c-rossgr@microsoft.com Subject: Response re DIET (PC) >From: turtle@darkside.com (Fred Waller) > > I wasn't familiar with the DCOMPRESS program, so I went and >downloaded its source code before posting this reply. I found that >DCOMPRESS is a rather different animal from DIET: it requires that >the user prepare ahead of time special ASCII files for each >subdirectoruy, listing files to be excluded, requires that a >separate utility (PCMANAGE) be used to compress the files also >ahead of time, and has other requirements, including the need to >change the system date by one week, then run PCMANAGE, then reset >the system date back again, etc., which render it much less elegant >and more difficult to use than DIET. I intend no offense against >DCOMPRESS' author, but DCOMPRESS is not equivalent to DIET v1.10a. No offense taken, Fred. When I wrote it, it was designed as a method for me to compress files on my hard disk that hadn;t been accessed within a certain time frame. The initial time you ran it, you'd *either* set the date back a week or so so that all compressable files would be marked as "old" and a compression attempt would be made, *or* you'd not play with the clock/calendar at all and simply run the stand-alone compressor later and let the compression take place on a "regular" basis. Since this was a program for publication and the mag it was printed in (PCMag) often requires lots of sizzle to what they print: it wouldn't look good to print an article for one of the free utilities that says "it won't do much for the first two weeks, and then there's no way of telling what it will do". So, the date change stuff allowed people to get something immediately -- immediate gratification is something that PC owners are used to, right? It was a program intended for an entirely different purpose than DIET, but it could be used, sorta, in the same way. In fact, if not for some legal questions, chances are that a commercial version of it would have been available that included many of DIET's features: I have the code sitting here.... >Thus, I repeat that the functionality offered by DIET v1.10a is >not found in any other program that I know. Its usefulness and >functionality are achieved, to an important extent, by the use >of what might be termed `virus programming techniques'. All that said: I would not consider DIET to have "virus programming techniques anymore than a standard TSR that can have different functionality when it is invoked multiple times should be considered as having these programming techniques. Ross M. Greenberg Author and Copyright Holder, PCManage & DCOMPRES ------------------------------ Date: 14 Oct 91 15:55:53 +0000 >From: eric@zen.maths.uts.edu.au (Eric Lindsay) Subject: Re: Hardware I tried hardware protection against virus attack in a student PC lab. It only "sort of" worked. Basically pulled the write gate via a low ohm resistor. The "switch" was a self closing 3.5 mm audio socket. The "key" was a 3.5 mm plug. This was a while ago, and of course will only work on ST506 style interfaces (not on SCSI or IDE). I tried it only on XT clones. On most of them, the hard disk wouldn't work (the drive cards must test to see if they can get to the drive), and therefore the machine would only boot from floppy. I eventually found that an LCS brand card didn't check, and therefore could be used. A better solution (albeit more complex) that doesn't require two hard disks would be an add-on for an ST506 drive controller. It would have a comparator on board, and step-up step-down counters run from the direction and step pulse lines. Reset from the track zero line. You would set switches to the track area you want protected (eg, track 0 to 200) and partition your disk at that boundary. Thus the C: area would be protected and the D: drive open for R/W. I never built it - we moved to a Novell system instead, with diskless PCs (ptui!) - -- eric@zen.maths.uts.edu.au Eric Lindsay, Sch of Maths, Uni of Tech Don't take life too seriously. It is only temporary. ------------------------------ Date: Mon, 14 Oct 91 14:47:00 +1300 >From: "Mark Aitchison, U of Canty; Physics" Subject: Help needed in testing software security combination, please. Could someone with a good library of viruses (incl DIR II, hopefully) please try the following combination of software virus protection, and tell me what file infectors still get through? Disk Manager's DMDRVR.BIN with the partition set to be READ-ONLY, *PLUS* Digital Research's DRDOS 6 with all executables password-protected, *PLUS* The partition(s) compressed with DRDOS 6's SSTOR (or STACKER, maybe). Notice that none of the software products are advertised as trying to fight viruses. What I want to end up with (and I know others would like too) is to have a read/write partition for user files, and another read-only partition with executables. I guess that at least one type of virus would still work in such an environment, that would be defeated if all partitions were read-only, but I'm looking for a reasonable compromise of convenience and protection, and can put some proper anti-virus software after I know what sort of viruses get through the cracks. If any kind person would like to try out the setup, I can chat about configuartion details, like using JOIN, and DCONFIG.SYS and so on, which might be important to improve security and convenience. What I predict is that the combination of simple software will defeat file infectors that go through DOS and those that go through int 13. Obviously, the software combination wouldn't prevent boot sector viruses, but I expect it would stop all the common file infectors (including DIR II & Dark Avenger). Thanks in advance, Mark Aitchison, Physics, University of Canterbury, New Zealand. ------------------------------ Date: Sun, 13 Oct 91 19:29:35 -0700 >From: turtle@darkside.com (Fred Waller) Subject: Measures AGUTOWS@WAYNEST1.BITNET (Arthur Gutowski), quotes from my post: > I disagree. Even hardware solutions can be overcome by the > ignorant or impatient. Of course. We weren't discussing idiot-proof measures. But the virus problem is not user-created, although a recent article by Mark Aitchison makes an extremely good point that public education must play a very important part in any attempt to solve it. Rather, my view is based on (Waller's) Pecking-order Metaphysics: 1. Generally considered, symbolic entities can overcome other symbolic entities. (A symbolic defense can always be devised against any symbolic attack and viceversa). 2. Consequently, a given symbolic entity cannot permanently overcome the class of symbolic entities. (Viruses and antiviruses will battle each other till doomsday, without conclusive results, as they have been doing). 3. However, physical entities can always overcome symbolic entities, unconditionally. (There is no way for a symbolic entity to resist a physical entity, unless intended or allowed by the physical entity as an operation within its capabilities or intent). 4. As a result of (3), symbolic entities cannot overcome physical entities, either temporarily or permanently, unless the physical entity is first disabled. (Thus, no software, no matter how sophisticated, can ever defeat even the simplest hardware protection. This is also why viruses can affect hard disks: they are ALLOWED to do so). 5. Generally considered, physical entities can overcome symbolic as well as other physical entities. (While their overcoming symbolic entities is permanent, their control of other physical entities may be only temporary, similarly to (1), above). 6. Neither software nor hardware, being creatures of humans, are ultimately able to resist a user's damaging or careless action when exercised at the hardware level. (Users are a higher class of agent that can overcome both symbolic and physical entities by exercising access to (5) above, or by using a hammer). So the "pecking order" is: 1. software; 2. hardware; 3. operator, ascending. (Above that, you have to resort to mythical or government forces... both yield unpredictable results). From (Waller's) Pecking-order Metaphysics, one readily derives the Waller Principle: "If You Really Want to Stop a Software Virus, Stop it with Hardware". and the Waller Afterthought: "Anything Else is Probably a Waste of Time and Effort, and May Actually Be Counterproductive". > ...who's to say that there won't come a day when we can create > (or encounter) infectious "beings"?).. We encounter them all the time, in the air we breathe, the water we drink, the food we eat and the things we touch. Zillions of them live within and on our bodies - permanently. As for creating new ones, we've done that, too. > ...there are reasonable approaches, both hardware and software. > (See Padgett Peterson's paper on the six-byte integrity checking, > and some of the programs he and others have written, as well as > his recent "Interesting Laptop Configuration" posting). I feel Peterson's post was extremely interesting and appropriate. The ideas presented there (and in several other articles of his) were the sort of thing that should be explored seeking more definite solutions to this problem. They should be pursued further. > ...What's the use of having a computer if you can't share data? > Answer: There is none. NOBODY is proposing to make it impossible to share data. Such a suggestion was NEVER advanced by me. And the ideas I did mention were NOT equivalent to `not sharing data'. Stopping viruses by hardware means DOES NOT equal `stop the flow of data'. It does, however, restrict and regulate the *uncontrolled* flow of executables, which is the main thing that enables virus spread. (It may also restrict a kind of programming that's becoming very popular, but will not eliminate it; only a modification is needed). > You might as well power it down, lock it in a lead safe, cast > it in cement, and drop it to the bottom of a shark-infested > sea (cf. Gene Spafford)... Oh, I really hope it won't be necessary to get rid of our beloved toys. Why not just send Spafford to a remote (but tropical) island instead? His rounded figure will get a suntan that should go well with the beard... :-) > .... and we're back in the forties. Todo tiempo pasado fue' mejor. > Yes, but the Apple OS is still *software*, isn't it. For that > matter, so is MVS and VM, and we see very few mainframe viruses, > too. Of course it's easier to infect MS DOS systems (Can many users write to a mainframe executable or system file?). But another (not minor) consideration is that there are some 60 million MS DOS PCs out there. That's a market. Both viruses and antiviruses must perceive that fact. It's likely to be a main motivator. In both cases. And a comment on etiquette: > Or, maybe we're just growing weary of going around and about > the same points over and over again. I for one, am growing > tired of paging through mounds of postings with nothing new > to say. People who feel this way should, by now, have grown very tired of the endless give and take between viruses and antiviruses, the pursuit of the unreachable Holy Grail, the eternal promise, never realized, and the eternal threat, always renewed. The monthly updates and the weekly new-virus reports, the more-ingenious attacks, and the ever-newer defenses. People who feel this way should perhaps try to contribute not just a negative comment, or not just comments whose intention is to stop discussion, but rather some encouragement of discussion in this forum. People who are tired of repetition should have grown EXTREMELY tired of the virus/antivirus repetition. Currently, that's the grandmother of all repetitions. Fred Waller turtle@darkside.com ------------------------------ Date: 14 Oct 91 11:30:42 +0000 >From: Reinhard Kirchner Subject: Experiences with hardware protection (Thunderbyte) Hello, I got produkt information about a hardware virus protector called 'Thunderbyte' which intercepts all mysterious writings to the disk, e.g. absolute ( not through dos ), writing to exe/com files etc. Such a thing costs appr. the same as a software package, and it does not depend on updates for new viruses. So I want to ask: Is there any experience with such devices, thunderbyte or others ? Is it worth the money ? >From the documentation I have such a hardware protector looks like a ultimate solution. Reinhard Kirchner Univ. of Kaiserslautern, Germany kirchner@uklirb.informatik.uni-kl.de ------------------------------ Date: Mon, 14 Oct 91 18:49:05 +0000 >From: greg@irl.ise.ufl.edu (Greg O'Rear) Subject: Re: is vshield working? (PC) RY01@ns.cc.lehigh.edu (Robert Yung) writes: >...How do I know that [vshield] is working up there without actually > getting my system infected??? >...Does vshield (v82) work right in UMBs? When my department purchased a site license (yes, we actually PAID for it!), one of the principal reasons I chose McAfee was VShield. I installed a test copy on a heavily used grad student PC, with the option to print a message saying to notify me in case of a virus. The other day, it happened. Someone came to my office saying that they tried to use the computer but it wouldn't let them continue, it just posted a message saying to contact me at once. When I got to the computer, I saw the VShield message advising the user to remove the floppy and run Scan on it. Even though it was only the Stoned virus, it could have been much worse if not for VShield, so I regard it as money well spent (good thing I bought it when I did, what with our governor's budget axe...). The usual disclaimers apply (even the ones for mutual funds (past performance does not necessarily indicate future performance) :-).... - -- Greg O'Rear Industrial and Systems Engineering Department, University of Florida Address: O'Rear@ise.ufl.edu ------------------------------ Date: Mon, 14 Oct 91 19:54:17 -0400 >From: Andrew Brennan Subject: Origins of "Worm" Last I knew ... The original 'worm' was from John Brunner's "Shockwave Rider." There has been at least one book on shutting down the phone 'network' - I can't remember the title right now, but it came out after Esquire's article (70's) on phreaking. As far as I know, there haven't been too many early novels on this type of stuff (pre-cyberpunk lit). Andrew. (brennaaa@duvm.ocs.drexel.edu) ------------------------------ Date: Mon, 14 Oct 91 22:21:17 -0700 >From: turtle@darkside.com (Fred Waller) Subject: More hardware! The origin of the following is lost to me in the thread but someone recently wrote, referring to hardware vs. software antiviral protection: > It is the classic and eternal argument between authority and > freedom, order and chaos. Hardly! In our case, it was an argument of hardware protection vs. software protection. Having seen over the last two years that (software) antiviruses cannot protect us against (software) viruses, the idea of using stronger medicine was being brought up by Padgett Peterson and others, including myself. > These problems do not lend themselves to global solutions. > While the hardware solution is attractive, it is Draconian. Few problems as complex as computer viruses truly lend themselves to "global" solutions. I dare say the people who propose hardware protection, as posted here, are aware that many practical problems have neither a "global" nor a "total" solution. Only naive software proponents continue seeking the Holy Grail of the "perfect virus detector".... :-) But hardware protection is a practical, much more effective overall solution, as has been demonstrated in actual practice. Still, it's not "global" or "perfect" or "absolute". Nor does it need to be! > The virus problem is not a problem of the machines that we sell > today, but of those that have already been deployed. Actually, the virus problem threatens BOTH new machines being sold and those already installed. > We have seen examples, described here, of machines that include > features that make them more resistant to viruses. While they may > offer some protection to the owners of those machines, they have > a vanishingly small effect on the global problem. Are there no means of hardware protection applicable to machines currently in use? I think there are. > Since the population of machines in the world is obviously large > enough to guarantee the success of some viruses, and since that > population is likely to persist for a decade or more, new hardware > features are not likely to have much effect. The world population of PCs is a main attractant for both viruses and antiviruses. However, most of them (about 60%), are located in the USA. With the advent of the IBM/Apple arrangement, and the announced introduction of a new hardware architecture AND a new kind of OS, it is not unreasonable to speculate that this nation of disposable-article lovers may junk/trade in their old PCs at a fast rate to acquire the new marvels. Anyway, old PCs are worth only a small fraction of the price of an automobile... and we regularly junk those every few years.... :-) > The issue is not preventing the execution of all viruses all > of the time. Rather, it is resisting the execution of most > viruses most of the time. Not really. That's never been the practical issue. If we could stop just 10 or 12 viruses, we would have stopped something like 90% of all viral attacks. The rest of the 1,000 or so "known viruses" are a very minor REAL threat. In fact, if we could just stop ONE virus, the old Stoned, we would have prevented the great majority of viral infections in the USA! But the Stoned continues on its merry way, and software antiviruses have not been able to stop its advance. In any case, it's important to note that, regardless of the kind or number of viruses we want to stop, hardware protection will stop them more effectively than software protection. And THAT's worth thinking about. > Software can contribute to this more readily than can hardware, > if only because it can be readily retrofitted to the existing > population, where the problem is. That is why it is called > "soft," rather than "hard," ware. The term "hardware" had been in use among electronics professionals for a very long time to refer to circuit components and related parts. Some of it could be "retrofitted", some could not. Later, when computers appeared, not very useful without programming, the term "software" was created, by analogy but in contrast, to indicate the other, symbolic half of the system. Unlike other machines, the IBM PC and its clones make it nearly as easy to modify the hardware as modifying the software, if not easier. That's one of its main virtues. People (most?) who have trouble installing complex software (e.g., Windows) can easily open the case and insert a new video card, attach a monitor or change keyboards, all major hardware components. And in many cases, new software cannot even be installed before making changes in hardware! So, in the case of the IBM PC and its clones, "retrofitting" of many hardware components is childishly simple and certainly not limited to software. It's understandable that software-oriented people might be less aware of this facility, but their restricted view should not be taken as indicative of general user's capability. But we might stop saying that software is easier to install than hardware - often it isn't. Fred Waller turtle@darkside.com ------------------------------ Date: Tue, 15 Oct 91 06:13:00 -0500 >From: KENNETH@ducvax.auburn.edu Subject: BGS9 (Amiga) Anybody know anything about an Amiga virus called BGS9? ------------------------------ Date: Tue, 15 Oct 91 16:55:30 +0000 >From: hoshor@mdd.comm.mot.com (Alan Hoshor) Subject: Rejects unformatted disk (Mac II) We are experiencing a localized problem on a network of MAC IIs and SE30s. The primary symptom is failure to recognize an inserted disk. Especially if it is unformatted. The symptom is never consistent. But once displayed, requires the rebooting of the MAC to resolve. These MACs are protected by current releases of either Disinfectant 2.5.1 or SAM 3.0. Has there been anyone else experiencing similar problems? Alan ------------------------------ Date: Mon, 14 Oct 91 11:35:18 -0500 >From: James Ford Subject: New files on risc (PC) The following files have been placed on risc.ua.edu (130.160.4.7) for anonymous FTP in the directory pub/ibm-antivirus: vsumx109.zip - Virus Summary Listing, v1.09 wscan84b.zip - McAfee's Scan for Windows validate.crc - Latest listing of validation codes from McAfee's BBS. - ---------- The advantage to being a pessimist is that all your surprises are pleasant. - ---------- James Ford - Consultant II, Seebeck Computer Center The University of Alabama (in Tuscaloosa, Alabama) jford@ua1vm.ua.edu, jford@risc.ua.edu ------------------------------ Date: Tue, 15 Oct 91 12:17:00 -0400 >From: HAYES@urvax.urich.edu Subject: safembr 1.3 available (PC) Just received and made available A. Padgett Peterson new version of SAFEMBR. file name: SAFEMBR.ZIP Site address: University of Richmond IP# 141.166.1.6 directory: [.msdos.antivirus] This is a VAX/VMS system! login user= anonymous, password . .ZIP file must be downloaded as binary (one user though we were as big as simtel20 and using TENEX ). Best, Claude - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Claude Bersano-Hayes HAYES @ URVAX (Vanilla BITNET) University of Richmond hayes@urvax.urich.edu (Bitnet or Internet) Richmond, VA 23173 ------------------------------ Date: Mon, 14 Oct 91 15:11:35 -0700 >From: p1@arkham.wimsey.bc.ca (Rob Slade) Subject: Appenders and prependers FUNPIV3.CVP 911013 Viral code addition In order to avoid damage to the original program, which might lead to detection of the infection, the viral code can be added to the beginning or end of the program. (Or not attached at all.) Adding code at the beginning of the original program ensures that the viral code is run whenever the program is run. (This also ensures that the virus is run before the program runs. The virus thus has priority in terms of operation, possible conflicts and detection.) With the addition of code to the beginning of the program, it is possible to avoid any change to the original code. It *is* necessary to alter the file/disk allocation table, at least, in order to ensure that the program "call" starts with the viral code, and that the viral code is not overwritten by other changes to the disk or files. While the original code may be left unchanged, the file will be, essentially, altered, and, unless techniques are used to disguise this, will show a different creation date, size and image. It is also, however, possible to add viral code to the end of the original program, and still ensure that the viral code is run before that of the original program. All that is necessary is to alter the file header information to reflect the fact that you want to start executing the file towards the end, rather than at the normal location. At the end of the viral code another jump returns operation to the original program. (This kind of operation is not as odd as it may sound. It is not even uncommon. A legacy from the days of mainframe "paging" of memory, it is used in a great many MS-DOS executables, either in single .EXE files or in overlays. It is, therefore, not a coding indication that can be used to identify viral type programs or infected files.) Appending, or prepending, viral code to an existing program therefore avoids the problems of damage and potential failure to run which plague overwriting viral programs. Even these viral programs, however, are not foolproof. Programs which load in very non-standard ways, such as KEA's "Zstem" terminal emulation program, use the header information which the viral programs alter. Although not originally designed for virus detection, the "Program abort - invalid file header" message thus generated is an indication of viral infection. Sometimes the first indication that users have. copyright Robert M. Slade, 1991 FUNPIV3.CVP 911014 ============= Vancouver p1@arkham.wimsey.bc.ca | "Power users think Institute for Robert_Slade@mtsg.sfu.ca | 'Your PC is now Research into CyberStore | Stoned' is part of User (Datapac 3020 8530 1030)| the DOS copyright Security Canada V7K 2G6 | line." R. Murnane ------------------------------ End of VIRUS-L Digest [Volume 4 Issue 191] ****************************************** Downloaded From P-80 International Information Systems 304-744-2253