Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue
Read all about it! Virus authors caught and charged! We also have some important details about the MBDF virus. In addition, check out the news about an impressive new Unix-based Macintosh emulator, a review of the DataClub virtual server software, why you should wait to buy a LaserWriter IIf or IIg, Apple/IBM news from France, and hopefully the last word on the new StyleWriter driver.
Copyright 1992 TidBITS Electronic Publishing. All rights reserved.
Information: <firstname.lastname@example.org> Comments: <email@example.com>
Thanks to all of you who have returned the survey to us. I'm sure results will continue to flow in, so send yours in if you want to be in the running for a button. Some people have expressed confusion about some of the numbers asked for in the survey - don't worry about these too much, estimates are fine if you don't know the specific answer. If you've already discarded that issue, I've posted the text of the survey form on the fileserver, so you can easily request it, fill it out, and return it. Just send email to <firstname.lastname@example.org> with the single word "survey" (no quotes) in the Subject: line. Thanks!
Errors of the week -- I originally had a funny mailer message slated for this spot, and I will include that below for a little relief from the real error of the week, which was the SFU mailing list. It has had problems on and off for a while now, but this week the underlying mailer software kept crashing and every time it crashed, it sent another copy of whatever it had to everyone on the list. My sincere apologies for this, and I would recommend switching to the TIDBITS LISTSERV at Rice if you are terribly frustrated with SFU. To do that, send email to LISTSERV@RICEVM1.RICE.EDU including this line in the body of the mail:
SUBSCRIBE TIDBITS your full name
I made a mistake last week in saying that the name should be in parentheses - if you do that, the LISTSERV will think they belong in your name. You can also have more than two words in your name, but must have at least two. If you wish to delete those parentheses or change your name on the LISTSERV, just send in another SUBSCRIBE mailfile with your name sans parentheses.
If you do subscribe to the LISTSERV, please remove yourself from the SFU list (you may have to wait a day or two until the site comes back up - it's been shut down temporarily) by sending email to email@example.com with the single word "remove" (in lower case and without the quotes) in the Subject: line. Thanks!
Funny error -- I occasionally receive messages from mailers when requests to the fileserver bounce, and this one takes the cake. This is the sort of thing that artificial intelligence researchers should watch out for from the very beginning - a little introspection is a good thing. :-)
<<< 553 poly.polytechnique.fr I refuse to talk to myself
SuperClock! error -- Steve Christensen, the author of SuperClock!, writes in regard to Mark H. Anbinder's article "Quadra Vampires", "Well, that's news to me. SuperClock! was written completely in assembly language, so there aren't any compiler-oriented issues to deal with. SuperClock! also doesn't do any cache-oriented operations directly - I make no hardware assumptions of that type since I want to be able to run on 68000s as well as 68040s. It's possible that some of the system code I call may flush the cache as part of its operation, though. I work with the software guys that did the ROMs for both Quadras, both use SuperClock! on their Quadras, and I haven't heard any complaints."
[Adam: Our apologies, Steve. Mark said that the information came from someone within Apple, so it's more curious that you weren't contacted first. Though SuperClock! is off the hook as a Quadra vampire, I would warn people to generally be aware of older programs that might exhibit this behavior. We don't know of any specifically, but older compilers had no way of knowing about the Quadras and certain out-of-date code may run slowly on them.]
Steve Christensen -- firstname.lastname@example.org
Mark H. Anbinder -- email@example.com
We've heard a rumor from the illustrious Pythaeus that the LaserWriter IIf and IIg will soon ship with more memory, standard. My spec sheets claim that the IIf ships with 2 MB and the IIg with 5 MB, but there have been rumblings from early purchasers that 2 MB in particular is just not enough, especially when printing legal size pages or pages with complicated graphics. Apparently, the low-memory symptom will often appear as a PostScript error, which might also imply something wrong with the document or printer driver, thus making the issue even harder to diagnose.
We have no word on whether or not existing printers will be retrofitted with extra memory in the US, and there is conflicting evidence. When the IIsi went to 3 MB standard, 2 MB machines were not upgraded. However, in the UK there is currently a free upgrade program that will take a IIf to 4 MB and a IIg to 8 MB. The upgrade is not advertised, but dealers should have known about it since a service note dated the 10th of February explained the situation. Again, this upgrade is only in the UK.
In other non-US news, we've heard that Apple Canada is running a special deal through March 31st in which you can get a PowerBook 100 bundled with Salient's AutoDoubler for a mere $1799 list. I have absolutely no idea how expensive the PowerBook 100 normally is in Canada, and AutoDoubler certainly isn't that expensive ($45 mail order), but it seemed worth mentioning for our Canadian readers.
Oh, if anyone wishes to contribute information or rumors anonymously (and I always honor such requests in mail I personally get), send it to firstname.lastname@example.org and Pythaeus will check it out for inclusion in an issue.
Pythaeus -- email@example.com
Just after I wrote last week that the Mac world hasn't seen a virus in some time, one has to pop up. The latest and slimiest entry into the virus hall of infamy (I know some people who are in a kneecap-breaking mood over this one) is called MBDF after the resource that it uses to infect System files and applications. MBDF resources are normally present in some files, so do not be alarmed if you see them while poking around with ResEdit.
The MBDF virus was discovered in Wales. Early detection was made possible by the foresight of Claris programmers who included integrity checking code in their applications, something which other application programmers would do well to add. As a suggestion, perhaps someone (at Claris perhaps?) could release some integrity checking code into the public domain so that it would be easy for all programmers to add such capabilities to their applications.
Several popular Internet archive sites contained some infected games for a short period of time, so a number of people around the world were affected. The games were named "10 Tile Puzzle" and "Obnoxious Tetris." In addition to these two games, a third game named "Tetricycle" or "tetris-rotating" was a Trojan horse which installed the virus. If you have any of these programs sitting around, do everyone a favor and delete them. It's all too easy to release these viruses again.
I don't think that MBDF was as widespread as some of the earlier viruses, such as nVIR, but there is a possibility that your Mac has been infected by a completely different program so it is worth checking your Mac with the latest virus checking software. We recommend Disinfectant 2.6 because it is free and easy to use, but new versions of Virus Detective (5.0.1), Gatekeeper (1.2.4), or any of the updated commercial programs should also do the trick.
Disinfectant identifies both infected files and the Trojan horse as being infected by the MBDF virus and can repair any infected files, which removes the virus and returns the file to its original clean state. Repairing the Trojan horse renders it ineffective and inoperable. Shucks.
The MBDF virus is not malicious, but it can cause damage in certain instances. In particular, the virus takes quite a long time to infect the System file when it first attacks a system. The delay is so long that people often think that their Mac is hung, so they do a restart. Restarting the Mac while the virus is in the process of writing the System file very often results in a damaged System file which cannot be repaired. The only solution in this situation is to reinstall a new System file from scratch. There have also been reports of directory damage which may or may not be related to the restart process.
Special thanks to John Norstad, as usual, for his excellent and timely response to the new virus, and to the folks at Claris for providing the defensive code that helped find this virus early on before it had a chance to spread its evil tentacles even further. Ooo, there's not much like a virus for evoking some good imagery.
Now that you've read the technical details, here's the human interest side. We just heard that two arrests have been made at Cornell University in Ithaca, New York. The virus had been traced to Cornell fairly quickly, and an internal investigation led to the arrests of two sophomores, David Blumenthal and Mark Pilgrim. The two suspects have been charged with second degree computer tampering and are spending the night in Tompkins County jail. Further charges are pending based on the distribution of the files to sumex-aim.stanford.edu and to its mirror sites around the world, specifically Japan. The legal system will decide whether or not these two are guilty of the charges and what the punishment will be, but if I were them, I'd be watching my kneecaps.
John Norstad -- firstname.lastname@example.org
Chris Johnson -- email@example.com
Jeff Shulman -- firstname.lastname@example.org
Mark H. Anbinder -- email@example.com
and many others around the world who helped nail
this virus to the wall.
Apple and IBM announced today the appointment of the senior officers and board of directors for Taligent, the joint operating system company the two formed last October. Joseph M. Guglielmi of IBM was named CEO, and Edward W. Birss of Apple was named COO of the independent company. Finally.
Why is this interesting? Well, it's not really. What is interesting about the announcement is that it shows that Apple and IBM are truly serious about this alliance and have passed all the major hurdles, the first of which was getting FTC approval. Taligent has over 160 employees already, and will be working in space leased from Apple until more permanent offices can be obtained. Similar and equally ambitious mergers have fallen by the wayside when two large and opinionated companies tried to work together, and the odds were good that this alliance would go that way too. Apple and IBM have risen above the infighting to set Taligent up, and now it's up to the programmers to produce the next generation of operating systems. Full speed ahead!
Mark H. Anbinder -- firstname.lastname@example.org
Latest in the line of Macintosh emulators to be announced (but how many of those have actually shipped - two?) come two programs from a small developer called Quorum. First, and most interesting is Latitude, a set of libraries that can be used to compile ANSI standard C code into a binary application that will run on several different RISC-based Unix workstations, including machines from Sun, IBM and Silicon Graphics. This method of porting the code directly (more or less, anyway, depending on how the code has been customized) from the Mac to the Unix environment has a number of positive features. Since Latitude replaces the Macintosh interface specifics with whatever windowing system is in use on the Unix box - Open Look, Motif, etc. - you don't need Apple's proprietary System and Finder. That has been a big barrier for many of the other emulators, because to create a true Macintosh clone, you have to have the real System and Finder. Latitude uses the operating system and windowing environment of the RISC machine, so that's no problem. In addition, since Latitude creates a native binary application on the Unix box, that application will run at speeds you would expect out of a RISC machine - fast, to understate the situation. Finally, since Quorum based the display parts of Latitude on Adobe's Display PostScript, there is no conflict Apple's patented QuickDraw software (which is why most other emulators have required that you find some Mac ROMs to pop in).
Equal is the second product from Quorum, and unlike Latitude, it does not require that Macintosh applications be recompiled to run. It will run standard applications by rerouting the Mac Toolbox calls to the equivalent in the RISC machine's OS and windowing system. Of course, the price for not having to recompile the application is speed, since Equal has to intercept almost everything a normal application does and translate it into the appropriate calls for the host machine. Nonetheless, remember that a typical RISC machine will still stomp on a normal Mac in performance, so the end result may still be decent. We've heard a couple of off-the-cuff comparisons, one comparing Equal to a Plus, the other to a IIci. I'd like the IIci personally. No idea what pricing will be on these programs yet, but people have been tossing around $1000 as a price for Equal, a typical price for Unix applications.
Of course, lots of Mac software will not work with these products because it simply won't make sense. For instance, if you have no System and Finder, extensions that modify them in some way won't work. Similarly, any program that touches hardware directly won't work because that hardware can't be there in the same way. Nonetheless, Latitude and Equal seem like the strongest entries in the Mac emulation market (such as it is) so far. One plus of Latitude particularly is that it will allow third party developers like Aldus, Quark, Adobe, and Microsoft (nah, probably not Microsoft - they don't use standard programming practices anyway as far as anyone can tell) to recompile their flagship products for Unix, thus allowing them to quickly broaden their markets without the trouble of rewriting the entire program and maintaining two completely different sets of source code. Only really Lotus, WordPerfect, Frame and perhaps a few others have managed to provide much in the way of multiple platform versions in the past, but that might change with Latitude.
I'm sure that Apple is not terribly happy about this state of affairs, but from what I've heard, Quorum is on fairly stable legal ground (considering they're in California, it's nice to have some sort of stable ground). Even still, Apple has a lot of money and a lot of clout and might even buy Quorum outright if they so choose. That's assuming that Quorum is selling, but everyone has a price. Time will tell, and we'll keep you posted.
Quorum -- 415/323-3111
Don Sleeter -- email@example.com
MacWEEK -- 20-Jan-92, Vol. 6, #3, pg. 1
PC WEEK -- 03-Feb-92, Vol. 9, #5, pg. 47
There has been some incorrect information flying around, and I may have even aided it in a posting I made to the Info-Mac digest, so let me see if I can explain what is really happening with the new StyleWriter driver, 7.2.2. People have been having some problems with the new driver, and although it's certainly possible that those problems are related to bugs in the driver or conflicts with existing software, other problems may be due to the fact that using version 7.2.2 of the driver can be a bit confusing if you are using system 6.0.7 or 6.0.8.
If you use System 7, you shouldn't have any problems. If you use 6.0.7 then you must upgrade Print Monitor and Backgrounder to their System 7 versions. If you use System 6.0.8, you should already have these files installed. The System 7 versions can be found on a set of 6.0.8 System disks or on the System 7 Printing Tools disk. StyleWriter driver version 7.2.2 also works with the versions of Print Monitor and Backgrounder that recently shipped with StyleWriter driver version 7.2.
Since version 7.2.2 of the StyleWriter driver makes the StyleWriter print faster than the previous version of the driver, try using it if you can. Those of you on the Internet can get the driver via anonymous FTP on ftp.apple.com as:
It is also available on the System 7 Tune-Up disk, but I gather that the Tune-Up installer will refuse to work on a System 6 disk (not too surprising, consider that it is a tune-up for System 7). However, if you can get your hands on that disk, all you have to do is look in the Tuner Parts folder and snag the StyleWriter driver from there.
Murph Sewall -- SEWALL@UCONNVM.UCONN.EDU
DataClub is one of those programs that people thought would die a horrible death when System 7's FileSharing appeared. From what I gather from talking to the folks at International Business Software and from using it on our Macs here, DataClub is still doing well, and for good reason.
Before System 7 came out, people usually used TOPS to share files among several Macs. AppleShare was too pricey and required a dedicated Mac, and some of the shareware and freeware applications didn't have quite enough in the way of a feature set. Then came DataClub, which has a completely different way of looking at the concept of file sharing, one which just might finish TOPS as a peer-to-peer networking program (although Sitka is currently pushing the inter-platform connectivity TOPS offers to Macs, PCs, Sun workstations, and soon, pen-based palmtops).
Most file sharing programs try to replicate the original disk or folder on the secondary machine, which is a fine way of doing it in most cases. However, this gets confusing when you have ten or twelve folders mounted as volumes via TOPS. If nothing else, on a compact Mac screen, it can be hard to find the right one. In contrast, DataClub takes all the space you allot to it and creates a single virtual volume. On our network, Tonya's Classic had about 4 MB free and my SE/30 had about 7 MB free, so the DataClub volume appears to have about 11 MB free. As you add more Macs to the virtual volume, the virtual volume size increases, so a normal five or six Mac network could easily appear to have a DataClub volume of several hundred megabytes.
There are a couple of advantages to this system. First of all, it is less confusing. There is only one network volume, not ten or twelve, and everyone has the same one, although users and groups can be set up with AppleShare-like access privileges so there's no need to worry about security. Second, because everyone with a hard disk contributes space, everyone can opt to create new folders on his or her Mac's hard disk, thus making those folders available to everyone else while retaining the speed of having files available on the local hard disk. Of course, using a file over the network is not nearly as fast.
As a rough estimate, I'd say that DataClub is a little faster than TOPS and about the speed of a dedicated AppleShare server. Actually, I'm talking about DataClub Classic, which is the version used only in peer-to-peer networks. If you can spare a dedicated server, DataClub Elite will provide even better performance on that dedicated Mac. In addition, the manual, which is clear and generally helpful (though I do admit that I didn't look at it until I wrote this review) suggests that you avoid putting applications in the DataClub volume if possible since running an application over a common LocalTalk network is slow and frustrating. For Macs without hard drives (or PCs for that matter) the AppleShare client software that ships with the computer will allow any floppy-only Mac to act as a client and mount the DataClub volume over the network - a good way to squeeze a few more months from those aging Pluses.
Back to the advantages of the virtual server system. Third, because no single Mac has to provide all the disk space for the network, any one of the Macs can be shut down or even removed from the network without causing a major hubbub (assuming no one is using the files on that Mac's hard disk at the time). Files in DataClub that are on a disconnected Mac's hard disk merely look greyed out; similarly, when a Mac is away from the network, all the other files are greyed out and only local DataClub files can be accessed.
This may sound like DataClub sits in the background making sure that files are where it wants them to be. That's true to an extent, but you have a fair amount of control over where you put files, and the administrator can also move folders around to more evenly distribute the load. Oh, didn't I mention the administrator before? You don't actually need one unless you are planning a relatively complex DataClub network with users and groups or want to get load statistics, etc. If you just want to share files, you can simply run the installer, restart, and be on your way. IBS has even sold various cheap DataClub packages in the past without the administrator software, in part to make DataClub available at a reduced price, and in part because you don't need it in every case.
There are some limitations to DataClub. It does conflict with some other extensions, but IBS lists known conflicts in the release notes. You can't move an item from the DataClub volume on to the desktop, but if you try, DataClub will helpfully tell you that it can't do that and ask if you want to copy the item to the startup disk's desktop. Even under System 7, you can't leave files and folders in the Trash; DataClub will tell you that it has to delete them immediately and ask for confirmation. I assume that these last two limitations have to do with not implementing the System 7 invisible folders in some way, but it's not really a big deal, and it might be a common problem with networking software. You might also cause some confusion if one person tried to delete a file while another person tried to copy it, or something like that, but I wasn't able to seriously confuse DataClub in my tests. You do need 2.5 MB of RAM and 2 MB of disk space to run DataClub, and 2.5 MB of RAM with DataClub and System 7 will leave little for the application you want to run. But you knew you were going to need more memory anyway, and it's cheap these days. One final caveat: you have to be a bit careful with your disk space because if you have 5 MB free on your hard disk and your DataClub volume claims it has 6 MB free, only 1 MB of that 6 MB is coming from another machine. The other 5 MB is the same 5 MB your hard disk has free. So it might seem as though you had 11 MB available, when in reality, you only have 6 MB.
All this said, who should buy DataClub? As clever and useful as it is, it can't come close to System 7's FileSharing in cost, and in the case of very small networks, like our two-Mac network, DataClub is overkill. It's simply easier to turn FileSharing on and off when we need it and not worry about it most of the time. However, I see the ideal DataClub network as one made up of five to ten Macs in a small office situation. In that situation, DataClub provides storage to everyone on the network and does so in such a way that everyone can easily access public files. There's no hassling with multiple network volumes or asking if so-and-so has turned on FileSharing, both problems likely to occur otherwise. Larger networks will probably need the added speed of DataClub Elite, in which each of the user machines can still contribute disk space to the virtual volume, or even AppleShare 3.0, but I'm not going to make any sweeping statements about large networks, since I'm no expert. Suffice it to say that for a small office networks, DataClub will provide an admirable file sharing service. Oh, a 3-pack of DataClub Classic will run about $265 mail order, but keep an eye out for one of those great deals IBS occasionally has.
International Business Software -- 408/522-8000
by Jean-Philippe Nicaise -- firstname.lastname@example.org
On January 28th, the French government chose IBM to be the technological partner of Bull, the state-owned mini, workstation, and microcomputer builder. IBM and HP were competing for a few months to provide Bull with RISC technology that Bull had failed to develop internally. This is interesting because IBM will undoubtedly provide their RS/6000 architecture, the same chips that are destined for the PowerPCs that will eventually be a major hardware platform for the Pink OS coming from Taligent. Since Apple is the pre-eminent microcomputer company in France (and Apple France is the second largest market for Apple after the US), it's nice to see that the Bull RISC machines, which mainly end up in the government and schools, will be ready for Pink when (and if, of course) it arrives in 1993/4/5/6.
Apple has been doing well in France directly too. The same day, the French government also announced the creation of a workgroup to do a study on future popular multimedia terminals. Organizations involved in the workgroup will include Apple, Thomson Consumer Electronics (TCE), CNET (national PTT research center), France Telecom (national PTT), and possibly Kaleida. Apple is interested for obvious reasons given their recent announcement about entering the consumer electronics industry. CNET and France Telecom have ten years of experience in popular terminals (the Minitel) and 150,000 B-channel ISDN lines have already been installed in France, running a wide range of character-based applications. [Adam: Go ahead, make us ISDN-less folks in the US jealous!] TCE has developed the D2-MAC high quality TV standard in the past years and has a lot of experience in TV technology.
Non-profit, non-commercial publications and Web sites may reprint or link to articles if full credit is given. Others please contact us. We do not guarantee accuracy of articles. Caveat lector. Publication, product, and company names may be registered trademarks of their companies. TidBITS ISSN 1090-7017.
Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue