chuck@trantor.harris-atd.com (Chuck Musciano) (12/03/88)
> Now, can anyone tell me whether we should have bought > FrameMaker instead, what it does that we can't now do, and what it costs? After all my postings about Frame, I should make clear that I don't work for Frame, but I do like their product. I found Publisher to be a backwards step from Frame. First, it failed Musciano's Law of New Software: I sat down without the manual and tried to do something productive. I got nowhere. I couldn't figure out how to create a simple document. With Frame, I did all sorts of things, multi-column documents, line art, different text flows, and never picked up the manual. Strike One against Publisher. Publisher is not a WYSISYG package. It is a compose/preview package. This is completely unacceptable in my book. I find it annoying to continuously move back and forth between compose and preview windows to check my work. Frame is true WYSIWYG, and has good response to even complex document modifications. Strike Two. The user interface is poor. Commands which do one thing in the compose window do another in the preview window. For example, I seem to recall that in compose, you used ^H, ^J, ^K, and ^L to move around (which is bad enough) but in preview you used B, F, P, and N (backwards, forwards, previous, and next). I don't want to learn two tools in one! I found the Frame interface intuitive, with a nice balance of menus for novices and keyboard commands for experts. Strike Three. The drawing programs are separate tools. Again, I don't want to learn N tools, I want to learn one. Frame is fully integrated, except for table of contents and index generation (which bothers me, but is outweighed by other features). Strike Four. Overall, Publisher seems targetted to people who know TeX. Why do you want to hang on to old technology when all this wonderful new stuff is coming out? Do you really care what the internal representation of your document is? What does it matter if it is TeX, PostScript, or C/A/T? I can insert PostScript into my Frame documents (and do occasionally) to accomplish the few things that Frame cannot. Frame is, I believe, $995/station at educational rates. With the floating license server, you can actually get away with much less. For example, suppose, you have 15 stations, but actual use of Maker is about four simultaneous users. Just buy four licenses, and share them among the 15 Suns. The license server idea is one which needs to be picked up by other companies. Finally, why don't you just check it out for yourself? Send a 1/4" tape to Frame, and they'll send you a demo copy. Call 1-408-433-3311 for more information. Chuck Musciano Advanced Technology Department Harris Corporation (407) 727-6131 ARPA: chuck@trantor.harris-atd.com
bernhold@orange.qtp.ufl.edu (David E. Bernholdt) (12/15/88)
I have never used Frame, but our Project has been using Publisher for well over a year now. I'm afraid the Chuck Musciano has presented a somewhat lopsided view of the Publisher software. I don't want to start a net-wide argument, but I do want to address some of the comments that were made: chuck@trantor.harris-atd.com (Chuck Musciano) writes: (edited for brevity) >X-Sun-Spots-Digest: Volume 7, Issue 34, message 7 of 12 > >I found Publisher to be a backwards step from Frame. First, it failed >Musciano's Law of New Software: I sat down without the manual and tried to >do something productive. I got nowhere. I too sat down with Publisher without the manual and had little problem getting things done. I guess different people "adapt" to new things differently. Most of our users (> 60) haven't looked at the manuals. >Publisher is not a WYSISYG package. It is a compose/preview package. The present version of Publisher is very nearly WYSIWYG (admittedly previous versions of Publisher were not as good in this regard). It has a special table editor and a special equation editor to simplify putting in those kinds of objects, but everything shows up in the edit window pretty much like its going to look in the final document. I find that you really only have to preview just before printing to check all of the formatting. >The user interface is poor. Commands which do one thing in the compose >window do another in the preview window. For example, I seem to recall >that in compose, you used ^H, ^J, ^K, and ^L to move around (which is bad >enough) but in preview you used B, F, P, and N (backwards, forwards, >previous, and next). I don't want to learn two tools in one! In all of the versions of Publisher we've had, from the beta release to the present one, this has *not* been true. The editor is modeled after EMACS as far as key assignments go, but the key, mouse, and all other definitions are fully configurable by the user if desired. If you don't like ^P, ^N, ^F, & ^B you can change them. >The drawing programs are separate tools. Again, I don't want to learn N >tools, I want to learn one. Does it really matter if the tools are distinct executables or not? Certainly there will be different "command" structures for creating graphics, etc. - are they easier to learn because you don't have to bring up a separate tool to use them? >Overall, Publisher seems targetted to people who know TeX. That may have been true to *some* extent at the start, but it isn't any longer. Publisher makes use of TeX and SGML, but you don't have to know either to use it. Publisher also has the advantage of being able to import and export TeX, LaTeX, and SGML format documents for compatibility with other systems. This is very useful when, for example, you an a collaborator (who uses LaTeX, say) can both work on a paper easily. Publisher also accepts TeX input, so that if you have something unusual to do & know how to do it in TeX, you can do it. >I can insert PostScript into my Frame documents (and do occasionally) to >accomplish the few things that Frame cannot. You can do this in Publisher as well; also Sun bitmaps, TeX code, and PubPaint, and PubDraw objects. >Frame is, I believe, $995/station at educational rates. With the floating >license server, you can actually get away with much less. For example, >suppose, you have 15 stations, but actual use of Maker is about four >simultaneous users. Just buy four licenses, and share them among the 15 >Suns. The license server idea is one which needs to be picked up by other >companies. nI do not know what the Publisher costs, but we have it licensed for half of our 60 workstations - we tend to have a lot of people writing papers. This sounds pretty much like Frame's license server to me. When we were choosing a desktop publishing package, one of the important considerations for our Project was the ease of entering equations. AT THE TIME, Publisher was the only package we found that had a "reasonable" method for entering equations. Things may have changed now, but it remains an important consideration to us. Arbortex, who produce Publisher, have been very responsive to us - both in answering (sometimes stupid) questions, as well as fixing any problems we might encounter. They have also been very responsive to requests/suggestions for additional features. They are really interested in the needs of their customers. I'm quite sure that Arbortext would also be glad to answer questions or send a demo. They can be reached at (313) 996 3566 or FAX (313) 996 3573. DISCLAIMER: I have no connection with Arbortext (who produce the Publisher) other than having used their software for > 1 year now. Dave -- David Bernholdt bernhold@qtp.ufl.edu Quantum Theory Project bernhold@ufpine.bitnet University of Florida Gainesville, FL 32611 904/392 9306
Kent_Wada@mtsg.ubc.ca (12/15/88)
I would like to address the points made in v7n34 by <chuck@trantor.harris-atd.com> on the Publisher vs. Frame Maker question, and in the original item by <root@helios.ucsc.edu> in v7n20. Mr. Musciano's comments were clearly made in accordance with a set of criteria which he uses to judge how effective a given product will be for the type of work he does. That is of course how we all function, but I think that what is not reflected in his item is the recognition that there is an astonishingly diverse set of text processing applications, needs, and requirements on this planet, and that no one product is going to solve all the problems effectively, let alone optimally. I like Publisher because it suits the types of work I do, and because I believe in the basis upon which it was built; but for Heaven's sake if PC-WRITE produces results that are satisfactory in terms of time spent, effort involved, and results produced, use it! An observation about the points Mr. Musciano raises: many of them transcend product-specific issues. Publisher and Frame Maker just happen to embody many of the qualities that people argue about... like `WYSIWYG or not.' >Do you really care what the internal representation of your >document is? What does it matter if it is TeX, PostScript, or C/A/T? On to the specifics. The simple answer to Mr. Musciano's first question above is `yes'. Not perhaps that it uses PostScript, or TeX, or SGML as its underlying paradigm, but that it follows a philosophy of adhering to standards--ISO, ANSI, de facto, whatever--as far as possible. In this context, it means that authors are able to share documents, and all that that implies. If co-authoring or otherwise sharing documents in geographically and/or electronically disparate environments is not a requirement, this may not be as much of a concern. There are, however, many people to whom this is very important. I might add that this is one the reasons the use of TeX has become so widespread (in the domains for which it was designed). >Publisher is not a WYSISYG package. It is a compose/preview package. >This is completely unacceptable in my book. >... >Overall, Publisher seems targetted to people who know TeX. Why do you >want to hang on to old technology when all this wonderful new stuff is >coming out? Do you really care what the internal representation of your I am intrigued by Mr. Musciano's implication that TeX is old technology. Is `all this wonderful new stuff' a reference to his earlier mention of WYSIWYG systems? There is certainly no consensus on whether WYSIWYG authoring systems best solve all text production needs. Likely there would be unanimity for the case that WYSIWYG interfaces are wonderful for many, but not all, applications. As for Publisher being targetted to people who know TeX... One of the reasons I like Publisher is because it offers me an interactive, TeX-independent presentation interface, while retaining the benefits derived from using TeX--such as availability on a large number of systems and printing devices, use of a de facto standard allowing portability between authors, quality of output, and markup language capabilities. Of course there are import/export facilities for those who use TeX, LaTeX, or SGML directly, but it does not diminish the capabilities of the software as a stand-alone text production tool if they are not used. >I found Publisher to be a backwards step from Frame. First, it failed >Musciano's Law of New Software: I sat down without the manual and tried to >do something productive. I got nowhere. I couldn't figure out how to >create a simple document. With Frame, I did all sorts of things, >multi-column documents, line art, different text flows, and never picked >up the manual. Strike One against Publisher. Another of the philosophical differences that people have is how averse they are to reading documentation. I do not particularly enjoy reading instruction manuals, but do not begrudge having to read a certain amount of it--if only to acquire a `feel' for how the product works. If the software is well designed, only a moderate amount of documentation need be read at first: the rest should follow intuitively, and the documentation used primarly as a reference tool. Both points are important: it should not be necessary to look up a manual for every little thing, but some formal basis in the operation of a product (if only to be able to use the product fully and effectively) is critical. >The user interface is poor. Commands which do one thing in the compose >window do another in the preview window. ... >I don't want to learn two tools in one! >... >The drawing programs are separate tools. Again, I don't want to learn N >tools, I want to learn one. Frame is fully integrated, except for table >of contents and index generation (which bothers me, but is outweighed by >other features). Strike Four. I do not argue about the merits of having a uniform user interface; patently, it is a desirable goal. However, I think it is too much to expect a single tool to do everything. Publisher's approach using separate graphics, table, and equation editors is perhaps imperfect in implementation, but not in intent. I would rather have ArborText spend the time expanding their integration capabilities with other packages--so that, for example, I can use my favourite graphics package to generate my pictures, as opposed to being limited to using what is provided--instead of trying to retrofit more and more capabilities onto a single piece of software. Would it not be wonderful if there was seamless integration between packages like Mathematica, Leonardo, and Publisher, and--dare I say it--all on a NeXT machine? Just _think_ of the possibilities! But one could not even begin to consider the thought without a basis rooted in standards... >Frame is, I believe, $995/station at educational rates. With the floating >license server, you can actually get away with much less. For example, >suppose, you have 15 stations, but actual use of Maker is about four >simultaneous users. Just buy four licenses, and share them among the 15 >Suns. The license server idea is one which needs to be picked up by other >companies. There is no reason to believe that the floating licence server concept is implemented only by Frame Technologies. The copy of Publisher I use resides on our server, but is usable from any of the Sun workstations that are connected to our network. >Finally, why don't you just check it out for yourself? I could not agree more! We all know computer products tend to evolve stressfully apace, particularly new products (one need only look at the quantum leap between what I call Publisher's `concept prototype', version 1.0 from only a year ago, and their latest, version 2.1). Mr. Musciano's suggestion about evaluating such competing software is really the only rational strategy: analyse what is required (and what is desired), see what is out there, and try them out! I will give into the temptation to include a (very abbreviated) Publisher features list: based on standards (TeX, SGML, and PostScript); great table and equation editors; TeX-quality output (`for the creation of beautiful documents'); multilevel `undo' facility; import/export of TeX, LaTeX, and SGML documents; tons of fonts (including complete math fonts); bibliography support; graphics editors, screen capture and scanner support, import facilities for Sun bitmaps, PostScript graphics, MacPaint, MacDraw and Excel graphics; and ASCII terminal support. -Kent >After all my postings about Frame, I should make clear that I don't work >for Frame, but I do like their product. I suppose I should mention that I do not work for ArborText. I do like Publisher an awful lot, but not exclusively! I would be glad to continue this discussion, but perhaps elsewhere? Maybe the desktop-publishing list? kent_wada@mtsg.ubc.ca (Internet) |Computing Centre/The University of USERWADA@UBCMTSG (BITNET) |British Columbia/6356 Agricultural Telephone: (604) 228-6496 |Road/Vancouver, British Columbia/ Facsimile: (604) 228-5116 |Canada V6T 1W5
chuck@trantor.harris-atd.com (Chuck Musciano) (12/15/88)
I received some attention from ArborText regarding my recent postings
about the relative merit of Publisher vs Frame Maker. I was able to spend
some time with Melanie Kessler of ArborText at the SUG, and looked at
version 2.0 of Publisher. Tony Camozzi of Arbor Text had chastised me for
basing my comments on version 1.0. So in the interest of fairness, here
is an updated look at Publisher, based upon my original four complaints:
1) I couldn't do anything without reading the manual first.
I didn't directly use Publisher, so I can't comment. The
interface is fundamentally the same, although more aestecthically
pleasing, with more use of dialog boxes rather than pull-right
menus. A personal opinion: the appearance of these dialog boxes
is rather unattractive, and makes finding important elements in
the dialog box difficult. A review by a talented graphics artist
would do wonders for the interface.
2) It isn't WYSIWYG.
This is still true. The edit-print-examine cycle is not tolerable
for me. I want to see what I do, as I do it. Another interesting
thing: table of contents generation is a two pass operation, and
Publisher does the first pass, and then reminds you to do the
second. That could be better automated.
3) The preview/edit interfaces are different.
This is still true, although you can rebind the keys to make them
match. For novice users, who don't know how to bind keys, this is
still unacceptable.
4) The drawing tools are separate tools.
Still true. Although these tools are quite powerful, and do nice
things, they are not integrated into your document, and you have
to learn multiple tools to use Publisher effectively.
Publisher does an excellent job with equations, and has an acceptable
table editor. Version 2.0 is an improvement over 1.0, but I still find
fault with basic design decisions within Publisher. Edit/preview is just
not state of the art in my book. It seems that ArborText is targetting
scientific publishing, and feels that Frame cannot penetrate this niche
right now. I wonder how ArborText will feel when Maker 2.0 with equations
comes out.
Publisher does not allow free form documents, like newsletters. It can
handle up to four columns of text per page (why such an arbitrary
restriction?) and cannot handle mutiple text flows in a single document.
I really believe that Frame, while not up to Publisher in equation
handling, is by far the more versatile and powerful tool. I also find it
much easier to use.
Again, I want to emphasize that these are my !opinions!. I think there is
a tremendous interest in doc-gen right now, and that lots of people are
puzzling over which tool to purchase. I would love to compare and
contrast issues with users who have tried both tools and like Publisher
better.
Chuck Musciano
Advanced Technology Department
Harris Corporation
(407) 727-6131
ARPA: chuck@trantor.harris-atd.comho@tis-w.arpa (Hilarie K. Orman) (12/22/88)
The discussion about these excellent products for Sun workstations has brought out some good points. I tried both of them last year, liked them both, and chose Publisher. One of the reasons is that it is NOT purely WYSIWYG. I have often found WYSIWYG distressing because it obscures the semantics of the document structure. It is sometimes the case that what you see on the screen looks OK, but it is wrong at printer resolutions, or it is wrong when modified slightly. This can result from having text in the wrong environment (paragraph instead of list, indented paragraph instead of block paragraph, etc.). With the Publisher, I can see the structure with explicit names and markers on one side of the screen, and I can preview it WYSIWYG on the other side. This saves me "debugging" time. Another benefit of this is that I can edit the document in a larger point size than will be actually used on the printed page. B&W screen resolution is still not good enough to make 10 point type easily readable, and this indicates to me that literal WYSIWYG is impossible today. As far as I know, the Publisher is the only system with good (any?) bibliography support, and this is precisely because it uses TeX's very well developed cabilities for accessing bibliographic databases. Last year ArborText was the only company I knew of with "floating" licenses. Frame has shown good sense in picking up that idea. One thing we found in the time we spent evaluating various WYSIWYG products was that they do put strain on 3/50's. Initially we blamed the vendors for "flakey" software, but later we found that we were running out of swap space, process slots, and text entries. Beefing up the configurations removed most of the problems, except for the generic one of things being a little bit slower than one would like. There are lots of grounds for comparing these products, but some things that are clear are that tastes vary a lot, no one likes learning to use something new, and no one likes spending money on software. Hilarie Orman Trusted Information Systems, Inc. Los Angeles CA (ho@la.tis.com, ...!trwrb!aero!trusted!ho)
carter@uunet.uu.net (Mike Carter - iccad) (12/23/88)
bernhold@orange.qtp.ufl.edu (David E. Bernholdt) writes: >I do not know what the Publisher costs, but we have it licensed for half >of our 60 workstations - we tend to have a lot of people writing papers. >This sounds pretty much like Frame's license server to me. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ I couldn't let this one go by! There is a profound difference between buying many copies of a software package, and having floating (network) licensing. When you have floating licenses, users in the ENTIRE network can share access to useful software, instead of just a privileged few. With floating licenses, it doesn't matter WHICH nodes in the network use the software, so long as no more than the licensed NUMBER of nodes use the software at any one time. A "license server" handles the software check-in and check-out process which monitors the number of currently active users. Companies like Frame (for documentation software), Cadence (for IC design tools), and an ever-increasing number of other software vendors, deserve a lot of credit for their support of floating licenses (also called network licensing, or software brokering). For software purchasers like Mitel, network licensing is the only approach we're willing to consider; there simply isn't any other approach which is practical when you have a large number of workstations. Mike Carter Mitel Semiconductor uucp: uunet!mitel!carter Phone: (613) 592-2122 x3326 FAX: (613) 592-4784
dela@ee.rochester.edu (Del Armstrong) (12/25/88)
chuck@trantor.harris-atd.com (Chuck Musciano) writes: [[ I have removed most of the included text. Anyone wanting to read it can go back and look at v7n54 again. --wnl ]] >I would love to compare and contrast issues with users who have tried >both tools and like Publisher better. I'm one! I did the initial evaluation of document processing systems for our department back in the spring of '87. After looking at a production version of Frame and Interleaf and a pre-Beta version of The Publisher, we selected The Publisher. Some of the reasons for our decision (and current observations): - We ruled out Interleaf due to it's interface. Interleaf takes over your entire screen, prohibiting the use of SunTool based applications while running Interleaf. - Frame Maker was ok, but the things we really want to do it was weak in, or didn't support at all. (Notice the use of the past tense, it has been a while since I looked at Frame Maker ... Caveat Reader!) - The Publisher was (still is) the only one based on TeX. For us this turns out to have been one of the deciding factors. TeX is probably the most developed document formatting system available, simply because so many people use it. I believe that this gives ArborText a real advantage over other developers. Also by writting TeX macros, we can add local hacks to The Publisher in a way that we couldn't to other systems. - Many of our faculty and students have already been using TeX for years, and have numerous documents in TeX. With The Publisher they can continue to use those documents, and their collaborations don't have to go through the "suddenly we're using incompatible software" crisis. Note: we also had some old troff hackers, for them TeX compatibilty didn't help much. This turns out to have been more important then I expected, many of the new faculty I've dealt with had previously written TeX files they wanted to use. If we had gone with Frame Maker, I'd be supporting troff, TeX (no easy task!), as well as Frame Maker. - For similar reasons, being able to "write" TeX files is very usefull. - Since we wanted to do scientific papers, equations and being able to access a references database were crucial to us. At the time Frame Maker failed miserably in these respects. Chuck says that equations will be available in version 2.0. If it's a structured equation editor, if it knows about equation numbers, and how to do inline equations, then it might be worthwhile to compare with The Publisher's equation editor. Keep in mind though, ArborText has been working on their's for a long time, it really is pretty good. As an academic department, we must have the bibliography (references) database ability. The Publisher gives us this with built in support for BibTeX. With The Publisher, we can access all the BibTeX databases people have, and the BibTeX tools people use. - I just didn't (don't) buy the "but it's not WYSIWYG" argument. The editor window in The Publisher is certainly close enough for me to know what the document will look like. Big fonts are big, bold fonts are bold, equations and pictures appear in the document. Granted, I can't tweak how thing line up along pixel boundries until I preview. But when I'm composing prose I find that I don't care about those things, spelling and "prettying up" the document always require another pass anyway (for me at least). This is true even on "real" WYSIWYG editors, such at those I use on the Mac. [[ It's not clear that you should care about things like pixel boundary tweaking. Ask Leslie Lamport about that. --wnl ]] - Although The Publisher is a large complicated program, I don't belive that it's user interface is too complicated. Most of my users are using it without benefit of the manual (they're students working in public labs). Certainly if you want to use the more advanced facilities, you'll want the manual, but that's true with any powerful utility. On the other hand I do agree, it'd be nice if somehow the different aspects of The Publisher all had exactly the same look and feel. Frankly though, I can't get myself to lose much sleep over it. My users seem to be able to handle Publisher's current interface without much problem. Finally I really do agree with your comment about ArborText's intended market. The Publisher is a tool that seems to be nicely crafted for our specific environment. It's not suitable for all things, but it's probably still the best tool around for those with the types of priorities we have. If you have different priorities, then other features will matter more to you. That's the way it should be, heaven help us if someone ever writes the PL/1 of document processing systems, one that's all things to all people. [[ And I am glad this discussion is continuing, because it should help people decide what their priorities and requirements are and what will be best suited to their needs. --wnl ]] Del Armstrong Internet : dela@ee.rochester.edu UUCP : ...allegra!rochester!ur-valhalla!dela Twisted pair: (716) 275-5342 Last resort : Hopeman 407 Electrical Engineering University of Rochester Rochester, N.Y. 14627
bwhittak@libra.uucp (Brian Whittaker) (01/14/89)
dela@ee.rochester.edu (Del Armstrong) writes: > - We ruled out Interleaf due to it's interface. Interleaf takes over > your entire screen, prohibiting the use of SunTool based applications > while running Interleaf. Just to set the record straight (or up to date) - we have Interleaf 3.0 and it does run quite happily *under* SunView allowing full access to SunTool based applications while running Interleaf. I'm not sure of the details, but I think it requires SunOs 3.2 or later (don't know about 4.0). I'm also told they recommend at least 24M of swap space, but we've been using it happily with only 12M. Brian Whittaker, Prime Computervision, Amersham, HP7 0PX, UK bwhittak@cvedg.prime.com bwhittak@uk.co.cv.edg +44 494 714771 x 304
rodgers@maxwell.mmwb.ucsf.edu (01/14/89)
In v7 issue 70, Hilarie Orman (ho@la.tis.com, ...!trwrb!aero!trusted!ho) writes: > The discussion about these excellent products for Sun workstations has > brought out some good points. I tried both of them last year, liked them > both, and chose Publisher. One of the reasons is that it is NOT purely > WYSIWYG. > > ...[many interesting remarks] > > There are lots of grounds for comparing these products, but some things > that are clear are that tastes vary a lot, no one likes learning to use > something new, and no one likes spending money on software. I find Hilarie's remarks the best thus far posted on this topic. I would add a few remarks ca. WYSIWYG vs. batch-mode processors like troff and TeX: 1) As Hilary points out, nothing short of specialized publishing hardware is TRULY "WYSIWYG," and this still involves considerable expense. 2) Although the psychological appeal of seeing a representation of the final document are undeniable, there are costs associated with this: for example, many WYSIWIG systems expose the user to delays while reformatting on the fly. In a batch-mode environment, the user goes off and does better things with his time while the actual formatting takes place. 3) Many users (myself among them) find the act of using WYSIWYG procesors a distraction from the writing process. Leslie lamport (author of LaTeX) has made a similar point in one of his essays. I actually PREFER to sharply separate the editing of a file containing my thoughts from the act of formatting them. When I am engaged in the former activity, I am concentrating on what I am saying and how I am saying it, on the intellectual organization of the presentation and grammer. This must all be in good order before I start worrying about where page breaks occur, etc. Hilarie's flexible attitude is important. I work alongside people who produce our monthly departmental newsletter, and it is clear that for that activity, the Mac-based WYSIWYG system they are using is well suited. It is equally clear to us that for the grant proposals, technical manuals, and manuscripts we produce, our present troff- and TeX-based tools are better suited than any of the WYSIWYG systems we have used, which include both The Publisher and FrameMaker. Troff and TeX are also very portable and incredibly cheap. R. P. C. Rodgers, M.D. Telephone: Statistical Mechanics of Biomolecules (415)476-8910 (work) Department of Pharmaceutical Chemistry (415)664-0560 (home) University of California, Box 1204 E-mail: Laurel Heights Campus, Room 102 ARPA: rodgers@cca.ucsf.edu 3333 California St. rodgers@maxwell.mmwb.ucsf.edu San Francisco CA 94118 BITNET: rodgers@ucsfcca USA UUCP: ...ucbvax.berkeley.edu!cca.ucsf.edu!rodgers