Big Blue Disk 23/bbd23.zip/DESK.TXT

Text File | 1988-06-21 | 7KB | 102 lines

|D╔══════════════════════╗════════════════════════════════════════════════════════
|D║ |5On the Editor's Desk |D║════════════════════════════════════════════════════════
|D╚══════════════════════╝════════════════════════════════════════════════════════

^C^1On the Editor's Desk
^Cby
^CDaniel Tobias

   In this column, I give news and reviews of products and developments that
have reached my desk.

   In a few earlier columns, I have mentioned Ted Nelson and his visionary
Xanadu project.  Now, it looks like a version of this system may actually come
out as a product in the near future: a company called Autodesk, Inc. has just
purchased an 80% interest in the Xanadu Operating Co., which is presently
readying a commercial version of this hypertext-based system.  Xanadu is
basically a system for storing massive quantities of documents (letters, books,
program code, etc.), and allowing interlinkings to be made between any part of
any document.  Documents can also be created that contain "windows" quoting
parts, large or small, from previously-created documents.  This can be used, for
instance, to produce a new version of a document without erasing the old
version; the new version would "window" to all unchanged portions of the old,
and this unchanged text would hence only have to be stored once, saving disk
space.  Anthologies can be made of previously-stored material by simply linking
to the desired portions, adding whatever new introductory or transitional text
is desired.  The whole process can be made transparent to the user, who can read
through a document without regard to the origins of the separate parts (unless
he's curious, in which case he can press a key to see the original context of
any "quoted" material, then press another key to get right back where he was in
the original document).

   Nelson has dreams of turning this system into a publicly-accessible network
that would ultimately replace the printed word as the major repository of human
knowledge.  This is still the ultimate plan for the project.  However, it is
being initially marketed in its narrower application as strictly a system for
office automation, which would be licenced to various corporations for use in
their inhouse systems.  There, Xanadu would have many obvious applications in
handling the office tasks normally performed using reams of paperwork.  For
instance, memos can be sent electronically with linkages to earlier memos and
other documents to which they refer.  This office version of Xanadu is expected
to be released sometime next year.

   Should Xanadu catch on in the office environment, the project will gain much
respectability and money that can be used to launch the global network Ted
Nelson has been dreaming of for all these years.  I hope it does; I can't wait
to see it.


   Another company to watch:  Activision, which was started as a video game
cartridge manufacturer and later expanded to entertainment software for
personal computers, is now diversifying to include productivity-oriented
software products as well.  To reflect this change, they're renaming themselves
"Mediagenic."  The Activision name will still be used as a trademark on their
entertainment products.  Mediagenic has recently pulled itself out of a rut they
were in following the collapse of the video game craze, and are showing profits
again.  They have acquired several other software companies, including Infocom,
makers of some of the best interactive fiction products around.  With all these
acquisitions, they have some very talented programming staff working for them.
With a new name dispelling their "game" image, it will be interesting to see if
Mediagenic can make it in the productivity arena in addition to their successes
in entertainment software.  They certainly seem to be a company to watch.

   One productivity-oriented program I hope they decide to market is
Cornerstone, a database program released a few years ago by Infocom.  It flopped
in the market, mostly because Infocom had no experience marketing non-game
programs, and their name was mostly associated with adventure games.  (Actually,
writing adventure parsers has much in common with devising database query
handlers.)  I haven't seen a copy of Cornerstone (they're not presently giving
review copies, since they're officially not marketing it anymore), but everyone
I speak to who's heard of it at all thinks it's very well done.  I've run a demo
of it, and its features seem impressive for such a low-priced package.  (The
demo itself is rather slow and tedious, but that's the fault of the demo
presentation, which doesn't actually use the Cornerstone program itself.)  It's
a full relational database, letting you link different data files together in
various ways, like to link the people in your personal address book to the
companies they work for in a database of companies, and to link your personal
notes on the outcome of phone calls to both the person called and the company
represented.  I suppose Xanadu (mentioned above) will let you do all that stuff
and more, but Cornerstone is available now at low cost for the PC.  You can buy
it from Infocom, but they're not making any effort at all to sell it.  I hope
that Mediagenic picks it up and re-releases it under the umbrella of their
productivity software division.  It could be a real hot product if marketed
right.  (Write to Infocom, 125 Cambridge Park Drive, Cambridge, MA 02140 or call
them at [phone], to request more information about this product.)


   Update department:  Last month, I reported a few bugs in the script language
of Procomm+ from Datastorm Technologies.  Here's another.  I used the following
sequence of commands in a script file:

           dos "if exist cis.obk del cis.obk"
           dos "if exist cis.out rename cis.out cis.obk"

What they should do is call DOS twice, first to delete any file named CIS.OBK
that may exist, then second to rename the current CIS.OUT to CIS.OBK.  What it
actually does is to ignore the first line entirely, then fail at the second
command because CIS.OBK exists, and you can't rename a file to a duplicate name.
The funny part is that sometimes this sequence of commands works fine, but other
times, when I've made some trivial change at a completely different part of the
script, it fails.  I can't find any rhyme or reason.  Can somebody at Datastorm
explain this one?