Text preview for : 19770830_Janus_Functional_Specification.pdf part of xerox 19770830 Janus Functional Specification xerox sdd memos_1977 19770830_Janus_Functional_Specification.pdf
Back to : 19770830_Janus_Functional | Home
Inter-OHice Mcmonmdum
To J. Szclong Date Augllst 30, 1977
From Wendell Shultz Location Palo Alto
Subject Janus Functional Specification Organization SO~/SO
XEROX XEtOX SDD ARCHIVES
I have read and unue~stuod
Pages _________ To ______ ~_
Filed on: (Lopez)Janllsreview.memo
Reviewer Date _ _ __
# of Pages Ref " 11~/)t>-'3fJ
These comments are relative to the July 22 Janus Functional Specification. Many points
have been covered in a number of other memos. Let me try to list some that may not have
been covered. Sorry that these comments did not get to you sooner, but I have had a hard
time getting responses collected from my section. J will summarize most of the comments
and then attach three writeups from my people that you may not have seen yet. (These
cover many detailed points.)
"Discussion Required" ltem~
1. In its present form, I do not see how this product can span the whole range of users--
from unsophisticated lIsers doing letter or short memo writing to those editing complex
technical or business documents or doing page layout. It is too rich and complex for the
simple lIser--especially the abstractions of option sheets. That is, a sophisticated user
can lise the system to do simple things, but a simple user may be completely lost.
Several other suggestions have been made to try to improve this area--and the document
does do many good things in this regard. Maybe a revised version will solve all this.
But I question the idea that one "universal" editor can serve all users. It has never
worked before. Icons might help, but maybe not totally. It still may be necessary to
"package" this as several different "application" packages, with increasing levels of power
and sophistication.
2. There are many comments on file naming. (See the memo from Kimball, item #3;
Bishop memo; Redell memo; Burr memo.) Basically these all ask:
a. Should string names be unique?
b. String names vs. icons
c. Multiple index (name) fields versus single (unique) names
d. Versions or other modifiers
e. Cataloging dismounted volumes
f. Naming of remote files (over the Xerox wire)
3. The question of filing on the floppy disk versus a future rigid disk generates a lot of
potential issues. For example:
a. What does it do to performance or allowable file size to "promote" files from a
user floppy to a system floppy? Why not have this only as an option, not a
standard? Why force always copying files, since they are not updated in place
anyway?
Janus Functional SI>,ecification 2
b. Why make the decision about "edit" or "view" when starling to work on a file?
Why not make it a/fer looking at or editing a file--i.e. by then saving what was
done, or not?
c. Would we catalog dismounted files or file drawer with a rigid-disk-based system
but not a floppy-based system?
d. What does "protected" status mean on a floppy? If it contains sensitive data, just
keep the floppy physically secure. It seems unnecessarily complex to try to
protect some document but not others in a floppy based file drawer. (A Janus
"B" or rigid-disk based system will have different protection needs, however.)
e. What sorts of messages appear to the user if his files don't fit on the scratch
space on the system floppy? On a user floppy (after updating)?
f. What is the effect of these file structures on UNDO or REPLAY?
4. The whole question of storage media other than standard OIS floppy is not covered.
a. What about tape cassettes?
b. What about magnetic cards?
c. What about Troy floppies?
d. What about IBM (OS-6) floppies?
5. What happens to the function keys in an lAS based application? For business control vs
editing in general? This is one of the reasons to reduce the number of function keys.
6. Should the functional specification contain reasons or rationale for why some things are
done this way? This document should help "sell" these concepts; or is there another
document to do that?
Attachments (Memos from Bishop, Burr, Redell)
c: L. Bersteinsson
P. Heinrich
C. Irby
D. Liddle
W. Lynch
R. Metcalfe
D. Stottlemyre
T. Townsend