From:	D0SFT::BNLD0::SERBAN       "Serban Protopopescu (516) 282-3721" 11-SEP-1995 10:09:21.54
To:	D0SFT::D0CD03::QZLI
CC:	SERBAN
Subj:	RE: minutes software tools meeting


        MINUTES OF SOFTWARE TOOLS MEETING (as heard and seen by SP)
        =================================

               August 30,1995

Present: D. Hedin, J. Hobbs, A. Jonckheere, K. Genser, L. Lueking,
Q. Li-Demarteau, S. Protopopescu, A. Para and L. Paterno


    The meeting was spent discussing  the recommendations by Jonckheere and
Hobbs  (as   interpreted in a  mail  message   distributed  by  Qizhong and
Harrison)  for the  handling of  D0  documents and  search for  an official
replacement of EVEDT as a D0 editor (for text and code).

    It is generally agreed  that documents should  be accessible on WWW and
managed by  WWW. One  problem is  D0NEWS, a tool  needs to be  developed to
handle notification of messages. It was suggested that TEX/LATEX be used as
standard for documents.  Jonckheere argued  against it, we should only list
requirements:

1) Viewable
2) Printable
3) Searchable by text

These requirements are met by HTML, TEX/LATEX, Postcript and plain ASCII text.
Pure graphics applications need not satisfy 3).

    To  make  files   accessible they  must  reside  in a  central  server.
Documents can be written  in anything the author  wishes but source file on
server should  be HTML. It is  a problem to  keep files in  synch without a
single  source.   Manipulation of  documents  by reader  should  be through
Mosaic.  There  was a  great  deal of  discussion   concerning  printing of
documents.   Hobbs  emphasized  that  printing  should be   specific to the
portions  of  the   document the  user  wants  and  not  generate  pages of
extraneous  auxiliary files.  Nobody  disagreed with that  statement. It is
believed that  this is possible in  Mosaic but was not  clear the tools are
sufficiently refined  yet. The simplest  short  term solution is to provide
easy access to postcript files to copy locally and print.

    There was  a  discussion on the  need to  catalog  documents but it was
argued that  we should  be careful  not to invent  some  cataloguing scheme
dependent on keys. Such systems have been tried in the past and it was felt
they  never work  well,  users  cannot  remember  obscure keys.  Search for
documents should be by strings of meaningful information.

    A question raised by Lueking was whether we should provide template for
systems used in  preparations of documents (such  as FRAMEWORK, LATEX files,
etc). This brings  up the question  of how many systems  should D0 support.
What was agreed upon was  that D0 should set  format standards that must be
followed before a document is accepted on the central server and that tools
are needed to enforce  standards. Well defined  procedures that are easy to
follow must be implemented for installing documents on the central server.

    Some of the  standard  templates can be  provided via  the D0 supported
editor. Adam Para made a  request for stronger  language sensitive features
in the future  D0 editor than  are now  available in the  present one. This
lead to a discussion on the list of  requirements that should be made on an
editor.  The  choice  should  be  dictated  by a  practical   solution that
satisfies most  (if not all) the  requirements on that  list. Hobbs in fact
had  already  prepared such a  table. His  view is  that EMACS  is the most
widely  available editor that  can easily  accomodate D0  standards. No one
disputed that view.

    In addition to documents it was  argued that we should encourage use of
computer generated transparencies,  this makes the transparencies as easily
available as any  document. There  was then a  discussion about encouraging
storing information from  meetings ranging from  minutes and transparencies
to audio tapes  and videos. The last  items were felt  to go beyond what we
know how to do  with present  technology but  could become  feasible in the
future. It was  agreed that  we need to  generate a list of  what we really
want  and  Adam  voiced  some   concern  that in  the   process of  listing
requirements we do not close the door to future technological developments.

    At the end of the meeting it was  agreed that Jonckheere and Hobbs will
send a  written set of  requirements  and  recommendations to  Harrison and
Qizhong  who will  generate a  second  draft of a  document for  discussion
during the next meeting.