Out of the 41 people that answered the survey, 10 said they did not have
SCRIPT/VS, and I'm inclined to think that at least 2 of them have Waterloo
SCRIPT but had erased the exec and did not re-run it when I said Waterloo
script was ok too. Sorry about this confusion, I just did not even know that
there WAS a Waterloo script :-)
8 people said they do have SCRIPT but would prefer the docs to remain in
plain format -- it is funny to see that these people are mostly of the "real
programmer" category, if I may draw the old sword of the "real programmers"
versus "quiche eaters" croisade ;-) That leaves us with more than 20 people
who would prefer SCRIPT/VS output. Here are their answers to the following
questions, expressed as a percentage of the nodes which have SCRIPT and would
prefer SCRIPT docs:
'Do you have a page printer (3820, 3800-3, 3812, et al) available?'
78% yes, other answers from people who only a XEROX-like printer which
can't be considered as compatible with the IBM page-printer standard.
'Do you have any local mods to the SCRIPT GML Starter Set?'
34% yes -- funny, I'd have thought it would be less, but of course "vanilla
IBM code is not fun" as Jeff says :-)
'Do you think these mods could severely impair the formatting of a
0% yes -- people trust their mods :-)
' 1. Would you take opportunity on this to reformat them to your'
' institution''s document printout conventions, eg by using'
' modified GML APAs, or to add your own local addenda to the'
' documents (using .IM for instance), etc?'
60% yes -- SCRIPT really makes it much easier to add you own preface or
appendix
' 2. Would you distribute copies of the manuals to members of your'
' user community?'
91% yes
'How many copies (per document), very approximately?'
--> flame <--- How can my stats mangler process answers like "???", "20?"
or "10-20"? Next time I'll add an "IFY2012R FORMAT USED FOR THIS
INPUT IS 'I4'" :-) Anyway it was between 10 and 150, usually around
20-30 but a lot of replies mentioned people being able to request
their own copy be printed thanks to an online -> printout server or
suchlike, so it really doesn't mean much.
'To which category of people (any 80-chars string)?'
mostly system staff and consultants, plus of course list owners.
' 3. Do you think this would significantly help your users'
' understand how to use LISTSERV?'
86% yes -- my, people don't like reading info from a CRT!!!
' 4. Do you think this would make LISTSERV more popular at your'
' institution?'
82% yes
'On the overall, how would you rate the migration to SCRIPT/GML on'
'the basis of the particular needs of your institution?'
' 1. Rather undesirable'
' 2. Neither desirable nor undesirable'
' 3. Desirable, but should not be a primary concern'
' 4. Very desirable'
Everybody answered 3 except 2 persons.
*-----------------------------------------------------------------------------*
In light of these replies I have decided to start migrating the LISTSERV
docs to SCRIPT format. The docs will be available both in SCRIPT 'source'
format and in pre-formatted form (as per a SCRIPT (DEV(2741) FILE)).
Pros:
- All that was said in the survey
- Makes it easier for me to make changes to the docs. You probably don't
realize it, but UPDATING the docs takes an awful lot of time, 50% of the
time needed to completely rewrite all the updated sections. Whenever
something is no longer applicable, I must remove a few sentences from a
paragraph and here you are, justification is
completely broken like that and
I must completely
reformat the paragraph, same thing when I want to add a sentence. I end up
adding sentences AFTER the paragraph to save time but it makes the doc a
little messy. :-)
- I'm really missing an index at the end of the docs. I'm used to the
"polys" distributed by our school (500 pages of endless blahblah about
some 10-hours of courses; you MUST purchase these books from the school,
otherwise you're automatically fired, but as a compensation they allow you
to use the Official Books from the School at some exams). These books
NEVER have an index and it's a real pain to find something in them during
an exam. One of the best features of SCRIPT is its ability to
automatically create an index and a TOC and it will help people a lot.
- It will be possible to make HELPCMS files out of the SCRIPT sources. At
least, I saw an IBM doc that suggested to use SCRIPT to create HELPCMS
files so it is probably possible, otherwise they would not document it :-)
- It will be able to prepare decent handouts out of the doc sources, to be
distributed to (various categories of people who can't read matrix-printer
printout and won't open the doc if there is not a nice-looking drawing on
the front page :-) ) at meetings.
Cons:
- Plain files are more comfortable to read at a terminal than SCRIPT/VS
output which contains page numbers, page skips (--> blank lines) and
suchlike. This will look complicated to the average end user. It's not a
major problem though.
- These files will be larger than plain files, because of the blank lines,
title pages, TOC, etc (but I believe this can be solved with a few .if
&$PDEV = 2741 .go ...)
- Compatibility problems between IBM and Waterloo script.
- More space wasted on the LISTSERV disk since both source and object files
will be needed.
Well, I intend to migrate to SCRIPT format slowly. I'll start with the new
docs I still have to write (ie I won't start plain-format docs any more) and
will then slowly proceed to convert the existing docs when I have important
changes to make to a doc which would anyway require a long time. Docs
converted to SCRIPT format will no longer be available in the old plain
format and will be replaced by their corresponding 2741 output. Although most
of the existing files will only be a subset of the future manuals (eg an
appendix or a chapter), they will still be available individually for
efficiency reasons.
Distribution for the SCRIPT source files will be as a separate shipment, so
that nodes which don't have SCRIPT can opt not to receive the SCRIPT source
files (--> disk space savings). However, major hub nodes which don't have
SCRIPT are encouraged to provide the files if they have enough disk space so
that people don't have to go in the boonies to order the source files.
Finally, I need to find a volunteer using Waterloo script to find out what
the major incompatibility problems may be so that I don't start docs using
options that Waterloo script can't handle (for example I like using .DC GML <
< - and .DC MCS > so that GML tags become 'This is an <hp1> hilighted <-hp1>
word' -- I don't like :hp1.xxx). The nearer this person is to FRECP11, the
better, of course.
Eric
|