Archive Magazine CD 1995/Archive Magazine CD 1995.iso/discs/prog_disc/volume_4/issue_03/pipeline/antonmans/colton1

Text File | 1990-04-13 | 9.6 KB | 689 lines

%OP%LM8
%OP%FO//- page @P@ -//
%CO:A,4,72%%R%%H7%%H8%Addington Hospital
%R%PO Box 977
%R%4000 DURBAN
%R%REPUBLIC of SOUTH AFRICA
%R%3 April 1990

Messrs Colton Software Limited
Broadway House
149-151 St. Neots Road
Hardwick
Cambridge CB3 7QJ
UNITED KINGDOM

Dear Sirs,

I recently purchased PipeDream 3 (3.07R 15-Jan-90), and I am very happy
with the program.

I use a 310, with no memory expansion, and as configured, it starts up
with about 610000-odd bytes of free memory. I have RISC OS, and dual
3½" floppies, with an external 5¼" drive. I also have a two slot
backplane with Watford's scanner attached.

I have a few queries:

1.


2.




3.


































4.



















































5.










6.





















7.









8.












9.








I hope I have not created the impression that I am knocking PipeDream.
I am most certainly %H2%not!%H2% I am highly impressed with the product, but,
as a registered user, I feel that I would be remiss if I were not to
offer these suggestions.

I look forward to hearing from you.

Yours faithfully,







Dr Anton L. Mans
Registration number: R9103 4510 3717 2578
%CO:B,12,68%
























I cannot find any explanation of Macro Parameters, bar the mention
on page 179.

Is it possible to direct output from the Spelling checker to
printer or file? I cannot find a way of doing so. Subgrams can be
particularly numerous, and it would be convenient to be able to
scrutinise a hardcopy at one's leisure.

I was spell-checking a very long file (which occupied most of the
memory - it is a section of an even longer database), and it was
frightfully slow. I am happy to accept your claim of 40000 wpm, and
happily concede that a floppy-based spelling checker is unlikely to
impress with speed.

This all acknowledged, can anything be done to ease the pain of
waiting in front of the screen, finger poised above a mouse button,
to tell the wretched thing what to do with every word it cannot
identify?

May I suggest an option to spool all unknown words to a disc file,
which can then later be edited and merged with a user dictionary?
One could then go and do something else while the checker is
checking.

Another way would be to hold the dictionary in memory and the
document to be checked on disc on a 1Mb machine, instead of the
other way around, as is currently the case (or such an option
might be user-selectable). The words in the document would be
accessed sequentially as stored on disc (or even buffered to the
maximum extent of free RAM), while random dictionary accesses are
in memory and thus quicker. I believe a considerable saving in time
could be realised.

On the subject of spelling checking: it would be nice to be able to
choose to have hyphenatable words hyphenated. I imagine this
feature would involve a disproportionately great deal of effort,
though?

It would also be very useful to be able to spool anagrams and
subgrams to a file or printer, so that one could then examine them
at one's leisure, instead of having to keep up with the
one-directional scrolling window provided.

A major problem - I cannot get my printer to produce the cent sign.
I have a Star LC10 (not colour), which is Epson LX-800, FX-86e,
EX-800, LX-86 and LX-80, and IBM compatible, according to the
manual. I keep the printer's DIP switches set to USA character set,
mainly because I have little use for the pound sign (except for
occasional letters to the UK!), and usually set to Standard mode,
i.e., all DIP switches ON in both banks.

I enclose a printout of the fx80 printer driver, as supplied by
Colton. It appears that the character immediately preceding the
cent sign, as well as the cent sign itself, are swallowed up by
some mysterious printer sink. To further confound matters, the
reaction is inconsistent. Sometimes it converts cent signs to 6's
(probably due to a misinterpretation of the leading 'ESC "6"' in
the translation), sometimes it turns a pound sign in the same line
into a capital 'R' (I cannot explain that), and sometimes it indeed
produces a cent sign.

I have tried a variety of remedies. I am a reasonably competent
amateur, and I have little difficulty with printer manuals. I tried
replacing the trailing 'ESC "R" 3''s throughout with 'ESC "R" 0'
(except preceding the pound translation, to which I appended
'ESC "R" 0' to restore USA characters). I tried setting my printer
up to start up with the UK character set, and using the fx80 driver
as is. Nothing cured the problem.

Defining any other character to produce the cent sign produced
similarly wierd and unpredictable results. This only affects
production of ASCII character 155 (the cent sign). No other
characters in that range are affected.

I chose a hex dump from the printer's frnot panel. This, as you
probavly know, causes the printer to produce a hex and ASCII dump
of all characters it receives, in a format similar to Acorn's
*Dump. This indicated that all codes, whether control or otherwise,
are received correctly. Is there something wrong with my printer?
So I tried sending character 155 in isolation (*Echo <2><155><3>) -
and produced a cent sign, as expected. Also, using this printer
with Master 128-Edit, produces cent signs correctly. I usually set
up the printer in Master 128-Edit to choose the IBM set #2. This is
done by sending the required ESC sequence, and not by setting DIP
switches, which remain in Standard mode.

I am utterly mystified. Can you help?

While on this subject: I usually find the IBM character set to be
more useful than the Latin or other ISO sets, as it contains most
of the characters I use, as well as the graphics characters used in
drawing boxes around tables. If some way could be found to have an
IBM set defined on the Archimedes, and a suitable printer driver
for PipeDream supplied, it would be most gratefully received!

It does not seem possible to have multi-line headers or footers? I
have used Edit as supplied with the Master 128 almost exclusively.
I never got to like VIEW, and the lack of multi-line headers and
footers was one of the reasons. I personally like to have a line
drawn immediately above my footers (i.e., a string of underline
characters). This implies a two-line footer, and was possible with
Edit on the Master (and still is, under the BBC Emulator). It could
conceivably also be useful for headers, such as in legal documents.
If it is not already possible, can it be introduced in future
developments?

How does one go about printing wide spreadsheets, other than as a
wide landscape with a RISC OS printer driver (which is too slow)?
I used ViewSheet to do my domestic bank account. This was about 30
columns wide, to aid in budgetting. The first 3 columns were cheque
number, date and item. I would divide the remaining columns into
three more or less equal sections, and arranged print windows so
that the first three columns were printed in front of each section.
The spreadsheet was thus on three pages, with a fourth page
containing the bank reconciliation statement.

I see that it is possible to print a range of columns, but it
appears that only one range may be specified at one time. I would
like to be able to specify not only ranges of columns, but also a
selection of blocks. This could be something like:






Is there a better way of printing wide spreadsheets with PipeDream?

Edit (Master 128) gave a very flexible page numbering system. One
could use Arabic numerals, Roman numerals (upper or lower case),
section and page, etc. Could this be implemented in PipeDream?

I might add that I am very fond of the Master 128's Edit. If one
does not lay too great a store by WYSIWIG-ness, it is probably one
of the most flexible word-processors around. This is not an implied
criticism of PipeDream - I trust that, as I become accustomed to
using PipeDream, I shall become equally addicted to it!

Is it possible for a way to be provided to introduce special
printer codes into a document? The facilities provided do not offer
sufficient scope to implement all of the facilities in modern
printers. The Star LC10 can produce several different NLQ fonts at
four different pitches and various different heights, for example.
This is probably not a standard feature, however, but I would like
to use it from time to time. There are also other features. Using a
new highlight for thus tailoring one's own individual (eccentric?)
purposes seems appropriate. Obviously, some codes would need to be
counted for layout purposes, while others would be exclusively
printer control codes, and would not appear as characters in the
printed document.

As I understand from the Manual, if the same document is loaded
into two different windows, then making changes to version 1 will
not be reflected in version 2, and %H4%vice versa%H4%. While this is
undoubtedly useful in certain situations, it would also useful if a
"New view" option were offered, as in the applications provided by
Acorn. One would thus be able to have two views of the same wide
spreadsheet, for example, and be able to see changes in view 1
being reflected in expression slots visible in view 2.
%CO:C,14,56%

















































































































































A1-C25,D1-H25
A1-C25,I1-M25
A1-C25,N1-R25
AA1-AE15
%CO:D,12,44%

















































































































































(date cheque# item-name, 1st 5 columns)
(date cheque# item-name, 2nd 5 columns)
(date cheque# item-name, 3rd 5 columns)
(bank reconcilitation statement)
%CO:E,12,32%%CO:F,12,20%