T.R | Title | User | Personal Name | Date | Lines |
---|
3654.1 | | BRUMMY::MARTIN::BELL | Martin Bell, NETCC, Birmingham UK | Sat Dec 11 1993 12:00 | 15 |
| This doen't actually answer your question, but if you want to ensure
that you have entered a real POUND sign in your document, and not a
HASH that may or may-not display as a pound, then use the dead-key
sequence ...
L-{left arrow}{gold D}
or on a terminal with a compose key
{compose}L-
which generates a true 8-bit pound-sign character.
mb
|
3654.2 | Compose ++ doesn't work!! | KERNEL::VANRIXTELE | Emma van Rixtel | Mon Dec 13 1993 09:12 | 6 |
| Well, this is another problem, a compose ++ actually produces a pound sign
instead of a Hash sign!!!
Some more help please......
Emma.
|
3654.3 | Looks like an emulator (SETHOST) setup problem? | BRUMMY::MARTIN::BELL | Martin Bell, NETCC, Birmingham UK | Tue Dec 14 1993 12:36 | 11 |
| If compose ++ produces a pound sign, then the terminal emulator has been
configured (reset or whatever) to display the HASH character (ascii
code 35) as a POUND. As far as the VAX is concerned, it is a hash,
but the display shows pound. (Unless emulator is extra clever, but
we will assume not for now).
Do you have access to a bog-standard VT terminal, rather than an
emulator, so that you can check exactly what character is stored
in the file?
mb
|
3654.4 | check character sets | SUOSWS::HAMANN | | Wed Dec 15 1993 07:43 | 8 |
| Hi,
check the settings of the terminal emulator, it should use the
ISO-LATIN or MCS character sets, but seems to use a national
replacement set.
Klaus
|
3654.5 | HELP - URGENT!! | KERNEL::VANRIXTELE | Emma van Rixtel | Thu Dec 16 1993 09:23 | 15 |
|
Hi,
I have spoken to the customer and he is still seeing exactly the same
problems on a VT Terminal as he does on his PC. I have checked all his
terminal emulator settings and they all look ok. He says compose ++
still produces a pound sign, even on a VT.
I also got him to try the ISO-LATIN character set, but this didn't make
any difference either. Please can anyone suggest ANYTHING else, because
he is getting very angry!!!
Thank you,
Emma.
|
3654.6 | should be easy now | SUOSWS::HAMANN | | Fri Dec 17 1993 08:22 | 18 |
| This is a very clear case: some program the user is running changes the
terminal setting from ISO-LATIN-1 to British.
Stay with the plain VT Terminal.
Make sure the SET-UP is ISO-LATIN-1 and have hime save the setting.
Test the COMPOSE ++.
Let him test COMPOSE ++ after any program he uses.
As soon as the � arrives he has located the guilty program.
Add a reset terminal sequence somehow after this program exits.
Good luck.
Klaus
|
3654.7 | WPS-PLUS changes the settings | KERNEL::VANRIXTELE | Emma van Rixtel | Fri Dec 17 1993 09:05 | 17 |
| The thing is we know which program is causing the problem, because it
only happens when he enters the WPS-PLUS editor, and ok he could put
some sort of reset after he enters WPS-PLUS but it is actually within
WPS-PLUS where he wants to see the # sign but see's a pound instead.
Something needs to be set, so that when he uses compose ++ in WPS-PLUS
editor it produces a # sign instead of a pound sign.
Why does WPS-PLUS suddenly change the ISO-LATIN to British? Is there
any way of making it more permanent??
Sorry to keep going over this, but I am getting really desperate for
answers now!
Thank you
Emma.
|
3654.8 | Check a few more things ... | BRUMMY::MARTIN::BELL | Martin Bell, NETCC, Birmingham UK | Fri Dec 17 1993 10:01 | 30 |
| Emma,
A few more questions that may trigger an answer ..
which LANGUAGE of ALL-IN-1 are you using - BRITISH or ENGLISH?
When it fails, have you run A1WPSPLUS_LOGIN.COM?
Have you done a SET TERM/INQUIRE?
What else is in LOGIN.COM (or SYLOGIN.COM)?
What is the result of SHOW TERM?
Is the terminal set to be EIGHTBIT?
Do you include any command qualifiers when you start ALL-IN-1?
Is there an OAINI.SCP or an OAINI.COM in either the user or OA$LIB area?
Is the WP menu or WPEDIT script customised?
Exactly what model of VT terminal are you using in .3?
There are many things that _could_ be causing the problem in some direct
or indirect way. We will probably all kick ourselves when we find out!
mb
|
3654.10 | one more question to ask | SUOSWS::HAMANN | | Mon Dec 20 1993 07:17 | 8 |
| re .8
one more question:
Does he run the OA$LIB:WPKEYBOARD procedure and what terminal does he
indicate?
Klaus
|
3654.11 | Getting there...slowly!! | KERNEL::VANRIXTELE | Emma van Rixtel | Tue Dec 21 1993 16:28 | 41 |
| Ok, I have spoken to the customer and have managed to find out a bit
more.
He is definately running the British ALL-IN-1 kit. He tried the
following out on his VT 300 and the results were as follows..
Running A1WPSPLUS_LOGIN.COM after it failed managed to somehow resolve
the problem on the VT. In other words before he ran this, compose ++
was producing a pound sign, but after running the file it continued to
produce a hash sign. So he is happy with that for the time being, but
this same procedure doesn't actually make any difference on the PC.
Why would that be?
SET TERM/INQUIRE made no difference on either the PC or the VT.
LOGIN.COM doesn't contain anything out of the ordinary. He even tried
with the most basic login.com, but still got the same results.
SHOW TERM has always been set to EIGHTBIT, so no change there.
He said he has found an OAINI.COM in OA$LIB, but the contents of it
appear to be commented out, so he doesn't think that that is the
problem.
Entering ALL-IN-1 with a /NOCUSTOM made no difference either!
With regards to .9 I am not too sure about running the
OA$LIB:WPKEYBOARD. Could you please give me some pointers as to the
context in which I should be running it, and which keyboards I should
be advising him to try.
He has mentioned something about customizing his keyboard mapping as
explained in the SET HOST TERMINAL EMULATOR GUIDE p3-28. I have read
it, but do not know how this will react to his other settings. Can
someone advise me on the consequences of this, and what I should advise
my customer.
Thank you very much again!!
Emma.
|
3654.12 | look and try | SUOSWS::HAMANN | | Wed Dec 22 1993 08:05 | 8 |
| Emma,
the procedure OA$LIB:WPKEYBOARD.COM should be run immediately before
entering ALL-IN-1. Look at the first 50 lines for available options.
The British variants will provide the trouble you have seen, so try the
US variants.
Klaus
|
3654.13 | ... but surely! | BRUMMY::MARTIN::BELL | Martin Bell, NETCC, Birmingham UK | Wed Dec 22 1993 09:17 | 37 |
| So just to confirm things, on a VT300 terminal, if you run A1WPSPLUS_LOGIN
_before_ entering ALL-IN-1, COMPOSE ++ displays as a HASH. Is this the case
both within and outside WPS-PLUS? Also, does COMPOSE l- produce a POUND,
within and without? If so, then hopefully setting the correct character
set in terminal setup will make SHIFT-3 produce the required character!
If the VT300 behaves this way, then all that we have to do is get the
terminal emulator to behave in the same way!!!
From the VT300, you can see what VMS and WPS-PLUS think about the
terminal, both via SHOW TERM and looking at the logical names set
when A1WPSPLUS_LOGIN is run.
Now use the emulator and see if VMS and WPS-PLUS still think that
the terminal is the same.
If the problem still exists, but the VMS/WPS-PLUS settings are the
same, then _maybe_ you need to change the keyboard mapping, because
everything is the same except for the behaviour of the emulator.
If the problem still exists, but the VMS/WPS-PLUS settings are different,
then somehow the answerback from the emulator when the SET TERM/INQ is
performed, so maybe some other setup in the emulator needs adjusting.
Regards
mb
p.s.
Check that a SET TERM/INQ is performed in the v2.4 A1WPSPLUS_LOGIN,
because i only have v3.0 and i am sure that it was missing in an
earlier version of ALL-IN-1.
p.p.s.
Sorry about the delay, i have been out of the office for a couple of days.
|
3654.14 | yes I know, but.... | KERNEL::VANRIXTELE | Emma van Rixtel | Wed Dec 22 1993 09:50 | 25 |
|
Well from what I know so far, the A1WPSPLUS_LOGIN.COM definately
includes the SET TERM/INQUIRE for v2.4, so that is ok. I am in the
process of waiting for the customer to get back to me with his results
of running the OA$LIB:WPKEYBOARD.
It is not simply a case of changing everything until we get the shift/3
key to produce what we want, because he already has his shift/3 key set
up to produce a pound sign, and he doesn't want that changed. It is the
# key itself (equivilent to "\" key on DEC keyboards) that is playing
up. At the moment my main aim, if for him to at least be able to
produce a # in WPS-PLUS using the compose key, after that I can look
into changing the keyboard mapping. At the moment compose ++ is
producing a pound sign and compose L- is also producing a pound sign on
the pc, this is no longer a problem on the VT.
I will continue to get him to compare his VT settings to his emulator
settings, but I am a little unsure of exactly what I am looking for. I
know you said SHO TERM will help, but how do the logicals help??
I will get the results in here as soon as I hear from the customer.
Thank you again!
Emma.
|
3654.15 | | BRUMMY::MARTIN::BELL | Martin Bell, NETCC, Birmingham UK | Wed Dec 22 1993 14:08 | 44 |
| Emma,
When A1WPSPLUS_LOGIN runs, it (as you confirm) performs a SET TERM/INQ
which allows VMS to determine what kind of device you are using. This
device can be seen on the first line of the SHOW TERM output.
Depending on what device VMS thinks that it is talking to, the command
file then conditionally jumps to code to setup things properly for
WPS-PLUS (eg downline loading the Technical Character Set). Have a
look at your v2.4 A1WPSPLUS_LOGIN, and see how it processes the
VT300 terminal - in v3.0 it defines a logical name which (i guess)
is used by WPS-PLUS so that it knows how the terminal behaves. If for
some reason it gets the wrong terminal type, then that _may_ cause
the problem that you are experiencing.
If both the standard VT300 and the terminal emulator have the same
device type in SHOW TERM, and they both have the same KOA$TERMINAL...
logicals, then it more likely to be the emulator that is causing
the wrong character to be echoed - because WPS-PLUS thinks that the
emulator is exactly the same as the VT300!
Bear in mind that SHIFT 3 could behave in one of 3 ways, depending
on various terminal setups.
1) Send an ASCII 35 character code to the VAX, and echo this as a HASH
2) Send an ASCII 35 character code to the VAX, and echo this as a POUND
3) Send an ASCII 163 character code to the VAX, and echo this as a POUND
where ASCII 35 is _typically_ a HASH in US terminal setup and a POUND in
UK setup, but that depends on the exact model of VT terminal. ASCII 163
should always be a POUND. unless the terminal or emulator doesn't
understand eightbit characters properly.
Thus it sounds like your terminal and emulator is doing 2).
If you want to force the correct ASCII code to see how it is displayed,
use EDT and enter GOLD 35 GOLD KEYPAD-3 or GOLD 163 GOLD KEYPAD-3, and
see how they appear on the screen for a given terminal setup.
Hope this is helping!
Cheers,
mb
|