T.R | Title | User | Personal Name | Date | Lines |
---|
443.1 | hmmm | VAXUUM::KOHLBRENNER | | Fri May 29 1987 13:46 | 2 |
| Imagine the OTHER possibilities, if it is not the LN03 buffer size!
|
443.2 | We're Russian to see what's the matter | VAXUUM::DEVRIES | Those are features, not bugs | Fri May 29 1987 14:57 | 24 |
| >Does this mean that the LN03's buffer is too small and I need an upgrade ?
There is no "upgrade" for the LN03 font memory (except the plug-in
RAM cartridges that have been around for years). The PLUS board
upgrade buys you additional space for graphics storage (essentially)
and some other features, but not more text or font capacity.
>Anyone ever get an LN03 finished result with a mixture of Russian
>and English?
I assume that, in the context of this note, "English" means "the
right characters" and "Russian" means "the wrong characters".
Yes, it is possible for the document to carry more font data than
the LN03 (of any flavor) can handle, and this could cause you to
get the wrong characters for certain fonts. However, the BL08 version
should "never" do this. If you were using BL08, please contact
me at VAXUUM:: or CLOSET::DEVRIES. I'd like to look at your files.
(And if you weren't using BL08, contact me anyway. It would be
nice to prove that this file *works* in V1.0.)
Thanks,
Mark
|
443.3 | | CUPOLA::HAKKARAINEN | Albatross! | Fri May 29 1987 15:28 | 11 |
| We've encountered situations which caused garbage (aka Cyrillic)
on fully RAMmed LN03s. The cases I've seen have been when a) there
has not been a reset sequence between print jobs and b) whene there
has.
> . . . However, the BL08 version
> should "never" do this.
Does this mean that a LN03 reset sequence is sent with each job
that's been produced under bl8? Or something even better?
All for world peace ...
|
443.4 | Hasn't a reset always been included | BUNSUP::LITTLE | Todd Little NJCD SWS 323-4475 | Fri May 29 1987 15:37 | 10 |
| I may be mistaken, but I think the LN03 files produced by DOCUMENT
already do reset the font memory.
I also thought that if you overflowed the RAM font memory, that
you typically got output that looked as though someone had used
white-out all over the place, i.e. missing characters etc., and
not "bad" characters. This sounds more like a printer or communication
line problem.
-tl
|
443.5 | | CUPOLA::HAKKARAINEN | Albatross! | Fri May 29 1987 16:40 | 4 |
| Yep. A reset is included in ln03 output produced by earlier editions
of Document.
Oh, well, back to sprinkling garlic on the ln03.
|
443.6 | Thanks to Mark Devries | 16410::HEISERM | Celtics in '87 | Fri May 29 1987 16:50 | 7 |
| My problem has been cleared up. I currently have 0 RAM cartridges
and I was told for DOCUMENT you need at least 1 maybe 2 for the
large files. The cartridges are at 128K each while, without them,
the LN03 is at 28K.
Mike
|
443.7 | error 24 should have also occurred | ATLAST::BOUKNIGHT | Everything has an outline | Fri May 29 1987 23:44 | 4 |
| if you had no RAM cartridges, you should have also seen the 24 error
MISSING FONT.
jack
|
443.8 | amplifying a couple of points | VAXUUM::DEVRIES | Those are features, not bugs | Mon Jun 01 1987 10:55 | 37 |
| RE: When is a RAM not a RAM?
I have seen a case in which an LN03 had both RAMs in its mouth(s),
but they were not properly seated and so were effectively not there.
When you push the (T) button on the back of the printer, the test page
will tell you how much font memory the LN03 thinks it has.
__________
RE: this should "never" occur... [where "this" is "too much font data
for an LN03 with both RAM cartridges"]
The LN03 converter has always checked for a maximum number of different
characters in different sizes in different fonts, and in rare cases
that the threshold was reached, aborted the job (because the threshold
was intended to mean that you'd already run out of space, and the
program had no way to output part of a job).
As of BL08, the threshold is now considerably lower, but once the
software detects that complexity, it continues to read to the end
of the current page; at that point it writes a font load into the
output file, then writes out the contents of the document since the
start (or since the last font load), then continues with a "clean
slate" at the next page.
So the goal is that you should "never" have too much font data for
an LN03 with 2 RAMs to swallow. But the predictors are not precise
so, while encountering even this threshold should be pretty rare,
I cannot absolutely guarantee that this problem will never occur
again.
Please send me details if you get the LN03 error "24" for MISSING
FONT with BL08. (First, please press the (T) button on back of
the LN03 to see that it "sees" both RAMs.)
Thanks,
Mark
|
443.9 | DOCUMENT does this itself ? | CHANI::HEISERM | 1987 World Champion Boston Celtics | Mon Jun 01 1987 12:22 | 7 |
| Excuse me if this is ignorant but I'm assuming that DOCUMENT does
not use the LN03 font cartridges CG TIMES & CG TRIUMVIRATE for the
extra small and large font settings. There is only 2 slots so they
would have to be removed for the RAM cartridges.
Mike
|
443.10 | You're right. | VAXUUM::DEVRIES | Those are features, not bugs | Mon Jun 01 1987 13:48 | 4 |
| That is correct. DOCUMENT does not use any LN03 font ROMs -- it
downloads its own fonts.
Mark
|
443.11 | font memory NOT reset? | GLINKA::GREENE | | Thu Jun 04 1987 10:07 | 16 |
| re: .4
" I may be mistaken, but I think the LN03 files produced by DOCUMENT
already do reset the font memory."
I think not, and this has caused numerous problems for us. After
sending DOCUMENT output through the LN03 (yes, with both RAM's firmly
in place), certain OTHER software (non-DEC) output gets screwed
up. And hitting the test button to see how much memory is available,
it is NEVER full memory after DOCUMENT output gets printed.
Is there some better way for us to set up the LN03 queue or ?
Thanks,
Penelope
|
443.12 | | CUPOLA::HAKKARAINEN | with hasty reverence | Thu Jun 04 1987 10:47 | 14 |
| The Document files reset the printer before starting the print job,
not after. So, it's the responsibility of the print queue (or the
next file) to clear out the buffers.
A reset module of some sort needs to be specified in the separator.
(See below under /separate.) This module need only contain an <ESC>c,
the LN03 reset sequence. It can contain other things, if you desire,
to set up the page for other non-Document jobs.
$ show que /full sys$print
Terminal queue SYS$PRINT, on CUPOLA::$PRINTER, mounted form DEFAULT
/BASE_PRIORITY=4 /CHAR=(52) /DEFAULT=(FEED,FLAG,FORM=DEFAULT) /NOENABLE_GENERIC /LIBRARY=LN03CTL Lowercase /OWNER=[SYSTEM]
/PROTECTION=(S:E,O:D,G:R,W:W) /SEPARATE=(RESET=(LN03$RESET))
|
443.13 | the last thing it does is clear font memory | VAXUUM::DEVRIES | Those are features, not bugs | Thu Jun 04 1987 14:53 | 25 |
| The last things written into the LN03 output file are
<esc>c RIS (reset internal state), which resets everything
except fonts
<esc>P0;1;0y DECLFF (load font files), with these parameters:
Ps1 = 0 means Digital format
Ps2 = 1 means do not print summary sheet
Ps3 = 0 means delete all fonts
<esc>\ ST (string terminator) for the DECLFF sequence
The DECLFF ... ST sequence is the one recommended in the LN03
programmer's reference manual (except that they use Ps2 = 0 to get
a summary sheet at the same time).
My LN03 is broken right now, so I can't check this -- but at the
time I implemented this (several months ago), I checked several
jobs and got the desired results. This has been distributed at
least as long as BL07 (first field test version) -- longer, I think.
Are you referring to baselevel 06?
Mark
|
443.14 | 2 � worth | BUNSUP::LITTLE | Todd Little NJCD SWS 323-4475 | Fri Jun 05 1987 00:08 | 6 |
| By observation baselevel 8 includes those escape sequences mentioned in
.13 at both the beginning AND the end of the .LN03 file. Seems little
doubt about DOCUMENT clearing the fonts out and leaving the printer in
a reset state.
-tl
|