| Title: | DOCUMENT T1.0 |
| Notice: | **New notesfile (DOCUMENT.NOTE) now available (see note 897)** |
| Moderator: | CLOSET::ADLER |
| Created: | Mon Feb 09 1987 |
| Last Modified: | Thu Oct 31 1991 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 897 |
| Total number of notes: | 4397 |
Two problems -- memory capacity and printing individual pages.
1. Memory capacity
I have a manual that is over 500 pages long. When I processed the
profile in BL07, I exceeded memory after 200+ pages. When I processed
the profile in BL08, I exceeded memory after 302 pages. Here is the
end of my log file
%TEX-E-NOROOM, Exceeded memory capacity:
-TEX-I-MAXHASH, Maximum number of entries in hash table is: 4250
%TEX-I-SHOWCONTEXT, '...annumeral \tempcs \endcsname '
%TEX-I-SHOWCONTEXT, ' {#2}\relax \origmark \expan...'
%TEX-I-SHOWCONTEXT, '...revskip \fi \MarkOutsidefig '
%TEX-I-SHOWCONTEXT, ' \ifdim \pagegoal <\maxdimen...'
%TEX-I-LINE, Error occurred on or around line number: 495
%TEX-I-SHOWCONTEXT, '\endfigure'
%TEX-I-SHOWCONTEXT, ' '
%TEX-I-FILENAME, '07_UVMG'
-TEX-I-ONPAGE, on page [7-16]
%TEX-I-PAGESOUT, 302 pages written.
-TEX-I-OUTFILENAME, 'user3:[ilsley.uvmg.new]UVMG_PRO.DVI_LN03'
%DOC-F-ERROR_FORMATTER, Errors found by the Text Formatter.
$ exit ($status + (0 * f$verify(verify_context))) ! SYS$SCRATCH:DOC$UVMG_PRO.TMP - End
ILSLEY job terminated at 28-MAY-1987 13:21:45.31
Accounting information:
Buffered I/O count: 726 Peak working set size: 3000
Direct I/O count: 2298 Peak page file size: 10250
Page faults: 6505 Mounted volumes: 0
Charged CPU time: 0 01:17:25.40 Elapsed time: 0 01:31:28.43
2. Printing individual pages
I've had different results printing individual pages from a .DVI_LN03 file
in BL08. Sometimes I get 0 pages, although I know the specifed page numbers
exist. Other times the specified pages print correctly, but the unspecified
pages print as well -- as garbage. I'm using the following command format.
DOCUMENT/NOTAG/NOTEXT file.dvi_ln03 doctype ln03 /device=(s:n,e:n)
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 436.1 | Need to see the files | CLOSET::ANKLAM | Thu May 28 1987 17:31 | 9 | |
Looks like we're up to another limit. Would it be possible to copy
all the files ? Since we've never seen a file that exceeds the hash
table size, I'm not sure I understand what's causing it, to decide
how to fix it. Also, Mark will be interested in the .DVI_LN03 file
to look at what's going on the with page numbers.
patti
| |||||
| 436.2 | S:, E:, N: are not quite there yet | VAXUUM::DEVRIES | Those are features, not bugs | Fri May 29 1987 14:12 | 12 |
Re: problems with starting-page, ending-page, number-of-pages
The condition of the support for these /DEVICE_CONVERTER keywords
currently differs from one destination to another (in BL08). I
am currently concentrating on these features, and will ensure that
they are consistent and correct for V1.0.
(In the meantime: if you use starting-page with the LN03 destination,
you may well get *parts* of the pages which preceed your desired
starting page, as referred to in the base note.)
Mark
| |||||
| 436.3 | 166 <figure>s | CLOSET::ANKLAM | Mon Jun 01 1987 17:57 | 4 | |
re: exceeding memory. we did hit another limit, and can up TeX a
bit more...the file in question has 166 figures, so TeX ran out
of room to store temporary strings to handle them.
| |||||
| 436.4 | it ran in bl6 | OILCAN::KEEFE | Wed Jun 03 1987 13:58 | 9 | |
We did run this manual successfully using bl 6, to get the table
of contents and index. So it looked like the amount of memory available
to TeX got reduced in the new version, not increased. It really
was increased though (right?), so the failure to process this manual
in bl8 seems like a side effect caused by something else, in addition
to the number of figures. ?
-Neil
| |||||
| 436.5 | different limit | CLOSET::ANKLAM | Thu Jun 04 1987 14:29 | 12 | |
TeX has a number of limits, one is memory for holding big chunks
of pages while figuring out to page it, one is for holding the
names of character string control sequences, and so on. We upped
the main memory between bl6 and bl7, upped the poolsize between
7 and 8, and just now ran into the string memory. It means that
we were close to a limit in BL7, but didn't hit up against. Any
control sequence names that get added (in internal macros) count,
so even a small number of additional names may have run us up over
the limit.
| |||||