LAYERS & MEMORY
9 messages in this thread
LAYERS & MEMORY — declared parent_msg_num=(none), resolved parent_id points to #(none)
I work in the facility department here at Intel and we use the AEC packages.
I have been given information that you all there are going to develop a
Planning AEC package. I would like to tell you of a problem we are having
with the present packages. It has to do with the layering of the packages.
Most of our engineers take existing entities and differentiates them with new
or removed entities. With the restriction of the LISP files this becomes
more work than starting from scratch. As a solution I believe it is possible
to use a different nomenclature for the layers. This would conform to
established standards and that system was pointed out in the CADalyst article
"Escape from the Labyrinth" by Robert Hartman of the Hawaii A.I.A. (v5#2).
In the article the layering follows the CSI standard for layering. From a
programming standpoint numerics ar easier to manipulate and combined with
user modified LISP files that are no more than associated lists, the AEC LISP
files could still provide what they at present do so well and allow user
modification, provide a standard that will transport between packages, and by
using layering labeling in extract files provide the standard ordering coding
be built in. The second question I have has to do with the memory allocasion
that AutoCAD does when one SHELL's out. My system is a 386 based one with an
Intel above board giving me 3Meg, we have another identicle system that has a
MEM20 board and QEMM driver (mine has no driver as it is all extended). On
my system I can get 400K SHELLed out while on the other I'm lucky to get 170K
and it has 4Meg of extended. I later configured it to have 2Meg of extended
and managed to squeze 175K SHELLed. How does AutoCAD deal with this
SHELLing? In order to make these computers real Workstations we need more
SHELL memory. Any suggestions? Answers? Godsends? Thanks,
— Kevin
#22082-#LAYERS & MEMORY — declared parent_msg_num=22082, resolved parent_id points to #22082
I think the first question was more if a wish. I will pass that part along to
the Product Managers. As for the second question, you were on the right
track, one thing that can effect the amount of shell you can perform is the
available system memory and what it is doing. But more likely it is some TSR
that may be effecting what AutoCAD sees as available memory.
You say both of these machines are identical? System for system? Are you
using an ADI on either/both?
On test I conducted with a Compaq 386/20, EGA, MicroSoft Mouse, I was able to
get shells of 300k even under the lowest of memory conditions. If you are not
sure what is getting loaded into your system memory, stop by the IBMHW forum.
There are plenty of memory mapping utilities that may help in finding the
difference between the machines.
Jamie
#22130-LAYERS & MEMORY — declared parent_msg_num=22130, resolved parent_id points to #22130
Jamie,
I've been doing some system re-arranging to find the REAL bugs with my
Verticom Twin Focus, and it appears that the SHELL command is affected by the
ACADFREERAM setting…..ie more freeram….more shell space available……
Y/N>: ?
#22140-#LAYERS & MEMORY — declared parent_msg_num=22140, resolved parent_id points to #22140
That is an interesting observation, but I am not sure that it is true. The
things that are known to effect the shell space are TSR programs (ADIs,
Sidekick, CEMM), disk space, graphic cards (like EGA) and available system
memory.
What kind of results are you getting and what system it this?
Jamie
#22201-#LAYERS & MEMORY — declared parent_msg_num=22201, resolved parent_id points to #22201
Jamie,
In speaking to the Verticom tech I was told to use ACADFREERAM at 12,
and to make my LISPSTACK >20000 in an attempt to solve some problems I have
been experiencing with the Twin Focus setup……..so I noticed that my
shells to PCTOOLS did'nt have enough memory released to operate properly (I
had it set at 200k in the PGP file…. a real minimum for the latest
release)…….(now at 210k, and appears to be working)……… and some
other shells did'nt have enough space to operate either…..
286 AT compat system w/Intel Above Board 4mb (device = emm.sys m5 i5 d),
Buffers=20, Files=20, Shell /e:360, MSDOS 3.1, Twin Focus ADI for Verticom
H-16………
There does seem to be a need to drop FREERAM down, but I could only get to
about 16 before I was getting out of space errors……. and sliding the
STACK up to 25000 so far seems to have helped also, in solving some apparent
Verticom problems…….
I'm not sure what the Verticom drivers are doing to affect these areas, but
there is something happening…..
#22263-LAYERS & MEMORY — declared parent_msg_num=22263, resolved parent_id points to #22263
I really don't see where adjusting LISPSTACK could effect the system shell.
I am not surprised that an ADI is effecting the limit of your SHELL, however
we have the Twin Focus here on a 386, 2meg Vdisk get a 512k shell.
It may be your EMM.SYS effecting your environment.
Also –
What is your CHKDSK showing for bytes free?
Jamie
#22130-#LAYERS & MEMORY — declared parent_msg_num=22130, resolved parent_id points to #22130
Yup they are identicel system for system. The only difference being one uses
an AboveBoard switched to backfill the 512 to 640K while the other uses
QEMM.SYS to backfill using the MEM20 board extended.
No TSR's are allowed, and besides the only trouble I find with those is on
initial ACAD bootup, not having enough memory for extended I/O or AutoLISP.
This problem occures when the SHELL memory parameters are increased. Our
Compac's can only get 200K SHELLed, what driver do you use for the
extended/expanded memory? This might help.
As for my first wish well… "When you wish upon a star…"
—Kevin
#22142-LAYERS & MEMORY — declared parent_msg_num=22142, resolved parent_id points to #22142
Kevin —
Isn't QEMM.SYS a TSR driver, and isn't that the machine that frees
up less memory during a SHELL? In some cases, I think that both the size
and the location of the TSR in conventional memory affect the amount of
memory available for an AutoCAD SHELL.
— Brad
..
#22142-LAYERS & MEMORY — declared parent_msg_num=22142, resolved parent_id points to #22142
Kevin, are you using the QEMM.SYS to build the 640K? It is very possible
that method is causing AutoCAD to think there isn't enough room for the size
shell you want.
Is there a way to set up the memory by switch? I would try that.
Let me know your results.
Jamie