Return-Path: <john.harrison-request@uk.ac.cam.cl>
Delivery-Date: 
Received: from ted.cs.uidaho.edu by swan.cl.cam.ac.uk with SMTP (PP-6.2);
          Sat, 26 Dec 1992 04:42:36 +0000
Received: by ted.cs.uidaho.edu (16.6/1.34) id AA11553;
          Fri, 25 Dec 92 20:32:43 -0800
Sender: info-hol-request@edu.uidaho.cs.ted
Errors-To: info-hol-request@edu.uidaho.cs.ted
Precedence: bulk
Received: from etlpost.etl.go.jp by ted.cs.uidaho.edu (16.6/1.34) id AA11548;
          Fri, 25 Dec 92 20:32:35 -0800
Received: from etlpom.etl.go.jp by etlpost.etl.go.jp (5.67+1.6W/2.7W) 
          id AA21417; Sat, 26 Dec 92 13:32:01 JST
Received: by etlpom.etl.go.jp (4.1/6.4J.6-ETLpom.MASTER) id AA25629;
          Sat, 26 Dec 92 13:32:00 JST
Received: by etlhit.etl.go.jp (4.1/6.4J.6-ETL.SLAVE) id AA18346;
          Sat, 26 Dec 92 13:26:45 JST
Date: Sat, 26 Dec 92 13:26:45 JST
From: weigert@jp.go.etl (Thomas Weigert)
Message-Id: <9212260426.AA18346@etlhit.etl.go.jp>
To: chou@edu.ucla.cs, info-hol@edu.uidaho.cs.ted (INFO-HOL mailing list)
Subject: Re: AKCL questions
References: <9212260252.AA16759@maui.cs.ucla.edu>


On Fri, 25 Dec 92 18:52:16 PST, chou@cs.ucla.edu said:

chou> #lisp `(room)` ;;
 450/450   55.7% 13 CONS BIGNUM RATIO LONG-FLOAT COMPLEX STRUCTURE NIL
chou> .... stuff deleted ....

chou> My first question is: What do these numbers mean?

AKCL has different pools from which it takes its memory. The command "room"
tells you how those pools are allocated. For example, the first line tells
you that the space from which objects of the indicated type (cons, bignum,
etc.) are taken has 450 units available, and all is taken up. There have
been 13 GC's so far for this pool. Currently 55.7% of those 450 are free. 

chou> Then I execute the Lisp code suggested in the contrib `AKCL-mods`:
chou> #lisp `(progn ()
chou> #        (allocate 'cons 900)
chou> #        (allocate 'string 100)
chou> #        (system:allocate-relocatable-pages 100)
chou> #        (system:set-hole-size 2048))` ;;
chou> () : void

chou> My second question is: What does the code in `AKCL-mods` do?
chou> What is changed by it and why can it increase the speed of HOL?

This command gives you more space. For example, the space for cons will be
twice as large as before, as you can see, when you reevaluate "room", as
you did.

chou> #lisp `(room)` ;;
chou>  450/900   57.1% 13 CONS BIGNUM RATIO LONG-FLOAT COMPLEX STRUCTURE NIL

It gives you more space, therefore (hopefully) less GC, therefore faster
execution.

chou> My third question is: Since the problem I had was:

chou>     Error: The storage for STREAM is exhausted.
chou>            Currently, 150 pages are allocated.
chou>            Use ALLOCATE to expand the space.
chou> can I fix it by doing:
chou> #lisp `(allocate 'symbol 200)` ;;

Yes (unless the problem is something different, like the code not releasing
used stream cells). Or use a higher number, if you still need more.

By the way, if you look in the on-line manual that comes with your AKCL
system, you will find all this information and more. So please, RTFM.
Further, if you have a AKCL specific question, there is a mailing list
(kcl@cli.com) also.

Cheers, Thomas.

Note: The address below is valid until Jan 15, 1993.

+--------------------------------+---------------------------------+
| Thomas Weigert                 |                                 |
| Machine Inference Section      |                                 |
| Electrotechnical Laboratory    |                                 |
| Umezono 1-1-4, Tukuba-shi      | weigert@{mcs.anl.gov,etl.go.jp} |
| Ibaraki 305, Japan             | +81-298-58-5918 (phone+fax)     |
+--------------------------------+---------------------------------+
