Unit 1 Code Description
This document (and the corresponding documents in future units) will describe any ACT-R commands used in the unit¡¯s models which are not described in the main unit text, the code that implements the task/experiment for the models in that unit, how ACT-R is interfaced with the task/experiment, and describe additional commands that one can use interactively when working with ACT-R (most of the information available through the tools of the Environment can also be obtained directly from commands).
For this unit there are no tasks for the models and all of them run simply by using the ACT-R run command. They start with a predefined goal that is specifically placed into the goal buffer, they have a given set of declarative memories, and the result is a modification of the goal chunk representing the processing that took place. Later units will include tasks which have much more involved perception and action, but for this unit all of the models have the same basic structure as shown here in a solution to the multi-column addition model one was to write for unit1:
(clear-all)
(define-model tutor-model
(sgp :esc t :lf .05 :trace-detail medium)
;; Add Chunk-types here
(chunk-type addition-fact addend1 addend2 sum)
(chunk-type add-pair one1 ten1 one2 ten2 ten-ans one-ans carry)
;; Add Chunks here
(add-dm
(fact17 isa addition-fact addend1 1 addend2 7 sum 8)
(fact34 isa addition-fact addend1 3 addend2 4 sum 7)
(fact67 isa addition-fact addend1 6 addend2 7 sum 13)
(fact103 isa addition-fact addend1 10 addend2 3 sum 13)
(goal isa add-pair ten1 3 one1 6 ten2 4 one2 7))
;; Add productions here
(p start-pair
=goal>
ISA add-pair
one1 =num1
one2 =num2
one-ans nil
==> =goal>
one-ans busy
+retrieval>
ISA addition-fact
addend1 =num1
addend2 =num2)
(p add-ones
=goal>
ISA add-pair
one-ans busy
one1 =num1
one2 =num2
=retrieval>
ISA addition-fact
addend1 =num1
addend2 =num2
sum =sum
==> =goal>
one-ans =sum
carry busy
+retrieval>
ISA addition-fact
addend1 10
sum =sum)
(p process-carry
=goal>
ISA add-pair
ten1 =num1
ten2 =num2
carry busy
one-ans =ones
=retrieval>
ISA addition-fact
addend1 10
sum =ones
addend2 =remainder
==> =goal>
carry 1
ten-ans busy
one-ans =remainder
+retrieval>
ISA addition-fact
addend1 =num1
addend2 =num2)
(p no-carry
=goal>
ISA add-pair
ten1 =num1
ten2 =num2
one-ans =ones
carry busy
?retrieval>
buffer failure
==>
=goal>
carry nil
ten-ans busy
+retrieval>
ISA addition-fact
addend1 =num1
addend2 =num2)
(p add-tens-done
=goal>
ISA add-pair
ten-ans busy
carry nil
=retrieval>
ISA addition-fact
sum =sum
==> =goal>
ten-ans =sum)
(p add-tens-carry
=goal>
ISA add-pair
carry 1
ten-ans busy
=retrieval>
ISA addition-fact
sum =sum
==> =goal>
carry nil
+retrieval>
ISA addition-fact
addend1 1
addend2 =sum)
(goal-focus goal)
)
It starts with a call to a function called clear-all followed by a call to the command define-model. Inside the call to define-model there is a call to a command called sgp, then all of the components of the model are defined (chunk-types, chunks, and productions) and finally the starting goal chunk gets set using the goal-focus command.
Most of these commands were described in the main unit text, which also described some tools in the ACT-R Environment for inspecting and debugging models. Here we will provide some
additional details for some of those commands, describe the sgp command, and show how one can also perform some of the actions that were done using the Environment tools from the prompt in the ACT-R window. In later units, we will also show how those commands could be accessed remotely using the included Python module as an example.
Additional ACT-R command details
Clear-all
The clear-all command is used to set the ACT-R software to its initial state. It removes all of the models that are currently defined, returns the clock to time 0, and removes any events which are on the event queue. If a model (or set of models) is contained within a single file, then one probably should call clear-all in that file to make sure ACT-R is initialized before defining the model components.
An additional side effect of the clear-all command is that it records the name of the file that contains it when it is loaded. That is how the reload command knows which file to ¡°reload¡±.
Define-model
The define-model command takes one required parameter which is the name of a new model to create and then an arbitrary number of other parameters which are the commands which create the initial conditions for the model. The name should be a symbol and must be unique with respect to other models that are currently defined. When a reset happens, all of the commands specified inside of the define-model are reevaluated for that model in the order they were specified. Each call to define-model creates a new model which is independent of the other models, but all of the models will run in parallel when the run command is called.
SGP
The sgp command is used to set or show the model specific parameters (it stands for set/show general parameters). The parameters for a model control many different things. Some are used in equations that control the performance of the model¡¯s cognitive modules, others are there to help the modeler with debugging by changing the outputting of information or the seed of the pseudorandom number generator, and others are available to provide ways that the modeler can extend or modify internal ACT-R mechanisms. The details of all of the parameters can be found in the reference manual.
When using sgp to set parameters the syntax is to specify a parameter and then the new value you want to assign to that parameter. Any number of parameters and values may be specified in a single call to sgp. All of the parameters in ACT-R begin with a ¡°:¡± (in Lisp syntax they are keywords). All of the unit1 models have a call to sgp similar to this:
(sgp :esc t :lf .05 :trace-detail medium)
That is setting three parameters: :esc, :lf, and :trace-detail. The first two together are specifying that retrieval requests will always take 50ms to complete, but further details on those are beyond the scope of this unit and will be discussed fully in later units. The third parameter being set, :trace-detail, controls how much information is shown in the trace when a model runs. The default value is medium, and that is also how it is being set in the example above. The other values that it can have are high and low. When it is set to high, effectively every action the model does shows in the trace. Here is the trace of the two-digit addition model with :trace-detail set to high:
0.000 GOAL
0.000 PROCEDURAL
0.000 PROCEDURAL
0.000 PROCEDURAL
0.050 PROCEDURAL
0.050 PROCEDURAL
0.050 PROCEDURAL
0.050 PROCEDURAL
0.050 DECLARATIVE
0.050 PROCEDURAL
0.100 DECLARATIVE
0.100 DECLARATIVE
0.100 PROCEDURAL
0.100 PROCEDURAL
0.100 PROCEDURAL
0.100 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 DECLARATIVE
0.150 PROCEDURAL
0.200 DECLARATIVE
0.200 DECLARATIVE
0.200 PROCEDURAL
0.200 PROCEDURAL
0.200 PROCEDURAL
0.200 PROCEDURAL
0.250 PROCEDURAL
0.250 PROCEDURAL
0.250 PROCEDURAL
0.250 PROCEDURAL
0.250 DECLARATIVE
0.250 PROCEDURAL
0.300 DECLARATIVE
0.300 DECLARATIVE
0.300 PROCEDURAL
0.300 PROCEDURAL
0.300 PROCEDURAL
0.300 PROCEDURAL
0.350 PROCEDURAL
SET-BUFFER-CHUNK GOAL GOAL REQUESTED NIL
CONFLICT-RESOLUTION
PRODUCTION-SELECTED START-PAIR
BUFFER-READ-ACTION GOAL
PRODUCTION-FIRED START-PAIR
MOD-BUFFER-CHUNK GOAL
MODULE-REQUEST RETRIEVAL
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT67
SET-BUFFER-CHUNK RETRIEVAL FACT67
CONFLICT-RESOLUTION
PRODUCTION-SELECTED ADD-ONES
BUFFER-READ-ACTION GOAL
BUFFER-READ-ACTION RETRIEVAL
PRODUCTION-FIRED ADD-ONES
MOD-BUFFER-CHUNK GOAL
MODULE-REQUEST RETRIEVAL
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT103
SET-BUFFER-CHUNK RETRIEVAL FACT103
CONFLICT-RESOLUTION
PRODUCTION-SELECTED PROCESS-CARRY
BUFFER-READ-ACTION GOAL
BUFFER-READ-ACTION RETRIEVAL
PRODUCTION-FIRED PROCESS-CARRY
MOD-BUFFER-CHUNK GOAL
MODULE-REQUEST RETRIEVAL
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT34
SET-BUFFER-CHUNK RETRIEVAL FACT34
CONFLICT-RESOLUTION
PRODUCTION-SELECTED ADD-TENS-CARRY
BUFFER-READ-ACTION GOAL
BUFFER-READ-ACTION RETRIEVAL
PRODUCTION-FIRED ADD-TENS-CARRY
0.350 PROCEDURAL
0.350 PROCEDURAL
0.350 PROCEDURAL
0.350 DECLARATIVE
0.350 PROCEDURAL
0.400 DECLARATIVE
0.400 DECLARATIVE
0.400 PROCEDURAL
0.400 PROCEDURAL
0.400 PROCEDURAL
0.400 PROCEDURAL
0.450 PROCEDURAL
0.450 PROCEDURAL
0.450 PROCEDURAL
0.450 PROCEDURAL
0.450 ——
MOD-BUFFER-CHUNK GOAL
MODULE-REQUEST RETRIEVAL
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT17
SET-BUFFER-CHUNK RETRIEVAL FACT17
CONFLICT-RESOLUTION
PRODUCTION-SELECTED ADD-TENS-DONE
BUFFER-READ-ACTION GOAL
BUFFER-READ-ACTION RETRIEVAL
PRODUCTION-FIRED ADD-TENS-DONE
MOD-BUFFER-CHUNK GOAL
CLEAR-BUFFER RETRIEVAL
CONFLICT-RESOLUTION
Stopped because no events left to process
That can be very useful when debugging a model but it can be a bit too much at other times. Here is the same model running with a medium level of :trace-detail (which is the default value):
0.000 GOAL
0.000 PROCEDURAL
0.050 PROCEDURAL
0.050 PROCEDURAL
0.050 DECLARATIVE
0.050 PROCEDURAL
0.100 DECLARATIVE
0.100 DECLARATIVE
0.100 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 DECLARATIVE
0.150 PROCEDURAL
0.200 DECLARATIVE
0.200 DECLARATIVE
0.200 PROCEDURAL
0.250 PROCEDURAL
0.250 PROCEDURAL
0.250 DECLARATIVE
0.250 PROCEDURAL
0.300 DECLARATIVE
0.300 DECLARATIVE
0.300 PROCEDURAL
0.350 PROCEDURAL
0.350 PROCEDURAL
0.350 DECLARATIVE
0.350 PROCEDURAL
0.400 DECLARATIVE
0.400 DECLARATIVE
SET-BUFFER-CHUNK GOAL GOAL REQUESTED NIL
CONFLICT-RESOLUTION
PRODUCTION-FIRED START-PAIR
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT67
SET-BUFFER-CHUNK RETRIEVAL FACT67
CONFLICT-RESOLUTION
PRODUCTION-FIRED ADD-ONES
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT103
SET-BUFFER-CHUNK RETRIEVAL FACT103
CONFLICT-RESOLUTION
PRODUCTION-FIRED PROCESS-CARRY
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT34
SET-BUFFER-CHUNK RETRIEVAL FACT34
CONFLICT-RESOLUTION
PRODUCTION-FIRED ADD-TENS-CARRY
CLEAR-BUFFER RETRIEVAL
START-RETRIEVAL
CONFLICT-RESOLUTION
RETRIEVED-CHUNK FACT17
SET-BUFFER-CHUNK RETRIEVAL FACT17
0.400 PROCEDURAL
0.450 PROCEDURAL
0.450 PROCEDURAL
0.450 PROCEDURAL
0.450 ——
CONFLICT-RESOLUTION
PRODUCTION-FIRED ADD-TENS-DONE
CLEAR-BUFFER RETRIEVAL
CONFLICT-RESOLUTION
Stopped because no events left to process
In that trace we no longer see the individual condition tests and only some of the actions of the productions. Now, here is the same model run with a low setting for :trace-detail:
0.000 GOAL
0.050 PROCEDURAL
0.100 DECLARATIVE
0.150 PROCEDURAL
0.200 DECLARATIVE
0.250 PROCEDURAL
0.300 DECLARATIVE
0.350 PROCEDURAL
0.400 DECLARATIVE
0.450 PROCEDURAL
0.450 ——
SET-BUFFER-CHUNK GOAL GOAL REQUESTED NIL
PRODUCTION-FIRED START-PAIR
SET-BUFFER-CHUNK RETRIEVAL FACT67
PRODUCTION-FIRED ADD-ONES
SET-BUFFER-CHUNK RETRIEVAL FACT103
PRODUCTION-FIRED PROCESS-CARRY
SET-BUFFER-CHUNK RETRIEVAL FACT34
PRODUCTION-FIRED ADD-TENS-CARRY
SET-BUFFER-CHUNK RETRIEVAL FACT17
PRODUCTION-FIRED ADD-TENS-DONE
Stopped because no events left to process
At low we only see production firings and buffer settings. The setting of :trace-detail does not change how the model actually runs. It only affects how the trace is displayed to the modeler and which events are available to ¡°step¡± to using the Stepper tool in the ACT-R Environment.
If one wants to completely turn off the model trace, there is another parameter which can be set to do so, and that will be described in a future unit.
To get a parameter¡¯s current value using sgp only the name of the parameter (or parameters) should be specified. When all of the values passed to sgp are the names of parameters it will print out the details of those parameters and return a list of their current values. Typically that is not necessary when defining a model (with one exception which can be very helpful for debugging a model and that will be described in a later unit), but it can be used at the ACT-R prompt to inspect the model settings (there is also an inspector in the Environment for doing so). Here is an example which is checking the values of the :trace-detail and :lf parameters:
? (sgp :trace-detail :lf)
:TRACE-DETAIL MEDIUM (default MEDIUM) : Determines which events show in the trace
:LF 0.05 (default 1.0) : Latency Factor
(MEDIUM 0.05)
If no parameters are provided to sgp then it will print out all of the parameters and their details.
Interacting with a model
In this unit we used the ACT-R Environment tools to load, run, reset, and inspect the model components. Most of the GUI tools in the Environment are using commands which are also available to the modeler for interacting with the model from the command prompt or in experiment or task code. Here we will describe some of the commands that correspond to the Environment tools used for this unit with respect to using them from the ACT-R command prompt (as was shown above for sgp). In future units we will discuss how those commands can also be used in experiments or tasks written in Lisp or Python because those are the languages for which an interface to the commands is included with the software (other languages could also be supported but the details needed to do so are beyond the scope of this tutorial).
Loading a model
To load the models in this unit we used the ¡°Load ACT-R code¡± button to pick a file to load. There is a command in ACT-R which can also be used to load a model file which is called load- act-r-model. It requires one parameter which is a string specifying the pathname of the file to load. Typically, the full pathname of the file must be specified, but it does accept a simplified specification for files located in the directory containing the ACT-R files which is based on the Lisp logical pathname convention. The format for those relative pathnames is to start with ACT- R: and then follow that with the file name or subdirectories separated by semicolons and then the file name. Here is an example of loading the count.lisp model from the default ACT-R distribution:
? (load-act-r-model “ACT-R:tutorial;unit1;count.lisp”)
T
The T printed after the call is the return value from the command, and T is the Lisp symbol for true, which for this command means it was successful in loading the file. If the file is not found or there is an error in loading it then it will print a warning indicating the issue and the return value would be nil instead.
Resetting and Reloading
Instead of pressing the buttons to reset or reload a model one can call the corresponding commands, which are named reset and reload. They require no parameters and will return T if successful:
? (reset)
T
? (reload)
T
Running the model
There are actually multiple commands which can be used to run a model based on how it should determine when to stop running. The one that corresponds to the button on the Control Panel is called run. It requires one parameter which is the maximum number of seconds to run the model, but it will stop earlier if the model has no more actions to perform. In later units we will introduce more of the running commands and show them being used in creating tasks that automatically run the model as needed. Here is an example of using the run command at the prompt after loading the semantic model:
? (run 1)
0.000 GOAL
0.000 PROCEDURAL
0.050 PROCEDURAL
0.050 PROCEDURAL
0.050 DECLARATIVE
0.050 PROCEDURAL
0.100 DECLARATIVE
0.100 DECLARATIVE
0.100 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 ——
0.15 24 NIL
After the trace there are multiple return values shown from the run command which indicate how long it ran, how many events occurred during the run, and whether it ended unexpectedly or not (a value of nil means a successful ending because it was not unexpected due to a break action or other problem).
Inspecting model components
During the unit we inspected the contents of the buffers, the model¡¯s declarative memory, and checked why productions did not match. All of those can also be done using commands.
Buffer Contents
You can use the command named buffer-chunk to find the names of the chunks in the buffers and inspect their contents. Calling it without any parameters will show all of the buffers and the chunks they contain returning a list of the buffer name and buffer chunk lists:
? (buffer-chunk)
GOAL: G1-0
TEMPORAL: NIL
AURAL: NIL
AURAL-LOCATION: NIL
MANUAL: NIL
VOCAL: NIL
RETRIEVAL: NIL
PRODUCTION: NIL
SET-BUFFER-CHUNK GOAL G1 NIL
CONFLICT-RESOLUTION
PRODUCTION-FIRED INITIAL-RETRIEVE
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVED-CHUNK P14
SET-BUFFER-CHUNK RETRIEVAL P14
CONFLICT-RESOLUTION
PRODUCTION-FIRED DIRECT-VERIFY
CLEAR-BUFFER RETRIEVAL
CONFLICT-RESOLUTION
Stopped because no events left to process
IMAGINAL: NIL
VISUAL: NIL
VISUAL-LOCATION: NIL
IMAGINAL-ACTION: NIL
((GOAL G1-0) (TEMPORAL) (AURAL) (AURAL-LOCATION) (MANUAL) (VOCAL) (RETRIEVAL)
(PRODUCTION) (IMAGINAL) (VISUAL) (VISUAL-LOCATION) (IMAGINAL-ACTION))
If you call it with the name of a buffer (or multiple buffers), then it will print out the chunks in the named buffers and return the list of the names of those chunks:
? (buffer-chunk goal)
GOAL: G1-0
G1-0
OBJECT CANARY
CATEGORY BIRD
JUDGMENT YES
(G1-0)
Declarative Memory
You can also inspect declarative memory from the command prompt. The command dm will print out all of the chunks in the model¡¯s declarative memory and return a list with the names of those chunks. You can also specify the names of chunks as parameters to the dm command and only those chunks will be printed. Here are some examples using the count model:
? (dm)
FIRST-GOAL
START 2 END 4
F
FIRST 5
SECOND 6
E
FIRST 4
SECOND 5
D
FIRST 3
SECOND 4
C
FIRST 2
SECOND 3
B
FIRST 1
SECOND 2
(FIRST-GOAL F E D C B)
? (dm b e f) B
FIRST 1
SECOND 2
E
FIRST 4
SECOND 5
F
FIRST 5
SECOND 6
(B E F) ?
It is also possible to search declarative memory using the command sdm. Its parameters are a chunk specification just as one would specify in a retrieval request in a production, but without the +retrieval> indicator at the beginning. It prints out only those chunks from the model¡¯s declarative memory which match that specification and returns the list of their names. Here are some examples using the semantic model:
? (sdm object shark)
P1
OBJECT SHARK
VALUE TRUE
ATTRIBUTE DANGEROUS
P2
OBJECT SHARK
VALUE SWIMMING
ATTRIBUTE LOCOMOTION
P3
OBJECT SHARK
VALUE FISH
ATTRIBUTE CATEGORY
(P1 P2 P3)
? (sdm – attribute category – attribute nil)
P1
OBJECT SHARK
VALUE TRUE
ATTRIBUTE DANGEROUS
P2
OBJECT SHARK
VALUE SWIMMING
ATTRIBUTE LOCOMOTION
P4
OBJECT SALMON
VALUE TRUE
ATTRIBUTE EDIBLE
P5
OBJECT SALMON
VALUE SWIMMING
ATTRIBUTE LOCOMOTION
P7
OBJECT FISH
VALUE GILLS
ATTRIBUTE BREATHE
P8
OBJECT FISH
VALUE SWIMMING
ATTRIBUTE LOCOMOTION
P10
OBJECT ANIMAL
VALUE TRUE
ATTRIBUTE MOVES
P11
OBJECT ANIMAL
VALUE TRUE
ATTRIBUTE SKIN
P12
OBJECT CANARY
VALUE YELLOW
ATTRIBUTE COLOR
P13
OBJECT CANARY
VALUE TRUE
ATTRIBUTE SINGS
P15
OBJECT OSTRICH
VALUE FALSE
ATTRIBUTE FLIES
P16
OBJECT OSTRICH
VALUE TALL
ATTRIBUTE HEIGHT
P18
OBJECT BIRD
VALUE TRUE
ATTRIBUTE WINGS
P19
OBJECT BIRD
VALUE FLYING
ATTRIBUTE LOCOMOTION
(P1 P2 P4 P5 P7 P8 P10 P11 P12 P13 P15 P16 P18 P19)
One thing to note about the second example is that it specifies both that the attribute slot is not category and also not nil (which means the chunk must have some value for the slot). Here is what it returns if it were to only specify that the attribute slot is not category:
? (sdm – attribute category)
…
(G1 G2 G3 P1 P2 P4 P5 P7 P8 P10 P11 P12 P13 P15 P16 P18 P19 SHARK DANGEROUS
LOCOMOTION SWIMMING FISH SALMON EDIBLE BREATHE GILLS ANIMAL MOVES SKIN CANARY
COLOR SINGS BIRD OSTRICH FLIES HEIGHT TALL WINGS FLYING TRUE FALSE)
In this case it also finds all of the chunks which do not have an attribute slot because chunks without the slot fail the ¡°attribute category¡± test and thus the negation of that is then true. That is something to be careful about when using the negation modifier in specifying retrieval requests in your productions as well.
Testing why not? for productions
The a command to test whether a production matches the current state is called whynot. If you pass it no parameters it will print out each chunk in the model¡¯s procedural memory along with either an indication that it matches or a reason why it does not match. You can also provide it with specific production names to test. It returns a list of the names of the productions which do match the current state. Here is an example using productions from the semantic model after it has been reset:
? (whynot direct-verify fail)
Production DIRECT-VERIFY does NOT match.
(P DIRECT-VERIFY
=GOAL>
OBJECT =OBJ
CATEGORY =CAT
JUDGMENT PENDING
=RETRIEVAL>
OBJECT =OBJ
ATTRIBUTE CATEGORY
VALUE =CAT
==> =GOAL>
JUDGMENT YES
)
It fails because:
The GOAL buffer is empty.
Production FAIL does NOT match.
(P FAIL
=GOAL>
OBJECT =OBJ1
CATEGORY =CAT
JUDGMENT PENDING
?RETRIEVAL>
BUFFER FAILURE
==>
=GOAL>
JUDGMENT NO
)
It fails because:
The GOAL buffer is empty.
NIL
Other model commands
Any of the commands that are specified in the model definition can also be called from the ACT-R command prompt. The add-dm and p commands are not usually called from the prompt since you want that knowledge to be included in the model when it starts running, but a command like goal- focus can occasionally be useful if you want to change the goal chunk for a model and run it again. That could have been convenient for the semantic model because instead of changing the file and reloading it to switch the goals one could have run the different goals sequentially by just calling goal-focus to change the chunk in the goal buffer and then run it again:
? (run 1)
0.000 GOAL
0.000 PROCEDURAL
0.050 PROCEDURAL
0.050 PROCEDURAL
0.050 DECLARATIVE
0.050 PROCEDURAL
0.100 DECLARATIVE
0.100 DECLARATIVE
0.100 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 PROCEDURAL
0.150 ——
0.15
24
NIL
? (goal-focus g2)
G2
? (run 1)
0.150 GOAL
0.150 PROCEDURAL
0.200 PROCEDURAL
0.200 PROCEDURAL
0.200 DECLARATIVE
0.200 PROCEDURAL
0.250 DECLARATIVE
0.250 DECLARATIVE
0.250 PROCEDURAL
0.300 PROCEDURAL
0.300 PROCEDURAL
0.300 DECLARATIVE
0.300 PROCEDURAL
0.350 DECLARATIVE
0.350 DECLARATIVE
0.350 PROCEDURAL
SET-BUFFER-CHUNK GOAL G1 NIL
CONFLICT-RESOLUTION
PRODUCTION-FIRED INITIAL-RETRIEVE
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVED-CHUNK P14
SET-BUFFER-CHUNK RETRIEVAL P14
CONFLICT-RESOLUTION
PRODUCTION-FIRED DIRECT-VERIFY
CLEAR-BUFFER RETRIEVAL
CONFLICT-RESOLUTION
Stopped because no events left to process
SET-BUFFER-CHUNK GOAL G2 NIL
CONFLICT-RESOLUTION
PRODUCTION-FIRED INITIAL-RETRIEVE
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVED-CHUNK P14
SET-BUFFER-CHUNK RETRIEVAL P14
CONFLICT-RESOLUTION
PRODUCTION-FIRED CHAIN-CATEGORY
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVED-CHUNK P20
SET-BUFFER-CHUNK RETRIEVAL P20
CONFLICT-RESOLUTION
0.400 PROCEDURAL
0.400 PROCEDURAL
0.400 PROCEDURAL
0.400 ——
0.25
36
NIL
? (goal-focus g3)
G3
? (run 1)
0.400 GOAL
PRODUCTION-FIRED DIRECT-VERIFY
CLEAR-BUFFER RETRIEVAL
CONFLICT-RESOLUTION
Stopped because no events left to process
SET-BUFFER-CHUNK GOAL G3 NIL
CONFLICT-RESOLUTION
PRODUCTION-FIRED INITIAL-RETRIEVE
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVED-CHUNK P14
SET-BUFFER-CHUNK RETRIEVAL P14
CONFLICT-RESOLUTION
PRODUCTION-FIRED CHAIN-CATEGORY
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVED-CHUNK P20
SET-BUFFER-CHUNK RETRIEVAL P20
CONFLICT-RESOLUTION
PRODUCTION-FIRED CHAIN-CATEGORY
CLEAR-BUFFER RETRIEVAL
start-retrieval
CONFLICT-RESOLUTION
RETRIEVAL-FAILURE
CONFLICT-RESOLUTION
PRODUCTION-FIRED FAIL
CLEAR-BUFFER RETRIEVAL
CONFLICT-RESOLUTION
Stopped because no events left to process
0.35 48 NIL
0.400 PROCEDURAL
0.450 PROCEDURAL
0.450 PROCEDURAL
0.450 DECLARATIVE
0.450 PROCEDURAL
0.500 DECLARATIVE
0.500 DECLARATIVE
0.500 PROCEDURAL
0.550 PROCEDURAL
0.550 PROCEDURAL
0.550 DECLARATIVE
0.550 PROCEDURAL
0.600 DECLARATIVE
0.600 DECLARATIVE
0.600 PROCEDURAL
0.650 PROCEDURAL
0.650 PROCEDURAL
0.650 DECLARATIVE
0.650 PROCEDURAL
0.700 DECLARATIVE
0.700 PROCEDURAL
0.750 PROCEDURAL
0.750 PROCEDURAL
0.750 PROCEDURAL
0.750 ——