Instead of looking at all these history things that I had prepared, we looked at the shaky

video together and that was kind of the idea is there that we look at the strips planning

formalism which is a very simple thing.

We have essentially as a description language a Boolean logic except that we have predicates

and objects but since we don't have any quantifiers we're really using Boolean logic to do this.

We have preconditions and effects and basically we only use positive literals as a description

that makes things easy.

Think of Prologue.

One of the things Prologue grew out of is actually strips.

People were playing around with strips and found out that indeed we can actually program

with this if you have a planner in the background.

In a way Prologue is one way of implementing planners.

The idea here is that we do search on states that are given as sets of positive literals

and the actions, that's the only difference, are given by a set of preconditions, an add

list and a delete list and we use the fact that we have positive and negative literals

by just saying well we have the delete and add list encoded by just having plus and minus

on them.

If there's a plus you have a positive literal, we add it.

If we have a negative literal we just delete it.

And so that strips.

Very very simple if you remember that we're lifting the search into a description state

and therefore need to have other actions.

Remember the actions in the black box state is just action 317 goes from state 2011 to

state 350,001.

Here we do something different.

Here we apply kind of an action as a family of straight transitions, namely it acts on

all the states that are supersets on all the description that are supersets of the preconditions

remember we have more literals than a description it's a finer description on the world.

So we have a superset and then it actually determines all the states that can be reached

by the add and the delete list.

And that's this intuition of lifting here gives us the semantics.

And essentially this is the part you want to look at and the idea is that you have the

states which are sets of propositions and you can always apply an action if your state

is a superset of the preconditions and then the results you get by just taking the state

adding the add list to it and deleting the delete list from it.

Very very simple but this actually gives us the semantics.

Semantics of something means we're mapping it into something we already understand and

as of mid-November 2018 you understand search with states.

So we map strips into that formalism and we're done.

We don't have to say anything more mathematically.

So we can actually say what a state a plan is a plan is just a solution in terms of the

underlying black box semantics.

Good.

And then of course we kind of looked at examples as always.

Very simple planning problem going between three cities in Australia with a truck making

sure that you've been at all of them and ending up in Sydney again which is something that

is very easy to do.

And if you look at all the kind of state space this is the reachable part of the state space

and we have two kinds two solutions here and I've just basically made it clear that I do

the mapping here right.