## What's all the fuss about?
+![Alan Turing](alan-turing.jpg)
+
---
layout: true
* As much notepaper as you wanted
* Input → output
+.float-right[![right-aligned Growth rates ](turing-machine.jpg)]
## Turing machines
Simplified human computer.
Repetition == iteration or recursion
+What questions can we ask about algorithms?
+
+* Is it correct?
+* Does it terminate?
+* How long does it take?
+
+Different hardware works at different rates, so ask about how the run time changes with input size.
+
---
# Growth rates
# Anagrams version 1a: checking off (fixed)
+<table width="100%">
+<tr>
+<td>
```
Given word1, word2
anagram? = False
return anagram?
```
+</td>
+<td>
+```
+Given word1, word2
+
+if len(word1) == len(word2):
+ anagram? = True
+ for character in word1:
+ for pointer in range(len(word2)):
+ if character == word2[pointer]:
+ word2[pointer] = None
+ else:
+ anagram? = False
+ return anagram?
+else:
+ return False
+```
+</td>
+</tr>
+</table>
---
* Often leads to logarithmic growth
"Guess the number" game
- * What if I don't tell you the upper limit?
+
+* What if I don't tell you the upper limit?
## Dynamic programming
* Greedy algorithms
* Backtracking
+# Parallelism vs concurrency
+
+* Parallelism is implementation
+ * spread the load
+* Concurrency is domain
+ * new task arrives before you've finished this one
+
---
# Making change
This is the Halting problem.
+```python
+while i > 0:
+ print(i)
+ i -= 1
+```
+
+Will this halt? When will it not?
+
---
# Universal Turing machine
* Virtual machines
* ...
+---
+
+# Back to halting
+
+* Universal Turing machine can do any possible computation
+* Has the same halting behaviour of the emulated machine
+
+Can we build another machine that detects if a machine halts when given some input?
+
+Assume we can...
+
+```python
+def does_it_stop(program, input):
+ if very_clever_decision:
+ return True
+ else:
+ return False
+```
+
+But I can use the program listing as an input:
+
+```python
+def stops_on_self(program):
+ return does_it_stop(program, program)
+```
+
+---
+
+# Halting: the clever bit
+
+<table width="100%">
+<tr>
+<td>
+```python
+def does_it_stop(program, input):
+ if very_clever_decision:
+ return True
+ else:
+ return False
+```
+</td>
+<td>
+```python
+def stops_on_self(program):
+ return does_it_stop(program, program)
+```
+</td>
+</tr>
+</table>
+
+Let's put an infinite loop in a program that detects infinite loops!
+
+```python
+def bobs_yer_uncle(program):
+ if stops_on_self(program):
+ while True:
+ pass
+ else:
+ return True
+```
+
+If a `program` halts on on itself, `bobs_yer_uncle` doesn't halt. If `program` doesn't halt, `bobs_yer_uncle` returns `True`.
+
+--
+
+What does this do?
+```python
+bobs_yer_uncle(bobs_yer_uncle)
+```
+
+---
+
+<table width="100%">
+<tr>
+<td>
+```python
+def does_it_stop(program, input):
+ if very_clever_decision:
+ return True
+ else:
+ return False
+```
+</td>
+<td>
+```python
+def stops_on_self(program):
+ return does_it_stop(program, program)
+```
+</td>
+<td>
+```python
+def bobs_yer_uncle(program):
+ if stops_on_self(program):
+ while True:
+ pass
+ else:
+ return True
+```
+</td>
+</tr>
+</table>
+
+What does this do?
+```python
+bobs_yer_uncle(bobs_yer_uncle)
+```
+<table>
+<tr>
+<th>If it halts...</th>
+<th>If doesn't halt...</th>
+</td>
+<tr>
+<td>
+<ul>
+<li>`stops_on_self(bobs_yer_uncle)` returns `False`</li>
+<li>`does_it_stop(bobs_yer_uncle, bobs_yer_uncle)` returns `False`</li>
+<li>`bobs_yer_uncle(bobs_yer_uncle)` must loop forever</li>
+</ul>
+<b>Contradiction!</b>
+</td>
+<td>
+<ul>
+<li>`stops_on_self(bobs_yer_uncle)` returns `True`</li>
+<li>`does_it_stop(bobs_yer_uncle, bobs_yer_uncle)` returns `True`</li>
+<li>`bobs_yer_uncle(bobs_yer_uncle)` must halt</li>
+</ul>
+<b>Contradiction!</b>
+</td>
+</tr>
+</table>
+
+No flaw in the logic. The only place there could be a mistake is our assumption that `does_it_stop` is possible.
+
+(M269 has a different proof based on different sizes of infinity, and showing that there are infinitely more problems than there are Turing machines to solve them.)
</textarea>