CSC104H1F
Wednesday, December 5, 2012
【Memoirs】 Week#10 Nov 14th
Term Test Two
Today we had our term test two, the last term test for this course. Honestly, this test was much more difficult then the last test. I dont know why they made the second test so hard? maybe the average of first test was so high? I dont get it...
Lecture
Work
Experience
Hardware effects
strong information gets smaller, cheaper, faster bu the decades
strong information gets smaller, cheaper, faster bu the decades
【Memoirs】 Week#9 Nov 7th
Lecture
Networksroundly connected
(ring topology)
Way that messages passed between machines on this network
Each machine as a designated receiver for its messages, and designated sender of messages to it, and they are thus connected in a ring. A bit pattern, called a token, is passed from machine to machine, and a machine wishing to send a message to another machine retains the token, sends the message with the address of recipient machine, and waits until the message it is sending travels around the entire ring and it can confirm that is was sent. Then the token passed to the next machine.
centrally connected
(staar configuratution)
all connected
(bus configuration)
eg. ethernet
Way that messages passed machines on this network
Each machine listens for messages addressed to it. when a machine needs to communicate with another machine on the same ethernet bus, is broadcasts a message with other machine's adress. If a machine detects a collision ( another machine trying to broadcast at the same time0 it stops broadcast for random time interval and re-broadcast.
【Memoirs】 Week#8 Oct 31st
Lecture
Operators and operating systems
Machines take over in batches ( data paper cards tapes moved around fed to reader)
Machines began to take over seeting the program counter to a new job, collecting the output, fetching memory... but it was still one job at a time.
one user, one program, one computer
dedicated -> just one job ( game typesetting)
-> programme know all details disk access
-> practically no op. sys needed
unix (mostly) to the desktop
|
|__ mainframe migrate
> several program
> networking
> GUIs
An operating system should have kernel (shell, sheield access to hardware, referee sharing); untilities (repair, maintain machine, list files, format disk)
one user, one program, one computer
dedicated -> just one job ( game typesetting)
-> programme know all details disk access
-> practically no op. sys needed
unix (mostly) to the desktop
|
|__ mainframe migrate
> several program
> networking
> GUIs
An operating system should have kernel (shell, sheield access to hardware, referee sharing); untilities (repair, maintain machine, list files, format disk)
【Memoirs】 Week#7 Oct 24th
Lecture
Now we learned about the hardware architecture.
Where Von Neuman's going
There are some issues
> Von Neumann bottlenck
( speed+size chockel at bow)
> mortality of Moore's Law
( notten, as gott smaller)
Where Von Neuman's going
There are some issues
> Von Neumann bottlenck
( speed+size chockel at bow)
> mortality of Moore's Law
( notten, as gott smaller)
DrRacket cond codes
; number-hunch : number -> string
; Is n really, really big?
; that is, is it more than 101
(define (number-hunch n)
(cond
[(> n 101) "Oh, it's *really* big"]
[else "Not so big"]))
(number-hunch 5)
(number-hunch 102)
(define (GCD m n)
(cond
[(zero? n) m]
[else (GCD n (remainder m n))]))
(GCD 35 15)
(GCD 15 35)
Tuesday, December 4, 2012
【Memoirs】 Week#6 Oct 17th
Tutorial
Practised the codes we learned and prepare for the quiz, we could get our test back in today's lecture.
Lecture
I had been announced that the test paper would be returned at the end of the lecture.Today we kept studying in algorithms questions.
There is NO potentially an algorithm to solve every problem
If there are two or meow algorithms solving the same problem, we compare the efficiency to decide which algorithm to use.
How do we discover new algorithms?
A: By heuristics + tips, but they do not guarntee you can find a new algorithm.
Before electronnic, programmable computers, Alonzo Church and Alan Turing showed there were many unsolvable algorithms, also showed class of solvable algorithms.
Before Canada-411, we used to look up phone numbers in white pages. There are (at least) two different, correct ways to find the leaf ( 2-sided sheet) with the business you're looking for ( conclude it;s not there).
Linear search -----------> order of thousand steps
Binary search -----------> order 8 steps
Clearly there's no fool-proof method, but there's some techniques that often make progress. It helps to write down the whole process:
> Understand the problem
> Devise ( one or moew) plan (s)
> Try the plan
> Look back
Wikipedia Project Part II was due on Oct 19th, Part II spent much more time than part I, we had to write a specfic plan to help to improve and edit the topics we selected.
【Memoirs】 Week#5 Oct 10th
Term Test One
Today is first term test of this course, so nervous indeed! And there is no tutorials this week, we started the test at 6:10pm.
We finished the paper at 7:00pm. I did think this test was difficult for me, after the test, I asked my friends, they also did think in that way. Based on the test, I noticed I had lots of things that I forgot to review for this term test. I should be more careful next time. Praying, hope I would get a low mark in this test >_<...
Lecture
After the test, we still had to stay and listion to the lecture. Today's lecture is about the Algorithms.simple sequence of feasible steps to solve a problem deterministic ( in this course) credit AI-Khwarizmi
Ex.
Multiplication
PBJ ( Sticky algorithm)
Google page rank
Paper folding
fold over upper surface of paper strip after one fold, is has a downward crease fold the once-folded strip again and it had one upward, two downward there are good physical reasons you can't experment far beyong 6 folds given the number of folds, predict the pattern.
In DrRacket, we learned about mapping.
Like, (map add1 (list 1 2 3 4 5 6))
(map sqrt (list 1 2 3 4 5 6 7))
we can get (list 2 3 4 5 6 7)
(list
1
#i1.4142135623730951
#i1.7320508075688772
2
#i2.23606797749979
#i2.449489742783178
#i2.6457513110645907)
aslo we have map-image, allows you to map image's color.
Subscribe to:
Posts (Atom)