Midterm-Review

• I haven't heard of many questions from reading, hopefully this means you

understand all of it. Questions may be pulled from reading that were not directly covered by slides.

• Browser
  • DNS hostname lookup
  • send receive data over TCP/IP via HTTP(S)
  • discover and download dependencies
  • create DOM
  • apply CSS
  • Display
  • executing code
  • Respond to user actions
• Tree structure used to model HTML

• Purpose of homework was to make sure every one tried different browsers
• Think of the trade-offs the browsers made
• Think about what technologies were behind that flexibility and features

• blink, font
• Easier to compose, can't get confused when you want to use markup character
• wiki text
• body, head

• Why separate style and content?
• What does cascading mean?
• What are some CSS selectors?
• When should you use the style HTML attribute?

• You can present content in different ways (like these slides)
• Styles can "cascade" over one another, applying the result of all stylesheets to the elements
• .class #id parent descendent
• When you don't have access to the stylesheet

• Purpose was to use different elements appropriately
• table used for tabular data, div for layout
• use CSS instead of font tag to set text styles

• HTTP method, GET or POST
• text checkbox password
• Helpful to provide instant validation
• No, must re-validate to avoid manual attacks

• What are two conceptual ways to remove elements from a page?

• Purpose was to practice using Javascript
• Get used to event driven nature (stuff happens when you interact with page)
• Pracice updating a page after an interaction
• Steps to change an element (select element, modify some attribute like

text, or css)

• parts:

  foo

  schema

  example.com

  domain / authority

  8042

  port

  /over/there/index.dtb

  path

  type=animal&name=narwhal

  query

  nose

  fragment

• both represented by trees
• absolute
• URLs, representations

• What are the OK and NOT FOUND return codes?
• How do we create a custom NOT FOUND page?
• Why is HTTP/1.1 connection reuse helpful?

• 200, 404
• Write HTML content in the response body
• Setting up connections causes page load latency

• purpose was to practice speaking HTTP
• HTTP request consists of a method, absolute path, version, headers, and body/content
• Browser helps by "following" redirects, running searches, etc.

PUT /foo HTTP/1.1
Host: localhost

HTTP/1.0 500 INTERNAL SERVER ERROR
Content-Type: text/html; charset=utf-8
Connection: close
Server: Werkzeug/0.8.3 Python/2.7.3
Date: Fri, 05 Oct 2012 07:20:41 GMT

• No
• It must resend all context for each request
• Usually slower than static
• Static since they do not change

• Where are query arguments in a GET request?
• Where are query arguements in a POST request?
• Why is Content-Length needed to send data in a PUT request?
• What are some example MIME types?

• Purpose was to practice sending data to the server, see how the server can respond dynamically
• Understand how Content-Type affects the interpretation of data

• Why are cookies used?
• In normal usage, how is the cookie data set?
• Cookies are structured as _ - _ pairs
• How do we prevent important cookies from being spied and intercepted?
• Without cookies, can we tell if a user is logged in with a typical GET request?
• Why set cookie expire times in 2037?

• What is an example of a RESTful system?
• Is a Noun endpoint / URL an indicator of a non-RESTful, or RESTful system?
• Is using POST for all commands an indicator of a non-RESTful, or RESTful system?
• What is the primary trade-off for composable systems like REST?

• Correctness, reliability with retries or error, ordering of

packets, congestion control

• Only one packet typically needed for request, one for response
• No

• It needs to know the IP address to which it will send the HTTP request
• Caching can improve the performance of a client, and at a network level

reduce traffic and improve performance of all clients on the network

• Top Level Domain, such as .com .edu

• Needed to establish a TCP/IP connection, it requires 3 round trips so can

be slow depending on server

• Assuming same bandwidth, the 100 files would be improved more since you'd have to wait for the latency of sending the request and getting the first byte back
• Accept-Encoding header