Browser and Server Setup

Mobile Application Design and Development [./]
Spring 2010 — INFO 152 (CCN 42504)

Erik WildeUC Berkeley School of Information, UC Berkeley School of Information
2010-01-29

[http://creativecommons.org/licenses/by/3.0/]

This work is licensed under a CC
Attribution 3.0 Unported License [http://creativecommons.org/licenses/by/3.0/]

Contents Erik Wilde: Browser and Server Setup

Contents

2Abstract
1 Browsers
- 4Sending HTTP Requests
- 5Browsers and Platforms
- 6HTTP Resources and Web Pages
2 Servers
- 8Handling HTTP Requests
- 9Serving HTTP Requests
3 Practical Issues
- 11File Access vs. Server Access
- 12Local Servers
- 13XAMPP
- 14Accessing Files
- 15Server Names
- 16Relative URIs

Erik Wilde: Browser and Server Setup

(2) Abstract

As a starting point for the assignments in this course, we are setting up a server and files on the server for accessing them with (mobile) clients. This lecture looks at some of the practical issues such as how file access works through a file system and through server software, how server software can even provide multiple access paths (virtual hosts), and how working with Web-oriented content can be made more convenient by setting up a local environment (a local Web server).

Browsers

Outline (Browsers)

Browsers [3]
Servers [2]
Practical Issues [6]

Browsers Erik Wilde: Browser and Server Setup

(4) Sending HTTP Requests

Clicking a link creates an HTTP request
- HTTP request to the host specified in the URI
- adding information about the client (HTTP metadata)
- adding information about client state (cookies)
Ideally, a local HTTP library should take care of HTTP
- a common cache among all HTTP clients on the device
- HTTP should be considered as essential as TCP/IP
Practically, many Web applications have their own HTTP handling
- different browsers all come with their own HTTP code
- caching across browser cannot be shared

Browsers Erik Wilde: Browser and Server Setup

(5) Browsers and Platforms

Browsers use the platform's Internet support
- requesting a connection from the OS
- Internet connectivity handling is handled by the platform
Mobile devices can be offline
- native applications should properly handle offline situations
- HTML5 adds a lot of offline capabilities to Web applications

Browsers Erik Wilde: Browser and Server Setup

(6) HTTP Resources and Web Pages

Web pages almost always need more than one resource to be displayed
- the page itself as HTML code rendered in the browser
- images that have to be included in the HTML rendering
- stylesheets that are required for the page layout
- scripting code that is required for dynamic HTML (DHTML)
- additional content embedded in the page (e.g., Flash)
Ajax refers to pages that dynamically load new resources
- adding new content based on user interaction or other events
- closest possible way to resemble a real application

Servers

Outline (Servers)

Browsers [3]
Servers [2]
Practical Issues [6]

Servers Erik Wilde: Browser and Server Setup

(8) Handling HTTP Requests

Incoming connections are targeted at a specific program
- IP specifies the target machine (identified by an IP address [Geolocation; Internet Addressing (1)])
- TCP specifies the target port (80 is the default)
Web servers typically listen on port 80 for incoming requests
- if a connection is opened to port 80, it is assumed to carry HTTP
- servers getting non-HTTP traffic will respond with an error
The request contains the request URI and metadata
Servers must respond with a proper HTTP response
- in most cases the response contains the requested resource
- it may also contain an error message or a redirect response

Servers Erik Wilde: Browser and Server Setup

(9) Serving HTTP Requests

GET requests ask to get a certain resource's representation
URI identify resources to be manipulated through HTTP requests
URI can identify various kinds of server-side resources
- a file to be retrieved from the file system
- a file to be retrieved and parsed and processed before being served
- a program to be invoked which will produce the resource
HTTP servers can be standalone or integrated
- standalone servers must have a way to invoke server-side programs
- integrated servers run in the environment that executes the program

Practical Issues

Outline (Practical Issues)

Browsers [3]
Servers [2]
Practical Issues [6]

Practical Issues Erik Wilde: Browser and Server Setup

(11) File Access vs. Server Access

Files are available through different methods
- access through the file system (command shell or file browser)
- server access through HTTP server (incoming HTTP requests)
- server access through FTP server (through FTP sessions)
Accessing files is done by different users
- logged in user through the file system
- configured account for server access
- server access must be granted by the file owner

Practical Issues Erik Wilde: Browser and Server Setup

(12) Local Servers

HTTP servers respond to incoming requests from the Internet
- localhost [http://en.wikipedia.org/wiki/Localhost] always refers to the host's own interface
HTTP servers are available from various open source projects
Better local testing of Web-oriented projects
- local servers replicate what will happen over the Web
- file-based access (file URI) can have unexpected side-effects

Practical Issues Erik Wilde: Browser and Server Setup

(13) XAMPP

XAMPP [http://www.apachefriends.org/en/xampp.html] is a software installation package
- available for a variety of platforms (Windows, Linux, MacOS, Solaris)
- Apache as Web server
- MySQL as a database
- PHP and Perl as programming languages
- control panel for ease of use
Why running a Web server on the local computer?
- browsing local files using HTTP
- testing server configurations locally
- testing server-based code locally
- running a local database and a Web-based interface for it
XAMPP is not required but recommended for this course

Practical Issues Erik Wilde: Browser and Server Setup

(14) Accessing Files

The lecture pages are available at four different places
- file:///c:/Documents%20and%20Settings/dret/Desktop/drectures/mobapp-spring10 [file:///c:/Documents%20and%20Settings/dret/Desktop/drectures/mobapp-spring10]
- http://localhost/drectures/mobapp-spring10 [http://localhost/drectures/mobapp-spring10]
- ftp://tik50x.ethz.ch/home/dret/public_html/lectures/mobapp-spring10 [ftp://tik50x.ethz.ch/home/dret/public_html/lectures/mobapp-spring10]
- http://dret.net/lectures/mobapp-spring10 [http://dret.net/lectures/mobapp-spring10]
Using a local server makes local development easier
- uploading the working result to the server should just work
- however, to really make it just work, URIs must be used responsibly

Practical Issues Erik Wilde: Browser and Server Setup

(15) Server Names

One server can have more than one name
- names are resolved to IP address [Geolocation; Internet Addressing (1)]es for establishing connections
- multiple names can be very useful for hosting environments
Requests to the same server go to the same IP address
- servers use a special request field to learn about the host
- requests then are routed to the configuration for that host
Server space is available through people and people
- both servers are configured to look into ~/public_html
- the mobapp configuration writes to log files in ~/mobapp-logs

Practical Issues Erik Wilde: Browser and Server Setup

(16) Relative URIs

URIs are identifications of resources on the Web
- they can use a variety of access protocols (HTTP, HTTPS, FTP, …)
- they can be relative or absolute
Relative URIs are interpreted relative to their context
- http://dret.net/netdret [http://dret.net/netdret] points to aquadret.jpg [http://dret.net/netdret/aquadret.jpg]
- dereferencing the image link works regardless of the page's base URI
- HTML and JPEG always have to be managed together
Deciding on how to use absolute and relative URIs is important
- relative: for references to resources in the same management domain
- absolute: for references to resources that are managed differently

2010-01-29 Mobile Application Design and Development [./]
Spring 2010 — INFO 152 (CCN 42504)