DITA module 09

Applications Development

JavaScript like Flash and SVG are all client-side approaches for dealing with information which is sent en-mass to the user, rather than being processed centrally and only the result viewed in a client browser. The adding of a java script to the HTML, containing only a few questions and options can allow an end user to get quickly to the part of a website which is of interest to them.
I built the script below in small sections which I checked worked and then knitted them together. The biggest problem I had was getting all the 'if' and 'else' selection statements to close correctly. The script is accessing the BBC News website and then directing you to relevant subject pages of your choice.
The script has four layers, which I have shown in this diagram, the first is news in general, the second the options for either regional news or sport, the third being an option within either the new or sports sections and the final layer is the relevant web pages for a news-region or sport-type choice.


I have added text lines within the java script to describe what it is doing.

http://www.student.city.ac.uk/~abhj012/java-bbc-text.txt

DITA module 08

Information Retrieval

Many databases work with an indexing technique. A small example of such an indexing I've included in my web space, on this occasion it is only accessing two documents finding key repeated words relevant to the historic subject matter.

http://www.student.city.ac.uk/~abhj012/dita-8-exercise.html

But sites like Google build up massive tables of millions of key words with the most frequently/recently visited sites at the top of the list.
To find an HTML tutor site I typed in 'HTML TUTOR' and the one mentioned on my web space was in the first page of suggested sites. My choice of sites was one which went straight to html rather than through lists of different programming languages.
To get a Google image of the Dubai World Trade Centre, I only had to type in 'Dubai W' and Google-Image directly offered me the full name.
In many ways the search for Kohl, carried out for DITA module 10, sums up better the pros and cons of these web searches. While doing a search on the Waitrose website for the vegetable Kohl an over enthusiastic over stemmed search came up with two phonetics equivalents bread 'rolls' and vegetable 'oil'. Joking apart if this happened often you would choose one of the other supermarket sites, which at least admits it can't find what you are looking for.
Drifting away from searching for my site, the Amazon on seems to be strong on indexing removal of small words, for example if you type 'Information Architecture WWW' Morville and Rosenfeld's book is the first on the list, though it's full name is 'Information Architecture for the World Wide Web'.

DITA module 07

Databases

The database approach to holding data allows data to be stored centrally once, with access being given where needed.
This central approach unlike the file one, also means that if a single piece of information needs updating ie a person's address it need only be done the once. Which in itself reduces not only the time for the update, but also possible errors between the databases.
With the database approach back-up and recovery is also undertaken centrally.
Relational databases are sets of tables which have a unique identifier called a 'primary key' for each entry.
The biblio library database can be accessed using Structured Query Language (SQL).
The display below shows an sql accessing just the 'titles' table so as to print out the ISBN number and title of a specific book.



The relational tables contain the total information held on a book split into smaller, quicker accessed bite sized pieces.
The following page, also on my web space, contains 10 additional SQL queries of multiple tables within the biblio book database.
http://www.student.city.ac.uk/~abhj012/lesson_7_all_sql.mht

DITA module 06

CSS

Style as an html tag can sit within each individual web page. But, cascading style sheets (CSS) allow for a standard format to all or some of your web pages. This not only gives a clearer message to the end user that these are all your sites pages, but it also allows for the HTML scripts to be much less cluttered and easier to read or update.

Within my web space I have used two different basic styles, neither were hard to generate and the second in particular has left the HTML script page it is applied too, much easier to read, and if applied when originally generated would have made the page build much quicker.

http://www.student.city.ac.uk/~abhj012/styleas1.css
used for index, first and GIS pages, sets the size and colour of the body text and the background colour for the screens.

http://www.student.city.ac.uk/~abhj012/styleas2.css
used for GI-Greenwich page only, this also contains table style information and two forms of headers.

The CSS also allows for just a single change if for example you change the header font colour for all your pages, rather than having to update each individual line of header tags or each page's embedded style tags.

The only thing really against CSS is that different browsers use the styling requests in slightly different ways. But these are well documented in sites like the following

http://freespace.virgin.net/sizzling.jalfrezi/frames/fstyles.htm compatability tables, danger list.

Such that, any dodgy tags can be kept at arms length.