Linux-SourceCase Studies
Here, we provide a few examples of the type of work which we perform.
Here, we provide a few examples of the type of work which we perform.
BAR Garages is a light commercial vehicle hire company in the North-West of England. They also buy and sell vehicles, as management of their fleet requires.
BAR Garages is a family run firm, and were not entirely comfortable with the idea of fully automating management of their hire fleet. They felt they needed some manual control of the process at some stage.
It was determined that many of BAR Garages customers were still using slow modem links, so large amounts of graphics and complex pages should be strictly limited.
The Solution: Linux Source designed a lightweight site to allow online hire queries, and also to display the vehicles that they had available for sale. The site uses Javascript for the fleet hire pages and PHP with a MySQL database for management of vehicle sales. Apart from the customer site, Linux source also designed pages to add, edit and remove vehicles from the database.
Due to international and national companies, Linux Source were unable to ensure that the pages rank highly in searches for car hire companies, but have ensured that queries with the locality attached place BAR Garages in the top 5 queries
Result: The site has been extremely successful and is responsible for approximately 50% of their vehicle sales.
GlaxoSmithKline is a multinational chemical concern. Their chemists have a need to select compounds based on their properties in order to determine which merit further (physical) study. In order to do this, they have developed a website to interrogate databases and modelling packages to find such compounds. The problem: The software has not been documented, is written by many different people in different programming languages, and is not properly managed.
The Solution: Linux Source rewrote the software, converting scripts written in TCL, PERL, awk, sed, grep, bash, csh all into a unified PERL CGI application, using SOAP to interface with the modelling packages, and the PERL standard DBI modules with SQL queries to perform database queries. In order to minimise the risk of loss of documentation, all modules included POD documentation with the source, and the application was version controlled in a CVS repository, where releases were tagged.
Result: GlaxoSmithKline chemists routinely filter batches of 50,000 compounds through the package, eliminating large amounts of laboratory effort and manual research.