About site

I created my first web site at work to make my work on parallel computing easily available to colleagues over the Web. The pages were just short descriptions of the projects I was involved in along with links to reports and software that could be downloaded. The pages were hand-edited using emacs and served by Apache on a Sun workstation.

Since then my site has been slowly growing...

Site design and development

The site is now developed and maintained on a laptop dual-booted between Windows 2000 and Fedora Linux. (But, when I’m rich and famous, the laptop will run both Win2000 and Linux simultaneously under VMware.) In this environment, the site is developed under Linux using the Zope web application server. Emacs is still used to edit all the HTML code, but Zope provides a lot of excellent features for dynamic generation of content and sharing of source code. Pages are still served via Apache to my mozilla browser. To check my pages are OK under Internet Explorer I currently have to boot up my noisy old desktop system under Windows and view the pages over the network. This is why I want VMware: so I can view pages under Linux and Windows at the same time on the laptop.

The version of my site accessible on the Web is a static version of my Zope site uploaded to my webspace provider from time to time. I use wget to suck out a copy of my Zope site to disk, followed by a bunch of Perl scripts to massage things for the Web. I also generate a CD-ROM version of the site this way.

The present design of the site owes a lot to the design of the Zope site. Many thanks guys! I have also stolen good ideas from many other places on the Web. Since I write all the pages directly in HTML code I’ve kept the layout reasonably simple; there’s no fancy graphics here!

Photo diary

The most complicated part of the development process has been getting the photographs onto the site. This seems to have taken up months of my spare time! I got a lot of useful information off the Web, especially from the photo.net site.

My photographs are mainly slides which I scan in at maximum resolution (2438dpi giving images around 2300x3500 pixels) and maximum bit-depth (10bit RGB) using a Minolta Scan Dual film scanner and VueScan software under Windows. I archive the raw scanned image (about 30-40MB compressed TIFF) to CD, before film, gamma and colour correction have been applied. I then use VueScan to apply film correction, a gamma correction of 2.2, and reduce the image to a bit-depth of 8. This is followed by using Adobe Photoshop to adjust the levels and perhaps make other corrections. I have found it very difficult to get images that look like the original slides! The image is then cropped in Photoshop and reduced to half size which removes a lot of the dust marks and gives a more manageable 4MB compressed TIFF file which is saved to disk.

All the photographs on the web site are then generated from these final TIFF files using a Perl script. The script generates JPEG images at a selectable number of sizes including the thumbnail size and full size, and takes care of landscape or portrait orientation and borders among other things.

Almost all the photo diary HTML pages are then generated dynamically by Zope using information about individual photographs and groups of photographs contained in a PostgreSQL database.

The maps in the photo diaries were originally produced using ArcExplorer from ESRI with embelishments added in Photoshop. These days I use the free GRASS GIS to produce them.

Documents

Get Acrobat Reader There are a number of documents on this site stored in Adobe Acrobat format. These documents can be viewed and printed with the free Acrobat Reader. The documents have been produced from postscript using the free postscript viewer, Ghostscript.

Feedback

If you have any questions, comments or suggestions about this site please contact me.

Last updated May 30, 2004
Copyright © Tim Oliver. All rights reserved.