Plans have been announced to remove the domain name
There is a new directory called "themes" containing themed
photographs. These are photographs already in the album, but
linked by a common theme. See
G'day. This site started out as a way of cataloging my numerous railway pictures, which I first starting building up as a set of root window backgrounds. That was back in 1992, before the web. But once I became aware of the web (in 1994), I realised that a web page would not only allow me to peruse my catalog easily, but might also prove useful to others.
This site first went on-line in the week starting 3 Jul 1994,
and I claim it as one of the first railway web pages! Proof of
this lies in the school's server log pages, at
Anyway, I hope you enjoy these pages. Many people have written
to me to say they have enjoyed them, and you are welcome to
Whether you are visiting for the first time, or revisiting, you
will probably find the
The catalog is restricted to photographs where either I own the
copyright, or permission has been given to include photographs
here. Permission is given to copy any these images for personal
or research use, but not for commercial gain or profit without
written permission from me or the original copyright owner. I
would appreciate
Work has started on archiving photos. Access to the larger images is restricted so that they can only be accessed through a cgi interface. This allows me to track the actual use of the images more easily than through log files. (One of the consequences of this is that Internet Explorer browsers cannot access large images, due to a fault in IE with how it handles cgi parameters. Sorry, but if you use IE, you deserve everything you get!) In the longer term, images will be moved to and from archival storage according to demand. It is intended that thumbnails will continue to remain on the web pages, to show what is available.
I currently also have a rather large backlog of photos to put up, and while I am always most pleased to hear from new correspondents, I would ask you to note that if you send my photos for inclusion in these pages, there may be some delay in placing them online.
This page describes the extensive collection of railway images
maintained at
The numbering scheme used for the images themselves is relatively straightforward, and provides the basis for referencing all images and their descriptions. Where there is only one locomotive in the scene, its class/road number is used as the title. Multiple images of the same loco are reflected in a "-1", "-2", etc., suffix.
Where there is more than one locomotive in the scene, and the locomotives are coupled together, the road numbers are appended in the form "leading+trailing+.." as appropriate, again with a "-1" suffix to distinguish multiple images. If the locomotives are on separate tracks, the notation "loco=loco-1" is used. Multiple parallel running, with multiple coupled locos is handled by "loco+loco=loco-1" and so on.
Occasionally, where many locomotives appear in a scene, only the main ones may be identified (such as at roundhouses, etc.).
Please mail
There is a dynamic ranking process that uses access to the
full images as a voting mechanism. This does not use the web log
files, but only records those images accessed by clicking on the
thumbnails. See the
In an ongoing project like this, managing the large amount of
information that is continually changing can rapidly become
intractable. Hence I recognized quite early on that some sort of
management system was imperative. Here are the various approaches
that have been used:
Each image has its own description XML file. The main pages
are defined as XML documents called index.xml that
"include" the description files. Such files are then
translated directly to the corresponding HTML page using an
XSLT script. The shunting yards, or navigation bars
at the head of each page are generated by a series of separate
xml scripts, such as
Requests for files from the server can either be to the base XML file, or the derived HTML file. In both cases, a .htaccess file gives the server additional information on how to render the file.
For an XML file, there is a handler index.py that is invoked, and it is passed a parameter string indicating the XSLT stylesheet to be used for transforming the document into HTML, which is then delivered to the client. Dynamic data (such as file modification times) are delivered to the XSLT script at this point.
For an HTML file, if the file exists, it is retrieved and returned. This is a transition phenomenon only, where the HTML file is left over from the previous (static) method of rendering (see above section). Where the HTML file has been replaced by the XML file however, an Error 404 is generated (since the HTML file does not exist), and this is captured by the handler. The corresponding XML file is retrieved, the stylesheet name recovered from the nearest upward .htaccess file, and the process of the previous paragraph completed. All without any obvious action observed by the user (well, there is a bit of a speed penalty, but I'm working on that). Neat, huh?
Given any image reference (e.g., Gerogery-1) and its
directory (e.g., misc), all the information required to
present this image can be constructed from them, relative to the
user home directory on the server, viz.:
A taxonomy of railway subjects is not straightforward. I
have taken the approach of cataloguing by attribute of
ownership, although the epoch is not necessarily consistent.
Thus for example,