Web site refresh

After retiring from full-time employment earlier this year and successfully completing a computer programming training course at the O'Reilly School of Technology I've decided to pick up my genealogy research where I left off a couple of years ago. This 'kills several birds with one stone', it means I can use my technical writing skills to communicate via this blog, it means I can put my web design skills to use developing, maintaining and operating the MyRellies.com website whilst at the same time continuing my personal research and adding data to the archive.

As part of this re-focus of activity I decided to refresh the web site which is beginning to show its age, wasn't delivering on its design goals and which could well do with some of my relatively recently learned design skills courtesy of The Open University.

In 2008 when the site was developed three original design goals were:

• To promote the genealogy project and recruit volunteers to contribute time and effort
• An entry point for the webtrees application
• An entry point for the 'Rellie Finder' - data-grid application.

Performance against the first design goal was disastrous, in two years not a single volunteer was recruited. The second and third goals are only required if the first is successfully achieved. Webtrees is designed to act as it's own entry point and promotion of use of the data-grid is not a primary purpose of the exercise.

In 2013 / 2014 it is time to implement a site to better meet those goals with an appreciation of two further issues:

• Genealogists generally don't like to let data out of their direct control and let it be hosted by other people.
• There is a general dis-trust in putting data up on Internet, the fear is that it will be compromised or downloaded and sold.

I intend to return to these issues in later blog articles when the subjects of succession planning and collaborative projects are addressed

Warning - the following text is technical IT 'speak' and as such may cause damage to health

I thought it was worthwhile sharing some of the technical issues that have arisen and the conclusions reached. The new site is not yet completed as I write this, but sufficient is in place to describe what will be live over the next few days.

The original site was fairly conventional – a simple horizontal menu near the top of each page with four, single level options and four more links laid out as vertical blocks down the right-hand side. Terms of use, contact details and a privacy policy appeared at the bottom of each page with the area between header and footer left for variable amounts of text. There was no attempt to limit the amount of text on an individual page meaning that a users had to scroll to access hidden text and the footer. The site predated both the HTML5 and CSS movements and it delivered static information, but used Javascript to generate Google Analytics. The 'engine' of the site was a single page which acted as a customised entry point to the webtrees genealogy package and also as an entry point to a to a PHP driven data-grid of BMD records I had transcribed.

I decided that the menu system was inappropriate, and overkill for a very small site with four pages and a few hyperlinks, that had little or no potential to grow and wasn't likely to experience high traffic levels. I figured that the menu's could be replaced by standalone buttons and the hyperlinks rationalised to provide a rudimentary help system.

In the medium term I plan to stop using Google Analytics to record traffic, rather to use post-processing of the raw logs maintained by my hosting provider. This means I can strip all the Javascript code out of my pages and avoid the tracking and third party cookie issues which be-devil my privacy policies. I personally believe that Javascript will go the same way as Adobe Flash, in a similar timescale, but development of that argument is for another day.

I looked at a couple of responsive design frameworks – Bootstrap 3 from Twitter and Foundation from Zurb and whilst they have attractions their prime focus is on mobile and cross platform development and the 'code bloat' and steep learning curve they both bring is not a good trade-off in my position. In fact I've gone away from Adobe Dreamweaver CS5 as my web coding tool of preference, again I think it is overkill for a small site – I've hand-crafted the code using the Sublime 3 text editor from Sublime Pty and the good old Chrome browser

Undoubtably the most contentious decision (and one I could end up regretting, and having to do an 'about face' ) has been to adopt an HTML5 / CSS3 only coding standard without degraded support for non-compliant browsers.  I can't imagine there are many people out there, even genealogists, still browsing on Microsoft XP - but I guess I'm about to find out !!