Website Update: Complete!

Finally my revised website is now up and running, and I would like to review some of the lessons which I learned during the process of revising a genealogy site with some 7000 - 8000 pages. This was the first major rewrite for 5 years, during which time coding techniques have changed, and visitors are looking for increasingly greater interaction and usability.

I started thinking about the design in November 2009, with the major changes being to the index pages and applying a common theme throughout the site. The headers for the Pedigree pages and some others I decided to leave, except for colour changes. I have said before, but I will repeat, design is not one of my strengths, so it took two or three weeks before I decided to use a scheme based on the colours of nature, mainly blue, green and brown, with an index page design to suit.

To write and verify the coding took another two months of work (actually three months but one was lost through illness and holidays). I like to work to dead lines, so in November I settled on 1st March for the launch, revised to 1st April, and it actually went on-line on 2nd April.

Relaxation

I recommend taking regular breaks! In the early hours of one morning I was working away trying to get rid of a bug, and nothing which I did seemed to make any difference. I simply could not understand it! Only after a short break did I realise that the screen I was looking at was on-line and not the one I was working on using my PC.

Tools used

The Pedigree pages were created using Legacy modified to CSS with the excellent LTools designed for use with Legacy. For my text (HTML, PHP and CSS) editor I use PSPad which allows batch search and replace. Because I have inserted some PHP in all my Pedigree pages I have to change all the file extensions from .HTML to .PHP using Ken Rename, and, finally, I also have to insert some script into the coding. This can only be done after the previous processing and hence requires the use of Regular Expressions to locate the point of insertion. To batch process I use Actual Search and Replace, but note that it cannot write the search and replace the strings for you!

Testing

Arguably this is the most important bit, it is certainly the most time consuming! Because of the use of PHP, I cannot test the Pedigree pages in IE unless it is actually on the web, which is something of a pain. I do, of course, have a folder on the server which I only use for testing purposes, but there always seems to be a difference between how something behaves in that folder and real life!

Links can be tested on-line using the W3C Link Checker, and do remember that for a 404 Error Page all links have to comprise the full URL ie. "http://www...."otherwise even your CSS won't work. This will not quite be the end of it, though, as you, or your visitors, will probably find a few missed bugs for a few days afterwards - make sure they can readily find a way of contacting you!

Finally

Ensure a plentiful supply of coffee, and enjoy!

Website Updates: Latest

During the past month it feels like the updating of my website has taken control of my life! The important bits, like the navbars, are working correctly, although they have yet to be tested in browsers other than Firefox and IE.

I am happy with the Index page, my Pedigree pages work, the Tutorials have been completed, as have the Books and Links pages. Yesterday I commenced work on the Photo Albums. During the past year I have taken a number of photos in England and Europe which need to be added to the collections, so I have have decided to split them, on a fairly arbitrary basis to Northern & Southern England, and Central & Southern Europe.

My existing albums were created in flash using SWF Quicker and one of their templates. I considered writing Dynamic CSS based albums, but after consideration, I decided that I couldn't really improve on the Album which I am using. So apart from changing the contents and the background colour I am sticking with what I have.

After this is completed, the next major task will be to transfer my Ferguson One-Name Site to my main site. On the one-name site are over 2000 names, about two thirds of which are Fergusons from northern England, and the format of the pages is that of Family Groups. It also has quite a different background style, so there is much work to be done here.

At present, the remaining eleven days for completion and testing appears to be a remarkably short space of time, so I can imagine a number of sleepless nights between now and the end of the month!

Genealogy Website: The Nav Bars

 A web page can be considered as being built on layers, the first being the base, or blank page, usually white but it can be given a colour. For my new site I have put another, slightly smaller, background layer on top, followed by the data and images. In my case I consider the nav bars as being a new layer since the drop-downs must sit on top of any data and images. One point about images, IE6, and earlier, do not recognise the ".png" transparency. Google will find many scripts to cure this, one of which is to be found here.

To date, I have used javascript for my drop-down menus, however it is now possible to write these in CSS. Unfortunately, because to do this the latest CSS standard is used there are problems with older browsers, in particular with IE6. The image on the right is a sneak preview of a section of my new index page, and as can be seen the drop-down sits on top.

To make them display on top of images is easy, simply use the "z-index" CSS style. However for other image forms it can be quite tricky. For Flash, within the "Object" tag the line < param name="wmode" value="transparent" > needs to be included. IFrames present a different problem, and to date, despite having tested varying "solutions", I have yet to find a way of stopping the nav bars from going behind the image for IE7 and earlier. Still, I've three weeks left!

Now, the above are the easy problems, again it is IE6 which causes the real head ache, because CSS drop-down menus simply do not work! To the rescue comes Suckerfish and the Sons of Suckerfish, which for my purposes I had to modify even further.

In my first Blog in this series I mentioned that I wished to eliminate/minimise the use of javascript, and for the navigation this has largely been achieved. As you will see to get them to work in the early versions of IE, and I no longer bother writing for IE5.5 and earlier, javascript has been a necessity, this I feel is a fair compromise. If those using IE6 do not have javascript enabled the navigation will not work, but it is still possible for visitors to get round the site albeit with more difficulty. (If you are viewing this using IE6, please upgrade, especially as you can now see our headaches!)

My final point concerns testing, I use Firefox for most of my development work and test in IE8, Opera, Safari, Chrome, Sea Monkey and Bing, these I have on my PC. As I do not have a Mac. I rather hope that if it works with Safari it will be OK (not really satisfactory, but there we go!). So how do I test in earlier versions of IE? A nice free piece of kit called IE Tester enables this to be acheived.; although still in Alpha version, and not every bit works fully, it does do the job, and for me is a life saver.

© Ron Ferguson 2010

Genealogy Website, New, Improved, & How 2

Everything had been going swimmingly with my new site, until I received from Dennis of Ltools a sneak preview of his forthcoming program update. He has made a couple of alterations to this program which changes the Legacy Pedigree web page output into validating CSS.

One of the functions of this program enables your own custom HTML, which Legacy allows you to incorporate into their code, to be exempted from processing by Ltools. I use this twice on a Pedigree page, for the header and footer. The changes which Dennis has made improve the logic behind the program, but in so doing tighten the way way in which we must write our code. I have detailed the couple of restrictions involved in my tutorial.

Naturally, my coding didn't comply, and since I work on the principle of "only as a last resort read the instructions", it took me a while to sort out what was going on! Basically they are that the start and end tags of an HTML section eg. a <DIV> must both be either inside or outside the custom section, and nesting of comments when using Javascript should be avoided.

The race to get my site update completed by the end of March is now definitely on!

Genealogy Websites: New, Improved, Updated & How

My current site Fergy's Website has been in its current form for over a year, except for the Pedigree pages having been converted to CSS. I commenced work on the new site in November 2009 with a view to launch on 1 March 2010, although this was pushed back to 1 April 2010 mainly due to illness.

What I hope to do over my next few Blogs is to illustrate the objectives and problems and how these were overcome. Please do not think that this is going to be so technical it will over your head, some of it perhaps, but most can be applied with little knowledge of programming.

It is essential to draw up the criteria which need to be met for the revised site, mine are:

• It must be based on the Legacy Pedigree Web creation function, for no other reason that I really like this layout. I have looked at formats such as The Next Generation, and others, most of which require a GEDCOM upload, but, for me, Legacy was the one.
• To maximise the use of CSS to separate out the design from the HTML data.
• To replace the Javascript navigation with CSS, and in general, minimise the use of Javascript.
• Where possible apply the most modern CSS and HTML standards.
• To ensure cross-browser compliance, including IE6 - which continues to be used by a surprisingly high percentage of genealogists!
• To redesign the Index page, which given that I do not pretend to be a designer is no mean feat!
• To ensure W3C standards are met.

Some of the above aims obviously conflict with each other, so compromises have to be made. These will be described later.

The first problem is that the Legacy Pedigree web pages are not W3C compliant and are written exclusively in HTML. Whilst I had converted the major design components to CSS, to convert all would have meant writing a program to automate this, or using a program which would use Regular Expressions to convert each tag individually - ugh! Unfortunately the first alternative I thought to be beyond my VB.Net programming skills.

However, in 2009 Ltools released a program for Legacy users which validates the HTML and converts all to CSS, and it was this which I used last year to overcome this main stumbling block to achieving a major aim. Some information on the use of this tool is also given in my tutorial here.

The other items will be considered in future Blogs, so watch this space!

Genealogy: Out of a test tube!

I have many reservations about prescriptive answers to genealogy presentations, much preferring them to reflect the rich tapestry of life. However with relationships I am less than certain. This question is not an ethical question, as many types of relationship exist in the real world, as we well know.

At present there is much discussion on how modern relationships, such as same sex partnerships, a surrogate parent and test tube children should be recorded by genealogy software, and much of this discussion has revolved around the differences between "true" genealogy and family history. I would ask are we asking the right question to the right people?

Let us consider the question of web browsers. Until recently (if now) browsers were designed in accordance with the ideas of the company designers. As a consequence web designers have for years suffered in trying to design sites which look the same in all varieties of browser, notwithstanding the existence of world wide web standards. Similarly, are we not in danger of genealogy software companies deciding for us what, in their varied opinions, is the best solution as to how the many types of relationships should be handled? I would suggest that this is not the way forward.

The laws regarding varying relationships differ from country to country, and  it would be wrong, therefore, for any standard to be based on what is legal in any one country, but the laws in most, if not all countries, should be considered. These days genealogy software is international, and surely the time has now arrived for a new approach to be made.

I would suggest that a cooperative venture between the national genealogy societies and the software companies be initiated to develop a format acceptable to most, but preferably all, participants. And who should pay for this? Why not the software companies by releasing their staff to attend meetings and providing the administrative back up, and perhaps supporting the attendance of the voluntary officials of the genealogy societies.

What I would like to see now is a proper international debate on this subject by genealogists and software companies, rather than comments, complaints and suggestions made by users of specific software to their supplier. As we saw from browser development, a standard may not resolve all the problems, but it would, I believe, be a big improvement on the ad-hoc arrangement we have today.

© Ron Ferguson 2010

Lies, Damn Lies and Genealogy

As we say up north, Martha Lydia seems to have been a bit of of a lass! She was the illegitimate daughter of Margaret who later married Joe, and had three illegitimate children herself before marrying Bill, with whom she had another six children. Whilst the details which follow relate to England, the principles are generally applicable.

Certificates

One point to remember is that although registration started in 1827, it was voluntary for births and deaths until 1875, so the absence of a certificate is not unusual. For births it is also possible to give a child any name one likes. Our Martha Lydia used her own surname ie. her mother's maiden name, for her first child and that of her step father for the next two. However, She did marry using her own name and left the father's name blank, unlike another illegitimate relative who decided to invent a deceased father when he got married.

In the great English tradition of telling officials what one thinks they would like to hear, where there was a disparity in ages between the couple, or maybe one of them is a little too young for marriage without parental consent, then some age adjustment was perfectly acceptable. Whilst talking about ages, an age at death is that which the reporter of the death thinks it is, which need not necessarily be the chronological age.

Censuses

A census provides an excellent opportunity for the head of a household to demonstrate their creativity. To come back to Martha Lydia, her stepfather came from a very religious family and it simply would not do for illegitimacy to be 'exposed' on a census return! This was resolved by giving the children the same surname as his, and describing them as "son" and "daughters". It was only after Joe died that Margaret recorded the true relationship.

Ages on censuses suffer from many variables, there are those whose increasing age is not related to the passage of time, and others where an accurate disclosure would reveal an illegal act. In particular, during the mid 1800s there were the Factory Acts which increased the minimum working age and restricted the hours of work. Naturally, the age of working juveniles had to comply with the law.

During the 19th century, divorce was the prerogative of the rich, bigamy was not uncommon, and clearly did not show on censuses. Even where there was no bigamous marriage partners would be recorded as a spouse, or maybe a 'housekeeper'. It was, of course, incumbent on all to maintain Victorian standards of morality.

Armed Forces Registration

Surely the details given on registration for service in the armed forces will be accurate? Well, no, especially during the first world war. At this time there was much peer pressure to serve 'King and Country' and there are many recorded instances of people enlisting and inflating their age in order to be eligible to join up and serve overseas.

Transcriptions

Even excluding those so called genealogists whose aim to to prove that they are descended from royalty, rather than to establish the truth, personal family trees must be treated with extreme care. Genealogist may well publish trees in which they clearly state that a relationship is a probability rather than an established fact, but by the time it has been copied and recopied by those who exercise less care this probability has been transcribed into the definitive family history.

With respect to censuses the images which we see, prior to 1911, are transcriptions and are, therefore, subject to both enumerator's and transcriber's errors. Certificates are written by an official, who might be ecclesiastical or civil, and particularly during the period when illiteracy was high the spelling of names is likely to be phonetic.

© Ron Ferguson 2010