I have not been very active over the last week or so - basically I have asthma and a cold had gone to my chest - nothing too serious - but rather than hit the computer I have actually spent whole days dozing in a chair watching the TV. However if someone sends me a message I have been endevouring to answer it - and this has been an opportunity to test out the FindMyPast new interface on real problems, although the answers to the questions which triggered this search will not appear online until I feel better.
As far as the Newspaper interface is concerned there has been definite improvements. The interface is a lot better once you have got to grips with it, and you can get results in date order and select individual years - but not specific dates with a year. There is one silly design error. When you do a search there is a box saying the items have been selected in terms of "relevance" but you can select a from a drop-down list choosing between "relevance" and "Date Order". If you select "Date Order" it gives you the oldest first - but if you reclick the drop down list now includes an additional option of oldest first. Who but an system designer idiot - given a three way choice implements this as a two stage process starting with a two item list - forcing the customer to have to call down another list .... Several other limitations of have not been addressed but these may be related to limitations on the British Newspaper archive.
As to the census search the position is even worse than I thought. The immediately obvious fault is selecting the year range for the date which (at least when using Google Chrome) produces nonsensical dropdown lists presumably because the boxes should be bigger. However the real problem is with the "Where" line - where you enter the place name. My first search was for a name in the town of Aylesbury, Bucks, and the system insisted in giving me all the names in the much larger Aylesbury Union - which includes many villages of no interest to me. I could find no way of restricting the search to the town of Aylesbury only.
I then tried the town of Thame, Oxfordshire and not only got people living in the town or Thame, the villages around Thame (some in Buckinghamshire) but also people living in Kingston on Thames, (Surrey), Walton on Thames (Surrey), Thames Ditton (Surrey), Henley on Thames (Oxfordshire) and for reasons that I didn't investigate Oxford St Aldgate (Oxfordshire), South Shields (County Durham) and St Dunstan in the East (City of London). Only 3 people on the first page actually lived in Thame and try as I might I could find no way of listing people who were living in the town of Thame.
I then did a search for people living at Buckland - as there are villages in Buckinghamshire and Hertfordshire with just the name with no additions, part of the former parish only being about a mile from where I live. I searched the first 10 pages and the majority of responses were people listed as living in Portsmouth but in addition I found Buckland Brewer (Devon), Shoreditch (London), Dover (Kent), Buckland Monachorum (Devon), West Buckland (Devon), Hatford (Berkshire), Stanford in the Vale (Berkshire), Filleigh (Devon), West Buckland (Somerset), Hinton Waldrist (Berkshire), Buckland Brewer (Devon), Buckland Tout Saints (Devon), Shellingford (Berkshire), Uffington (Berkshire), Ashburton (Devon), Kingston Lisle (Berkshire), Egg Buckland (Devon), Longworth (Berkshire), Buckland Winchcombe (Gloucestershire),
Therfield (Hertfordshire), Buckland Farringdon (Berkshire),
Buckland (Buckinghamshire PAGE 9),
Buckland (Hertfordshire PAGE 10).
This result is devastatingly bad with the first relevant Buckinghamshire and Hertfortdshire entries on page 9 and 10. The reference to Therfield is also interesting because the census form had no reference to Buckland whatsoever but Buckland is a parish next to Therfield, and several of the other entries, for instance those in Berkshire, refer to parishes near one called Buckland. So it would appear that FindMyPast were trying to be helpful and provide a facility that would extend the search to adjacent parishes - except
- The matching routine does not allow for the fact that many parishes across the country have similar or identical names and as a result the search routine frequently "match" with parishes hundreds of miles away!
- The badly designed (but useful if it worked) matching routine has not been provided as an option but instead used as the main search routine. Was this deliberate or is there a proper unsophisticated search routine in the code and software simply has a wrong link. However if it was a simple software bug there have been complaints for a couple of weeks and it would only be a matter of minutes to correct it.
- It is far from clear that any thought has been given to the problem of Poor Law Unions -which were a problem in the simple old search, but which you could work your way around using the old advanced search.
If FindMyPast had a default search for the town/village name (NOT the Union Name), a search for the Union name (which includes nearby places not always in the same county) and as another option a WORKING version of their nearby parishes search they would have a killer. The present facility - which only provides a seriously flawed version of a nearby parishes search is totally unfit for purpose if you have a place name which matches with places in different parts of the country. It can also be unsuitable when you want to search a small village (population a few hundred) immediately adjacent to a large town (population say 10,000). The changes in this area totally ignore the fact that many people (like me) use the site for local history studies and when we are searching for a name (person name or places name) we want the option to search for a place name like "Thame" without the computer extending the search to include "Thames". (This "computer being too clever by half" is also a major problem with the Newspaper search.)