To search Westminster records go to search at top of page, Bottom item is A-Z of record sets, Just type westminster and you get 4 options. Hilary On 30 April 2014 19:21, David Beakhust <dave@beakhust.com> wrote: > Sorry, due to a finger problem, this post was partially sent a few > moments ago. This is the intended one. > One aspect that might have afflicted the old system, but for me came > into sharp focus as a navigation and search filter issue is the question > of inconsistent and inappropriate use of Counties. > Finding myself unable any more to select uniquely and only the records > from Westminster archives, a matter of moments on the old site, i > searched for advice. > Included (and it worked, to a point) was to select "Middlesex" as the > county. Trouble was, it did not single out Westminster Archives > material, but did indeed include it. > > This drew me to another aspect. On census and BMD records you can only > search for recent counties, so back in Westminster, I find my man when I > choose "London", but don't find him with Middlesex. > This is inconsistent with the advice and findings on Westminster > archives records, so FMP seems to be using two meanings for county. > > Also in transcripts of BMD Index records, county appears (and did > previously appear) in the results, though it is not part of the record > set. Registration districts did span county boundaries, so some of these > must surely be plain wrong. > The one thing that does not ever change till the end of the world is the > name of a place that ruled at the time the record was created. Yet you > cannot use that if it includes a pre 19th century county. > Example: Wokingham (or Oakingham) WILTSHIRE, this is in post County > Council but pre 1974 BERKSHIRE. It is a long way from Wiltshire, but > before county councils existed, such fragmented counties were common, > and records included them. > Where a FMP search includes county, you must enter Berkshire for that one. > > Indeed, a FH program I use that is American in origin will report if a > town, county, state combo are valid for a particular date, and this is > of value for example where one element is absent from the original > record, or a date is in doubt. > Would that a comprehensive "county validator" was available for the UK! > And FMP was consistent (or preferably allowed you to search for pre- CC > era counties as an alternative to more recent ones (and was clear about > which you were choosing) > > On 30/04/2014 18:11, Adrian Bruce wrote: > > > > <<snipped>> > > I don't know how long they spent testing the site but it clearly wasn't > > adequate. > > <<snipped>> > > > > While I agree that user testing wasn't adequate - plainly! - the fact is > > that the new system was tested for some months before full cut-over, as > > numerous people were using it in pilot mode. > > > > We also need to remember that the new screens were demonstrated at WDYTYA > > Live and, so far as we understand, were generally well received. Though > of > > course they would have been chosen to display the powerful aspects of the > > new system. > > > > My suspicion is that the testing and demonstrations focussed on > individual > > screens and the appalling navigation was never in focus. After all, if > you > > are demoing screens, you are just not going to go up and down, up and > down > > all the time. This is *not* an excuse - navigation should be part of the > > test - but it may be an explanation. > > > > Adrian B > > > > > > > > > > ------------------------------- > > To unsubscribe from the list, please send an email to > SOG-UK-request@rootsweb.com with the word 'unsubscribe' without the > quotes in the subject and the body of the message > > > ------------------------------- > To unsubscribe from the list, please send an email to > SOG-UK-request@rootsweb.com with the word 'unsubscribe' without the > quotes in the subject and the body of the message >

Thanks for the advice re a-z data sets...... I see it now. (I still think that county is being used to a degree inconsistently, but this is arguably not a ui issue but probable a genuine difference of opinion.) Also do others find what i have - possibly because of changed search form design: If you change to searching a different data set, you have to enter data again (sometimes there are naturally different fields). So far so obvious, but irritating if you want to follow one family through the 51/61/71 etc census. However, the browser helps you by remembering words, names, dates previously entered in similar fields, so retyping should be minimised in some cases. So far so good, and often useful in the old design. With searches involving a handful of surnames or place names this has been a time saver. However, with the new design, I am as likely to find that the browser remembers a bunch of place names for a date field, a surname for a place name and so on. I conclude that the underlying code has just "field1", " field2" etc, instead of meaningful names. As a result, when moving to another dataset, the remembered fields are not in useful places because the presence of an extra field moves the re membered data into the wrong place. Anyone else found this? Dave Beakhust