On Fri, 25 Jun 2004 16:12:10 +1200, you wrote: > >I have tested both image servers, my comments are; >Public server - It's definitely slower but works fine. Users may be frustrated having to click on so many pages to get to the sheets and then play the guessing game of where the "Mayall" page is somewhere in the I-S list of from 1200 to 2000+ pages. I have never been on 1837 online but if their system is easier users may opt to pay rather than go thru this exercise each time. >Would it be possible to target the pages a little better ie >1848D4-Mat-1234.tif >1848D4-Maz-1235.tif (examples) >rather than 1848D4-I-S-1234.tif. >Surely this would be easier on system resources too?. It took me 7 guesses to finally get to a target name. It would be a HUGE renaming job, and will in any case be redundant as we advance to better ways of accessing scans. >Another possible improvement, many websites nowadays display multiple list boxes on the first page and hold the data for each in the script code. >ie when the type list is selected (1-3 options) the Year list is instantly updated (3 sets of Years) and so on, select the Qtr list (1-4 options) and then go to another page for the huge individual page list. >An example site; >http://www.auto.co.nz/used.php try selecting a Make which updates Model We may look at something like this (it uses Javascript), but we are mindful of the fact that part of the reason for the interface is to punt enough adverts to pay for the service. >I randomly selected one Year. After going thru Deaths,1849,Dec ending up here; >http://images.freebmd.org.uk/GUS/1849/Deaths/December/LDS-211-000-0951 >only to find two files available; >00x0-000.tif - a useless message >0namefix.wbt - garbage >All 4 QTRS are the same. >What is the point of making this Year available at all . I wonder how many more of these are offered to waste users online time, which with some is costly. This is a known issue, and the dummy files are being removed. >Notwithstanding, I'm sure the public will be appreciative of the another free service being available, so well done team. > >Transcriber server - I find it quite elongated (Password,type,year,Qtr,page#) compared to the simple table we once had. >I would have tested further but I kept getting a Gateway Timeout error from the server. This bit is IDENTICAL to the existing facility, you are just accessing it directly instead of via the front end. Once it goes live, you will be going via the front end. -- Dave Mayall

Dave Mayall wrote: >>I have tested both image servers, my comments are; >>Public server - It's definitely slower but works fine. Users may be frustrated having to click on so many pages to get to the sheets and then play the guessing game of where the "Mayall" page is somewhere in the I-S list of from 1200 to 2000+ pages. I have never been on 1837 online but if their system is easier users may opt to pay rather than go thru this exercise each time. >>Would it be possible to target the pages a little better ie >>1848D4-Mat-1234.tif >>1848D4-Maz-1235.tif (examples) >>rather than 1848D4-I-S-1234.tif. >>Surely this would be easier on system resources too?. It took me 7 guesses to finally get to a target name. > > It would be a HUGE renaming job, and will in any case be redundant as > we advance to better ways of accessing scans. Though we could consider letting users provide such hints to as as they access pages. Cheers, Ben. -- http://www.apache-ssl.org/ben.html http://www.thebunker.net/ "There is no limit to what a man can do or how far he can go if he doesn't mind who gets the credit." - Robert Woodruff