Family Historian User Group

I use Gedcom Census on my 9Mb, 18000 person database and I find it frustrating having to wait nearly 40 seconds (1Gb RAM, 2Ghz Pentium, 160Gb HD) for GC to save the file after a census entry and re-load it before I can then work on the next task.

Since I do a batch of censuses (censi?) at a time, this delay is extremely intrusive. If it was only the odd census then it wouldn't be such a big deal.

Could I suggest an improvement where a census entry can be saved but the whole file is not written back to disk straight away. The current Save Entry button and option would be changed to Add Entry and a new option added to the File menu to save the whole file.

I realise that GC is extra careful with data and makes lots of backups [wink], but I would strongly argue that backing up a batch of census entries as opposed to after each individual one is no great a hardship if there was a catastrophy.

Are there any thoughts about why this can't or shouldn't be done?

Whilst I'm on the line, another improvement in terms of speeding up entry would be to allow entry of the record ID straight into the Rec ID field to the left of the name rather than having to click the Find button. This is mainly for the head of the household, the rest of the family can easily be selected from the immediate family pane. Since I have both GC and FH open at the same time, it would quicker and easier to get record ID's from FH and enter them into GC and tab to the age field rather than navigate with the mouse.

Also, a slight issue with the auto text feature - the marriage condition is set at married even when entering editing a person with no marriage details. The person was entered as the first person so therefore the head, but since there is no marriage record I would have thought GC should set the condition to unmarried.

ID:2460

Unfortunately it would require a major rewrite of large parts of Gedcom Census to incorporate the batch idea. Gedcom Census doesn't load in the whole Gedcom File into a data-structure then make changes to that, instead it just finds the relevant parts of the file and edits them. Therefore, it isn't as easy as just delaying writing the data back.

If I was starting afresh I would probably do it differently, but the initial version was just knocked together by me in a few hours on a wet sunday afternoon when I began to realise how time-consuming adding census entries was. The later versions have just been built onto existing code. While I appreciate that it must be frustrating, I have to be honest and say I'm unlikely to find the time to make such a major change to the coding.

The Record ID entering idea seems a sensible one and wouldn't take too long to implement so I'll certainly look at that for a future version (as well as probably making the backup system a bit more flexible to stop you moaning [wink])

The auto-text thing was only really designed to be a quick 'guess'. It doesn't currently look at marriage records though it probably wouldn't be too difficult to do so as it is looking at family records when it fills in the 'All Relatives' list so I'll add this to the 'things to do' list too.

Cheers

Nick

Thanks. I didn't realise that the file isn't 'in memory' so I can understand the problems involved in a re-write and I take into account that it is a 'hobby' program done in your spare time (but still very useful). So ignore my request.

In terms of the backup bit, just the ability to set the directory for backing up to as nothing to stop the backups is fine.