Thoughts: forward/backwards planning
I'm still working on my first database and as such I 'may' not have planned as well as I might have done with a little more experience.
My database is a data collection tool for energy requirements of commercial/industrial properties, the intention is to collect the data using an ipad. I initially assessed my goal as: each 'primary' record = a job/a survey/a property, (as you can see from the graph each job starts with the table 'preliminary data' and basically related records spring from it and whilst there are many related records each job only has one 'preliminary data' record)
At the outset, I looked at the limitations of FMGO (I had used bento previously, which has a sync feature after off-line use which FMGO doesn't) and realised that travelling from property to property (many without wifi at all, and others that because of their construction meant they were virtual Faraday cages) I would be unlikely to be able to use the wifi features of FMGO. As such I decided I would create a master copy of my database and use a copy of that master for each job, thereby creating a separate file for each job.
I believed I would ultimately be able to write another database that could bring all these individual files together in a single database on my desktop for search etc., should I need to do so, which I doubt I'd actually need to do anyway, but would be nice anyway.
I was actually quite happy with the one job-one file solution.
I now note however that requests are being made for the same sort of bento sync after off-line use for FMGO.
If this happens I'm wondering if i should change my strategy, should I redesign my database to make a new record for each job, rather than a new file for each job. What are the advantages and disadvantages of each system.
I have a lot of work left on my current database, however most of the scripting and calculations are now out of the way, the graph may expand about 3-fold but I believe I have learnt the majority of things I need to know to achieve this.
After completing my database, I intend carrying out several sweeps (so to speak) to improve efficiency, to improve consistency from one data entry interface to the next, should I also consider moving from a one job/one file model to a potential sync model, what are your thoughts, please let me know.