Dave, have you seen the Common Transfer Format document (it's a Word doc)? I found it by accident while Googling for something else; it looks like the format documentation to the now defunct MapMate/Recorder data exchange software.

Charles

The sheer number of MapMate users in South Wales has meant that we had to adopt it alongside Recorder - the thought of importing 400,000 records from MapMate was just far too scary!

So the system that we currently have is that any records that are purely Lepidoptera are entered directly into MapMate and sync'd with the Glamorgan Moth Recording Group. Other Lep records that come to us from within reports are input in Recorder with the aim (when I get round to it) of importing these into MapMate. These records will have their 'Reference' set to somthing like 'Imported from Recorder'. So for our general operation we combine records from both Recorder and MapMate into a third central Access database (based on something that Dave Cope uses in BIS). The duplicated MapMate records will be filtered out.

Of course this system is not yet fully tested (or indeed implemented for that matter!), but it seems to work at the moment.

Dave

Tony, you should ask Hannah for a test version of R6 -- I'm using v6.6.5 here and it's very stable. More so than any version of 2002!

I think bespoke reporting systems are fine, even inevitable, along with bespoke data manipulation tools. The important thing is to have compatible core databases (NBN standard), with other technologies spinning off from there. I like to think of the NBN standards as the HTML -- or, more accurately, web standards -- for biological recording. Indeed, they're similar in many respects. The internet would never have become what it is without the standard that is (X)HTML (and of course the TCP/IP protocol - another 'standard'). Companies tried to build their own, proprietory, networks (AOL, MSN, Compuserve, CIX), but these have faded as they were not open standards and couldn't talk to one another. Then the browser makers tried adding their own, proprietory, extensions to the browsers causing huge headaches all round, with a seperate version of a website having to be coded for each browser. It's taken well over ten years (a looong time, in internet time) to finally reach a stage where web designers are finally demanding that software for the web be strictly adherent to the defined standards; and for software developers to deliver on that promise. Even Microsoft is committed to web standards, and we all know what MS are like when it comes to non-proprietory stuff.

I know it sounds a little preachy, but if we don't make an effort to support and advocate standards ourselves, we're doomed to suffer the mess of proprietory, non-interoperable systems for a long, long time.

I agree with your analysis Charles. A standard framework is the key. Perhaps the NBN data model would seem like a fair place to start. As we are "preaching" a bit , I'd like to think of the framework as transferable as possible given that I would prefer to work with PostgreSQL then MS SQL server. If the majority of the data validation was carried out by stored procedures then the work of client applications would be easier.

I've asked for a copy of Record 6 to test ages ago, but still not here. Charles, I'd be interested to know what the status of the stored procs in Rec 6 if you get a moment. Thanks.

Thanks for the info Charles. Everything I was hoping for has been mentioned. This is great news as it means data access clients will be able to hand off work to the server and I may not have to code routines in the client. In particular the import validation procs will be very useful.