A little while ago Steve McWilliam asked how names that are given to me make it into various databases in the NBN family. I thought it might be helpful to repost it here so that people can refer to it and I can keep it updated if procedures change.

Anything that is given to me (via any route - here, email, iRecord forum, iSpot forum etc) will go:

* Firstly I make the changes to my master UKSI copy (an Access 2000 database) on my PC in the NHM (and backed up to multiple locations). This database has utilities that help me manage the data fairly easily but it is in need of improvement - something I have been working on.

* Then I copy the master database to my sandbox (a SQL database and PHP website on the NHM servers) on a daily basis so that I can quickly check that the additions work in the way that they should - and to let others check when I have been working on. Nothing is ever amended in the sandbox - it is just a copy used to check the data and it is completely wiped and reloaded each time.

* I usually respond by letting the requester know what I have added (tvks etc). They can add the species to Recorder as 'User Added Taxa until 'until the full dictionary upgrade is generated. If they have said that they are waiting to add records to iRecord I email the details to John van Breda who amends his copy locally. This just allows users to enter records without too much delay.

* When I work on checklists there are often periods when the database couldn't be sent out as a whole because only half the tables have been updated ... it's a rolling work-in-progress. But when there is a point at which everything has been updated and the data integrity checks come up with no errors then I can take a snapshot of the whole database and send that off to Mike Weideli to be added to Recorder6.

* Mike then does his own checks on the data, independently of mine, and if he spots any issues (very rare these days) I correct them and resend until it looks good. When Mike has given it the thumbs up a copy will go to John van Breda (iRecord) and Graham French (NBN Gateway) to be imported to their systems at the next available slot. They obviously work on their own projects so they have to find time to update the taxonomy but it usually happens fairly quickly.

I hope that explains everything :) It's quite a complex process when you write it down but it all works well and we have been working hard to reduce the time between amendments getting requested and the changes appearing in the data-entry (iRecord, Recorder6) and data-reporting (NBN Gateway) systems :)