LoginSign Up
← Back to all sparks
I

inbodb

DEVOPS
Velocity0.0

Connect to and Retrieve Data from Databases on the INBO Server

github.com/inbo/inbodb ↗

inbodb is plumbing — it exists to survive INBO's database migrations without breaking user code

databasesrbiodiversitydata accesssql serverinbo
◆Current state
inbodb opens connections to the Research Institute for Nature and Forest's SQL Server databases and wraps their contents in R functions — INBOVEG vegetation recordings, Florabank observations, Taxonlijsten species lists, Meetnetten monitoring visits, Watina groundwater. Version numbers have stayed in the 0.0.x range across five years. The recent releases are all reactive: a server migration, a database rename, an encoding failure, a load error.
◆Where it's heading
The package's job is absorbing institutional change so analysis scripts keep running. When INBO moved to a new database server, inbodb learned to try the new one and fall back to the old if a database had not migrated yet; when D0021_00_userFlora was deprecated in favour of D0152_00_Flora, the florabank functions were repointed rather than renamed. Feature work happens in bursts when a new database is opened up — Taxonlijsten and Meetnetten arrived together with vignettes — but the steady state is compatibility maintenance.
◆Prediction
With the server migration handled by a fallback that tries new-then-old, the fallback becomes dead weight once every database has moved; removing it is the natural next cleanup, though nothing in the notes commits to it.

◆Recent moves

  1. 1mo ago

    Watina encoding fix, autoconvert_utf8 deprecated

    An encoding problem against the Watina database resolved by deprecating the autoconvert_utf8 argument in connect_inbo_dbase(), plus the long-deprecated multiple argument finally made defunct. Housekeeping.

    View source ↗
  2. 3mo ago

    Fix for an error on package load

    A single-line release fixing a load-time error. The notes give no further detail.

    View source ↗
  3. 3mo ago

    inbodb follows INBO to a new database server

    The most consequential recent release: connections now target the new server and fall back to the old one for databases that have not migrated yet, so user scripts survive the transition untouched. get_taxonlijsten_items() is also fixed to return results without features under the default feature = '%'.

    View source ↗
  4. 1y ago

    Florabank functions repointed to the replacement database

    get_florabank_observations(), get_florabank_taxon_ifbl_year() and get_florabank_traits() move from the deprecated D0021_00_userFlora to D0152_00_Flora. The function names stay put, which is the whole point — the migration is invisible to calling code.

    View source ↗
  5. 1y ago

    inbodb opens up the Taxonlijsten and Meetnetten databases

    Two new databases reach R at once — three functions for Taxonlijsten species lists, four for Meetnetten schemes, locations, visits and observations, each with its own vignette. The package's feature work arrives in bursts like this rather than continuously.

    View source ↗
  6. 2y ago

    Point-plant-distance data and a simpler recording query

    get_inboveg_ppa() adds point-plant-distance data from INBOVEG, and get_inboveg_recording() stops requiring multiple = TRUE when several surveys or references are passed — the argument is deprecated with a warning. Removing a flag users had to remember is the kind of small API repair this package does well.

    View source ↗