Critical path to running CIPHA booster effectiveness study in Graphnet

[evansd]

These are the issues I think are absolutely critical to running anything:

• Port in changes from Graphnet PR #520
• Azure Synapse-compatible query engine #483
• Translate comparative booster effectiveness study into ehrQL #463
• Put together a list of contracts that are needed for the booster effectiveness study #485
• Support fetching results in batches in MSSQL engine #531

There are some other performance/reliability issues which we might be forced to address, or might turn out to be deferrable:

• Add exponential backoff to MSSQL engine #532
• More efficient implementation of codelists and categories #512

And there are some other issues which concern the API surface of databuilder which ideally we'd sort out before going too much further with "real study code" but aren't absolutely essential to do first:

• Publish a v1 tag for the databuilder docker image #564
• Binary output formats #507
• Remove backend registration system #561
• Placeholder: Decide what researchers need to import in order to write a dataset definition #527
• Remove need to register dataset #529

[inglesp]

Closing, as these issues are all labelled with the CIPHA label.