Makes it to cache only relevant directories, adds `dist-newstyle` to prevent needless rebuilding even harder and shortens cache key name by supplying `hashFiles`multiple arguments
Recent cabal adopted XDG guidelines and stores data
across multiple directories under $HOME.
Creating ~/.cabal manually returns old behavior and allows caching single directory again.
"render" is a loaded term than might be thought as the generation
of a full HTML page. While for this phase we only process the status
and the headers.
Changing it to "response" so is not misleading at least. Users can check the
docs for clarification.
* ci: try nix-community/cache-nix-action
* ci: add cache-id param to setup-nix action
* ci: tidy up cache keys for non-nix jobs
* ci: merge-nix-caches-linux job
* ci: merge caches other way around
* ci: reduce number of caches
Should prevent disk overflow
* ci: change cache id prefix for merge
* ci: comment out cache merging job
* ci: revert cache id prefix
* ci: setup-nix: use latest cache-nix-action
Among others, makes action logs look more tidy (see https://github.com/nix-community/cache-nix-action/commit/17d19d3d8be918757589635bc1a6830be0b129d2)
* ci: use test-pg cache key for loadtest
* PUT returns 201 instead of 200 when rows are inserted
* POST with "Prefer: resolution=merge-duplicates" returns 200 instead of 201 when no rows are inserted
* test text/html and drop HtmlRawOutputSpec.hs
* all tests passing, removed all pendingWith
* make functions compatible with pg <= 12
* move custom media types tests to own spec
* anyelement aggregate
* apply aggregates without a final function
* overriding works
* overriding anyelement with particular agg
* cannot override vendored media types
* plan spec works with custom aggregate
* renamed media types to make clear which ones are overridable
* correct content negotiation with same weight
* text/tab-separated-values media type
* text/csv with BOM plus content-disposition header
steve-chavez