maltzsama commented on issue #96:
URL: https://github.com/apache/arrow-swift/issues/96#issuecomment-5164723863

   Yes @willtemperley , AI-assisted. Writing my own from here.
   
   You're right that the executable takes arbitrary paths, that removes the 
dependency I claimed. And the XCTest in #185 does duplicate what Archery would 
cover; it's a draft so I'll strip it back rather than build on it.
   
   What I want to get right is the order.
   
   Running the whole arrow-ipc-stream/integration corpus through the reader, 
142 of 182 files fail — decimal, dictionary, union, map, interval, null and the 
view types all hit unknownType. So a first pass at Archery would skip most 
cases.
   
   Is landing that with a long skip list the right move, and shrinking it as 
types get implemented? Or would you rather see types land first so the 
integration starts out meaningful?
   
   And is there a smaller first step you'd prefer than going straight at the 
JSON reader?


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to