lidavidm opened a new issue, #37704:
URL: https://github.com/apache/arrow/issues/37704

   ### Describe the enhancement requested
   
   An Arrow schema in IPC is a base 'schema' message, which can then be 
encapsulated in a 'message' payload.
   
   In Python/C++, convenience methods like `Schema.serialize` generate the 
latter (schema + message), which is also how it would appear in an IPC stream 
or file.
   
   In Java, 
[`Schema#toByteArray`](https://github.com/apache/arrow/blob/15a8ac3ce4e3ac31f9f361770ad4a38c69102aa1/java/vector/src/main/java/org/apache/arrow/vector/types/pojo/Schema.java#L220-L231)
 generates the former, which is a footgun; you'll generate data that looks fine 
so long as you only use Java, but which will just not work if you bring in any 
other implementation, for no obvious reason. [This bites even Arrow core 
developers.](https://github.com/apache/arrow/issues/20793)
   
   We should deprecate the Java methods and add new conveniences that are in 
line with C++ et al.
   
   ### Component(s)
   
   Java


-- 
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