Hi all,

Thanks Nandor and others for bringing my comment up for discussuion. The 
failing tests were written by me, as I wanted to know how Avro really worked - 
and if we change the behaviour, naturally the tests should be updated as well.
The question is how we should go about this, as it breaks backwards 
compatibility?

Someone might be relying on the non-promotion (not expecting the conversion to 
work). On the other hand, the spec is quite clear on this, so it could be seen 
as a bug, violating the spec. Still, I think it is worthy a more general 
discussion/voting (though I must admit, I am not that well versed in the Avro 
release/contribution process)

On a personal note (though my application does not, to my knowledge, rely on 
floats), I actually would prefer changing the spec, not allowing conversion to 
float. But YMMV,

Thanks for your time,
/Anders Sundelin


-------- Original Message --------
Subject: [jira] [Commented] (AVRO-2072) ResolvingGrammarGenerator doesn't 
implement schema resolution correctly for unions
From: "Nandor Kollar (JIRA)" <[email protected]>
Date: 30 aug. 2017 13:55
To: [email protected]

    [ 
https://issues.apache.org/jira/browse/AVRO-2072?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16147116#comment-16147116
 ]

Nandor Kollar commented on AVRO-2072:
-------------------------------------

Thanks Zoltan, I'll have a look at the failing tests!

> ResolvingGrammarGenerator doesn't implement schema resolution correctly for 
> unions
> ----------------------------------------------------------------------------------
>
>                 Key: AVRO-2072
>                 URL: https://issues.apache.org/jira/browse/AVRO-2072
>             Project: Avro
>          Issue Type: Bug
>            Reporter: Nandor Kollar
>            Assignee: Nandor Kollar
>         Attachments: AVRO-2072.patch
>
>
> According to 
> [specification|https://avro.apache.org/docs/current/spec.html#Schema+Resolution],
>  int and long is promotable to float, but when using SchemaValidator, a union 
> with a single int or long branch is not readable by an union with a float 
> branch.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to