You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@geode.apache.org by "Jacob S. Barrett (Jira)" <ji...@apache.org> on 2019/08/23 17:10:00 UTC

[jira] [Commented] (GEODE-7119) Version ordinal constants are byte but should be short

    [ https://issues.apache.org/jira/browse/GEODE-7119?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16914471#comment-16914471 ] 

Jacob S. Barrett commented on GEODE-7119:
-----------------------------------------

What’s the rational for short rather than int? Is it saving us anything meaningful? If we go with an int we could align the byte values with the version number too. Maybe not necessary but worth rethinking if the potential need for a wider ordinal range is going to bite us later.

> Version ordinal constants are byte but should be short
> ------------------------------------------------------
>
>                 Key: GEODE-7119
>                 URL: https://issues.apache.org/jira/browse/GEODE-7119
>             Project: Geode
>          Issue Type: Bug
>          Components: serialization
>            Reporter: Bruce Schuchardt
>            Priority: Major
>
> The ordinal constants in Version.java are all of type "byte" but the "ordinal" instance variable is a "short".  The constructor for Version takes a "byte" but it should take a "short".
> Version.GEODE_1_11_0_ORDINAL is currently set at 107 but it should be increased to 110 to give us some spare ordinals in case we need to create a patch release of 1.11 that has on-wire changes.



--
This message was sent by Atlassian Jira
(v8.3.2#803003)