[
https://issues.apache.org/jira/browse/LUCENE-5850?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14079419#comment-14079419
]
Robert Muir commented on LUCENE-5850:
-------------------------------------
Right, to be clear, here we have calmed the symptoms (the back compat time-bomb
bugs in SegmentReader and codec).
But to me this isn't enough, i am afraid of what would happen if we released
such code in a bugfix release.
We need to cure the disease, and prevent such bugs from being easy to do.
Additionally we should probably change the release instructions to generate
back compat index for *every release* bugfix or not, index format change or
not, and retroactively generate them for all 4.x releases. This would provide
additional protection.
> Constants#LUCENE_MAIN_VERSION can have broken values
> -----------------------------------------------------
>
> Key: LUCENE-5850
> URL: https://issues.apache.org/jira/browse/LUCENE-5850
> Project: Lucene - Core
> Issue Type: Bug
> Components: general/build
> Affects Versions: 4.3.1, 4.5.1
> Reporter: Simon Willnauer
> Fix For: 5.0, 4.10
>
> Attachments: LUCENE-5850.patch, LUCENE-5850.patch,
> LUCENE-5850_bomb.patch, LUCENE-5850_smoketester.patch
>
>
> Constants#LUCENE_MAIN_VERSION is set to the Lucene Main version and should
> not contain minor versions. Well this is at least what I thought and to my
> knowledge what the comments say too. Yet in for instance 4.3.1 and 4.5.1 we
> broke this such that the version from SegmentsInfo can not be parsed with
> Version#parseLeniently. IMO we should really add an assertion that this
> constant doesn't throw an error and / or make the smoketester catch this. to
> me this is actually a index BWC break. Note that 4.8.1 doesn't have this
> problem...
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
