You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@maven.apache.org by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org> on 2008/10/28 13:01:51 UTC
[jira] Created: (MNG-3805) Prefer LinkedHashMap over HashMap in
ArtifactUtils
Prefer LinkedHashMap over HashMap in ArtifactUtils
--------------------------------------------------
Key: MNG-3805
URL: http://jira.codehaus.org/browse/MNG-3805
Project: Maven 2
Issue Type: Improvement
Components: Artifacts and Repositories, Dependencies
Affects Versions: 2.1.0-M1, 2.0.9
Reporter: Benjamin Bentmann
Priority: Trivial
Attachments: deterministic-map-ordering.patch
Sometimes order is important, Maven is learning that the hard way (MNG-1412, MNG-3118, SUREFIRE-61, ...). Methods that don't know for sure that ordering is irrelevant should follow "better safe than sorry" and keep the ordering of input collections.
This is basically a clone of MARTIFACT-9, this time aimed at the 2.0.x and 2.1.x branches.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Commented: (MNG-3805) Ordering of extension class path is
indeterministic
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=152189#action_152189 ]
Benjamin Bentmann commented on MNG-3805:
----------------------------------------
A (currently disabled) IT is online.
> Ordering of extension class path is indeterministic
> ---------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Attachments: deterministic-dependency-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Ordering of extension class path is
indeterministic
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Attachment: deterministic-dependency-ordering.patch
> Ordering of extension class path is indeterministic
> ---------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Attachments: deterministic-dependency-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Ordering of extension class path is
indeterministic
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Summary: Ordering of extension class path is indeterministic (was: Prefer LinkedHashMap/-Set over HashMap/-Set for artifacts/dependencies for determinsitic ordering)
> Ordering of extension class path is indeterministic
> ---------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Attachments: deterministic-dependency-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Prefer LinkedHashMap/-Set over
HashMap/-Set for artifacts/dependencies for determinsitic ordering
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Attachment: (was: deterministic-map-ordering.patch)
> Prefer LinkedHashMap/-Set over HashMap/-Set for artifacts/dependencies for determinsitic ordering
> -------------------------------------------------------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Attachments: deterministic-dependency-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Prefer LinkedHashMap over HashMap in
ArtifactUtils
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Attachment: deterministic-map-ordering.patch
> Prefer LinkedHashMap over HashMap in ArtifactUtils
> --------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Improvement
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Priority: Trivial
> Attachments: deterministic-map-ordering.patch
>
>
> Sometimes order is important, Maven is learning that the hard way (MNG-1412, MNG-3118, SUREFIRE-61, ...). Methods that don't know for sure that ordering is irrelevant should follow "better safe than sorry" and keep the ordering of input collections.
> This is basically a clone of MARTIFACT-9, this time aimed at the 2.0.x and 2.1.x branches.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Prefer LinkedHashMap/-Set over
HashMap/-Set for artifacts/dependencies for determinsitic ordering
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Priority: Major (was: Trivial)
Description: One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes. (was: Sometimes order is important, Maven is learning that the hard way (MNG-1412, MNG-3118, SUREFIRE-61, ...). Methods that don't know for sure that ordering is irrelevant should follow "better safe than sorry" and keep the ordering of input collections.
This is basically a clone of MARTIFACT-9, this time aimed at the 2.0.x and 2.1.x branches.)
Attachment: log-good.txt
log-bad.txt
deterministic-dependency-ordering.patch
Issue Type: Bug (was: Improvement)
Summary: Prefer LinkedHashMap/-Set over HashMap/-Set for artifacts/dependencies for determinsitic ordering (was: Prefer LinkedHashMap over HashMap in ArtifactUtils)
> Prefer LinkedHashMap/-Set over HashMap/-Set for artifacts/dependencies for determinsitic ordering
> -------------------------------------------------------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Attachments: deterministic-dependency-ordering.patch, deterministic-map-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Prefer LinkedHashMap over HashMap in
ArtifactUtils
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Attachment: (was: deterministic-map-ordering.patch)
> Prefer LinkedHashMap over HashMap in ArtifactUtils
> --------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Improvement
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Priority: Trivial
> Attachments: deterministic-map-ordering.patch
>
>
> Sometimes order is important, Maven is learning that the hard way (MNG-1412, MNG-3118, SUREFIRE-61, ...). Methods that don't know for sure that ordering is irrelevant should follow "better safe than sorry" and keep the ordering of input collections.
> This is basically a clone of MARTIFACT-9, this time aimed at the 2.0.x and 2.1.x branches.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Updated: (MNG-3805) Ordering of extension class path is
indeterministic
Posted by "Benjamin Bentmann (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Benjamin Bentmann updated MNG-3805:
-----------------------------------
Attachment: (was: deterministic-dependency-ordering.patch)
> Ordering of extension class path is indeterministic
> ---------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Attachments: deterministic-dependency-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] Closed: (MNG-3805) Ordering of extension class path is
indeterministic
Posted by "Brett Porter (JIRA)" <ji...@codehaus.org>.
[ http://jira.codehaus.org/browse/MNG-3805?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Brett Porter closed MNG-3805.
-----------------------------
Assignee: Brett Porter
Resolution: Fixed
Fix Version/s: 2.0.10
Applied and re-instated the IT
> Ordering of extension class path is indeterministic
> ---------------------------------------------------
>
> Key: MNG-3805
> URL: http://jira.codehaus.org/browse/MNG-3805
> Project: Maven 2
> Issue Type: Bug
> Components: Artifacts and Repositories, Dependencies
> Affects Versions: 2.0.9, 2.1.0-M1
> Reporter: Benjamin Bentmann
> Assignee: Brett Porter
> Fix For: 2.0.10
>
> Attachments: deterministic-dependency-ordering.patch, log-bad.txt, log-good.txt
>
>
> One part of Maven where class path ordering hasn't already been fixed in the past is the extension class path. Apart from a proposed patch, two logs from our IT testmng3599useHttpProxyForWebDAV are attached. Both were run against Maven 2.0.10-RC1, with the only difference being one using JDK 1.4 and the other one 1.6. Not only do these logs substantially differ, the build on JDK 1.6 even fails due to picking up the wrong extension classes.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira