[
https://issues.apache.org/jira/browse/MSITE-453?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hervé Boutemy updated MSITE-453:
--------------------------------
Summary: Add new lifecycle bindings for "maven-skin" packaging (was: Add
new lifecylce mapping "maven-skin")
> Add new lifecycle bindings for "maven-skin" packaging
> -----------------------------------------------------
>
> Key: MSITE-453
> URL: https://issues.apache.org/jira/browse/MSITE-453
> Project: Maven Site Plugin
> Issue Type: New Feature
> Reporter: Benjamin Bentmann
> Priority: Minor
> Fix For: backlog
>
> Attachments: new-lifecycle-mappings.patch
>
>
> Currently, creating a custom skin for Maven is done by a project with
> packaging "jar". The attached patch intents to introduce an individual
> lifecycle mapping named "maven-skin" for this purpose.
> Why that? I consider the re-usage of the "jar" packaging an abuse for the
> case of building a Maven skin. On the one hand, the "jar" packaging does too
> much. Skins usually do not get compiled or unit-tested, do they? Since any
> unused plugin invocation is an unnecessary risk of a build failure (sorry to
> say), I would appreciate a lifecycle mapping that is not overdressed. On the
> other hand, I could image that skins required some additional processing some
> day like a check whether all required images are present in the skin or
> whether the CSS references unknown IDs/names. Having a distinct lifecylcle
> mapping in the Maven Core would allow for a central definition of the build
> steps instead of requiring all users to extend the "jar" packaging.
> Especially for the first reason, i.e. having a packaging that does not more
> than required, the patch also defines a "resources" packaging. Such a
> packaging is intended for JARs that just contain resources one wants to share
> with other projects like rulesets for PMD, Checkstyle, etc. The lifecylcle
> mappings "resources" and "maven-skin" are identiical (now) but I consider it
> a bad practice to merge different use-cases just because they happen to be
> equal by coindicence.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
