[
https://issues.apache.org/jira/browse/TAP5-2742?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17639676#comment-17639676
]
ASF subversion and git services commented on TAP5-2742:
-------------------------------------------------------
Commit 298a4709e485660dbadcaec1841a4a19788e2801 in tapestry-5's branch
refs/heads/better-page-invalidation from Thiago H. de Paula Figueiredo
[ https://gitbox.apache.org/repos/asf?p=tapestry-5.git;h=298a4709e ]
TAP5-2742: refactoring URLChangeTracker to make it parameterized
> Smarter page cache invalidation
> -------------------------------
>
> Key: TAP5-2742
> URL: https://issues.apache.org/jira/browse/TAP5-2742
> Project: Tapestry 5
> Issue Type: Improvement
> Components: tapestry-core
> Reporter: Thiago Henrique De Paula Figueiredo
> Priority: Major
>
> Since Tapestry 5's inception, it throws the whole set of assembled page
> instances when anything related is changed, be it the class itself, its
> template and maybe also associated messages and assets. In very large
> projects with large pages, this can reach a point it slows down the user
> (programmer) productivity, forced to wait for unchanged pages to be
> reassambled.
> Tapestry should provide some way for users to segment page, component, mixin
> and base classes to separate regions, one for each classloader, to avoid
> clearing out cached page instances that don't have themselves or the classes
> they use changed.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
