aboutsummaryrefslogtreecommitdiff
path: root/website/templates/features/experimental/StandardException.html
diff options
context:
space:
mode:
Diffstat (limited to 'website/templates/features/experimental/StandardException.html')
-rw-r--r--website/templates/features/experimental/StandardException.html2
1 files changed, 2 insertions, 0 deletions
diff --git a/website/templates/features/experimental/StandardException.html b/website/templates/features/experimental/StandardException.html
index 1484154d..c7fc40fe 100644
--- a/website/templates/features/experimental/StandardException.html
+++ b/website/templates/features/experimental/StandardException.html
@@ -48,6 +48,8 @@
There is a very slight functional difference: Normally, invoking <code>new SomeException(message, null)</code> will initialize
the cause to be <em>no cause</em>, and this cannot be later changed by invoking <code>initCause</code>. However, lombok's
standard exceptions <strong>do</strong> let you overwrite an explicit no-cause with <code>initCause</code> later.
+ </p><p>
+ A second slight functional difference: Normally, invoking <code>new SomeException(cause)</code>, if implemented as <code>super(cause);</code>, will set the message to be equal to the message of the cause. However, lombok does not do this - it leaves the exception as having no message at all. We think inheriting the message is fundamentally wrong - messages are not guaranteed to be sensible in the absence of the context of the exception-type. The cause ought to be listed anywhere where it is relevant; if you are using messages as direct user feedback (which is rare, in the java community), <code>@StandardException</code> can't really help you anyway; the infrastructure of e.g. <code>getLocalizedMessage()</code> is too complicated.
</p>
</@f.smallPrint>
</@f.scaffold>