<!DOCTYPE html><html><head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body>
<p>Weighing in on this, I (too) don't think we need to adjust this
for Serialization.</p>
<div class="moz-cite-prefix">On 2026-02-25 18:52, Brian Goetz wrote:<br>
</div>
<blockquote type="cite" cite="mid:7b3fb512-36a7-4b12-961c-662209a6da36@oracle.com">
<br>
<blockquote type="cite" cite="mid:995555563.49010127.1772041382846.JavaMail.zimbra@univ-eiffel.fr">
<div style="font-family: arial, helvetica, sans-serif; font-size: 12pt; color: #000000">
<div>One missing discussion is serialization or exatcly what
about the easy serialization we get with records ?</div>
</div>
</blockquote>
<br>
Records that are marked Serializable get special treatment in
serialization. They also are excellent candidates for use as
serial proxies. Deconstructible classes with canonical
constructors would also be good candidates for use as serial
proxes. I am doubtful we want to grant the special serialization
treatment of records to deconstructible classes; I don't see a
principled basis for doing so, since deconstructibility is about
taking things apart, not representation. <br>
<br>
<blockquote type="cite" cite="mid:995555563.49010127.1772041382846.JavaMail.zimbra@univ-eiffel.fr">
<div style="font-family: arial, helvetica, sans-serif; font-size: 12pt; color: #000000">
<div>If a deconstructible class implement Serializable, should
a canonical constructor, with the same visibility as the
class, required * ?</div>
</div>
</blockquote>
<br>
I think we are better off leaving the legacy serialization
mechanism alone, and directing such questions to Marshalling. <br>
<br>
(Serialization bullet dodged!) </blockquote>
<pre class="moz-signature" cols="72">--
Cheers,
√
Viktor Klang
Software Architect, Java Platform Group
Oracle</pre>
</body>
</html>