<!DOCTYPE html><html><head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
</head>
<body>
<br>
<div class="moz-cite-prefix">On 3/21/2024 4:07 AM, Severin Gehwolf
wrote:<br>
</div>
<blockquote type="cite" cite="mid:7f95dd63bff94ca764d58226c2be464fe006f0e0.camel@redhat.com">
<pre class="moz-quote-pre" wrap="">On Thu, 2024-03-21 at 11:42 +0100, Volker Simonis wrote:
</pre>
<blockquote type="cite">
<pre class="moz-quote-pre" wrap="">This proposal sounds interesting and reasonable.It would be especially
beneficial if the JDK Updates Project (added to CC) would adopt this
new model.
</pre>
</blockquote>
<pre class="moz-quote-pre" wrap="">
We can consider it for jdk-updates at a later point. I don't think it's
good to conflate those two at this point. Lets get mainline JDK
development switched, first. See how it pans out for mainline and
corresponding downstream consumers.
Once we have that experience, we could model it for jdk-updates further
down the line.
</pre>
</blockquote>
<br>
I agree. This proposal deliberately does not address update
releases. Any changes to update releases would necessarily need a
separate proposal with its own discussion. Waiting to see how it
works in practice for the JDK Project seems wise.<br>
<br>
If and when the JDK Updates Project makes a proposal to switch to
branches, the repository and branch structure -- specifically,
whether to continue to have a per-release-family jdkNNu repo vs
moving to a single jdk-updates repo -- would be a big part of that
discussion. I have an opinion on that, based on experience with our
internal update releases and with the GitHub workflow, but will
leave it for that later discussion.<br>
<br>
-- Kevin<br>
<br>
</body>
</html>