diff --git a/index.html b/index.html
index 5acce5eb..8b1cec53 100644
--- a/index.html
+++ b/index.html
@@ -3043,9 +3043,13 @@ <h3>Algorithm</h3>
               element from the value of <code>@preserve</code> as <var>value</var>.</li>
             <li>Initialize <var>containers</var> to an empty <a>array</a>. This
               <a>array</a> will be used to keep track of an ordered list of
-              preferred <a>container mapping</a>
-              for a <a>term</a>, based on what is compatible with
-              <var>value</var>.</li>
+              preferred <a>container mapping</a> for a <a>term</a>,
+              based on what is compatible with <var>value</var>.
+              <div class="note">
+                Algorithm steps may append the same value to <var>containers</var>,
+                but the order in which they are added is significant for choosing the most appropriate term.
+              </div>
+            </li>
             <li>Initialize <var>type/language</var> to <code>@language</code>,
               and <var>type/language value</var> to <code>@null</code>. These two
               variables will keep track of the preferred
@@ -5559,16 +5563,11 @@ <h3>Security Considerations</h3>
 <section class="appendix informative preserve">
   <h2>Open Issues</h2>
   <p>The following is a list of issues open at the time of publication.</p>
-  <p class="issue" data-number="3">Provide a way to compact <a>terms</a>, but not <a>value objects</a> or <a>node references</a>.</p>
   <p class="issue" data-number="4">Allow <code>@type</code> to have a <code>@container</code> specification, such as <code>@set</code>.</p>
   <p class="issue" data-number="5">Define a "streaming profile" for JSON-LD to help parse data from a stream, rather than require the entire document to be in memory. Also to generate documents that can be so streamed. This would aid in using JSON-LD as a dataset dump format where there are a very large number of quads.</p>
-  <p class="issue" data-number="6">Provide a way to not compact <a>node references</a>.</p>
-  <p class="issue" data-number="8">Remove requirement that dictionary memberes to either processed in lexographical order or generated in lexographical order.</p>
   <p class="issue" data-number="14">Problems importing contexts which reference other common contexts and result in an error.</p>
-  <p class="issue" data-number="22"></p>
-  <p class="issue" data-number="25">Currently, the algorithms are quite proscriptive on how blank node identifiers are named, principally to make it easier to test the results, however for some things this becomes difficult or impossible, and inhibits algorithms from being conformant based on a strict textual/structural validation of the results.</p>
   <p class="issue" data-number="26">For <code>@graph</code> containers, use the generated graph name identifier as the default subject for nodes in that graph.</p>
-  <p class="issue" data-number="27"></p>
+  <p class="issue" data-number="33">Option to specify level of compaction for literals &amp; resources.</p>
 </section>
 
 <section class="appendix informative">