Style ordered list to use nested counters.

diff --git a/index.html b/index.html
index 26f9828d09d9b5b5e655ed561f3c817698558f34..0f3a599d88346c0349d3238d50fd81a8329ef52e 100644 (file)
--- a/index.html
+++ b/index.html
@@ -314,6 +314,10 @@ SPDX-License-Identifier: GPL-3.0-or-later
                div.hex{display:inline-block;font-family:monospace;}
                div.hex::before{color:grey;content:'0x';display:inline-block;font-family:monospace;font-size:90%;left:0.5em;position:relative;width:0;}
                div.hex>input{font-family:monospace;padding-left:1.5em;}
+               ol{counter-reset:i;list-style-type:none;padding-left:1em;}
+               ol li{display:grid;gap:0.25em;grid-template-columns:auto 1fr;}
+               ol li::before{content:counters(i,'.')'.';counter-increment:i;}
+               ol li>ol{grid-column:2;}
                .warning::before{content:'\0026a0\00FE0F';}
        </style>
 </head>
@@ -333,9 +337,9 @@ SPDX-License-Identifier: GPL-3.0-or-later
        </ul>
        <p>The final score is displayed as "work-per-second" and is calculated as follows:</p>
        <ol>
-               <li>For each test run:
+               <li><span>For each test run:</span>
                        <ol>
-                               <li>Collect all the generate-validate sequence timings for the test run.</li>
+                               <li><span>Collect all the generate-validate sequence timings for the test run.</span></li>
                                <li>Truncate the results by eliminating the fastest 10% and slowest 10% of sequence timings.</li>
                                <li>Calculate the arithmetic mean (average) of the truncated results.</li>
                        </ol>