<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<p>Oh you mean the 'orange' book? Yeah, they were great.</p>
<p>Your observation that not all months have 31 days is, of course,
correct. I think what it's telling me is that DATE75 is <i>not</i>
the monitor internal format, but rather a method of encoding and
conveying the current date. Explicit, the monitor is never going
to have the DATE UUO give February 30<sup>th</sup> and 31<sup>st</sup>
and maybe not the 29<sup>th</sup>. It would be worth a look if
you know what module it's in (I don't have Tops-10 sources).<br>
</p>
<div class="moz-cite-prefix">Internally, Tops-10 is likely
maintaining two clocks, a JIFFY based uptime counter and a time of
day based on 18 bits of a day count from 1858 and eighteen bits of
a fixed fraction of day/(2^<sup><font size="-1">18</font></sup>),
which yields a resolution of about ⅓ of a second. <br>
</div>
<div class="moz-cite-prefix"><br>
</div>
<div class="moz-cite-prefix">Tops-10 is almost the same except that
it uses milliseconds as an uptime counter.
</div>
<blockquote type="cite"
cite="mid:008701d6dfca$acb48d70$061da850$@com">
<hr width="100%" size="2">
<pre class="moz-quote-pre" wrap="">On 12/31/20 6:14 PM, Robert Armstrong wrote:
Yeah, but the arithmetic doesn't work out. Between 1/1/64 and 1/4/75 are only 4021 days.
Hmmm - 31 days * 12 months * 11 years = 4092... That works out about right for 1/4/75.
The date calculation must have assumed that there were 31 days in all months.
I should dig out my old TOPS-10 "telephone book" reference manual and see what it says...
Bob
<blockquote type="cite"><hr width="100%" size="2"><pre class="moz-quote-pre" wrap="">On 12/31/20 6:05 PM, Thomas DeBellis wrote:
For 12 bits, the largest date comes out to 1/4/1975 </pre></blockquote></pre>
</blockquote>
</body>
</html>