History of TIMESTAMP on Firebird
<p>Need to dig out the notes on just how a TIMESTAMP works in Firebird. It is essentially two 32 bit numbers, one being the date and the other time which I thought was a fraction of a day, but is apparently stored as a 100mS count.</p>
<p>The date element runs from January 1,100 AD to February 28, 32768 AD. The '0' date is November 17, 1858 and caclulations on date ranges prior to that are done with -ve numbers. This is based on the VMS operating system that the code originated in. ( <a href="https://groups.yahoo.com/neo/groups/firebird-architect/conversations/messages/8148">Jim's post on that</a> )</p>
<p>http://www.firebirdsql.org/en/firebird-date-literals/ for the conversion to convential date formats ...</p>
<p>The time element was originally only returning seconds, but later developments provided the option for millisecond accuracy.<br />
<a href="https://www.ibphoenix.com/resources/documents/design/doc_169">https://www.ibphoenix.com/resources/documents/design/doc_169</a> provides the fine</p>
<p>Some more historic material will be added later.</p>
<p> </p>
<p>The date element runs from January 1,100 AD to February 28, 32768 AD. The '0' date is November 17, 1858 and caclulations on date ranges prior to that are done with -ve numbers. This is based on the VMS operating system that the code originated in. ( <a href="https://groups.yahoo.com/neo/groups/firebird-architect/conversations/messages/8148">Jim's post on that</a> )</p>
<p>http://www.firebirdsql.org/en/firebird-date-literals/ for the conversion to convential date formats ...</p>
<p>The time element was originally only returning seconds, but later developments provided the option for millisecond accuracy.<br />
<a href="https://www.ibphoenix.com/resources/documents/design/doc_169">https://www.ibphoenix.com/resources/documents/design/doc_169</a> provides the fine</p>
<p>Some more historic material will be added later.</p>
<p> </p>