<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000001">
<br>
<br>
On 10/19/2011 8:11 PM, Michael T. Dean wrote:
<blockquote cite="mid:4E9F754E.9080607@thirdcontact.com" type="cite">
<pre wrap="">On 10/19/2011 11:59 AM, Brian J. Murrell wrote:
</pre>
<blockquote type="cite">
<pre wrap="">On 11-10-19 11:48 AM, Roger Horner wrote:
</pre>
<blockquote type="cite">
<pre wrap="">Since --dd-grab-all first deletes all your schedule data, if it starts just before you are about to record a program, wouldn't it cause you to miss the first 2 minutes of your program (assuming it takes 2 minutes to run as Kris B. says) since without schedule data your BE no longer knows to record it?
</pre>
</blockquote>
<pre wrap="">Hrm. Is this really true? This seems like a good case for building the
new schedule before deleting the old one and keeping the "no schedule"
time/lock to a very minimum while the already-built, new schedule is
simply swapped in and the old one swapped out.
</pre>
</blockquote>
<pre wrap="">
The only "potential danger" time comes between:
2011-10-18 06:05:46.852 Clearing data for source.
and
2011-10-18 06:05:56.909 Program table update complete.
in your mythfilldatabase log file. That is the period during which you
may be missing listings data for now. My mythfilldatabase runs with
--dd-grab-all take a total of 60s, and (as you can see, above,) only 10
seconds are in the "potential danger" period. So, for a user with a
2-minute mfdb run time, I'm guessing that the "potential danger" window
is much shorter than 2 minutes.
With a 10s "potential danger" window, --dd-grab-all isn't a concern to
me (and has many times saved recordings due to last-minute schedule
changes on the day of airing and has never caused any problems for me
since I started using --dd-grab-all on August 31, 2010 (
<a class="moz-txt-link-freetext" href="http://www.gossamer-threads.com/lists/mythtv/users/449426#449426">http://www.gossamer-threads.com/lists/mythtv/users/449426#449426</a> )).
This isn't to say it shouldn't be a concern to anyone--each user must
decide whether or not they trust it enough to run with it.
In truth, during that "potential danger" period--due to the way
mythfiilldatabase works, I'm estimating I actually have only a 5s window
of actual danger. Use the entire "potential danger" period for the
conservative estimate or about half to three-fourths of it for the
aggressive estimate (where 3/4 is more appropriate portion for much
larger "potential danger" windows).
Mike
</pre>
</blockquote>
<br>
What if you internet connection happened to be down when mfdb
--dd-grab-all ran? Would you lose all scheduling info and have
nothing till it ran again?
<br>
</body>
</html>