[mythtv-users] Guide updates for CBS and TNT this evening

[Gary Buhrmaster]

On Tue, Nov 17, 2015 at 4:08 AM, Robert Eden <rmeden at gmail.com> wrote:
....
> Checking right before prime time won't give better data.

Actually, by your own statement, this is not true, because
you see peeks at 2am(*), so checking at (say) 6pm local time
*may* give those people better data.  So, I think what you
perhaps meant to say was that IFF your suggested next time
was guaranteed to always be a good random time after the SD
next future data feed update has completed (regardless of
any issues on your end that may cause delays (**)), and
before the data was to be utilized for scheduling(***), then using
a suggested time may be best for Schedules Direct and the
customer.

I am not currently suggesting that the recommendation to
use "next suggested time" may not be good, only that the
absolute statement is demonstrably false, which means
any/all statements representing SD should be reexamined
in that light.  You did Schedules Direct a great disservice
here.

Gary




(*) It is not clear if by 2am you mean some local time, or
UTC.  If UTC, that is after prime time starts for (say) the
the east coast of the US, and would likely miss any recent
updates.  If you meant local time, then individuals would likely
miss "todays" updates which happen "late morning" (local
time or UTC)?  As someone whose customers cross time
zones, I would have expected that you would know to always
specify the time zone in communications to eliminate the
ambiguity and confusion that is introduced by not doing so.

(**) If your internal systems fail to complete the downloads
and back end processing, do the codes return a next
scheduled update time to be in an hour (or so) rather than
"tomorrow" (so that users can retry when you have completed
your fixes/retries/recoveries), or does the next suggested
time just move to the next window, regardless of whether
a confirmed and validated download has been made.  I do
recall that there have been cases of issues (between
Gracenote and SD) which delayed completion of the
download/reformatting of the data, and I do not remember
seeing a next suggested time in just a few more hours to
adjust for that.  I would appreciate confirmation that there
is no conceivable code path such that the next suggested
time would ever miss an update, and would be within (say)
two hours of the data being available.  Since I have seen
variances of the next suggested time in excess of two hours,
I am skeptical that the recommendation is actually good for
the customer if they want the most accurate data as soon
as it is available (it might be good for SD, but that is not
relevant for the customer).

(***) Which, of course, can be "prime time" or any time.
I do not think SD should be in the business of telling people
exactly what shows they are allowed to consider important
(some people want their "soaps", or "Dr. Phil", or "SpongeBob",
and others only care about the Big Game).