Re: [xmltv-users] UK Atlas load/usage warning email
Ian Campbell <[email protected]>
| Newsgroups | gmane.comp.tv.xmltv.devel |
|---|---|
| Message-ID | <[email protected]> |
On Fri, 2016-05-13 at 01:27 +0200, Karl Dietz wrote: > On 12.05.2016 16:15, [email protected] wrote: > > On Thu, 12 May 2016 14:03:41 +0100, Ian Campbell wrote: > > > >> I'm asking whether this sequence (which corresponds to the loop I > >> originally posted): > >> tv_grab_uk_atlas --days 1 --offset 0 > >> tv_grab_uk_atlas --days 1 --offset 1 > >> tv_grab_uk_atlas --days 1 --offset 2 > >> tv_grab_uk_atlas --days 1 --offset 3 > >> ... > >> tv_grab_uk_atlas --days 1 --offset 10 > >> > >> will produce a data series which fully covers the period from > today (at > >> some point, either now, previous midnight, previous 2300, it > doesn't > >> really matter) until roughly ten (or 11) days time, without any > gaps > >> (overlaps are fine). > > > > Ok. Yes it will. > > This is part of the "baseline" interface contract and being > automatically tested for. > See http://wiki.xmltv.org/index.php/XmltvCapabilities Perfect thanks. > > Although as a general rule one shouldn't ask for 10 days' data > every day. I usually get 'today' + the next 2 days to catch any last > minute changes to schedules, plus the 'new' day in the future (e.g. > day 10). i.e. I get days 0,1,2,10 (in two passes). The schedule in > between rarely changes and so it is not necessary to request days 3-9 > every day. This also helps to reduce the load on the Atlas server. > > Which is in principle what is the default behavior of MythTV's > mythfilldatabase. My setup predates this capability and I've never switched from the old cronjob -- partly because I also run an XSL transform on the data, but mainly because it if it works don't touch it. The current situation may be the catalyst for change, especially if it ends up raining over the weekend ;-) > It has a loop running from today until it gets no more data or hits a > limit, one day per iteration. > Tomorrow is always downloaded > and every day that has less then 90% of channels with guide data > compared to the day before (relevant e.g. when some channels provide > only 7 days of data while most provide 14 days.) > > That leads to grabbing tomorrow, the seventh and the fourteenth day. > (Days vary depending on the available guide) FWIW based on the files produced by my manual day-by-day loop (I tee each invocation to a file as well as feeding it to myth) it looks like for my line up (Freeview) atlas has full data up to day 10 or 11 and then channels start to drop out -- that's based entirely on the size of the daily files, not on any actual analysis so take with a pinch of salt. Net result is I'm now fetching days 0, 1, 2 and 10. Ian. ------------------------------------------------------------------------------ Mobile security can be enabling, not merely restricting. Employees who bring their own devices (BYOD) to work are irked by the imposition of MDM restrictions. Mobile Device Manager Plus allows you to control only the apps on BYO-devices by containerizing them, leaving personal data untouched! https://ad.doubleclick.net/ddm/clk/304595813;131938128;j