[mythtv] [mythtv-commits] mythtv commit: r24980 by robertm
robert.mcnamara at gmail.com
Mon Jun 7 18:46:14 UTC 2010
On Mon, Jun 7, 2010 at 11:22 AM, Scott Russell
<list-mythtv at bluecamel.eml.cc> wrote:
> On Jun 7, 2010, at 1:04 PM, ryan patterson wrote:
>> On Mon, Jun 7, 2010 at 12:13 PM, Robert McNamara
>> <robert.mcnamara at gmail.com> wrote:
>>> I am unlikely to ever support an option to go back trough and
>>> re-re-re-grab metadata from our sources. It's disrespectful and
>>> unfair to constantly be pulling the same metadata over and over. I
>>> personally feel like it's selfish and unfair to the unpaid, free data
>>> sources we use. . .
>> I don't entirely disagree with your position.
>> In my case I was usually grabbing new metadata that I had added to
>> themoviedb.org. So I wasn't just hammering their servers over and
>> over. But that would be the result of an automated "check for new
>> metadata" job.
> And it should support partial metadata. Often all the data isn't needed, just the missing art, description, etc. Having to clear the entry and then grab new data just to check if partial data is now present isn't exactly friendly to the end user or the services.
> -- Scott
I hasten to express what an unmitigated pleasure it is to get through
with a project like this, only to be met with a list of demands about
how it should work. You download all the XML data one way or another
when you do an "update," the source has no concept of a per-attribute
update. It's all or nothing. So yeah, I'd rather people have to make
a choice to do it, and I'm not out to design a mechanism to make it
easier for us to run the data sources into the ground.
A week ago there was no such thing as mass metadata update in
MythVideo. Now there is, and you can pick your artwork. I'm sorry it
doesn't meet with your requirements, but to *me*, it was the correct
balance between convenience and respect for the providers of data.
And I'm the one doing the work.
More information about the mythtv-dev