[mythtv-users] Zap2It needs our help

[Jason Keirstead]

Chris Petersen wrote:

>Request a listing?  It's not just a list of channels for a particular
>region/provider.  My listing has a bunch of channels removed (HSN, ESPN,
>etc) that I will never watch.  The only thing the request says is "hi,
>I'm xris, give me lineup number 1"..   Even if you were proxying the
>entire request, you'd end up storing EVERYONE'S personal lineups, since
>the specific data isn't passed back and forth.
>  
>
So you have several channels removed. Those would not be cached when you 
make your request then. When other guy uses the proxy, and he
is in your  region and using your provider, if he needs those channels 
then his lineup is downloaded. If he does not then it is not.

The point is you can still cache, because as soon as you get one guy who 
wants *all* the channels, no one in that region / provider needs to 
fetch all that data anymore, since it is all cached locally. I know from 
previous requests what channels you want, so I only send you those.

I don't think people are thinking through how easy it would be to store 
and correlate this data. It's not like you have to save it on a 'by 
lineup' basis.  You can save it on a 'by channel by provider' basis., 
thus if I am showing only channels 15-40 of a provider's 60 channel 
lineup, and you are showing channels 30-60, then when the third guy 
comes along requesting channels 10 - 50, it can*all* be retrieved from 
the cache.

>Oh, and the whole point of connecting datadirect and downloading stuff
>daily is because stuff changes on a daily basis.
>
I don't see what this has to do with anything. If 300 people are using 
the proxy, it doesn't matter if  they all request the data every day.  
You could have the proxy expire the data every 12 hours even and it 
would still save zap2it tons of bandwidth and requests.

>And if you're
>only saving it for a day, then all you've done is become a middleman for
>the download, and wasted bandwidth by proxying something I could just as
>easily have downloaded directly.
>  
>
Wasting whose bandwidth? *You* are using the exact same bandwidth. And 
in the absolute *worst* case ( every person on the proxy has
different non-overlapping channel lineups ), zap2it is using the same 
bandwidth it would anyways. But in the *average* case, the proxy would 
be saving zap2it a ton of hits and bandwidth.

I really don't get why everyone is making such a big deal about this - 
caching proxies are simple to implement. It's not rocket science.