More > Media Center 17

Carnac - Optional

<< < (6/6)

glynor:

--- Quote from: struct on November 04, 2011, 03:32:33 pm ---I hope it is not exactly like as you describe Glynor.  I hope that there a way to verify also.

--- End quote ---

I was talking broad strokes, not details.  For any future automatic cloud lookup system, you'd (of course) need a way to verify the info for reasons just like the ones you described.

rick.ca:

--- Quote from: glynor on November 04, 2011, 10:10:15 am ---To confirm:
1. It only changes completely blank tags.
2. It doesn't run "all the time".
--- End quote ---

Nice summary. To add a tidbit, and expand a bit on...


--- Quote from: Matt on November 03, 2011, 05:19:06 pm ---Tag on import rules trump tags or Carnac.
--- End quote ---

The applications might be a little obscure, but this means it's possible to use the fields the values of which have been determined by Carnac (or file attributes read from the file) in a rule that will change a particular value to something else. So if there's a need, you can do things like set [Name] to "S00 E00 [Name] • [Duration] • [Date]." Yes, that includes the ability to use [Name] in an expression that sets [Name]. 8)


--- Quote from: struct on November 04, 2011, 03:32:33 pm ---I hope it is not exactly like as you describe Glynor.  I hope that there a way to verify also.  For example, I would love a fully automatic lookup based on the parsed name/year for the first run through of the couple hundred existing movies.  After that, when I am doing the one by one import, I would like to flick a switch and have some interface that lets me verify that the Notorious movie it has looked up is the one with Cary Grant and not the rapper.
--- End quote ---

Yes, this this part of the reason I believe the meta data retrieval part needs to be a separate system. It would likely still be triggered as an auto-import function (making it appear to be a fully automatic single system—in the case of new files), but there needs to be a way to run it separately. Not only will there be identification errors that need to be fixed manually (e.g., by changing a URL to point to the correct version of Notorious), but to update information that changes over time. Information normally changes frequently in the months after release. Some data changes continuously (e.g., ratings, votes, box office, etc.). I think we need a system that allows us to select and configure meta data sources, and that will run on auto-import, on a specified schedule (e.g., run this TV source daily) and on demand for selected files.

BTW, getting meta data from an external source should not be confused with Tag on Import rules. It would be awesome to be able to change data coming from an external source using an expression. This would provide a means to automatically change, format or just clean up incoming data. But that too must be configurable by source, and happen after auto-import.


--- Quote ---TV shows are a bit different, you want it to bias its heuristics to what you already have.
--- End quote ---

In the case of Series, I hope we get something with the capability (even if not to be implemented immediately) of adding information about episodes for which there are no files (like those to be released in the next few weeks—for which there is usually data). Even without that, this is another reason why we need a separate configurable system. The easiest and most effective fix for this situation is the ability to modify the configuration of a series (in a particular source) as to the [Series] name it should be matching. In other words, things would happen automatically most of the time. But, when necessary, there's a configuration dialog standing in the middle—providing control over the mapping of [Series] to configuration data (for the retrieval of data from a particular source) and the mapping of that configuration to the source record (i.e., a reference id—should [Series] not work for any reason).

Navigation

[0] Message Index

[*] Previous page