2013-07-23, 10:28
There is no such thing as 100% accuracy when doing a mass-fetch, at best I can choose to skip an item when there's more then a certain number of results, leaving you to manually fetch information for those items, but the 'first=right-choice' ratio is a little higher than the 'only-one-result', that is to say, a lot of movies that most likely would have been guessed correctly might then end up not being automatically fetched information for at all.
Personally, I always try to fetch manually, since I can choose the correct result, it does have a 100% accuracy, and i can look for better images right away if the defaults don't suit me.
But then, I'm not working with 1000+ movies at a time...
1) if the information is showing in the main display upon selection, the .nfo is in there somewhere, otherwise maybe ViMM didn't have write permission to that folder?
2) I do have a little bit of hope on enhancing this in the future by giving a list of movies that had more then 1 result, with menu's to pick the right result from, but it wouldn't be all that different as to what i mentioned above.
Disabling automatic ordering and doing a 'rename' > 'all' after you're done sorting things out is probably the easiest way to get around this for now, you can also leave the movie file pattern field empty so as to at least not rename the movie file.
3) if it's an empty file with nothing but a straight link to the correct movie on IMDb, it should work, however it may be easier to select the item in ViMM, choose the 'Manual Search' button from the toolbar, enter the right title or IMDb ID in the search field, and once you've found the right title, press 'OK' and choose to replace existing images if needed.
4) When ViMM searches for a movie, it sends a search request to TMDb, which returns to ViMM a list of results ordered by relevance, which means that the first item in the list is what TMDb thinks is the most likely correct title. During a mass-search, ViMM ignores all the other titles and chooses the first result to fetch information.
TMDb has a relatively high chance of getting the correct result if the folder name includes both the title and year, the TVDB has a bit lower rate of success, so in it's case I do try not to fetch information if there's more then one result, since an auto-rename on a tv show with many episode files is even more disastrous than with a movie.
But any automated process is prone to mistakes when trying to match a real person's question to a computed correct answer, I can't do much to improve this, it's the way computing works. ^^;
With kind regards,
Vidal van Bergen.
Personally, I always try to fetch manually, since I can choose the correct result, it does have a 100% accuracy, and i can look for better images right away if the defaults don't suit me.
But then, I'm not working with 1000+ movies at a time...
1) if the information is showing in the main display upon selection, the .nfo is in there somewhere, otherwise maybe ViMM didn't have write permission to that folder?
2) I do have a little bit of hope on enhancing this in the future by giving a list of movies that had more then 1 result, with menu's to pick the right result from, but it wouldn't be all that different as to what i mentioned above.
Disabling automatic ordering and doing a 'rename' > 'all' after you're done sorting things out is probably the easiest way to get around this for now, you can also leave the movie file pattern field empty so as to at least not rename the movie file.
3) if it's an empty file with nothing but a straight link to the correct movie on IMDb, it should work, however it may be easier to select the item in ViMM, choose the 'Manual Search' button from the toolbar, enter the right title or IMDb ID in the search field, and once you've found the right title, press 'OK' and choose to replace existing images if needed.
4) When ViMM searches for a movie, it sends a search request to TMDb, which returns to ViMM a list of results ordered by relevance, which means that the first item in the list is what TMDb thinks is the most likely correct title. During a mass-search, ViMM ignores all the other titles and chooses the first result to fetch information.
TMDb has a relatively high chance of getting the correct result if the folder name includes both the title and year, the TVDB has a bit lower rate of success, so in it's case I do try not to fetch information if there's more then one result, since an auto-rename on a tv show with many episode files is even more disastrous than with a movie.
But any automated process is prone to mistakes when trying to match a real person's question to a computed correct answer, I can't do much to improve this, it's the way computing works. ^^;
With kind regards,
Vidal van Bergen.