While entering video files via 'Scan your video files', if a file name is proper but detected wrong (wrong IMDb movie), there is no option to enter the proper IMDb page, at least in the scan result window option, right?.. Regardless, there's more..

For example, my file 'A Christmas to Remember (2016)' is named proper and the IMDb page is here:
http://www.imdb.com/title/tt6288876

But alas, the 'recognized' auto scan/search results only provides this result:
A Christmas to Remember (1978)
http://www.imdb.com/title/tt0077333

Furthermore, there's also this movie of the same name that is not shown in the scan result:
A Christmas to Remember (2015)
http://www.imdb.com/title/tt4067684

So why did Coollector just suggest A Christmas to Remember (1978), am I missing something here?

Now, I know one could manually enter 'A Christmas to Remember (2016)' in later, but that's not an option as it's not in the Coollector data base, yet. Regardless, 'A Christmas to Remember (2015)' is there, but that choice was not even an option, at least from within the scan results window. As such, this makes me realize I have to look at every detected result to make sure, and then some. Hmm.

That said, things being as they are with Coollector, I see no easy way to detect files such as this, but I know other software can do it better.

TIP:
In the 'Scan Results' window, maybe there should be a another tab called something like 'Same Name but Different Year(s)' or 'Not Fully Recognized'.

So, IMO this is an obvious issue, and likely a deal breaker for me.

Except for this issue, I think Coollector works great, yet because of this issue, I will keep on looking for other solutions.

Any ideas?

Cheers

_________________
There's truth is the details, interlaced with in an enigma of illusions.

Yes indeed, when you can't edit the file name to make it match, you have to mark the file as "not recognized" and at the end of the scan the program will give you the list of all the files that were not recognized so you can try to understand what went wrong. If the movie is missing you can request its addition, and if it's there you can associate it manually by clicking the "play" icon.




The 2016 movie is missing from our database so you have to request its addition:

https://www.coollector.com/viewtopic.php?f=5&t=116

Because there was no exact year match, the program picked up the 1978 movie because it's more popular than the 2015 movie.



Absolutely, it's highly recommended to check the scan results because it's the best moment to fix an error.



Me neither. If only magic existed...

In my point of view, it's not the fault of the scanning engine if this 2016 movie was missing from the database. It should have been there and I'll keep working hard to improve the database even more.



Really? Please give names.



'Not Fully Recognized' is not possible because 90% of the time the file names don't contain any year at all, and the users only add the year to solve recognition issues.

'Same Name but Different Year(s)' => in that case the program picks up the most popular result, and because it's the most popular it's most of the time the right result.



Any ideas? Honestly none, sorry. I'm very curious what alternative you will find.

I insist, please come back to tell us what program you've found that scans the files better than Coollector. It's been a while that I haven't surveyed the competition, who knows, maybe they've made huge progress...

When the file name do contain a year but there's no exact year match, maybe the file should rather be sent to the "not recognized" tab. What do you think?

Thanks for the detailed summary and fast reply.. So as you see, I actually had more than one issue. I will just focus on what I think you would be interested the most here..

As for the program just picking up that 1978 movie because it's more popular than the 2015 movie: Well, as a HUGE time saver, I think it should pick them both up, providing a list to choose from, such as this:
A Christmas to Remember (1978)
http://www.imdb.com/title/tt0077333
A Christmas to Remember (2015)
http://www.imdb.com/title/tt4067684

So, this is what Eric's Movie Database (EMDB) does rather well with its scan/search process, no magic needed. Yet EMDB has other issues I won't go into. Also, in that process, as need be, EMDB allows the user to click on the given IMDb links for further confirmation. You should try it out.

TIP: As you may know, for testing you can make dummy files via making shortcuts and renaming them with Advanced Renamer.

That said, I think this is something worth adding to Coolletor, but I leave that to you.

And yes, having another tab for this seems to be the best idea, why didn't I think of that? Ha... Furhtermore, you can even make it a beta feature for us to test:
'Not Fully Recognized (beta)'

Cheers from Caliland

_________________
There's truth is the details, interlaced with in an enigma of illusions.

Of course no magic needed since it's replaced by user decision.

I had already tested EMDb in the past, and I've tested it again to make sure. Well, I suppose it's a matter of personal taste, but I really don't plan to transform our file scanning to make it work like EMDb. It's a deliberate UI and UX choice. The philosophy of Coollector is to relieve the user from the tedious task of choosing the correct match, and 98% of the time it makes the right choice, even if it's out of luck. The few times when the association is wrong, it should jump to your eyes that it's not the right poster (no doubt that you've reacted when you've been shown the poster of the 1978 movie).

The most common reason for the scan to fail is when the movie is missing from our database. To reduce this problem, my best option is to make the database better, and believe me it has greatly improved and keeps improving with each version. Yes, Coollector failed with "A Christmas to Remember (2016)", but I doubt that you have many other examples like that. If it's a deal breaker for you, then I'm sorry and I wish you good luck with an alternative.



Please note that it's also possible with Coollector, you can click a scan result to make appear the movie details.



Shortcuts or just empty files. That's how I do unit testing of the scanning engine. I have a tool that takes a coollection and transforms it into a dummy file system which I scan, and I'm warned when the user has associated the file with a different movie. The results are impressive, that's why I doubt that you may have a lot of examples where it fails.



I didn't say it would be a NEW tab. I said the file would be "not recognized" if the name has a year that doesn't exactly match a movie. For the users who have put the effort to add the year to their file names, it could indeed be an improvement, and for the others there wouldn't be any difference.

I kind of like this idea and I should do it with the next version, thank you for your suggestion!

Let me detail what I mean...

If there's the choice between 2 movies, I prefer to choose one of them without asking the user, because there's 50% chance that I'll choose the right movie and the user won't even notice. Furthermore, if I choose the most popularity movie, the odds are even higher than 50%.

You see the difference? Instead of bothering the user 100% of the times there's an ambiguity, the user is bothered only the few times when the program has chosen the wrong movie...

You are much more likely to have on your disk a popular movie rather than a rare movie. When in doubt, the program picks up the most popular movie, without asking, because it's very likely the right choice. Yes, some errors may slip in, but they're easy to spot and as a whole the process is far less tedious this way.

These numbers you throw around seem rather silly when I have found much different results.

In fact, my first scan with CMB had way too many detection errors to deal with that way, so I ran them through EMDB and it was much easier, not a matter of taste or whatever, although the GUI for that in EMDB sucks IMO. The errors it found were that the year was off, others were not errors but rather just making sure that all the movies with same names were confirmed to be correct.

Then thanks to that help of EMDB, I scanned them again in CMD and all went smooth, except for that one file AFAIK.

As for the idea of comparing every file with the poster in CMD, I find that absurd compared to what EMDB did for me.

As for my idea of a 'Not Fully Recognized' tab, I now see you had actually ignored that, with the idea of just putting such files in the 'Not Recognized' tab.. To that, yes I think that would at least be better than what you have now, perhaps.

In the end, this is the the sum of what I'm getting from you here:

I don't take issue with your issues, CMD is fine, I just don't care to fuss over it, if you don't like it move on.

To all that, all I can say is what I do now works fine, Whereas I have IMDb links with all my files. I found a program that did that for me.

I was just looking for something better, without Kodi or Plex, which I find to be too much for my simple needs.

Furthermore, as I do tech support, folks ask me for solutions, so I like to know the options.

As it is, I guess I can't recommend any free options other than Kodi or Plex.

Cheers

_________________
There's truth is the details, interlaced with in an enigma of illusions.

You confirm that you could easily spot the errors.



You say that you "had way too many detection errors", but you should realize that those errors occurred only when your file wasn't the most popular result, and there was no problem the many other times when your file actually was the most popular result. In other words, you would have had way much more checking to do if the program had asked you every time it stumbled on an ambiguity.



The right approach is to mark the files as "not recognized" and when the scan is finished you use the list of not recognized files to understand the issues. With the list, you can use copy & paste, you can search the IMDb, you can search Coollector itself, you can go to the forum and request missing movies to be added, etc...



I could not disagree more. With a cursory look over the posters, you don't even have to think, you immediately notice if there's an intruder. In the end, I have no doubt CMD is faster and easier. IMHO, EMDB is a PITA compared to CMD.

You don't like the approach of CMD, it's your right. I don't like the approach of EMDB, it's my right.

Clearly we have different taste, different personalities. Clearly you should find another program because you'll never like my solutions and I'll never like your suggestions. We're too different. Well, I respect your difference, but you're an EMDb user, it's in your gene.



Yep, it's a good summary, but you should not be surprised, what else did you expect? In my opinion, CMD has the best approach. Probably there's room for improvement, but certainly not by mimicking EMDb. The fact that you like EMDb better is beyond my understanding. But I respect, I respect. We're not all the same. When it comes to CMD, I'm the author, you did not convince me, I have the final word, that's life. Sorry, I can't imagine myself working on a solution that I consider inferior. I've explained in great length why I considered that my approach was better, but I never said that it was better for everyone in the world. There may be some persons like you who'll prefer a different approach.



Kodi doesn't ask confirmation either (don't know about Plex, don't use it) and it's harder to fix the errors, it's slower, etc...
I don't think CMD is worse than Kodi, several Kodi users even came to me asking if I could write a scrapper for Kodi, which says a lot...

Cheers.