Profiles that you intend to keep locally.
If you need to create a profile that you do not intend to upload to the online database, an alternate id offers no advantage over a manual id. In fact, the latter has one slight advantage - you don't risk uploading it by mistake.

Profiles that you intend to upload.
This is where I see a real problem. Let us assume that you have a disc with two movies on a single side. You have (or create) a parent profile with the UPC as id. You then create a child profile with a disc id for title A, and a profile with an alternate id for title B. So far, so good.

But at the same time, someone else has done the same thing, but used title B for the disc id and title A for the alternate id. And this gets uploaded. Now we have a mismatch. The ids for your local collection and the ids in the online database do not match. This is not a good thing! If you should mistakenly update one of your child profiles from the online, you would have two profiles for the same title. If you update both, you are in slightly better place, but any profile information or locked fields will now be in the wrong profile! Yes, you should check your updates. But if you do, and you reject the updates, then you just don't get the updated information.

Boxsets that contain the same child profile.
Again, if you intend to keep this local, you might as well use a manual id. What happens if you assign an alternate id and upload it? Well, to anyone who only downloads that boxset, but not the other, it's going to look strange that there is an alternate id where a normal disc id would have been appropriate. And again you could get mismatched ids. This would of course be less of a problem since it's the same disc in both cases, but it could still lead to some confusion.

To sum it up, from a pure database design viewpoint, alternate ids as implemented here is just a bad idea. That said, I have no suggestion as to how it should have been implemented. It's good that we can create box sets from discs with multiple title on a single side. I just wish it could have been done in a safer way. So - just be careful when you use them!

I even see people submitting "alternate"-profiles for perfectly normal, regular releases that *really* don't need (shouldn't have) an alternate, as literally nothing differs from the original profile. Inexplicably, it sometimes seems like people buy a new disc, enter the UPC/EAN, think: "hey, there's already a profile for that one, I don't need that, well, then I'll make a new one"... Or even stranger: there are occasions where an "alternate"-profile is submitted - and approved - before the original profile is submitted and approved. How can that happen? How can an "alternate" be needed if the original profile isn't even in the database yet? How does that not come up when that contribution is accepted? Example: EAN 4-020628-759117. A profile for that EAN was first submitted (by me) on October 13, 2018, but the "alternate" profile under EAN 4-020628-759117 #1 was already submitted by someone else five days earlier, on October 8, 2018. So the alternate, the #1 version, was there five days before a profile was actually submitted under the actual EAN. How can that be? How can you document the need for an alternate if no profile for that disc even exists yet? So now there are two profiles for one and the same disc, with one and the same EAN, and with absolutely no difference. It's not like one version has other artwork, not even a slipcover of some kind, it's not some kind of re-release, nothing. There really is just one version of this disc in existence. So now we have two profiles for exactly the same disc, with no reason whatsoever to justify an "alternate" profile. Yet the alternate is happily accepted into the database, and this keeps happening as we speak. This underlines that the concept isn't explained thoroughly enough, and that the validity of these contributions isn't sufficiently checked.

Quoting mediadogg:

Quote:
I left out mentioning Disc ID, because I don't understand when it is valid to be used as a variant for UPC/EAN

The answer is simple - never! Alternate UPC and alternate disc id are two distinctly different concepts that are used to solve two distinctly different problems. Simply stated alternate UPC is used at release level, whereas alternate disc id is used at disc level.

I think the rules are confusing, especially this sentence:

Quote:
Use for: Two or more releases of the same title with largely different cover images (e.g. Disc ID profiles shared among different box sets).

This mixes the two concepts into a single example. If you have two releases with the same UPC but different cover images, then an alternate UPC can be used. If you have two box sets (with different UPCs) but with the the same disc (same disc id) then an alternate disc id can be used for that disc.

One of my main objections against alternate UPCs is when it is used to differentiate releases with the same UPC and content, but just different cover images. One of the most requested features is "one film, one database entry" in order to minimize duplication of data. I support that principle, even if I can foresee some problems. Using alternate UPCs for different cover images goes totally against this principle. Here we duplicate everything except the cover scans. There should be a better solution for this.

Finally, quoting mediadogg:

Quote:
The advantage of the Alternate ID, is that that they can be cross-linked and correlated, since by definition, the base-id.locality is the same for all variants

Can be? In external programs, sure, but are they in Profiler? I'm not sure what you mean by correlated, but I don't think there is any cross-linking? Or am I missing something?