Quoting surfeur51:

Quote:
I always said that real name is not birth name, but name used by the actor (actress) while his (her) career, which include stage names.

What about Zoe Saldana -> Zoë Saldana -> Zoë Saldaña then, which you chose to ignore??
It would perfectly fit as a counterargument to your thesis that "most database(s) use "real name"".

But as usual you prefer to ignore the point made, which was:
No database I know of uses the real name as a basis, except if you define the "real name" as the name under which the person first appeared in the database.

Thanks Tim.
My point exactly, you just said it better.

Quoting surfeur51:

Quote:
Not joining me is just be sure to keep what we have for years.

The problem is that you haven't ever provided any kind of solution that would even be remotely workable. As soon as I see one I'll get behind it a.s.a.p. Note that in my book, supporting a good solution does not mean trying to derail each and every common name-finding thread with endless carping about real names and accents, so you won't be seeing me doing that - but hey: that's just me. If you want to know what I'd like - here goes:

As of yet, the best idea is a single, unique key for each and every actor and crew member, a key that has all known name variants associated to it, and then when I enter, say "John Doe" into Profiler, the system would ask me "Do you mean the John Doe that's the editor on films like A, B and C, or the actor that appears in films like X, Y and Z?" That I'll gladly support, but I just don't expect it to happen anytime soon, because it means a HUGE amount of work for Invelos.

What I would also support is the "one profile in the database for each film" principle, instead of dozens of separate profiles for each of them. That way, when someone submits a common name, a typo correction, or documents an (uncredited) cast entry, EVERYONE who owns the film on DVD/Blu-ray gets to vote on it, and subsequently EVERYONE benefits from the update, instead of just the few users that happen to own that particular disc in that particular locality. Of course, there will have to be a possibility for local devations, but generally, I see this as the best way to clean up the database.

Again, though, I'm not holding my breath just yet. As it stands, I try to make the best of what we've got.

Quoting T!M:

Quote:

As of yet, the best idea is a single, unique key for each and every actor and crew member, a key that has all known name variants associated to it, and then when I enter, say "John Doe" into Profiler, the system would ask me "Do you mean the John Doe that's the editor on films like A, B and C, or the actor that appears in films like X, Y and Z?" That I'll gladly support, but I just don't expect it to happen anytime soon, because it means a HUGE amount of work for Invelos.

As other people, you ask for a complete change of the structure of the database, that we have no chance to have before years. I just ask to use correct spelling rules for each actor's name. This can be implemented today, with two lines of Ken, and would solve 90% of non linking entries. I prefer my non complete solution than nothing.

Quoting surfeur51:

Quote:
I just ask to use correct spelling rules for each actor's name. This can be implemented today, with two lines of Ken, and would solve 90% of non linking entries. I prefer my non complete solution than nothing.

I'm sorry, but please: don't be silly. Your "non complete solution" doesn't solve anything. It makes things worse. "Use correct spelling rules for each actor's name" is decidedly not a rule that helps matters forward. It's also worth pointing out what you actually mean with this little bit of vagueness - you want the rules to allow you to ANDRE as "André" into DVD Profiler, and FRANCOIS as "François". That, you believe, would solve 90% of non linking entries. Trust me: it'll only create more problems - the current system (E = e, É = é) is far easier to understand for all users around the world: it's simple, universal and fool-proof. If there's any chance of consistency in dealing with accents, then it's in the "E = e, É = é" principle - anything else is inventing data that isn't on the screen, and you'll never get the entire userbase on the same page about when to invent an accent (and which one), and when not to. As such, linking between accented and non-accented name variants will be eternally broken. "E = e, É = é" was the only chance we had at solving this, and that's why Ken ruled that way. Case closed.

So there you go - that's exactly what I said earlier: you haven't ever provided any kind of solution that would even be remotely workable.

Quoting T!M:

Quote:
the current system (E = e, É = é) is far easier to understand for all users around the world: it's simple...

1x2=2
2x2=2
3x2=2
4x2=2
5x2=2
6x2=2
7x2=2
...
is far easier to learn for all pupils around the world: it's simple...  but it's wrong...

Quoting surfeur51:

Quote:
Quoting T!M:

Quote:
the current system (E = e, É = é) is far easier to understand for all users around the world: it's simple...

1x2=2
2x2=2
3x2=2
4x2=2
5x2=2
6x2=2
7x2=2
...
is far easier to learn for all pupils around the world: it's simple...  but it's wrong...

Wow!

Quoting TheDarkKnight:

Quote:

Wow!

E=e
Wow!