Quoting Prof. Kingsfield:

Quote:
You and your crap are one the big reasons I don't Contribute.

Ooh, that makes me feel all warm and fuzzy inside.     

Quoting lyonsden5:

Quote:
He has said more than once (and I'm sure you know this since you've been around for quite some time) that he has to balance perfection with user friendliness. He even stated he has been involved in software that demanded so much from the users it failed (not a direct quote).

Here's the quote:

Quoting Ken Cole:

Quote:
I will say that there's no way this database will ever be perfect to everyone's definition, nor do any of us really want that.  A system designed to track every nuance completely correctly would be cumbersome in the extreme.

DVD Profiler must walk a fine line between enough accuracy to keep one side of the aisle happy while maintaining ease of use and entry to keep the other, generally somewhat silent and large majority happy.  My development experience is not trivial, and I've been forced to design systems where one person's concept of accuracy has driven the project into the ground, completely unusable to the end users.  I don't plan to make that mistake here.

Quoting T!M:

Quote:
Quoting lyonsden5:

Quote:
He has said more than once (and I'm sure you know this since you've been around for quite some time) that he has to balance perfection with user friendliness. He even stated he has been involved in software that demanded so much from the users it failed (not a direct quote).

Here's the quote:

Quoting Ken Cole:

Quote:
DVD Profiler must walk a fine line between enough accuracy to keep one side of the aisle happy while maintaining ease of use and entry to keep the other, generally somewhat silent and large majority happy.  My development experience is not trivial, and I've been forced to design systems where one person's concept of accuracy has driven the project into the ground, completely unusable to the end users.  I don't plan to make that mistake here.

What do you have... "Ken Quote Profiler"? 

Thanks (again) for the link 

(glad to see I was close with what I said he said  )

Quoting CharlieM:

Quote:

While, I do agree, that a program for the general population that is overbearing on accuracy is a bit much, I feel, as far as linking goes, we are way to far on the other end.

Could be. Not sure if I agree it's too far as you say but I will admit it does lean that way.

Quote:
I do understand about contributing, cast is relatively easy, as long as I keep my 10 thumbs away from the keyboard, crew can itself be interesting (even for the small amount of crew that we actually track).  Does this belong when it says this or that.  Sometimes pulling the correct information is a mess.

Quote:
A DB is only as good as the accuracy that lies within. The less accurate the less usable.

I guess that depends on what your after in the db. In principle I agree with you. We don't know that Ken's goal is the most accurate db though. The main page says:

One-step entry by UPC via our massive online database

Looking further you will find:

With an enormous community of dedicated users worldwide, DVD Profiler has become the most compelling source of DVD information anywhere.

Massive and compelling seem to be what he is highlighting as a feature, not the most accurate.

Now before everyone attacks and says I want quantity over quality keep in mind I am not stating what "I" want. I'm just pointing out what Ken feels the selling points are to his software.

Quote:
While I would contend that the Cast and Crew DB's are getting better (still a long way to go), the linking system has been going backward.

I think there are those here who would disagree with you. I see progress being made.

Quote:
When we can arbitrarily link an 80 year old with a 30 year old, without question, then the data is no good to anybody.

I guess it depends on the overall numbers. If blind acceptance will correct most of the issues but fail on a few 80 year old / 30 year old people then overall it is an improvement. I don't know the statistics of just ho many of the type you refer to there actually are. I do know the majority of those I check (and yes, I do check them most of the time) are correct. Using random numbers as an example if corrcting 10 gives you 2 new errors then it's a win to me since overall 8 were corrected. Again, I don't know the numbers to see how much of an issue this is. If of course blindly correcting 10 gives you 7 new problems it not worth it (although you're still ahead by 3  )

Quote:
A fine line is true, but the proper place must be found for that fine line.

HEY! We ageee! 

Quote:
Maybe what needs to happen, is the linking system needs to be scrapped, until a more user friendly system is designed.  What that system is, is really hard to say, but the way the current one is implemented leaves a lot to be desired.

Once again we're back to Ken. 

It is nice actually discussing some of this for a change without my blood pressure going up  . Believe it or not I enjoy a good, rational debate. 

Quoting T!M:

Quote:
Quoting lyonsden5:

Quote:
Quoting CharlieM:

Quote:
While I would contend that the Cast and Crew DB's are getting better (still a long way to go), the linking system has been going backward.

I think there are those here who would disagree with you. I see progress being made.

I see progress being made, too. Sure, it's not perfect (because we're still stuck with a sizeable chunk of IMDb-mined data and because we're having to deal with incorrectly formatted titles, incorrectly entered production years, missing original titles, different and translated season and disc indicators for TV-show-profiles, people with the same names and so on), but I certainly do see things improving step by step. The endless carping by one or two individuals doesn't change that. Sure, there's still lots of work to do, and sure, mistakes will be made, but all in all, I do honestly believe things are slowly (let me stress slowly) improving. Yesterday we established Mr. Wisher's correct common name, and today we're well underway establishing the correct common names for Obba Babatunde, Doug Meerdink and Andre Maranne. And so things get a little better every day. Again, I'm not saying it's perfect, but I wouldn't say it's going backward.

The problem is that while your personal database may be slowly getting better, and perhaps those who just happen to read these forums  as well, the vast majority of the users out there are stuck with a system which requires them to fix each and every profile themselves, unless of course they blindly accept updates (which in my opinion would not make their local data better).

This is not a system (the "Credited As" system, that is) that will ever yield the level of linking that I believe is deserved from this application.

You do remember what the definition of insanity is, don't you??? 

backward may have been a little harsh.  I do like the term slowly.

There still must be a more user friendly approach.  To be honest, it probably is not the linking system itself that needs help, it probably originates from the crew and cast DB.

Right now, we do not have a proper way of individually identifying people.  BY's are only good when two of the same name, and only if we can find the BY's.  We actually need a way to distinctly ID people.

I don't mean to sound harsh, and as long as names are not too common, we will be able to move forward, as people correct profiles.  It is a lot of the names that have 3 or more people that are associated.  we need to be careful moving forward.

Ken brought all of this on with his wishy washy attitude in his comments as discussed already.

But to say the CLT results are totally worthless, wrong example it tells us what the user looked at, as in the Wisher case, the user only looked at Wiiliam Wisher and William Wisher, Jr. and completely ignored the other possibilities which change the entire result.

Skip