Bruji

sjk wrote: Pool what data? Any of mine in DVDpedia that seems worthwhile enough to publicly share already came/comes from IMDb and Amazon. It's not like I've got anything new to contribute aside from bits of personal metadata of insignificant general interest.

While it is true to say that 95% of the data in my pedias is drawn from existing sources, there are many errors and omissions in those existing sources, that I endeavour to resolve - after all, the data requires validity in order to be useful to me. Since I don't completely trust any one source I always cross-reference all data. I appreciate that there will be all sorts of Pedia users using the software for other reasons, but surely I cannot be alone in wanting truthful data. I also appreciate that certain information cannot be firmly established (e.g. Theatrical Release - does one use year of completion, year of first festival screening, year it's properly distributed to theatres? In some cases all these years will be different!). However, If our data was shared, we would at least have the option to cross-reference and act upon that, whether we choose to or not.

Personally I really like the idea of sites like Discogs (and to some extent AllMovie/AllMusic.com though these sites are less user-upload-friendly) since they make use of web 2.0 user-interaction to populate their databases. Validation comes from user-enthusiasm and the fact that collectors share information and cross-reference between themselves. I have entered my vinyl albums onto Discogs myself, alongside entering them into my CDpedia, that's more than 700 entries including 60 odd original contributions. I have received feedback and additions from fellow enthusiasts about those original entries which I find useful and vice versa. However, I cannot face the idea of entering my whole music collection to Discogs (another 2000+ CDs, hundreds of singles, etc). That data already exists in my CDpedia, if I could share it, with a few keyboard strokes, that would be great. I know these sorts of facilities exist (eg genealogy sites/software offer exporting Gedcom etc).

I apologize for any ignorance I may display, though not a technophobe, I'm not up on the technological issues, I'm into databases for the data, not the technology. I also apologize if this post has mutated into an other feature request rather than extending the discussion above.

sjk wrote:

adamhh wrote:The new database idea is very interesting. I've often thought that the pool of data we pedia users hold, if it could be shared, might be really valuable.

Pool what data? Any of mine in DVDpedia that seems worthwhile enough to publicly share already came/comes from IMDb and Amazon. It's not like I've got anything new to contribute aside from bits of personal metadata of insignificant general interest.

Exactly. The whole point of DVDpedia, to me, is that I don't have to enter a great deal of information myself, but instead draw it from IMDb and, occasionally, Amazon (I could live without Amazon very easily). I'd guess that this is the appeal to 99% of DVDpedia users too. Which means, as you say, that it is entirely pointless to ask users to contribute data, other than in some extremely rare cases where very obscure movies are added which cannot be located on IMDb. This is very rare and, by definition, any such movies will be of interest to only the tiniest minority of users. If Bruji's intention was to *replace* the IMDB as a source with its own data source, then, sadly, I would be looking for new movie/DVD cataloguing software: it is precisely the ability to populate the fields from IMDb that makes me want to use DVDpedia in the first place. I would not want "user data", the accuracy of which I couldn't guarantee and which, as I said before, is absolutely certain to be of inferior quality to IMDb.

sjk wrote:So, I'm curious what a "new database" is intended to offer other than unfettered access for 'pedia software (at least), including mobile client usage. Without replication of data that already exists in other public dbs I don't see how it could be sufficiently populated to actually be useful in practice.

I agree totally with you. The stumbling block here seems to be the notion that all 4 Pedias are the same and that if *one* needs Amazon, then they *all* need Amazon. It's not so. Bookpedia, for example, definitely needs Amazon (or something similar) as there is no equivalent to the IMDb for books. But just because Bookpedia needs Amazon does not mean that DVDpedia needs Amazon. This insistence on treating all 4 programs as one is why, I assume, we can't just have a DVDpedia that ignores Amazon entirely and relies on IMDb for its main data source. I've been running DVDpedia like that anyway for months now and haven't noticed a single disadvantage. Once you see DVDpedia as separate, because of the existence of IMDb, then that paves the way immediately for an iPad version. The other Pedias could adapt the DVDpedia code if and when Amazon ever give consent to their data being used on mobile devices. Personally, and I have expereince of small enterproses trying (futilely) to persuade massive enterprises to change their ways, I can't see a resolution to this any time soon. And, from Bruji's commitment now to establishing ther own, user, databases, it seems that Bruji recognise that they won't persuade Amazon to relent. So what we will be left with, for DVDpedia at least, is either a version that uses IMDb and ignores Amazon (easily done) or a version that uses IMDb and also a "user" database (pointless), or a version that *only* uses a "user" database (no use to me at all). Seems to me the first of those three options is a doozey.

sjk wrote:Puzzled by this … what am I overlooking?

I have no idea SJK - but whatever it is, I seem to be overlooking it too. I've even been checking out other cataloguing programs that are already on the iPad - none yet approach DVDpedia for quality and usability. But I want an iPad database of my DVD collection, I know that. It's one of the reasons I bought an iPad. I just hope it's Bruji who can provide it.

Kind regards,

Keith

adamhh wrote:
While it is true to say that 95% of the data in my pedias is drawn from existing sources, there are many errors and omissions in those existing sources, that I endeavour to resolve - after all, the data requires validity in order to be useful to me. Since I don't completely trust any one source I always cross-reference all data.

The problem is that you are a rare breed of user! Most users dont require, or can't be bothered, to drill down to the level of accuracy you demand. So any data that you cull from the majority of users will be inaccurate anyway, by your standards at least. IMDb is not a perfect source, I agree, but it is easily good enough and, I contend, better than the individual efforts of ordinary users. I agree that the other Pedias have a different problem but I ony use DVDpedia and all my comments are directed at that one program in the Suite.

adamhh wrote: I also appreciate that certain information cannot be firmly established (e.g. Theatrical Release - does one use year of completion, year of first festival screening, year it's properly distributed to theatres? In some cases all these years will be different!). However, If our data was shared, we would at least have the option to cross-reference and act upon that, whether we choose to or not.

You've said it there yourself. Other users will have other ideas and so you will still have to alter the data manually. In that case you may as well get the base data from IMDb and tweak it to your liking (I do this myself). Also, there are differences caused by location. For example, in Britain we have a totally different Ratings system for movies than America (and possibly other places too). It would be of no use to an American user to import data from me that classified a movie as "Suitable for 15 years and over" as America does not have that rating. DVDpedia currently looks at the American IMDb site for data, so this is a field I always have to check manually. But it would be possible for DVDpedia to look, optionally, at other IMDb sites and draw in the correct information for a locality - that seems to me a better option that either creating a user database which would not be accurate anyway or relying in some way on Amazon, which is more or less useless as a source for movie data and is wrong more often than it is right.

adamhh wrote:I have entered my vinyl albums onto Discogs myself, alongside entering them into my CDpedia, that's more than 700 entries including 60 odd original contributions.

I admire your dedication! But I can honestly promise you that you are in a minority Most people won't go to the lengths you do - and indeed, for most people, the major appeal of DVDpedia will be that the data fields can be so easily populated from data held at IMDb. So any shared data, as SJK comments elsewhere, is fairly pointless as it is already data that anyone can access easily. The attraction of DVDpedia for me, and I suspect for thousands of others, is that it gives me a sort of personalised and enhanced version of IMDb, relating only to the movies I am interested in or own, in a self-contained way, and hopefully, soon, on an iPad. Plus the fact that I can customise the data in various ways which I cannot do if, for example, I set up a movie collection on IMDb itself - eg when I last saw a movie, when and where I bought it, if it one of my 'favourites', etc, not to mention linking my preferred review sites directly to the movie.

By all means, let Bruji create a user datavase if that is what they and some users such as yourself want. But please let this be an adjunct to IMDb and not a replacement so that those who optionally wish to ignore it or to use it can do so.

Please don't get me wrong on this: I admire your dedication to your data (even though I don't share it to the same degree) - but for DVDpedia users at least, we already have a very good data source in IMDb and so a user database would be of limited appeal to us.

Kind regards,

Keith

Keith - you are totally correct in your assertion that IMDb is sufficient for 99% of users, including myself most of the time. Plus I also agree that DVDpedia and CDpedia could stand alone without Amazon. I'm sure Bruji are listening to you on that point.

However, the base assumption that user-populated databases might be less good in some way just isn't supported by experience. When was the last time you looked at the Encyclopedia Britannica (or similar) for information? Compare that to the last time you used Wikipedia (a user-populated database). 99% of us don't contribute to it, but are happy to make use of its data, accurate or not. And on that point, Discogs is far more accurate than any other music carrier database I've seen.

I guess the important issue is to offer a range of stable data retrieval sources to users. Perhaps part of Bruji's thinking is being proactive in maintaining such a range, in case of future changes in the webscape (e.g. Amazon pulling the plug on mobile access to IMDb too). If that's the way things stand, I endorse it.

adamhh wrote:Keith - you are totally correct in your assertion that IMDb is sufficient for 99% of users, including myself most of the time. Plus I also agree that DVDpedia and CDpedia could stand alone without Amazon. I'm sure Bruji are listening to you on that point.

However, the base assumption that user-populated databases might be less good in some way just isn't supported by experience. When was the last time you looked at the Encyclopedia Britannica (or similar) for information? Compare that to the last time you used Wikipedia (a user-populated database). 99% of us don't contribute to it, but are happy to make use of its data, accurate or not. And on that point, Discogs is far more accurate than any other music carrier database I've seen.

Yes that is a good point. But I suspect Wikipedia has hunfreds of thousands of users all over the world, millions maybe, whereas DVDpedia will inevitably have far fewer so the pool of knowledge will be a lot smaller. I agree with you re the 3 other Pedias - but with DVDpedia we have a superb database in the form of IMDb and I just can't see a database put together from DVDpedia users is going to be anything like as good or as reliable. I hope I am wrong. So long as DVDpedia continues to give access to IMDb, then an additional user-generated data set can't do any harm.

adamhh wrote:I guess the important issue is to offer a range of stable data retrieval sources to users. Perhaps part of Bruji's thinking is being proactive in maintaining such a range, in case of future changes in the webscape (e.g. Amazon pulling the plug on mobile access to IMDb too). If that's the case, I endorse it.

Yes I do agree with that sentiment. If IMDb did pull the plug though I suspect my DVDpedia would become redundant overnight, although hopefully they would never do so on the desktop version of DVDpedia, and as for the mobule version, well we haven't got one anyway!
The link with IMDb is what sets DVDpedia apart from the others and what makes it so superb to use. I would never have bothered entering my hundreds of DVDs by hand and would therefore never have enjoyed all that DVDpedia brings to my hobby. I have checked out every Mac-based DVD cataloguing program (how odd that looks to have an English and an Anerican spelling sit right next to each other like that!) and nothing comes even remotely close to DVDpedia. Delicious, for example, only pulls in the very limited (*and usually wrong) information from Amazon. No other sources at all. Moviecollectorz is better but the interface is horrible. DVDpedia rules! And that is beforew we even get to mention the *amazing* support from the guys at Bruji!

Kind regards,

Keith

We want users to have the best data possible. This is why we believe in having data in open standards so that users can work with DVDpedia in a million ways, even if that involves moving away from DVDpedia. We would like to extend that now to a database not to block other sources of data but to give us some control over the format of the data to fit better with our programs, and to leverage our incredible user base. It would be wonderful if there was a reliable source of information out there that we could all share. For the next version we are looking at integrating with some of them. For example Wikipedia is actually a good source of information on movies, it's just the search API for Media Wiki that Wikipedia is built on does not provide accurate matches (it's broad as Wikipedia is meant to be). Another good newcomer is Freebase that is looking to index all data for all. It's very exciting but the international support is not as extensive at the moment. Then there are very targeted websites like TheMovieDB that provide great structured data and are built on modern technologies that make working with them a snap. The more data we have to share and to work with the better, the more users that we can introduce to a new website (data source) the better.

Conor wrote:. For example Wikipedia is actually a good source of information on movies, it's just the search API for Media Wiki that Wikipedia is built on does not provide accurate matches (it's broad as Wikipedia is meant to be). Another good newcomer is Freebase that is looking to index all data for all. It's very exciting but the international support is not as extensive at the moment. Then there are very targeted websites like [url=http:www.themoviedb.org]TheMovieDB[/url] that provide great structured data and are built on modern technologies that make working with them a snap. The more data we have to share and to work with the better, the more users that we can introduce to a new website (data source) the better.

But none of those is remotely as good as IMDb (for movies). Why make life hard?

I'm not criticising, Conor. I admire what you do and the way you do it and I absolutely love DVDpedia as you know. I just can't for the life of me see the point in all this hard work to replicate what is already out there and working just fine. I've made my point so I'll leave it now - if and when you launch an iPad version of DVDpedia I will be first in line to buy it!

Kind regards,

Keith

Thanks for all the comments, it's what makes our user community so strong.

I keep hearing how wonderful IMDB is, but so far no one has mentioned another solution that does not use IMDB. As Conor mentioned, creating our own user-enhanced database makes sense from a number of perspectives (better data control, reliability, etc) and assures we have a source that will never go away. What would happen if tomorrow we don't have access to IMDB for whatever reason?

We've spent years and countless man/woman hours working on the pedias, it would be unwise to put all our eggs in one basket that doesn't even belong to us.

That said, the idea is not to remove sources but rater add one that we (as a whole community, including our users) have better control and access to.

Alex wrote:Thanks for all the comments, …

I enjoyed catching up on the interesting followup since my reply. Wish I'd had time to participate but didn't really need to considering how well the discussion was doing without me.

Maybe this topic could be renamed to something less general than "iPad", especially since most of it's related to non-iPad-specific "Bruji app database" issues?

sjk wrote: Maybe this topic could be renamed to something less general than "iPad", especially since most of it's related to non-iPad-specific "Bruji app database" issues?

Good point. Might command a wider audience too. Something like "Should Bruji develop its own database?".

Kind regards,

Keith

I agree with the skeptics; I paid for DVDpedia, not Bookpedia or any other -pedia, so it is frankly mind-boggling why an issue with syncing to what should be a completely separate app is blocking access to what could be one of the best apps around (if it existed), DVDpedia for iPad.