[aptalca]

after trying out dtrailer in my library, I found it not so useful due to speed. The quality is ok, def better than nothing, but nothing really streams for me for more than 5 seconds, then it's buffering forever (you actually see the buffer percentage go up one by one)

I really appreciate the efforts and don't want to sound ungrateful but I think yahoo trailers by queeup or trailer addict would work much much better.

[stacked]

after trying out dtrailer in my library, I found it not so useful due to speed. The quality is ok, def better than nothing, but nothing really streams for me for more than 5 seconds, then it's buffering forever (you actually see the buffer percentage go up one by one)

I really appreciate the efforts and don't want to sound ungrateful but I think yahoo trailers by queeup or trailer addict would work much much better.

I made a trailer addict scraper a few months ago, but I had some problems with it returning wrong trailers. For example if I scraped "American Beauty" and the trailer didn't exist, then the next movie in the search results would be used. So I would end up with a "American Gangster" trailer for "American Beauty". The problem was with Trailer Addict search engine returning similar movie titles and the scraper would chose the first movie found.

Maybe I can look at it again and solve the problem. Trailer addict quality and speeds are really good but their database of trailers are not that great on older movies.

[felixxx999]

I second the motion that themoviedb.com trailers should be used in XBMC. Dtrailer is not that high-quality anyway.

Themoviedb used Youtube but many are high quality. There are hundreds and contributors are adding dozens a day. I think it'd be great if XBMC had the option of scraping either Dtrailers or themoviedb.com for trailers.

[mkortstiege]

Scraping the youtube link is not the problem at all. It is the playback as a youtube url is no video (url).

[zosky]

getting them from youTube seems best ... i think google will soon BE the internet
but like vdrfan says i couldn't find a URL for the stream.

props on the new dtrailers scraper. its finding URLs for most of my new stuff.
its been said before -- we dont want to re-import (~400 in my case) to get trailers,
so with some linux-fu i added 80% of the missing URLs into my library
i think everyone will cringe at how it all comes together, but none-the-less i made it happen

1. instal sqlite3 - $sudo apt-get install sqlite3
   
2. copy the library from my htpc - $scp [email]xbmc@xbmc:~/.xbmc/userdata/Database/MyVideos34.db[/email] ./
   
3. create the script dtrailer.sh & make it executable
   
   Code:

   #!/bin/bash

# set location
videoDB=MyVideos34.db

# find all movies missing a trailers
dbDUMP=$( sqlite3 $videoDB 'select IdMovie,lower(c00) from movie where c19=""' |sed "s/ /-/g"|sed "s/[*',.:]//g" )

for x in $dbDUMP
do
    mvNUM=$(echo $x | cut -d\| -f1) #define ID.Num
    mvNAM=$(echo $x | cut -d\| -f2) #define Name

  #Parse for a dtrailer URL
    flvURL=$(wget -q http://en.dtrailer.com/movies/watch/$mvNAM -O - | grep '<embed src="http://dtrailer.com/dplayer.swf" width="470" height="280"' | cut -d'&' -f 4 | cut -d\= -f 2)

  #test if trailer found
    if [ -z "$flvURL" ]
    then
  # scraped URL empty ... FAIL ... do nothing
    let COUNTER=COUNTER+1
    echo 'not found('$COUNTER'): '$mvNAM
    else
    let NEW=NEW+1
      echo 'FOUND('$NEW')... '$mvNUM'|'$mvNAM|sed "s/-/ /g"
    flvURL='http://www.dtrailer.com/dupload/trailerz/'$flvURL
  # drop the payload
    sqlite3 $videoDB 'UPDATE movie SET c19="'$flvURL'" WHERE IdMovie='$mvNUM
    fi
  
done
4. OPTIONAL: using sqLite browser (for ubuntu)
   open ~/MyVideos34.db ... check all changes. test some URLs
   
5. OPTIONAL: place the new & improved library back into xbmc
   

something like this (done properly) would be handy to stay current with new FanArt, Banners & such. hope this gets the great minds thinking

[aptalca]

I just updated to camelot and my tmdb scraper stopped adding dtrailer links.
Just to test, I refreshed and older movie that previously had a trailer linked, and after the refresh there was no trailer link.

Is anybody else experiencing this?
(I checked the source settings and the dtrailer option is still enabled for tmdb scraper)


EDIT: Nevermind, I just saw that in revision 26165, vdrfan fixed this.
Thanks vdrfan

[Superorb]

I upgraded to Camelot Live and trailers still don't populate in the movie info screens.

[mkortstiege]

This was fixed after the Camelot release. You need to grab the updated scrapers from SVN or upgrade to a more recent nightly build of XBMC.

[chunkhead]

Just want to keep this thread alive so more people see it and spend just a few minutes to add some trailer links to TMDB. This will be done in no time if everyone does this!

[zag]

The new version Of TMDB will also have the option for links to quicktime trailers hopefully

[aptalca]

That is the best news I heard in a long time.

Lately, (I guess ever since camelot) tmdb scraper has been linking wrong trailers from dtrailer to my library for the majority of the movies.
(Really random movies, some are black and white from ages ago)