Punctuation Problem in Title/Description
Posted 20 December 2010 - 10:54 PM
I don't know if this is considered a bug but something in stripping out punctuation and other desired characters from the Youtube video titles and descriptions. We have text like "W.T. White School" that gets changed to "WT White School" and "http://ow.ly/3lz0v" changes to "ow.ly" ...I suspect you're doing something intentional to strip these out but I need to know how to turn it off or remove it.
Posted 21 December 2010 - 04:00 PM
TubePress doesn't do any kind of filtering on the video descriptions. The only exception is if you choose to limit the description length, TubePress will trim it.
Could you send along a link to a video that has this issue? I'd be interesting in seeing the raw data from YouTube. Thanks, as always!
Posted 21 December 2010 - 05:04 PM
Something in Tubepress is filtering, and not consistently, and you can see it here:
You'll see the videos in the grid all have W.T. White altered and the top video description only shows ow.ly
Now click on the first video in the grid (same as the main video) - the description now shows http://ow.ly/3lzNJ
Now click on one of the videos with WT White in the title. You'll see the title changes to W.T. White. Something is stripping out these things and it seems like it must be in Tubepress.
Posted 23 December 2010 - 04:41 PM
I did some investigation locally using the same videos on your site. As it turns out, YouTube seems to strip out some of the punctuation in the description of the video thumbnails, but not the video itself (as you also discovered). To see it for yourself, you can put your gallery into debug mode (http://gdata.youtube...er...e&format=5
What's not clear is why YouTube is doing this. I did some searching on the YouTube API forum as well as in their issues list, but couldn't find anything related. YouTube has always made it clear that what shows up on youtube.com is not an exact mirror image of what you see via the API. This must just be another example of the disconnect between the two.
In short, I'm not sure what we can do about this, as TubePress simply uses what it gets from the API. If you or your developers have any ideas, I'd be interested in pursuing them..
Posted 24 December 2010 - 10:28 AM
I did some investigation locally using the same videos on your site. As it turns out, YouTube seems to strip out some of the punctuation in the description of the video thumbnails, but not the video itself (as you also discovered).
Actually the description I'm using is from the thumbnails meta information - you can see the code fragment on this in the post I made on how to get a description for the video. Insofar as I can tell this has more to do with selecting a specific video to play via passing the video id in the URL. Which suggests a Tubepress issue I think.
Posted 05 January 2011 - 10:42 PM
Posted 10 January 2011 - 06:53 PM
Notice that by visiting http://i.imgur.com/yX3gn.png. We see the same missing punctuation from the second video titled "From hardship to success: WT White student Lester McGowan". Let's click on the second video.
Now the address bar reads: http://i.imgur.com/vFuki.png. There are a few things going on here:
- The video is duplicated in the thumbnail gallery. This is a known bug and will be fixed in the next release of TubePress: http://code.google.c...s/detail?id=248.
- The first thumb for the video in the gallery, along with the main video, has the correct title of "From hardship to success: W.T. White student Lester McGowan" (with all the punctuation)
- The second (duplicate) thumb in the gallery has the punctuation wrong again.
So the question is, why does the thumbnail have the correct info only when it's in the first slot of the gallery? The answer lies within the way that TubePress interprets the "tubepress_video" query parameter (e.g. ?tubepress_video=j2GwjgwmToI). If TubePress detects this parameter, it assumes that we need to go fetch the video and add it to the gallery. Currently it does not check to see if the video is already in the gallery (hence the duplicate video). Now, when TubePress fetches the info for this single video, it uses a different API call to YouTube. It's this different API call that produces different results with regards to punctuation.
To illustrate, let's look at all the API calls that TubePress makes to generate the page of http://gdata.youtube...e...e&format=5. Looking at the output of this, we see the wrong punctuation:
... <title>From hardship to success: WT White student Lester McGowan</title> ...Then, TubePress fetches the single video's information via the following URL: http://gdata.youtube...de...J9QO_QppMg. The output of this call shows the correct punctuation:
... <title>From hardship to success: W.T. White student Lester McGowan</title> ...Why the difference in punctuation? Only YouTube knows. It's clearly a bug on their end, and I'm not really sure what else TubePress can do to work around this. We don't want to make a network request for each individual video in the gallery, as that would make the response time unacceptable. If you have ideas, I'm definitely open to hearing them! I hope this clears up the issue, and I'd be happy to continuing the discussion with you. Thanks.
Posted 12 January 2011 - 08:52 PM
I'm glad you'll fix the duplicate thumbnail bug ... but PLEASE in that update also add the metatags for the currently playing video's ID tag and description so I can eliminate my workaround for this. Feel free to add everything else I've ever suggested as well. You know I'm completely genius on all this.
Thanks again Eric.