SEARCH
TOOLBOX
LANGUAGES
Create a book
Portal/WP/Cambridge SMS Issues

Portal/WP/Cambridge SMS Issues

From Steeple

< Portal | WP(Redirected from Cambridge SMS Issues)
Jump to: navigation, search

This details problems with the Cambridge SMS (http://rss.sms.csx.cam.ac.uk/) which need to be fixed in order to use it as a media source for the Steeple Portal

Contents

[edit] 1 Issues

[edit] 1.1 media:content medium is always video 

video (480x360) [http://platypus.sms.csx.cam.ac.uk:8080/649944/650064.flv] - OK
video (720x576) [http://platypus.sms.csx.cam.ac.uk:8080/sms-dspace/649944/649946.mp4] - OK
video [http://platypus.sms.csx.cam.ac.uk:8080/sms-dspace/649944/649945.aif] - not a video, should be audio
video [http://platypus.sms.csx.cam.ac.uk:8080/648376/648381.mp3] - not a video, should be audio

The type is also invalid on the audio files as it is set to: type="video/" which is not valid according to the spec.

channels is also blank which should be set for audio files.

isDefault is always set to false. It should be set to true for at least one item or completely omitted.

width and height should also be omitted for audio files rather than outputting blank values.

[edit] 1.2 media:content sizing data missing

There seems to be size info missing from a lot of videos and all streams (see example below). Every rm stream has no size data set. mms is the same. Some videos are also missing size info. 

Media: Enterprisers Highlights Promo
[http://rss.sms.csx.cam.ac.uk/rss/media/1413]
       --SUMMARY: Enterprisers Highlights Promo (Short promotional
video highlighting CfEL's Enterprisers Programme.)
       Content: video
[http://downloads.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-mov_medium_default_stream.mov]
       Content: video
[http://downloads.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-mov_small_default_dl.mov]
       Content: video
[rtsp://realmedia.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-rm_high_default_stream.rm]
       Content: video (720x576)
[http://downloads.sms.csx.cam.ac.uk/sms-dspace/118-1413-1220522319-0/118-1413-sms_archive.mpg]
       Content: video
[http://downloads.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-flv_medium_default_stream.flv]
       Content: video
[http://downloads.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-ipod_high_default_dl.m4v]
       Content: video
[http://downloads.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-wmv_medium_default_dl.wmv]
       Content: video
[mms://windowsmedia.sms.csx.cam.ac.uk/118-1413/118-1413-sms_archive-wmv_medium_default_stream.wmv]

If this data is not available then we probably will need to put in some default values when loading the videos from SMS. Can someone tell me what the defaults are or fix up the feeds to include the sizes?


[edit] 1.3 Invalid media URLs

Some of the media URLs are not functional (i.e. they return a 404). I am not sure how to handle this. We are assuming the data from the SMS is valid and if we get URLs that are invalid the user is going to see errors. I could test every URL when I do the import but this will be extremely costly and will slow down the import tremendously.

I think we need to run a process to remove invalid media URLs from the system or at least not return them in the feeds.

[edit] 1.4 No way to identify feed item

Nothing identifies what kind of data I am looking at (institution, collection, media item) It would help for incremental processing if it were possible to tell what kind of feed I am looking at when I get the XML. Right now they are all identical. I can sorta figure out the media items since they are the only ones that have media:group and/or media:content tags but that's not great either. If we could use the media-rss tag <media:category> to represent these then it would be major improvement.

Bjoern 12:09, 1 November 2009 (UTC): Possible solution: http://purl.org/steeple/feedtype.

[edit] 1.5 Empty collections

There are a suprisingly large number of collections with no media items in them. This is not really an error but I am not sure what we should do when loading them into the portal. No user wants to browse to an empty series, that will just upset them, so for now I will skip over empty collections. It might be worth purging out all the empty collections or at least not including them in the feeds.

[edit] 1.6 No Dates/Owners For Institutions/Collections

There are no publish dates or owners indicated in the feeds for institutions and collections. Without publish dates it is hard to know when these items have been updated. Without the owners or authors it is hard to know who created the content and should be the contact point for it.

Retrieved from "http://www.steeple.org.uk/wiki/Portal/WP/Cambridge_SMS_Issues"