Over the last week or so, exams aside, I’ve been trying to get the future ENTV site online. Drupal is the backbone as it’s something we at the NSS already use and is familiar both to myself and the Captain.

Drupal is, well, mostly good. There’s plenty it does quite well and plenty it utterly sucks at. For example, I can really easily add modules to it that do all sorts of things. Most of them not that well.

It does support blogs, feed aggregation and many other things, but most of them in the same way Linux supports high-end video editing. Yeah, it can be done, but it’s not easy or pleasant.

So, next comes the site construction. Thankfully it’s increasingly becoming the case that I don’t need to build things from scratch. Adobe Flash makes embedding video or audio a doddle, Feedburner and RSS feeds makes it easy to embed blogs and then there’s a few other things.

The other things are the problem. Twitter is causing me no end of headaches. I currently don’t use Twitter, although I may in future (if I ever get this bloody site working) but I can see a use for it in ENTV. Whilst I can’t imagine our viewers and society members give a rat’s arse about what Chris is having for lunch (soup, incidentally), I can see them being interested in the progress of the week’s episode or what a particular channel is up to.

In order to create Twitter feeds for any given topic we’re using “hash tags”. Any time someone “tweets” something with a hashtag in, it can be grouped and syndicated. In plainer English that means any time any of us are talking about the society in a twitter update, if we append #entv it’ll appear in the #entv hashtag feed.

So on each channel’s page One would see something like this:


Twitter updates


About the segment

To put that into an example:

A Grand Day Out

#agdo twitter feed

Feed from Blogger account

Blurb about A Grand Day Out


The current stumbling block is the Twitter feed. What I need to achieve is as follows:

  1. Pull the latest, say, 10 messages from Twitter’s Atom or RSS feed.
  2. Parse the XML and grab the description text, author and time of posting
  3. Output those three variables for each into a list in HTML

Depending on your perspective and knowledge, that might sound tough or easy. From what I understand, it should be a walk in the park for any vaguely competent coder. Well, let’s not beat about the bush – any coder brighter than a potato should be able to hack together something rough to do the job.

Sadly I’m not a coder, I simply cannot get my head around PHP, JavaScript or even CSS to anything more than a meagre degree.

Twitter provide “badges” that do a reasonable job of syndicating a user’s (*shudder*) “tweets”. Yeah, great. So why are there none for hashtags?

You see, the issue with most feed aggregation scripts I can get my hands on is this – they don’t allow me to choose which variables to output. Instead they try to be helpful by throwing away all the options I might want to tweak. Just because Apple do it doesn’t mean everyone else should.

Twitter messages are so short that both the title and description are the same text, the title simply has the addition of linking back to the original page. Given that twitter posts often contain links, this would mean our users would have to first click on the link, get taken to a different site and view the message in its natural habitat, then click the link there. Additionally, perhaps Twitter uses a non-standard “author” variable, but I’m fairly sure it doesn’t. Apparently most aggregation script writers don’t care who wrote the items they’re helping the syndication of.

Feedburner was able to do a reasonable job and for a short while I was happy. Then I realised that it would only update every half hour or so, sometimes longer. Given the immediate nature of Twitter, that was of no use to me. When a message is sent it should appear. Either instantly for preference or within five minutes.

So, this leaves me trying in vain to find a coder who can bash out a script for parsing Atom/RSS into HTML. Please, someone – help me.

0 responses to "Aggravation At Attempting Aggregation"

Leave a Reply