The switch last week from TerraMinds (the service appears to be dead) to TweetScan for the Yahoo Pipes Twitter Reply Sniffer had a couple of unintended side effects. For people with twitter account names that are also common words, the new sniffer delivered way too many false positives. I realized that TweetScan completely ignored the “@” sign, which I didn’t think was a big deal at the time. With a Twitter account like @geekygirldawn, I didn’t notice any issues; however @verso noticed. I was also having a hard time pulling the date out of the TweetScan RSS feed last week for some reason. It was obvious today, which means that they added the date to their feed, or I completely missed it last week when I was tweaking the pipe. Either are good possibilities, since I made the tweaks to the pipe in about 5 minutes while talking to a couple of Legion of Tech board members and waiting for the board meeting to start.
This new and improved Twitter Reply Sniffer explicitly includes only references to your twitter account name that are preceded by the “@” sign. For any of you who liked seeing every reference to your twitter account name, you can easily clone the pipe and remove the filter for @accountname.
I’ve also added a time stamp to the end of every title so that you can easily see exactly when each person replied. If you don’t like the time stamp, you can also clone the pipe and remove the loop right before the pipe output module.
Thanks to Ms. Fishbones for suggesting the improvements and for pointing out the typo in my blog post about the new version of the reply sniffer from last week!
I also wanted to thank Justin Kistner at Metafluence for creating the first rev of this pipe. He came up with the idea to do this and found the services that made it possible. I cloned his original version and have been making minor tweaks along the way that seem to have taken on a life of their own as things like this frequently do.
Here is the new version of the Twitter Reply Sniffer. I think that your rss reader should automagically pick up the changes if you were already using the old Twitter Reply Sniffer.
Related Fast Wonder Blog posts:
A while back, I modified Justin’s Twitter Reply Sniffer to add a little more functionality. The reply sniffer is really useful to help find Twitter @ replies directed toward you from everyone (whether you follow them or not). This is incredibly helpful for me in a couple of ways. First, it finds replies that I missed from people I follow. Second, I just can’t keep up with the noise if I follow everyone who follows me; however, I think that the conversations are one of the most important aspects of Twitter, and the reply sniffer helps me participate in those conversations.
My issue with the original reply sniffer is that it was based on TerraMinds.com, which was down all day yesterday, so I decided to modify it to use a similar service, TweetScan. Here is the link to the new and improved Twitter Reply Sniffer pipe.
Related Fast Wonder Blog posts:
I was a little bored sitting in the airport at 5am, so I thought that it would be fun to use Yahoo Pipes to help make sure that I didn’t miss any important sxsw tweets during my trip. I thought that other people might be interested in using it, too, so I built my Twitter Filter for sxsw using a user input field. You just enter your “with friends” Twitter feed URL, grab the RSS feed output, and put the feed in your favorite mobile RSS reader. Now you can make sure you don’t miss anything!
How it works:
- You enter your Twitter “with friends” feed (it defaults to mine).
- It searches the feed for these 2 keywords “sxsw” and “austin”
- It outputs only the tweets with one of those keywords
Keep in mind that you can clone this pipe and add some extra keywords. Good candidates might be “party” or “werewolf” 🙂
For those sxsw’ers on Twitter, a bunch of us are writing similar filters, so you might want to make sure you include the text or hash tag “sxsw” in your Tweets.
Related Fast Wonder blog posts:
I’ve been really excited about the potential of Yahoo Pipes recently, and as a result I’ve spent quite a bit of time playing with Yahoo Pipes over the past couple of weeks. I recently put together a pipe that I am finding really useful, and I thought a few others might find it interesting, too.
When reading my rss feeds, I tend to skip blog posts with titles that do not immediately catch my eye as something interesting. As a result, I sometimes miss important news or ideas that everyone else is talking about.
I decided to put together a pipe that takes some of my favorite blogs as inputs and sends the posts through AideRSS to find the ones with the most comments, discussion, bookmarks, etc.
- I put together a csv file with some of my favorite blogs formatted without the leading http:// to make them easier to process through AideRSS (alternatively, you could also bring in the complete URL and use pipes, to reformat the strings, but I was striving for simple). I then pulled this into the pipe as input using the Fetch CSV module.
- I then used the Loop module with an embedded URL Builder module to append the appropriate string (blog url from the csv file) to an AideRSS URL (filtering on only the “great” posts). The output from this module produces a bunch of URLs each looking something like this:
- I ran this output through another Loop module with an embedded Fetch Feed module to fetch each individual blog post from each URL built in the previous step.
- In order to filter out any duplicates, I then ran it through a Unique Filter module based on item link. You would only need this step if one or more of your original sources in the csv file aggregates feeds from other sources.
- I also wanted to limit my results to blog posts from the past 5 days, so I used the Filter Module along with the Date Builder module to restrict the dates.
- The result of the above steps gives you the basic information, but I decided that I also wanted to reformat the titles to add the AideRSS rating and post date directly into the title, so that I could easily see which ones were the most important. I used yet another Loop module with an embedded String Builder module to add additional data to the title. I then stored the output back into the item title, which results in titles like this:
Rank 10.0 1-19 This is the blog post title
- My final step was to sort the items using the Sort module to put the highest rated posts (using AideRSS rating) at the top with a secondary sort by date that puts the newest posts at the top when you have several posts with the same rating.
Viola! I have a pipe that finds the most important blog posts for me. Keep in mind that this will never help you find breaking news, since it usually takes a day or so for many posts to accumulate enough comments / links / etc. to have a high AideRSS rating, but it does keep you from missing really important news and ideas.
You can view the source of the Top Blog Posts pipe or get the RSS feed. You can also clone the pipe when viewing the source if you want to use it as a starting point for something else you want to do.
Related Fast Wonder Blog posts: