=============================================================== From: William Wade ------------------------------------------------------ Phil, How do I get an RSS/Atom feed of just your articles? I tried the author page but the only feed marked there goes to nothing...
=============================================================== From: David White ------------------------------------------------------ Phil, I just read through this article, and want to ask a question: What do you consider the difference between metadata and hash tags? Why can't the existing infrastructure of hash tags perform the function you are describing as needed? I think you have a very good point though. I, for one, use Twitter a good bit. But the thing that really gets me going into dozens+ of tweets & retweets a day is when some sort of emergency or natural disaster occurs (for example, I would almost pity the follower uninterested in the April tornadoes). I use hashtags in almost all of my tweets, and I think that something like this could be implemented very easily with hashtags - if a utility isn't already out there (which I doubt - Tweet Deck or Hoot Suite, anyone?). Just a few observations of mine. But all in all, I like the article, and agree with the sentiment.
=============================================================== From: Aaron Welch ------------------------------------------------------ Metadata goes beyond what is referenced in hash tags. Time of day, occuranc= e, relevance, and context can all be additional metadata. -AW onsider the difference between metadata and hash tags? Why can't the existin= g infrastructure of hash tags perform the function you are describing as nee= ded? it. But the thing that really gets me going into dozens+ of tweets & retweet= s a day is when some sort of emergency or natural disaster occurs (for examp= le, I would almost pity the follower uninterested in the April tornadoes).=20= this could be implemented very easily with hashtags - if a utility isn't al= ready out there (which I doubt - Tweet Deck or Hoot Suite, anyone?). gree with the sentiment.=20 : ge but the only feed marked there goes to nothing... latter part of the blog post describes some FOSS stuff yk=20
=============================================================== From: Ryan Bales ------------------------------------------------------ It seems like a pretty complex domain to reason about without any end-user intervention like managing complex meta hash tags in each tweet. I don't think it's impossible, but it presents a pretty interesting challenge. ~Ryan Bales
=============================================================== From: Chad Smith ------------------------------------------------------ The date, time, and tweeter are all already recorded. The Hashtag adds the category. I don't think Twitter wants or needs anything to make it more complicated to tweet. Having all twitterers everywhere go through more and more complicated steps just so you don't have to filter your incoming tweets - not going to happen. Most tweeters tweet via text message - i mean that's where the whole thing got started anyway. Adding metadata via text would suck. You are over thinking things. Just let it be. Twitter needs the noise to survive. They aren't going to add features which makes 80% of their tweeters stop tweeting. *- Chad W. Smith* *"I like a man who's middle name is W."* President George W. Bush - February 10, 2003 bit.ly/gwb-dubya
=============================================================== From: Dan Lyke ------------------------------------------------------ On Fri, 27 Jan 2012 21:26:14 -0500 David White wrote: It seems like what Phil's describing is something akin to channels, which gets me on to my rant about personas and identifiers and how nobody gets internet identity right. Except, perhaps, that Twitter doesn't stop me from creating multiple accounts to provide multiple channels or streams. (which reminds me, since my server crashed I need to rework a system which manages one of those streams) Getting metadata from users is hard. I've got systems which ask me for verification of automatically parsed metadata, and have hit "yeah, whatever" far too many times, so now that database may have as well just been derived from machine parsing keywords. It's easier when it's something you can measure automagically (ie: location data, post time), but it's still hard. Dan