Saving tweeted items for later

Boone Gorges — Sun, 08 Nov 2009 04:03:11 +0000

I get a ton of reading material through recommendations on Twitter. But Twitter has a few problems as a source of reading material (problems that, among other things, keep it from being the “RSS killer” that people like to yammer on about). Perhaps the most pressing problem is that my normal use of Twitter is more or less at odds with the way in which I like to consume reading materia online. Typically, for me, Twitter is a sort of attention dump: if I’m doing work that doesn’t require all of my attention (or if I’m doing work that is sufficiently boring that I don’t want to give it all my attention), I’ll often pour my excess attention into Twitter. Web reading, on the other hand (which for me is typified by the kind of reading I do in Google Reader or Read It Later) is usually quite different, in that I generally don’t multitask as I’m reading. As a result, I frequently see links in my Twitter stream that I’d like to look at but can’t at the moment.

There are a couple of different ways that I deal, or have dealt, with the problem of collecting Twitter recommendations for later.

Readtwit – There’s a service called Readtwit that collects every link from your Twitter stream and creates an RSS feed out of the linked items, which are also annotated with the name of the original linker and the text of the tweet containing the link. Subscribe to the link in your RSS reader, which presumably is a place much better suited for concentrating on reading.

I used Readtwit for about a week before I had to give it up. The problem is volume. I follow enough people, some of whom send out a huge number of links, that I found myself sifting through literally hundreds of items every day – on top of the hundreds of items I receive from all my other RSS feeds on a daily basis. Making things worse, I was only interested in a very small percentage of the items being linked to. I was wearing out my j key in Google Reader.

Readtwit provides a bit of filtering tools to prevent against overload. You can filter out links from certain individuals, or (I think) links in tweets containing certain text. If you’re following a relatively small number of people, and you think you’ll be interested in most of the links that they send out, then Readtwit might be a good option for you.
Subscribing to my Favorite Tweets feed – Twitter creates an RSS feed for any individual’s favorite tweets (located at http://twitter.com/favorites/yourtwitterhandle.rss). When I read a tweet containing a link that I want to check out later, I favorite the tweet. I’ve subscribed to my favorites feed in Google Reader, and a few times per day I receive the tweets that I’ve most recently favorited.

The huge advantage of this method over Readtwit is that it lets me select only those items that I want to read, instead of getting every linked item in my stream. And since the favorites function is part of Twitter itself, I can mark items in any Twitter client. (It’s especially handy when, for instance, someone links to a video or something that can’t be viewed on my phone – I favorite the tweet, and check the link out the next time I’m reading my feeds on a computer.)

The downside: What ends up in Google Reader is not the content of the linked article, but only the tweet containing the ink. That’s fine when I’m reading feeds on the computer and can easily click on the link. But it’s not a great way for me to read things on my phone while I’m in transit, which is what I like to do with longer-form web stuff.
Read It Later – That’s where Read It Later comes in handy. It’s a Firefox extension and a web service that allows you to mark individual web pages for later reading. (Instapaper is similar.) I use the iPhone Twitter app Twittelator Pro primarily because it has the built-in ability to save things to Read It Later. As you see at the right, I can click a button that will scour the current tweet for links and send the linked items to my Read It Later list (which I can then read in Firefox or in the RIL iPhone app or on the Kindle, as I have geeked out on before).

My workflow ends up being a combination of (2) and (3). If the link contains mostly text and images (which can be viewed in the Read It Later iPhone app, where I do almost all my RIL reading), I send it to Read It Later. If it’s got video or audio or Flash, I favorite the tweet and check the link out when it arrives when I’m reading my feeds in Google Reader at a computer.

On the cloud

Boone Gorges — Mon, 02 Feb 2009 20:43:49 +0000

Google freaked out this weekend, which, in turn, freaked me out. I’m a pretty ardent user of Google’s cloud services. Gmail is the most important to me, as it’s where all my email from the past four or five years resides. Reader has streamlined my online reading process so much that’s it’s hard for me to imagine how in the pre-Reader days I managed to read even a tenth of what I get through now. So when Google hiccups – even when the hiccup is apparently unrelated to where I store my data – I get scared.

via Reza Vaziri

These Google fears came just a week after I read Jason Scott‘s delightfully titled “Fuck the Cloud”. I don’t really buy into all the too-simple “you’re a sucker if you use cloud services” rhetoric, and I think (as urged in a Twitter conversation I had with @GeorgeReese) that a lot of what Scott is complaining about is more about backups than it is the cloud. Still, this piece, along with my Google woes, was enough to get me thinking about how wise it is to depend on web services like I do.

My first reaction on Saturday morning, when Google was acting up, was to back my stuff up. I saved all of my Reader subscriptions in a local OPML file, updated my POP3 backups of my Gmail messages in Thunderbird, and saved local copies of my important GDocs. I was able to make these backups because Google has allowed it by embracing the right kinds of standards. And this fact – that backups can be made and exports done – is one of the things that makes me relatively comfortable using Google’s services so extensively.

This relatively straightforward exportability stands in contrast to the situation at some of the other sites where I create and store content. I’ve used Tweetake to export my Twitter activity to a CSV file, but the solution is far from elegant. For one, I don’t really like giving my Twitter password out to a bunch of sites. Also, I’m not crazy about the fact that I can’t really do incremental backups. Ideally Twitter itself would offer some streamlined way to export one’s tweets. Facebook is even worse. I feel uncomfortable using Facebook’s message/email system because I know that there will probably come a day when I want access to those messages but cannot get them.

I don’t necessarily blame Twitter or Facebook for their total failure to provide content exporting. There is a sense in which the kind of content being created in these spaces – or, rather, the meaningful units of content to which we attach value and thus would want to save – is quite different from the most discrete units provided by email. What’s really valuable in Facebook is not just what I write, but what others write to and about me and my friends. Only a total snapshot of my entire immediate network would provide the kind of value for posterity that I want. With Twitter the situation is perhaps even more extreme: like in Facebook, the content I value is closely related to the content created by others, but in Twitter these people are not necessarily part of my immediate network at all (like when you @reply to someone you don’t follow because of some term you’re tracking). Pushed to the limit, you might even say that only a snapshot of all Twitter activity would really capture its value at any given time, since part of the value of Twitter lies in the potential you have to mine the collective consciousness, to get a sense of the zeitgeist. When the content that you value is so holistic, the details of backing it up become dicey.

On a more local scale, it’s probable that standard export formats will emerge as services like Twitter become more popular, in the way that something like Atom or RSS can be used to backup or restore a blog. In this sense, maybe my worries about certain kinds of cloud data storage are the kinds that will go away with time. Or at least until the next new kind of content is invented.

There are some other aspects of the cloud question that I find interesting, such as whether one should really feel more comfortable with local backups than with remote ones, and whether paying for a service really makes it more reasonable to feel comfortable keeping your stuff there, but I’ll save that for another day.