Since I began using Twitter actively, I’ve been looking for a reliable way to ensure my tweets will be forever accessible. In my experience, Twitter never offered an easy way to locate old tweets. From a profile page, a user can keep loading tweets by simply scrolling down the page. That scroll-to-load approach works, but it takes a lot of time, and there’s no guarantee that Twitter will allow this mechanism to load thousands of additional tweets. As far as I can tell, there’s no simple way to find a tweet from June 2010, for example – even if you try to search by text, it won’t necessarily be found. It seems to me that Twitter prioritizes recent tweets, and eventually removes the older ones from their index. While I find this is somewhat appropriate for Twitter, I would prefer having the same confidence I have in Gmail. Regardless of how many e-mails I write, I know that Gmail will always keep them around.
After trying a few potential solutions, it eventually hit me: what if I could have one WordPress post per tweet? I would have no problems locating old tweets, as the WordPress search feature works quite well. The resulting database would be easy to manipulate, and completely under my control. Not to mention the huge amount of plugins available at WordPress.org. A couple hours later, I had the WordPress solution implemented, and am still using it today.
To get it working, I used two key WordPress plugins:
I used Twitter Importer to retrieve all tweets that had already been published. This is necessary because the automatic tweet importer only handles new tweets.
This plugin, among other things, is capable of automatically importing new tweets. To get it working, I had to register an application with Twitter, which is a bit odd – it seems like something only a Twitter application developer would need to do. In any case, once I got that configured, Twitter Tools began to import new tweets automatically. The rest is really a matter of preference: for example, I decided to disable comments, pingbacks, and other features that I felt weren’t inline with my intentions. I also downloaded and installed a WordPress theme that looked suitable for microblogging: P2, by Automattic.
The only imperfection I noted is a bug that I’ve first seen in Google Buzz. Twitter recently improved the re-tweet action, allowing users to simply click a Retweet button, instead of manually replying with the “RT @username” prefix.
Being better off alone is super til you need to open the jar to trap the spider who’s crawling around where you’ve fallen & you can’t get up
— Liana Maeby (@lianamaeby) March 15, 2011
It is exactly 140 characters, the maximum allowed for a tweet. When re-tweeted, here’s how it looks:
Looks great; no “RT @username” pollution, instead, a clean icon in the top left corner. It seems like Twitter is handling it internally, rather than simply adding the “RT” prefix to the tweet. Yet, when you view the same re-tweet from outside Twitter, different story:
In these cases, the “RT @username” prefix is visible – and it actually uses up part of the 140 characters. If the tweet was 140 characters to begin with, this means the last few characters could be sliced off to make room for the prefix. Not only that, but an ellipsis is added to convey that the original text has been truncated. Since this is happening in Google Buzz, I’m ready to write this off as Twitter bug. It’s unfortunate, though, because it means my Twitter archive is truly imperfect. Tweets could be missing content, which sometimes means a URL gets chopped in half.
Imperfections aside, I believe I found a decent way to archive my tweets, all the while without changing the way I already use Twitter. Plus, my Twitter Archive includes links back to my other websites – so it could also be useful in bringing in new visitors.