Northern Heights
TweetDeck’s List-Centric Model Shits the Bed

I’m really starting to get pissed off with TweetDeck right now.

It’s bad enough that they canned the proprietary groups in favor of the API-draining lists, but now the damn application doesn’t even display lists properly.

Take a look at these snapshots below (click for bigger version).

Here’s a view of one of my lists at twitter.com. Lots of recent updates from various Tweeters.

And this is what TweetDeck’s been showing me for the last few hours. All my “public” lists are completely devoid of updates. The “private” lists actually load up fine (they’re off screen), but for some reason TweetDeck doesn’t want to show me updates from any of my public lists (which accounts for 90% of them).

The “all friends” group is fine, which means if they were still using proprietary groups I’d be seeing all my updates.

There’s an avalanche of complaints in TweetDeck’s support forums about all of this, all of which are currently falling on deaf ears. A bit of a shame. TweetDeck was one of the better desktop Twitter clients. Unfortunately the only realistic alternative is Seesmic and that’s even more problematic for slightly different but nevertheless frustrating reasons.

TweetDeck, it’s very simple. PUT PROPRIETARY GROUPS BACK IN. The code’s actually still in there, because TweetDeck 0.32.1 will play nice with your previously created groups. It’s when you delete them in favor of lists and discover you can’t recreate them that the problems begin.

Why TweetDeck’s New List-Only Model Fails

TweetDeck have really dropped the ball by removing group functionality. I can see it driving away power users in the long term.

Here’s my situation. I currently follow approx. 400 people (and have about the same number of followers in return). I have a pretty broad range of interests, so I currently have 10 lists set up and, ideally, would like to set up 5 more but can’t afford the resultant API drain. (As an aside, before TweetDeck went list-only I only had 2 lists. But now I’m pretty much forced to use lists in order to replicate the group functionality I previously enjoyed.)

With the old group setup I could set my “all friends” and “mentions” sliders pretty high (to give me updates every 90s and 5 mins respectively) and it would still only consume about 60% of my hourly API calls. I’d even occasionally crank the “all friends” up to a 60s turnaround.

With my groups now replaced by lists, the fasted turnaround I can possibly get by cranking the “lists” slider all the way up to 100% is about 4m 30s. Leaving some room for searches, “mentions” and “all friends”, it’s closer to 7m 30s.

7 minutes may still seem fairly zippy (and for a casual user who follows less than 50 people it still is), but having developed a Twitter usage pattern around the 90s turnaround for “all friends” (and thus, by extension, all my groups) it’s like night and day. For 90% of the people I follow, a 7m turnaround is fine, but the remaining 10% I follow (which includes many of my favorite Tweeters) often indulge in a rapid, quickfire pace of conversational tweeting for short term memes. Ie, someone could mention the movie Avatar and within the space of ten minutes there’s half-a-dozen of us who are suddenly engaged in a brisk exchange of tweets on the subject. With the new list model this isn’t possible; I’m now forced to completely revise the way I use Twitter.

I’ve currently settled on a compromise where I now have “all friends” updates coming in every 3m 30s, “mentions” every 10m and “lists” every 6m 30s. It’s not ideal but it’s as fast as I can get it and ensure I have enough API calls left over for searches (I have about six columns set up to search the Twittersphere for info relating to my blogs).

Because I can no longer get a fast turnaround of updates via my lists, I’m now forced to concentrate on my single “all friends” column. This isn’t ideal because I’m forced to wade through irrelevant material to get to the material that interests me at that moment in time. I’ve also noticed that the “mentions” aren’t updated very reliably. I’m seeing mentions via the Twitter web site GUI that never make it into TweetDeck. Similarly, mentions that did previously make it into TweetDeck randomly disappear and reappear. Sometimes I see just the old ones and not the new ones.

At this point I should mention that I actually have two Twitter accounts. Both of them are related to two different blog projects with two very different sets of followers (one’s a retro video game blog, the other’s a book review & writing critique blog). Thankfully each account carries its own set of API calls, so the number of lists I have in one doesn’t impede the other.

At some point in the future, I can see myself having to create another account or two. On one set of accounts I’ll just follow the news/information providers, while on the other set I’ll follow the conversationalists. If TweetDeck brought back groups I wouldn’t need to do that, but if they continue with their list-only model I’ll have no choice but to create additional accounts.

Ideally, I’d like to see TweetDeck bring groups back but give us the ability to either sync them with an existing list or just create a proprietary group not connected with any list. These would then just effectively function as a filter for our “all friends” (as it did in the past) and not rape the API as much as a pure list would. The mind boggles as to why this wasn’t implemented in the first place. It looks like TweetDeck got a bit too excited about the potential for integrating lists and rushed to the list-only model without considering the consequences.

If TweetDeck ignore our pleas to bring groups back and steadfastly remain a list-only app, I can see their base of power users eventually drift away to other applications. I would use Seesmic but its interface is diabolical and setting up proprietary groups rapes the API (the last time I used it I worked out that it cost you 3 API calls to add a single user to a group. I added 50 users to a group in the space of 2 minutes and used up all my hourly API allowance. With 400 users to add to groups I decided I didn’t want to spend 8 hours doing 15 minutes worth of work and uninstalled it).

Please, TweetDeck, reconsider your approach to lists and proprietary groups. They can co-exist and aren’t mutually incompatible.

Why I Subscribe to 721 Blogs in Google Reader

Every now and then, I’ll be engaged in a conversation with someone that invariably turns to the subject of Google Reader.

“How many feeds are you subscribed to?” I’ll be asked.

“721,” I currently reply.

What follows is usually a look of incredulity, amazement or just plain bewilderment.

Yeah, 721 seems like a lot. Most people I know who use Google Reader only subscribe to a few dozen. Some subscribe to hundreds. One person I know has over a thousand. How, people ask, can I possibly find time to read them all?

Well, here’s the thing: I don’t.

Let’s break down exactly what sorts of feeds I have.

Of the 721 feeds I’m currently subscribed to, only 94 have been tagged as “faves”. This is the elite group of feeds I make sure I read every day, without fail. A much more manageable number, I’m sure you’ll agree.

So what about the rest?

The remainder are all tagged with one or more of 95 tags, covering a broad range of subjects and topics of interest. Once I’ve made my way through the faves, I’ll then cherry pick these subjects based purely on whatever I’m interested in reading at that moment. Some days my interests will swing more towards film criticism or book reviews, whereas on others I may want to catch up on the world of web design or music production.

I do a lot of skim reading, so for the most part I’ll click on a new tag then quickly skip through articles, just reading the headline, only stopping to read articles that actually grab my attention. Occasionally I’ll switch to “show all” and hit the “mark all items older than a day” button to flush out the older articles that have been accumulating.

While Google Reader’s great for reading blog articles, it’s also pretty nifty as a general notifier. A good two dozen or so of the feeds I’m subscribed to are keyword alerts for places like Technorati, Delicious and Google. For example, because I run a retro gaming blog, I have a “retro gaming” keyword search set up on these sites that alerts me whenever someone bookmarks something on Delcious, posts something on a Technorati-monitored blog or Google indexes a new web site that relates to retro gaming.

I also have Twitter keyword searches monitored via Google Reader, so whenever someone mentions specific key phrases, I’ll check out the Tweeter in question and possibly find a new contact to follow. I also subscribe to a Twitter digest service that enables me to get a daily posting of every tweet specific people have posted during the previous 24 hours. I don’t have Tweetdeck running 24/7, but I don’t want to miss anything my favorite Tweeters have posted, so it’s a useful thing to have.

However, the single greatest benefit of subscribing to a large number of feeds is a very simple one that many Google Reader users overlook: Google’s ability to search your feeds.

When you type a keyword search into Google Reader’s search box, it not only searches all your feeds for articles you’ve previously read, but searches through as much of the blog as Google’s indexed and will even find articles that you haven’t yet downloaded/read.

In short, it’s a very nifty search engine that exclusively searches your favorite sources of information. For example, earlier today I wanted to read a review of Luke Vibert’s latest album. Normally you’d just fire up Google, search there and start clicking through the search results in the hope of finding a dependable source. Instead, I went into Google Reader, searched for the album and found five reviews posted on my favorite, trustworthy music blogs.

Don’t just think of Google Reader as a means of reading new blog articles and nothing else. While it’s great at doing that, it really starts coming into its own as a personalized search engine and keyword notification system. And the more web sites you subscribe to, the greater its potential for finding just the right information you need.

So, remind me. Why do you only subscribe to a few dozen blogs in Google Reader?