I’ve caved.

After all my rhetoric about reading blogs au naturel, I’ve switched to using an aggregator. But while I’m pleased as punch with my new setup, I still have serious reservations about aggregators as tools for “the rest of us”—at least right now.

After reading Cory Doctorow’s description of Shrook—in particular its ability to display a post in the context of its original web page, using the Safari rendering engine to show it with all accompanying styles and presentation—I decided to give it a try.

The result? I’m hooked. There’s no question that it’s streamlined my time online, reduced my at-times overly obsessive checking of favorite sites, expanded the number of sites that I’m able to monitor for interesting ideas, and improved my ability to search for, mark and return to thought-provoking items rather than losing track of where I saw them.

The feature set in Shrook is impressive, and it’s nicely integrated with OS X visually and functionally. I love the four-pane display (especially on my big 17” powerbook screen; I suspect it wouldn’t be as great on a 12”), which works like this:

1. Sources: A list of topic groups (“channels”) I’ve defined, from friends to links to tech bloggers
2. Channels: groups of feeds (weblogs, magazines, news sites) within each topic group.
3. Items: Posts (from weblogs) or stories (from magazines and news sources) for the selected source
4. Content: The content of the post or story. By default, this shows up as unstyled text, but I can toggle it to “Show Web Pages” so that when I’m online I can see the full page displayed inside Shrook.

For sources that provide excerpts, I can use either the “Show Web Page” option to see the whole thing, or I can click on the title of the item to launch my default browser.

When I find an item that I know I want to go back to later—to read in more detail, to load online (if I’m reading offline), to send to a friend, to comment on—I can mark the item, and it goes into a special channel called “marked items.” (Removing items from this list is very unintuitive at present; you can’t do it from the marked items list, you have to go back to the original source and choose “Mark Item” again—it’s a toggle, but that’s not at all clear.)

Shrook checks for new items regularly, in a way that’s both convenient for info junkies like me, and respectful to content publishers paying for bandwidth. And it makes it easy for me to see at a glance what’s new in my reading list—I get an unread items count in each of the first three panes, as well as a total unread items count in the Shrook icon down in my dock (very much like OS X Mail provides).

So yes, I’ll be ponying up my $19.95 registration fee before my eval copy expires, and I suspect I’ll be using this tool for quite some time.

But before the chorus of “I told you so’s” starts up from the folks who’ve been singing the praises of aggregators as the ultimate tool for end-users, let me add some caveats.

This is probably the best-designed aggregator Ive used, but still the interface has some serious flaws. Menus are confusingly named, options are hard to find, and nothing seems to be in the places I’d expect them to be. Even once I’ve discovered an option—like, say, how to toggle from plain text display to native web page display—I find myself consistently unable to find it again the next time I need it.

I can’t find a way to rename sources—which means I have to try to remember that “Jon’s Radio” is Jon Udell, not Jon Dunn or Jon Schull, and that “Reading & Writing” is really Joseph Duemer (aka “chujoe”). Since I tend to think of weblogs based on the people writing them rather than their title, this is a major inconvenience.

It took more than a few hours (I’d guess probably five or six, total) for me to set up the software with all of the blogs that I currently have on my blogroll. The “Import OPML” option in Shrook looked promising, but it failed to properly import the OPML that blogrolling.com generated. It wasn’t the OPML files—I checked them, and they looked fine. (It was the first release of Shrook 2.0, so it’s not surprising that it still has some bugs left to be fixed. I’ve just installed the second release, which fixed a few other problems—but I haven’t checked to see if the OPML problem was included.)

Based on my experiences with this and a variety of other aggregators, I don’t think we’re anywhere near the point where aggregators are good tools for non-geek information consumers. They’re wonderful for those of us who have voracious, insatiable information appetites, and need a way to control and manage the flow of content. But the learning curve is still too steep, without sufficient accompanying benefit, for most of the people I know outside the world of high geekery.

That said, it’s clear the tide is turning. I think Marc Canter’s really onto something with his “digital lifestyle aggregator” approach. As information becomes more available, tools for managing and displaying it (not to mention categorizing and filtering it) are going to become critical. It’s not just about getting more information—it’s about finding what you care about in that avalanche of content. (I saw an Accenture ad I really liked in the airport today; it showed a photo of Tiger Woods cupping his hands around his eyes to help him focus, and it read “Sometimes you have to see less to accomplish more.”)

In the new blog being written by the Social Computing folks at MS Resarch, Lili Cheng recently wrote about wanting a single “Media Control” so that she could give her kids one allotment of “media time” that they could as they like—game consoles, handhelds, tv, videos, web, IM. That fits nicely with an everything-aggregator idea. An integrated control panel for information and entertainment, regardless of its source. A screen is a screen is a screen—I don’t care if the box behind it is my local hard drive, a game console, or the distributed power of the net.

Critical to the development of these large-scale aggregators, however, will be an attention to varied information consumption styles, a sensitivity to the need for user-level configuration, and—perhaps most important—“ridiculously easy” interoperability. Drag-and-drop addition to an aggregator interface, auto-discovery of resources. Not to mention the ability to learn from and subscribe to others’ views of the content—why should my colleagues have to recreate my lists of social computing pundits? Or my carefully-chosen lists of web development resources? Why should my mother-in-law need to enter the family’s addresses into her book if my husband has already done it on his computer?

In my ideal future aggregator world, I should be able to drag a person from my address book into my aggregator, and be given the option to add views to their blog, the emails they’ve sent me, their IM accounts, and things they’ve written in other places. I should be able to integrate my local content (whats on my disk) with online content. (Probably worth an entire post on its own is my growing frustration with that online/offline issue—I want to capture far more of what I find online for later searching and reflection; Shrook is a start, but would be even better if it actually archived old posts after retrieving them, creating a long-term archive of “stuff I’ve seen”; coincidentally, the issue of Newseek that I’ve been reading on this flight talks about search engines heading in this direction, as well.) I should be able to search all content and display it in various ways—by person, by topic, by date, by rating, by frequency of use.

I saw the seeds of a lot of this in the brief demo of Wallop that Lili Cheng gave at ETech…I see pieces of it in other tools that are popping up online (person to photo to blog integration in Flickr, for example). And I’m increasingly excited about the possibilities in this space. But we’ve got a long way to go before I’m setting up aggregators for my mother and son, or recommending them to all of my students.