Gadling's resident pilot explains what life in the cockpit is like

REVIEW: Google Reader

The trend in RSS adoption sees consumers moving toward invisible services such as My Yahoo! and tools built into the browser such as Firefox's Live Bookmarks. Accordingly, dedicated services with power tools for aware users must bring added value to the table. The home-page+RSS is an added value package, as is browser+RSS. Stand-alone sites and desktop programs that offer sheer RSS functionality, powerful though they may be, are threatened with marginalization.

It is in this environment that Google launches its Google Reader. Though not bundled with Gmail as many people believed it would be, Google hooks Gmail into the RSS sorting and reading interface, as well as Blogger entry posting. Furthermore, Google search technology powers feed discovery. So, the seeds of added value are in place, and it's easy to imagine further powerful integration of Google's assorted engines, especially Local and Froogle.

Promising though Google Reader might be, the newly minted service lives down to its Beta status; it is buggy, partly broken, and excruciatingly slow (on its first day, at least). Pretty to look at but impossible to use effectively at the time of this post, Google Reader earns a mixed review. Read on…

Google Reader is a content search engine and feed aggregator that supports tagging. The Ajaxy look and feel punches through the screen with a distinct “Wow” factor. Content dividers slide around and feed entries display cleanly, without encumbrance. A mechanism for OPML imports is in place, enabling users to dump their existing feed lineups from another newsreader. Subscribing is slick and undisruptive to the ongoing search process. These niceties float the new user into the Google Reader experience for several minutes, before problems impose roadblocks. Looking over my test notes I see much more bad than good—all, presumably, issues that can be ironed out as Google focuses on its overall RSS direction. To itemize:

  • Google Reader is, to quote my colleague Todd Carter, “painfully slow.” The sluggishness is more than an inconvenience; it prohibits serious consideration and long-term testing of the site. Agonizing processes are occasionally abandoned by server errors. Yes, it’s the first day, but this is Google, with one of the largest distributed computing infrastructures in the galaxy.

  • Subscribed feeds display with the unnecessary prefix, “You have recently subscribed to…” which takes up so much room in the sidebar that the name of the feed is pushed out of view! Mousing over doesn’t help. No way to expand the sidebar. Logging out and in again doesn’t help. Just get rid of that foolish prefix.

  • Search results are not particularly good, and possibly quite poor (pending further slooow testing). More on this below.

  • The importing feature appears to be flat-out broken. A few Weblogs Inc bloggers reported Google’s inability to accept their OPML files. Google Reader gagged on my file, which contained about 225 feeds. Being unable to establish a large number of feeds quickly makes it difficult to test meaningfully.

It’s important to remember that Google did not ask to have Reader reviewed; the service is unannounced on the Google blog as of now. Google Reader is Beta, and it acts like Beta.

More on the search engine, which is a central feature: In a notoriously difficult arena of RSS search, my early testing reveals more inadequacies than strengths. First, Google Reader does not separate blog serch results from entry search results; I have found that distinction valuable in Bloglines. Second, I find myself with a peculiar feeling that good results are being blocked by PageRank. Clearly, some general-purpose algorithm is bringing up results that would be appropriate for a Web search, but are inappropriate for an RSS search. Google Reader even disclaims: “Not all sites have feeds yet…” which is an astonishing admission in this context. It goes without saying that Google Reader should implement an RSS search crawler, not a general Web crawler.

Here are the top ten results for a Google Reader search for the keyword yahoo:

1) games.yahoo.com - probably a useless result to most people
2) news.yahoo.com - a Web-search result; Yahoo! News is not about Yahoo!
3) ZDNet.com - very poor; ZDNet is a general tech-news publisher
4) ysearchblog.com - Yahoo!’s official search blog; excellent result
5) ysearchblog.com/.... - a page within Yahoo!’s search blog; duplicative
6) slashdot.com - poor; slashdot is a broad tech blog
7) yahoo.weblogsinc.com - our Yahoo! blog; any blog about Yahoo! is a good result
8) wired.com/news/culture/... - poor result; an article page
9) technorati/com/tags/yahoo - Technorati’s tag page for Yahoo!; defensible result
10) theregister.co.uk/... - an article page; very poor result

It seems clear that PageRank is a hindrance in Google Reader, forcing the delivery of high-ranking results with low usefulness as feed-discovery destinations.

On the important subject of feed tagging, Google Reader uses “labels,” the term meaning the same as in Gmail, where labels represent virtual folders. Also as in Gmail, individual feed items can be starred. Read/Unread functions work well and intuitively. the spaciousness of the interface gives me the same troubled feeling as with the personalized home page; while easy on the eyes, all the white space reduces functionality. I’m sure Google is thinking of marketing to the mainstream more than to the power user, but as I noted at the top of this review, dedicated feed aggregation bucks the current mainstream trend.

This is where integration comes in. Google Reader’s tie-in of Blogger is perhaps the biggest long-range story of this release. what similar integration of a major blog provider and feed aggregator exists, anywhere? Bloglines offers blogs, but I cannot think of Bloglines as a major blog provider. SixApart’s family of services (LoveJournal, TypePad, and Movable Type) are bereft of any feed-discovery hook. The major self-installed programs (WordPress, Movable Type, and Radio) have nothing. Yahoo! has not effectively bundled My Yahoo! feeds with the Yahoo! 360 blogging platform—though My Web 2.0 hints at a similar integration.

Overall, then, this launch bookmarks an interface style while hinting at intriguing future directions. It cannot be taken seriously as a usable tool right now, but that could change quickly. When I can deposit my pre-existing feeds, get better search results, and not be kept waiting for every function, a new round of assessment will begin.




Reader Comments

(Page 2)
Previous 20 Comments | Next 20 Comments

RESOURCES

RSS NEWSFEEDS

Powered by Blogsmith

Other Weblogs Inc. Network blogs you might be interested in: