There are a lot of blogs out there that push the web 2.0 hype about new services out there.
I personally think it’s just as interesting to see what isn’t working about web 2.0.
A good example of this is superglu – which I thought was a really cool concept.
From the horse’s mouth:Do you already use services like del.icio.us, flickr, blogger, typepad, etc? SuprGlu is a new way to gather all your content from those sites. In a nutshell, SuprGlu:
I thought this was a nice idea, I certainly like combing rss feeds and I find them an invaluable way to keep up with various feeds of things. I prefer my own tool: bozpages, but I could certainly see why someone might want to use superglu.
Except they dont.
This is the saddest picture I’ve ever seen. If they sink any lower Alexa won’t even bother tracking them.
Why? SuperGlu is nice looking, del.icio.us, flickr, and blogger have hundreds of thousands of users, surely there must be some overlap?
I suppose the first clue comes from the word content in the introduction to superglu. You know you are likely to get in trouble when you say words like framework and content. People don’t think about their “content”. They think about their blog maybe, or their photos, but they don’t think about them as a collection of content. So the right place to aggregate these services isn’t on a third party “content” aggregator: it’s on their blog, or maybe with their photos.
I think that’s what we’ve seen with the various blogging plugins and widgets, people are going to reuse their del.icio.us bookmarks and flickr images, but they aren’t going to put them on a third party site, they are going to put them on their blog, so poor superglu is stuck between a rock and a hard place.
:/
First there was SuperGlu – which I think is a cool idea, although I don’t use it.
I’m generating a fair amount of RSS feeds – my SWiK user page, my del.icio.us bookmarks, my blog, my netflix, my last.fm, my flickr feed and probably more that I don’t even realize.
With all these feeds, they seem to be begging to be used and aggregated together. I know this is useful in SWiK for projects that separate their file release feeds and their developer blog feeds, or for joining disparate web services like a flickr photo feed and a del.icio.us tag feed about a project.
Seth Godin unleased Squidoo, which was a general purpose riff on the same idea, although I was less enthused about this product.
Say hello to Ziki

Ziki sounds like it might be some open source wiki product, maybe Zardoz wiki or something, but it doesn’t seem to really have anything to do with a wiki beyond the similar iki postfix.
Ziki’s a 5 person team, but I’m not sure how they are going to support them, my guess would be they used ‘getting bought by Yahoo’ as collateral for a bank loan.
Here’s a view from Alexa (for some reason ziki maps to pierrecardin, not sure what that’s about)

Bon chance Ziki!
I’ve been looking a lot lately at what XSLT can add to web development, and this weekend I started playing around with Google’s AjaXSLT project.
I’m not that much of an XSLT expert, so to get started I coded up a simple project to generate XSLT templates for RSS feeds, called “Skinnerbox”.
Although Google didn’t provide much in the way of documentation for their XSLT library, it’s actually quite powerful and robust, working with XSLT in JavaScript turns out to be fairly easy. I especially like Google’s choice to be agnostic about namespaces, as XML namespaces make me want to tear my hair out.
I’ve been running into a lot of barriers with JavaScript when it comes to speed. JavaScript is just slow, and even modern processors (or my now obsolete PowerBook) start to choke very quickly under heavy load, resulting in a poor user experience, or worse, appearing to stall the browser completely.
This might be an area that the XSLT processors in browsers can assist, as seen in Johnvey’s del.icio.us director. Filtering in XSLT has the potential to sidestep a lot of the Javascript speed barriers, although unfortunately a significant minority of browsers don’t have that much support for XSLT.
Skinnerbox is just a project I did over the weekend, so it’s not quite as robust as I might like, but I think there definitely is a need for people to clean up the way their feed is presented, and if people like the idea I can expand on how you can customize the display of your RSS feed.
Nice feed style is part of the reason I switched my feed to FeedBurner, they offer handy subscription chiclets on their feeds, so that visitors can subscribe with the click of a button. I like this feature, because it means I don’t have to clutter up my pages with all the various chiclets, although I’m not quite as enthusiastic about having the big advertisement for Feedburner right at the top of my feed.
There are a variety of Blogsearch engines out there now, competing for the crown of the blog searchosphere. Blogsearch slowly but surely continues to heat up, and with Google and Yahoo getting into the mix, it’s anyone’s game now.
Here’s my review of the current mix:
Potential contenders: Google Blog Search, Technorati, Blogpulse, IceRocket, Feedster, Sphere (Double Secret Beta), Waypath, Yahoo News, Findory, Blog Digger, Daypop.
Disqualified list: Pubsub – you can’t search, you can just subscribe. Bloglines – often not working. Waypath – very few results Yahoo News – just an addon for Yahoo News and doesn’t return much Findory – Quite limited in scope Daypop – Very limited search results Sphere – Not publicly available
If you want to see all of the contenders together, you can search here and take a look at how each service compares.
I judge these to be good features: RSS, Nice URL, Depth of Search, Search Timeliness, Spam Prevention, Aesthetics, Speed, Site Availability, Citation Info. Relevance, Zeitgeist, Timeliness of entries.
I’m not putting “Searches outside of XML” in as a feature because a) people should be encouraged to syndicate a lot of their content into RSS and b) It’s not that easy to tell (I know at least Google BlogSearch doesn’t do it yet, and Technorati does). Also it creates a lot of garbage in the results IMO.
| Doesn’t get everything and has a short history, but gets the most posts. A- | Offers a choice of RSS/Atom, 100/10 results, includes author and blog-name in xml A | Gets spammed. B- | Very fast search, typically within the range of a regular Google search. It is also very quick to publish new blog content. A | URL as ugly as sin D |
| Gets a medium-high percentage of posts, with a high percentage of high quality posts and some posts no one else gets. A- | No public RSS feeds on basic blog search. You must register/subscribe and make a watchlist to subscribe to a search. D- | Low incidence of Spam A- | Medium speed, can be quite slow although mostly improved to reasonable levels. B+ | Beautiful URLs, searches are often picked highly in Google search results. A |
| Medium-high percentage of reasonable quality posts, including some unique to blogpulse. B | Public RSS on every search, and zeitgeist features. Includes source-blog url in xml. A | Quite low incidence of Spam A | Medium-High speed, quite acceptable. A- | URLs uglier than sin D- |
| Medium numbers of posts, hardly ever a unique one to IceRocket B- | Public RSS on every search, assigns a guid to every post. A | Low incidence of Spam B+ | Medium-High speed, acceptable. B | URL relatively clean B+ |
| Eclectic mix of results including dubious ones such as del.icio.us rss feeds and other non-blog feeds.B- | Public RSS on every search, currently disabled however. D | Low incidence of Spam B+ | Medium-High speed, although at times interminably slow. C | Ugly URLs D+ |
| Medium-Low Number of posts, very few original to Blogdigger C+ | Public RSS/Atom on every search, uniquely includes full text of searched posts, however also adds on a full item banner ad. B- | Not much spam A- | Medium-High speed, acceptable. B+ | URL relatively clean B+ |
Google has different aesthetics to choose from for blog search, such as plain jane and blogspot theme. They both do the same thing, but personally I prefer Melissa Mayer’s stark Google.com aesthetic. The one downside is that I find Google Blogsearch to be confusing visually with Google, I think making both products look nearly identical was a mistake. The default search: relevance, is completely wrong, especially given that most bloggers don’t have a huge pagerank. Relevance hides everything from anyone who isn’t fortunate enough to be blessed with high pagerank.
Google does give citation info, but it’s very difficult to see where, some links will have a tiny citation link next to them (it won’t indicate the number of citations however). Searching for a link on Google is generally very good.
Google Blogsearch currently has no Zeitgeist features, which is a shame, however it does have a ‘topic blogs’ feature that is right half the time, and pretty nice. There’s no way to claim a blog in Google blogsearch, so you can’t put your picture next to your blog or anything like that.
Generally, I rely on Google Blogsearch as my #1 blogsearch tool. Technorati in some ways is better, but not being able to subscribe to a search is basically crippling for me.
Technorati has improved by leaps and bounds, especially in the area of speed, and is quite strong at bringing in a lot of good results and keeping spam out. The new front page is pretty messy and they make me click in the search box before I can search, but generally the style is good. It also does a good job of connecting you to the blogosphere, which I think is an area ripe for innovation. I used Technorati before Google Blogsearch came out, but speed problems and the lack of RSS feeds have driven me to Google.
Technorati has a nice aesthetic, the ads are not intrusive (I’d probably never click on one though, when I’m searching blogs I am not looking to buy things). They offer a citation rank for a blog, so you can see how much authority a given blogger has based on people linking to them. This is a very noisy indicator, it’s often completely irrelevant or wrong but it can be nice.
Technorati also includes Avatar Pics for feeds, as it allows you to claim a feed. Technorati also has TagSearch, a brilliant scheme to get pagerank and spread virally, and feed finder (which unfortunately doesn’t work on feeds that aren’t manually added.) Overall, I really don’t recommend Technorati in principle, given that they search RSS feeds but offer no RSS feed of their own.
As far as I’m concerned, BlogPulse is the dark horse in blog search, they offer good results and are keeping pace with innovation. Their aesthetic, especially their front page, is godawful, there’s way too much going on for what is essentially a search engine. But it’s fast and relevant and not very spammy, and the search results are clean and nice looking.
They offer a nice feature where you can track a conversation, it’s essentially citations ala Google except they do a worse job letting you know if citations exist and a better job letting you know that the feature itself exists. A feature that is unique to Blogpulse is the trend tracker, you can search for a term and then see the graphed trend of postings for the term, it’s really a nicely featured tool in and of itself. There’s also a blog trends feature in beta and several zeitgeist features, with rss feeds.
There are no claim blog features on blogpulse, but the zeitgeist features and blog citations are good enough IMO. I use BlogPulse as a close second to Google now, because they too offer RSS feeds for a search.
There are a bunch of other blogsearch engines, the best of the worst being IceRocket, which until recently billed themselves as a straight search engine, but have given that up and focused on blog search. Generally, nothing stands out about these blog search engines, and I rarely use them.
Feedster was once my #2 blogsearch to Technorati, and they were very early movers, but innovation seems to have gone nowhere there, results have gone downhill and the speed is still often slow. Sometimes Feedster RSS feeds are unavailable.
Blogdigger I rarely use, it’s slow to get content and nothing really distinguishes it from the pack.
Generally I think that there is a lot of room for improvement in blog search. Right now it’s very noisy and spam is a serious challenge. I really want to see a lot more features in the way of tracking conversations and visualizing what people are talking about on their blogs.
I like blogsearch overall, it’s a useful niche search, especially for current events. For example, I never would have heard about this blowup at the Les Blogs conference: “Mena Trott (Co-Founder of SixApart), at the Les Blogs conference (organized by SixApart) interrupted her speech about civility to single out a blogger who had criticized her, called him an * and dropped the F word…”
Or about how GMail has finally let me turn the really annoying “WebClips” feature off, and how Dave Winer hates Atom getting a ‘free ride’ in the text RSS/Atom.
I definitely think RSS is a critical feature for blog search, it’s how I do a lot of my searching, since I maintain a sizable blogsearch bozPage shrine to my vanity – to see who is commenting on my blog outside of it.
I’m sure that all the blogsearch contenders are busy working on a bunch of new killer features and basic improvements, it will be interesting what happens.
Something I’ve been thinking about is the way I use my blog reader, and how it’s transformed the way I use the web.
In many ways it’s great to have information delivered to me on my terms. In a lot of the reading I do online, I don’t ever leave the RSS reader.
However there are problems with reading a lot of feeds, which makes me start to think maybe I need to rethink the classic model of RSS reading – create your blogroll and watch all the posts as they come in.
I’ve found myself often in a quandary about adding a feed to my blog reader. It’s already so choked with feeds that I constantly have to play gardener, pruning and trimming to make sure that I don’t get overloaded. (I almost always am anyways).
OK so a bunch of geniuses think they can solve my problem with blog overload by only finding me the ‘relevant’ posts.
Except that doesn’t work, that sucks. Taking management of my information consumption away from me is bad. Unless Google Reader is plugged into my subconscious, how will it really know what I want to read?
OK so here’s my solution for adding a zillion feeds. Add them to different blogrolls. I now maintain 5 blogrolls. I visit them in varying frequencies:
With many blogrolls, I don’t have to worry so much about overloading my queue of things to read every day. The question is: how do you keep a bunch of blogrolls? Well since there isn’t an easy way in Bloglines, Google Reader, or any other reader I could see, I rolled my own blogreader: BozPages.
It’s dead simple. Add a feed, make a blog roll, that’s it.
By the way, hearing about OPML at Seattle MindCamp really impressed on me of the usefulness of it. I have had generally a pretty negative opinion of OPML as people talking about it have made it seem too semantic-web-ey. Also it shoves data in xml attributes, and it’s of much less obvious use than RSS, but so far as having a standard format for blogrolls, it’s awesome.
I sometimes use BozPages as a kind of OPML viewer. I subscribe someone’s feeds to a BozPage, and then I can check out what the world looks like from their perspective. Unfortunately, OPML discovery is still in its infancy, people often don’t know and don’t publish their OPML feed urls. It’s too bad because there’s a lot of cool things that could be done if you could see who’s subscribing to you and who is subscribing to them and so on.
Skinnerbox is an XSLT generator for creating XSLT files to put a human readable style on RSS feeds.
Skinnerbox is alpha – a good deal of things are broken, due to the various combinations of web browsers and rss feeds.