SDForum Vertical Search Conference Program Overview.
Search and search engines are changing the way we live, work, and play. First Yahoo and now Google have created search empires and platforms generating billions of dollars of value, and others like AskJeeves, MSN Search, and Amazon's A9 are also competing for search market share. More recently, local search capabilities have become popular on Yahoo and Google, AOL, and AskJeeves. And in the past year, shopping search and travel search engines have established significant attention by focusing on popular niches for search. Now a whole new crop of venture-funded startups focused on vertical search are beginning to enter the field and draw attention. Search is going Vertical -- vertical search engines are available for local businesses, product & shopping search, travel search, weblog & news search, and search for classified & jobs listings, with more domain-specific search engines debuting every month.
So is Vertical Search just some new hype for Yahoo and Google? Or is there real opportunity for new startups to emerge as domain-killer search engines? Will the big players crush these new upstarts, or do they have a shot? And will Microsoft try to “extend and embrace” these new young Internet Turks, in order to LEAP over Y! and G? The answer to these and other burning Internet 2.0 questions can be found at Vertical LEAP, a 1-day special event on vertical search hosted by SDForum and your friendly neighborhood venture capitalists. Join us for a whirlwind tour of the world of vertical search, and an overview of leading search engines in both popular and emerging new markets for domain-specific search.
As if that weren't enough, launched mere minutes ago: Technorati: Live 8, which brings you:
...the latest conversations about the campaign to Make Poverty History. Read first hand accounts of the concerts and events, and get all the news and opinion from the blogosphere.
Monday just around noon, the folks who've been working hard developing and designing microformats launched microformats.org, a community site for the development and design of microformats. Read more about it on the microformats.org blog. Technorati has contributed all of the microformats specifications in progress from the Technorati Developer's wiki. Ryan King and Kevin Marks did a monumental job of copying and cleaning up the wiki pages. CommerceNet generously donated both the domain names, and funding of the design and setup of the site. The design and logo is the beautiful work of Dan 'Simplebits' Cederholm. Thanks especially to Adam Rifkin for helping kickoff and manage getting microformats.org started, and to Gilles, Kragen, and Rohit for all their hard work. And big thanks to anyone else on the IRC channel this past weekend that I'm forgetting to mention because it's past 4pm in the afternoon and I had two hours of sleep last night (because Greg Elin, Kevin Marks, Suw Charman, I, and a bunch of other remote folks were SubEthaEditing and IRCing until 3am PDT on a fotonotes microformat).
So go checkout microformats.org. Read all about it. Join us in the IRC channel and mailing lists. Read up on the microformats principles and current specifications. Blog what you think and tag it with the tag "microformats".
The Technorati team has been putting in a lot of long hard hours, and this past Monday afternoon (can you tell I've been a little busy since Monday?), flipped the switch and moved the recent public beta to the main site.
I'm really proud of the work that everyone has done in this site redesign. For one, I'm so glad to be working fulltime with such a kick-butt designer like Derek Powazek. When Derek first showed me the visual mockups of the direction he wanted to take Technorati with this redesign, it was like, oh wow. It was so much nicer, friendlier, and more inviting than previous visual design. I distinctly remember him showing it to me with a slightly worried look on his face, and asking me, "Can we do this with CSS? These rounded corners too?" , and I still remember what I said: "We'll figure it out. We'll make it work." I knew I could hack (ahem, carefully construct) the markup for what Derek showed me. And I knew that with enough long hours I could come up with the CSS. It looked theoretically possible. So I turned to Derek, and we got someone who eats and breathes this stuff (turning gorgeous graphic designs into elegantly coded style sheets) on a daily basis much more than I do, and who I knew would be up for a bit of a challenge.
Eric Meyer cranked on these templates and delivered some seriously clean markup and efficient style sheets. He's written a little about it here. I worked on parts of the design with Eric (markup and stylesheets), we developed some novel solutions (e.g. the XHTML design pattern of using four
<b> tags at the end of an element to do rounded borders) that I look forward to seeing a write up of. Those of you who have been peering behind the covers (viewing the source) might have noticed my first experiments with
<b> as a presentational <b>uilding </b>lock for the hexagons on my site map.
What was even more amazing though is how hard everyone on the team worked to get this all done. Ben Jenkins and Jason DeFillippo in particular worked the kind of hours that I remember doing back when we did last summer's redesign and politics site. These guys took the one-off XHTML+CSS pages that Eric crafted, and replicated and reworked them across the functionality of the entire site. I think you can see the wear marks on their keyboards.
Simultaneously with all this frontend work, the backend folks and operations folks worked very hard to put in all kinds of new levels of redundancy, efficiency and reliability. Some of this has been deployed in time for the new site, and enables some of the amazing display of rich contextual information when you do searches and the brilliant redesign of Technorati Watchlists. There's much more in this department to come. I'm looking forward to even more improvements that make our searches faster, fresher, and even more accurate in the coming months.
S5 1.1rc2 is a wonderful example of reuse. XOXO is ideal for publishing outlines on the web, and a presentation is essentially an enhanced outline. I have to confess that the original
<div class="slide"> markup in S5 is my fault, as that's what I used in my original XHTML+CSS presentation, back in early 2003, before Kevin Marks and I had the epiphany in Joi Ito's chat room on IRC about subsetting/profiling strict semantic XHTML, and came up with XOXO. But now we know better. Since a presentation is an essentially an ordered list of slides, rather than using semantically vanilla
<div>s, it make much more sense to use an
<ol> for the list of slides container, and
<li> elements for the slides. Well done with the integration Eric!
The CSS Working Group has just published the CSS 2.1 Specification as a W3C Last Call Working Draft. The new draft can be found at:
Yes, another round of edits where we (the CSS Working Group) took into account implementation experience from numerous implementers and web authors in response to the previous CR draft, made lots of fixes, and specified more details. Because of all the fixes, the spec now goes back to being a Last Call Working Draft to alert people to the fact that there have been some functional (minor, but functional) changes and fixes since the CR. The CSS Working Group is being very thorough with CSS 2.1, and as we see better and better interoperability among modern browsers, I think the the web community will benefit from their efforts for many years to come.
For those of you familiar with the CR, a diff marked version is provided for your convenience. Those who look closely at that diff marked version may notice that my email address changed, since I'm now an independent invited expert rather than a representative. Those who look even more closely at the cover of the latest draft, in particular, the source, may notice that editors have been marked up as hCards (it's just a few classes, what harm could that do?, and it took mere seconds to add), and therefore, using X2V, you can easily add them to any Address Book application that supports vCard RFC2426.
Bud Gibson wrote a post recently titled Microformats provide immediate search visibility. Bud compares microformats with other efforts and in the process explains why both publishing and indexing microformats has already proven to be a positive feedback loop.
If you're local to the San Francisco Bay Area, don't forget to check out Tag Tuesday tonight (6:30pm) at Gordon Biersch in San Francisco, where we'll discuss RelTag, one of the most popular microformats to date.
Picking up where I left off last week. I've completed uploading my photos of Tokyo DisneySEA. After the "Journey To The Center Of The Earth" ride, we checked out many other attractions, including the "Indiana Jones" ride where I got to drive the jeep. At around 8 or 8:30pm, they put on a huge show in the lake in the middle of the park with fire and water. Actually, fire on the water. It was quite a sight. Check out the larger versions of the photos to get a better feel for it.
KRON4 has posted a summary of yesterday's event, including a group photo. Lots of familiar shiny happy people there. It will be very interesting to see how a local TV station interacts with local bloggers. According to the video clip posted to that same article, KRON4 will be hosting both a directory of bay area blogs, and an aggregator for posts from the same.
Got there a bit late, but enjoyed it nonetheless. KRON is jumping into blogging with both feet, and figuring it out as they go. I admire them for their fearless engagement with this new medium, and holding an open event where perhaps 100-200 bloggers showed up, mingled, discussed, etc. All the stuff bloggers are good at. Afterwards a quick late lunch (brunch for some, ahem) at Mel's across the street with free wifi courtesy of a nearby La Quinta.
Next Tuesday, Kevin Marks of Technorati and Stewart Butterfield of Flickr will be kicking off the inaugural meeting of Tag Tuesday. Tag Tuesday (TaTu) is a monthly meeting organized by and for cutting edge implementers of tagging software, services, standards, and systems, what practical lessons we've learned, real-world challenges we've encountered, and how we can all work together to empower users to make the most of this tagolution.
Looking forward to some tag-talk next Tuesday!
Fresh from furious edits, the authors of hReview bring you hReview 0.2. The changes are small and few:
And that's it. I really don't foresee many more changes. As usual, there is a wiki page for feedback. Let us know what you think.
Antonio Fumero recently wrote some nice words about microformats, folksonomies, and xFolk. They're in Spanish though so the best I can do is to try running it through Babelfish to understand it.
The other cool thing I noticed about Antonio's blog is the Spanish Technorati Searchlet, so you can "Buscar" either "en este blog" or "en todos los blogs". Very nice!
Almost a month ago I left for a one week trip to Japan for WWW2005. With this trip, I decided to finally start uploading photos to Flickr. I've opted to keep the descriptions short, instead linking to the photos, which in many ways tell a whole story of their own.
I arrived the Saturday before the conference to give myself a couple of nights to adjust to the time change. That night Rohit, his wife, and I ate at a sushi place in a local mall. The sushi was beautiful as well as tasty.
The next day we took the train to Tokyo and walked around a bit, taking in sights at places like Shinjuku, Harajuku, the Meiji shrine. On the train ride back to Chiba (where the conference was held, and our hotels were), we stopped at the Tokyo Disney Resort and bought discounted (after 6pm) admission to Tokyo DisneySEA. The "Journey To The Center of The Earth" ride was a lot of fun, which was loosely based on the Jules Verne book that I enjoyed reading when I was 10 years old or so.
And that's as far as I've gotten with uploading my photos. I'm uploading about 6-12 per day using the web interface (I tried the "Uploadr" but that seemed to randomize the order of the photos that I uploaded, and iPhoto doesn't work for me not only for performance reasons - I take a lot of photos - but also for disk/file organization reasons - it forces me to adapt to its way of organizing photos rather than it adapting to my way of organizing photos). The pace of uploading 6-12 per day seems to actually work well for folks reading/browsing my photo stream however, as it gives them just a little bit each day, enough to enjoy, and not so much as to be a burden to get through. We'll see if my photo stream ever catches up to "now".
A few days ago, we quietly updated the Technorati Favelets page with an improved "Technorati This" favelet. Thanks to Jonas Luster for tracking down why the previous favelet was failing in Opera & the new Safari for OSX.4, and coming up with clever DOM workarounds, and thanks to Derek and Kevin for beating me up about it. :)
So if you're an Opera or Safari user, go grab the latest Technorati This favelet and give it a try.
More rapid iteration and evolution of microformats. Bud Gibson has published xFolk v0.4, which simplifies (and even removes) a few things from v0.3. (a fine example for microformat developers to follow). While RelTag is an open standard microformat for distributed/decentralized author tagging of posts and other content (e.g. as specified for tagging items in hReview, and as used by Upcoming.org to tag their hCalendar-savvy events), xFolk is an open standard microformat for decentralized tagging of arbitrary URLs. Keep up the good work Bud.
I quietly published this a while ago, but added a bunch more hCard fields recently. Check out the hCard creator. But that's easy stuff. For some real fun, take a look at the hCalendar creator and hReview creator that Ryan wrote.