May 2007 Archives

Shelley Powers is trying out Google Gears, and Dion Almaer already has an implementation for one of his web applications.  The simple explanation of Gears is  that it adds a two-way VCR to your browser.  The browser stores relevant data from the web application so that it can be accessed even when the user is not online, and it saves relevant user input to be played back to the web application when the user goes back online.

Just taking a quick look at Dion's sample code, it's clear that this functionality adds a whole new layer of complexity to browser coding.  Now suddenly you have to be concerned with creating a data storage model and maintaining data consistency between the browser client's long-term storage and the web server. There's a lot of low level code to write and maintain.

It's hard not to think of Rich Skrenta's observation that the more code you have to write, the more things you have that will eventually go wrong. Rich's comment suggests that the benefits of the two-way VCR capability will have to be pretty high for gears to attain long term viability.

If the Gears capability were somehow built transparently into the browser with a standard API, the prognosis would seem much brighter.  Rumblings suggest that that is indeed the direction things are taking.

Well, I started with Byrne Reese's port of the cutline theme and have made a bit of progress.  I think I'm going for a simplified light weight retro look.  For instance, all links are going to be styled the same.  That way, people know what to click on.  Eric Meyer's color blender is likely to come in handy.

Byrne notes that the original cutline is under a creative commons attribution and share alike license.  While I have to confess that my work is derivative, I feel I am going to wander so far away from the original that mine will bear little resemblance.  Once I get done, I'll decide whether to link and attribute.  At some stage, even your derivations are different enough to become just your own.

So, I've been working on the blog design, and you're getting to see a bunch of partially completed stuff.  One thing that's been holding me back is this tremendous jet lag.  It's 12:10 AM, and I want to be awake.  That's because it's twelve noon in China.

That's a real cost of coordination in this global economy.  The human body can only endure so much. 

Byrne Reese of Sixapart has been kind enough to crank out some WordPress themes for Movable Type, the blogging engine that powers this weblog.  Two look particularly interesting, cutline which he uses to power his blog and 2813 which has a couple of extra columns.  I'm going to playing with both of those as I recover from my China jet lag.  So, be prepared for oddity.

Sixapart sells the Movable Type blog publishing platform and also provides the typepad, livejournal, and vox blogging platforms utilized by many bloggers.  Recently, they've been outlining the details of an upcoming major revision to Movable Type, codenamed Athena.  One of the key new features is that they are going to make it possible for programmers to modify the core objects in the system and have those changes reflected throughout the system all the way to the interface.

This may seem like a rather technical addition, but the key point is that it enables using Movable Type as a versatile lightweight publishing platform with much less need for programming than before.  No need to learn multiple codebases as you create a community opinion sharing site or a multiblog site.  In other words, the move has the potential to vastly widen the applicability of their Movable Type code base and therefore Sixapart's opportunity to monetize it.

The thing I wonder though is whether they are too late.  In particular, the Django project has has had these advances since inception, and it has the advantage of being open source.  You don't have to worry about someone suddenly pulling the code, and lots of people contribute to it.  Furthermore, Django is receiving a real baptism of fire by the Lawrence Journal World and Washingpost Post in their interactive online properties.

Of course, sixapart also has big customers.  The problem they face is more one of re-architecting an existing application with what seems like less community involvement. 

Interestingly, modern China seems to have come about as a result of the destruction of Chinese culture during the Cultural Revolution.  During that period, from 1966 to 1976, everybody suffered, and the country seems to have generally been reduced to a subsistence level.  The good thing the Cultural Revolution did was remove many of the outmoded social structures that lingered on after the end of Imperial China in the early twentieth century.

One thing the Cultural Revolution did not remove was the top down nature of Chinese society.

Top down is the opposite of Web 2.0 which is characterized by user contribution.  Sure, those contributions may be aggregated and monetized by large web conglomerates, but they still represent individual contributions, not contributions programmed by a higher authority.  In the US, we seem mesmerized by this people-powered phenomenon, but I wonder about the extent to which the admiration is universal.  Even within the US, people find bottom-up contributions disruptive, and they really only seem to flourish in domains where hierarchical structure does not exist to moderate them.

So, I'm sitting here at Tianjin University of Commerce, trying to figure out Internet connectivity from China.  As reported by Natalie Pace at Forbes a little over a year ago, local estimates have placed Chinese Internet users at 150 million or more.  That suggests the number of Chinese users is equal to or greater than the US number.

The real kicker, though, is that chinese users spend far more time connected than US users.  Pace suggests it is chinese surfers hunger for news.  I have a simpler explanation.  If you're trying to access sites outside of China, access is dog slow most times of day.  People are staying online longer because it takes longer to get places.

Nick Carr points to an article he wrote for the Guardian in which he remarks the increasing dominance of a few sites in Google results and traffic.  The top ten sites now account for 40% of traffic whereas in 2001 they only accounted for 31%.

He's looking at global numbers.  I suspect it's a little different if you look at different countries or regions.  For instance, in China, the top search engine is Baidu, not Google.  The whole orientation here is just ... different.

Right now, I think most Westerners are suffering from a uniquely Western perspective when they consider the Internet.  We're nestled in their high speed networks with content delivery networks to speed up things like video.   We think that that is what the world is like.  Sitting here in China, not really.

So, you cannot access Google video from China, and that seems to be Google's choice. We're not being blocked by a firewall or anything.

So, why would Google do this? A lot of people cite potential political issues. Maybe there's something else. To what extent can Google really manage 1.3 Billion people trying to upload and view videos of themselves?

The title is a little provocative and maybe not true, but then how do you explain Microsoft's desire to obtain royalties from providers and users of free and open source software.  I had this to say on Burning Bird:

The evolution here is interesting. It's hard not to think that software itself has no value. What does have value is what you can do with the software. That idea has been around for a long time, but people always still paid for the software giving it the lie. Now, people don't want to pay for software anymore.

People don't pay for Google's advertising system, they pay for the sales it can bring.  And, that's what Google sells.  It's that kind of value proposition that pays for software development now.

We saw a lot of Beijing today. The scale of the place is unlike anything I've ever seen. You can put one million people into Tiananmen Square. It's the largest such square in the world. Click through to see many other China photos.

Well, I'm listening live to Prairie Home Companion, a Saturday night public radio show, at 7:30 AM Sunday in Beijing.  It's like doing a time machine or something.  Saturday night, wasn't that yesterday?  If I kept the curtains closed, I could be in Ann Arbor.

We're past Japan now on the way to Beijing.  Lots of bureaucracy.  I have to fill out a separate customs form for the laptop, and it is only available from customs officials once I get there.

Joy.

n.b. Well, none of that extra customs stuff we were warned about.  In fact smoother than I might have expected.  Rates a good relative to US immigration.

So, I tell students one thing that makes globalization possible is the Internet.  It's basically a set of interconnected networks built on one standard that makes worldwide communiation possilbe. Well, where we're headed in China, that's not really the case.  The chines block some sites.

Perhaps more importantly, maybe all networks really are in some sense local.  If you're not in the same space, it's hard to share context.  Over ten years ago, Robyn Dqwes exposed me to some research that demonstrated that all people needed to form a group was to share a physical space for a short duration of time.

On the Internet, even if sites are not blocked, you still have the issue of distance.  The data pipe over the Pacific is only so wide.  You can't get the same flow across it as you can from the guy down the hall.

So, I've found a value to Adobe Contribute.  You can blog offline. That means you can live blog your flight from Detroit to Tokyo Narita and then on to Beijing.  Not bad, perhaps worth the price, particularly with an educational discount.

At any rate, we're headed Northwest, the name of the airline I'm taking as well as our direction.  It's 4PM now, Eastern time, and we'll arrive about 5 PM in Tokyo with a flight of a little over 12 hours.  In other words, we're almost standing still against the earth's rotation and letting Japan spin over to us.  Remarkable.

It's like chasing the sun.

n. b. Contribute allows you to both Live Blog and re-edit afterwards.  This post benefited from that.

When I was University of Michigan, I used to get all of my teaching done in one semester and then basically have 8 months until I had to teach again.  That time could be spent on other projects.  In some ways, that's a fantastic job, paid R&D way beyond Google's famous 20%.  However, it leaves you with a big chunk of time to organize into something productive, and frankly can be a rather lonely pursuit, particularly if you focus it on trying to get into top tier academic journals as Michigan desires.  Lots of analyzing and writing down that track.

At Eastern Michigan, my current academic home, the emphasis is almost entirely on teaching with some afterthought given to writing and analyzing. 

Well, I like analyzing and writing, but I also need interactivity.  In some ways, blogging is the perfect thing for me since it's a writing medium that you can twist to your own ends.  So, this summer, I have a few exciting projects:

• Michigan Innovators:  I actually have that URL. This will be podcasts with tech entrepreneurs in SE Michigan.  We have quite a few.
• Learning Remix Reboot:  We have a dedicated server for this now.  I'm going to move the architecture more toward something like django integrating over multiple blogs.
• CASOS social network analysis institute:  We've been doing a lot of networking in our class blog system.  We've shown that contributing more to this network leads to higher overall academic achievement.  The next great frontier seems to be how students connect.
• Muscle Ventures optimization:  This is a hobby site devoted to fitness that earns revenue.  I've been using it as a playground to learn things.  I'm going to tie Google analytics in there.
• BudGibson.com hubbification:  This site links to the others and acts as the nerve center.  I'll include links to some of my now much less active projects like MichiganMuscleBoy and TheCommunityEngine.

Well, I'm off to China tomorrow with my Lonely Planet Guide.  Of the six my wife previewed, it seemed the best.  However, it turns out it sometimes gets confiscated at the border.  Apparently, the Chinese are mad that Taiwan is indicated as being separated from the main land by an international boundary.

To a non-Chinese, that seems odd.  However consider if New England was shown as independent, and we thought that could actually happen.  We did fight a civil war 145 years ago over a similar concept.

Still, I'm going to try to sneak the book in.  It really is the best and seems to have the most information on the places we're going.  I am however also taking a reserve book. I'll be back May 26. Posting during this period may be light.

And that reader is Shelley Powers from Burning Bird.  I took my inspiration to relaunch a blog of personal observations from her.  It seemed like a good place to pull everything I've got going on together.  A bunch of exciting stuff this summer.

Follows a live blogging account of my first blog post with Adobe Contribute.  It's not a reflective piece.  Take it for what it is worth. 

Well, this is a rather remarkable experience.  I'm editing a blog page, with some delays.  It's using the style sheets I have set up for this blog.  I'm using Adobe Contribute, an application I reviewed previously under tight confidentiality.  Plusses and minuses to this tool.  Neat to see your stuff in the format you created, but not as integrated for things like tags.  For instance, no type ahead completion for tags.  It's unclear to me that what I would call a tag is actually inserted into the blog as a tag with the right kind of URL.

Further, it seems to be a bit of a resource hog.  For instance, hitting return at the end of the last paragraph led to a five second delay.  My browser seems slower.  They want $149 for this thing.  Not really sure it is worth that.

I'm going to publish.  Let's see what that produces.

Update:  Unbelievable.  The tag at the top is a tag to a technorati tag space, not to the built-in MT (movable type, the blog software used here) tag space.  Apparently, this is a settable option, but it's not clear to me that this thing is submitted as an MT tag at all.

Update 2:  So, this might be useful for editing posts that are already there, but if you have a customized blog set up using MT plugins, you'll lose some functionality with this tool.  That's because the MT API does not allow access to all underlying features.

Update 3:  It doesn't know my time zone.  It thinks I'm on West Coast time where it was built.  Poor.  My system clock is set properly.

Y Combinator is Paul Graham's early stage VC platform.  He has a community-driven news site called Y Combinator Startup News.  Generally speaking, the site picks out some interesting stuff, but there's clearly some group dynamics going on that bias the stories that make it to the top.  I'm not sure that helps creativity and originality in sourcing stories.  Certainly not a lot of critical thought pieces.

For that, I'd recommend Om Malik's GigaOm site.  Like any individual site, it's going to be idiosyncratic.  But, at least his posts don't come off as playing to the crowd.  He has an actual point of view.

Over the weekend, Don Park published an interesting analysis of whether he should use Amazon's EC2 and S3 services for a new web service given the new pricing scheme S3 has come up with.  The killer is the one cent charge per 10,000 requests the service has instituted.  Don has calculated that hosting significant portions of his app on S3 would lead to prohibitive costs of $8000/month while he could host his own under similar loads for $1000/month.

Shelley Powers voices similar concerns for a site she is planning.

The rub seems to be the request load these sites represent to Amazon.  Both authors were essentially using Amazon as a secondary web host for serving images and other media files.  Amazon is finding that it's too costly to do that for the original price scheme that did not include a per request charge.  Based on my own personal usage for the sites hosted on this server, I've concluded that the Amazon fee would be higher than the hosting fee I'm currently paying.  So, S3 is not for me.

Well, who is it for?  Apparently people who serve files on average larger than 53.6 KB and who don't expose the service directly to the web.  They apparently want you to impose your own site as an intermediary as a sort of buffer.

As Shelley Powers notes in an indirect way, the real issue in web scaling is not back end storage but dealing with the interaction.  I suppose S3 could offer a nice storage solution, but to use it, you have to learn their API and then buffer them from too much user interaction.  That sounds like a fairly lmited market. 

Which is an odd thing to say because I haven't updated this blog in over two years.  But, the truth of the matter is that I am back, and I'm using this as an area to keep people updated on my doings in May 2007 and beyond.  A lot has happened since my last post here.  I've started a whole host of blogging projects involving many people and done other things that have moved me more and more squarely into Web 2.0.

My mother died, and my father had a stroke.

I moved from University of Michigan to Eastern Michigan University.

I'll be updating you on my projects and my thoughts about thngs going on in the wider world of the technologies I use.  I'll be redesigning the site in some sort of blog information architecture.  It should be fun.