Thursday, May 22, 2008

Twitter's scaling issues

I saw this article on Twitter's scaling problems earlier on my mobile and immediately (of course) Twittered it. The gist is this: Twitter's ever-more-frequent downtime has to do with the fact that early on, they greatly underestimated the complex issue of one-to-one and many-to-one connections. "The issue is that group messaging is very difficult to achieve at a grand scale," Nik Cubrilovic points out, and "Twitter is unique in that it needs to parse a large number of messages and deliver them to multiple recipients, with each user having unique connections to other users." Basically, Twitter is doing something that has never been done before at this scale. "Every new Twitter user and every new connection results in an exponentially greater computational requirement."

Cubrilovic goes on to say:
Some of the best web applications are able to efficiently solve very complex problems to produce simple results for users (Eg. Google). The success of these applications is due to the innovative efforts by developers to solve large technical challenges, where they have often had to break new ground for solutions. For Twitter to reach a similar point of reliability they too will need a very comprehensive, ground-breaking solution.

Okay, let's stop there. At about the same time Twitter came online, Jaiku was unveiled by two former researchers from Nokia. Like Twitter, Jaiku was conceived from the ground up to work with mobile phones through SMS. Like FriendFeed, Jaiku pulled in activities from other services and called them "activity streams." And like JotSpot, Jaiku has been acquired by Google and has entered what looks to be a long period of reengineering, rescaling, and rebranding similar to JotSpot's. (JotSpot recently reemerged as Google Pages.)

I wonder if Google is even now working on the same scaling issues that are plaguing Twitter. They certainly have the expertise and the hardware to scale. I just hope it doesn't disappear into the Google maw the way Dodgeball did.

(n.b., Jaiku was founded by Jyri Engeström and Petteri Koponen. Jyri Engeström is the son of Yrjo Engeström, a name familiar to anyone with a passing interest in activity theory.)

(n.b., I have not bothered to discuss services such as FriendFeed and Pownce as competitors because they do not update via SMS. They do what they do decently, but SMS is a must for this space.)

Twitter At Scale: Will It Work?
Blogged with the Flock Browser

Of course they do

I blogged yesterday about how Opera reports that a high percentage of Opera Mini use in the US goes to social networking sites. Maybe one explanation is the phenomenon in the article below. If people become concerned about their bosses reading their company emails, one obvious way to deal with it is with a relatively inexpensive mobile device that does not use the company's networking infrastructure at all -- a device like the iPhone, consumer Blackberry, Android phone, or even my aging Samsung A500M. To paraphrase Princess Leia, the more IT tightens its grip, the more packets slip through its fingers.

Slashdot | US Firms Read Employee E-mail On a Massive Scale
Blogged with the Flock Browser

Bad news for Blackboard? Google Pages is an easy way to generate public or private course sites

I did this one in about ten minutes. UT apparently has a Google Apps Professional Edition subscription, making it easy to make these pages public or confined to users. Will have to seriously think about using this in the fall.

Spinuzzi - sample course page (spinuzzi)
Blogged with the Flock Browser

Twitter went down several times, and they don't know why

I had assumed that it was just high demand, but "The truth is we're not sure what's happening." I should note that yesterday was also the "TwitOut" led by Robert Scoble, in which people boycotted Twitter for FriendFeed; doesn't look like that put a dent in usage at all.

Twitter Blog
Blogged with the Flock Browser

Google Sites open for business

The former JotSpot, which I had hoped would become a robust project management system, is instead a free collaboration space. As I speculated a while back, this might be useful for transient collaborative projects such as student group projects. I'll have to take it for a spin later, but here's the link.

Official Google Blog: Google Sites now open to everyone
Blogged with the Flock Browser

Wednesday, May 21, 2008

Am I a typical Opera Mini user?

I wonder because of these interesting survey results from Opera, which makes the browser that I use, Opera Mini. The results are in line with my usage:
According to Opera's survey of the more 11.9 million Opera Mini users in March, almost 41% of mobile traffic now goes to social networking -- up to 60% in some countries, including the US.
Yes, I spend a lot of time checking Slandr (a Twitter web-based client), FriendFeed, Facebook, MySpace, etc. in addition to reading news and checking email and calendar. All this usage comes at odd moments, such as riding the bus, waiting for an event, or, er, the moment I get up. My sense has been that I use the mobile web more than others, but I wonder what Opera Mobile users (who I imagine to be more continuously connected than most) are up to and how that group resembles others who will hit the market soon via their iPhones, Android phones, and consumer-end Blackberries.  I imagine Opera's raw data would be fascinating to look through.

Report: The Mobile Web is the New Hangout - ReadWriteWeb
Blogged with the Flock Browser

Email failing

Quote of the day:
There's too much wrong with email. 14-year-old girls know this. I'm just trying to catch up.
Sure. I'm finding myself using Twitter's direct messaging more for on-the-fly contacts (and I really wish Twitter would buy some new servers to compensate). Email is good for moving large texts and small attachments, but it doesn't work well for short conversations, group conversations, public conversations, or especially mobile conversations. That's why people are expanding their media ecosystems; email no longer serves as a best-fit solution for the many communication tasks that we've funneled through it.

Faster Future: Publishing possibilities now and beyond: How many 14 year olds use email?
Blogged with the Flock Browser

Tuesday, May 20, 2008

Reading :: From Teams to Knots

From Teams to Knots: Studies of Collaboration and Learning at Work
By Yrjo Engestrom

Yrjo Engestrom is, of course, well known for his work in developing activity theory via theorization and empirical cases. I've reviewed several of his pieces on this blog as I've worked through my own thoughts on AT and my own studies. When I found out that he was publishing a new book with Cambridge University Press this year, I immediately preordered it and waited for it with anticipation -- and dread.

The anticipation came, of course, from the fact that Engestrom is such a leading thinker in AT and produces both theoretical constructs and empirical research with expertise. The dread comes from the fact that I had just sent the final manuscript for my upcoming book, Network, to Cambridge UP. Network is my own attempt to theorize (and criticize) AT by putting it into dialogue with actor-network theory via a study of a telecommunications company. It also seeks to develop AT to better address knowledge work. And it discusses and (gently, I think) criticizes Engestrom's own attempts along these lines. So while one part of me was eager to see how Engestrom has developed his work, another part of me worried that his new book would make mine outdated before it could even be published.

I'm happy to report that Network is still relevant -- and that Engestrom's From Teams to Knots is a new essential text for activity theorists.

Engestrom's book revolves around the question of how to understand, study, and theorize collaborative work and collaborative learning in knowledge work. The book is based on the many articles and conference presentations published by Engestrom and his research teams over the last few years, and almost every chapter is firmly grounded in empirical work. (One partial exception is Chapter 9, the bulk of which involves the analysis of a Tony Hillerman novel.) As Engestrom moves through the chapters, he moves away from the notion of teams and towards the notion of knots, "rapidly pulsating, distributed, and partially improvised orchestration of collaborative performance between otherwise loosely connected actors and activity systems" (p.194). He persuasively argues that this form of collaboration has become more prevalent than traditional teams due to ongoing changes in organizations, particularly the phenomenon of co-configuration (in which products adapt to customers, supported by ongoing conversations between customer-product pairs and the company) (p.195).

To deal with such changes, Engestrom develops some elements of activity theory.

One example is that of "runaway objects," partially-shared large-scale object(ive)s that bind together multiorganizational, multiactivity fields and that "require new forms of distributed and coordinated agency" (p.208).

Another example is that of mycorrhizae, or subterranean, "horizontal and multidirectional connections in human lives" (p.228), something that is "simultaneously a living, expanding process (or bundle of developing connections) and a relatively durable, stabilized structure; both a mental landscape or mindscape (Zerubavel, 1997) and a material infrastructure" (p.229). Engestrom characterizes this concept as used "in the same general sense" as Deleuze and Guattari's rhizome (p.228). However, Engestrom tends to fixate on the metaphor here (mycorrhizae as the "invisible organic texture underneath visible fungi" vs. rhizome as "a horizontal underground stem," p.228) rather than the theoretical concept that underlies the metaphor. Consequently, Engestrom reads Deleuze and Guattari's concept of rhizome as developmental and substitutes his own as a more developed and empirically grounded developmental account.

And this is where I breathed a little, selfish sigh of relief. Engestrom's work here is definitely important and useful work, and I plan to use these theoretical constructs in my future work. But one of the criticisms of AT that I make in Network is that activity theorists tend to focus on development to the exclusion of the sort of antigenealogical splicing that is described in Deleuze and Guattari and in actor-network theory. Here, Engestrom continues that path by taking development -- or weaving, to use the term I coin in Network -- as the central feature to investigate. So, although I really wish I had been able to read this book beforehand and incorporate its concepts into my own manuscript, I still think Network will have a contribution to make after all.

But enough about me. If you are at all interested in AT, you really should pick up From Teams to Knots. It develops AT in solid ways and backs the work up with solid case studies.

Wisdom = broader search of larger database

When older people struggle to remember names, phone numbers, etc. they sometimes assume that they are merely "getting old" and the mind is not working as well as it used to. But according to this article,
Instead, the research finds, the aging brain is simply taking in more data and trying to sift through a clutter of information, often to its long-term benefit.
The clutter of information makes older brains process information more slowly, but also more thoroughly, making them better problem solvers.

Memory Loss - Aging - Alzheimer's Disease - Aging Brains Take In More Information, Studies Show - Health - New York Times
Blogged with the Flock Browser

CCCC Theming Contest

Update: I changed the deadline to May 23 because I realized that a monthlong contest would be crazy. I now have 19 entries as of 5:52 AM May 20, 2008.

On Twitter, I noted that CCCC's theme this year ("Making Waves") led many to employ strained analogies in their panel proposals. This is not a new phenomenon: a couple of years back, as Colin pointed out, the theme "Taking it to the Streets" launched hundreds of similarly named panels.

This is a recurrent issue. When I was SIGDOC program chair, we decided to forego a conference theme because of these tendencies. But CCCC is prety well locked into supplying conference themes. And the broad nature of the conference means that these themes tend toward the broadly metaphorical rather than the specific ("Making Waves" rather than "Improving Editing"). It's the broad metaphors with no specific referents that cause the problem.

But surely we can do better than "Making Waves" or "Taking it to the Streets." So I announce the 2008 CCCC Theming Contest. Here's the ground rules:
  • Entries should be one sentence or phrase that supplies a metaphor appropriate for a CCCC theme. It's got to be a metaphor. Make it fit within 140 characters, like a tweet.
  • Metaphors should be broadly applicable to writing, or at least as broadly applicable as the actual themes above.
  • The more outrageous, the better.
  • But obscene, bigoted, or otherwise unacceptable entries will be deleted.
  • You don't have to be a CCCC member to participate.
  • Entries are due by June 18. (CHANGED: May 23.)
  • Entries can be submitted in the comments of this post, via Twitter (@spinuzzi), or in a comment to this blog on FriendFeed. Twitter preferred.
All right? Let's take it to the streets and make some waves!

Monday, May 19, 2008

Scoble on FriendFeed vs. Facebook

He sees the two services as analogous, but one is open and the other is closed. He also thinks that Facebook's walled garden + Microsoft's search = locking Google out of the web. I think he goes a little over the top here, but you be the judge:
Can the open public Web fight back? Yes. It’s called FriendFeed. Notice that FriendFeed replaces almost all of Facebook’s killer features with open ones that are open to Google’s search.

So, now, do you see why I’m so interested in FriendFeed? It’s our only hope to compete with Microsoft’s new “buy enough and keep it closed” search strategy.
Scobleizer — Tech geek blogger » Blog Archive Why Microsoft will buy Facebook and keep it closed «
Blogged with the Flock Browser

Sunday, May 18, 2008

Special issue of TCQ: Posthuman Rhetorics and Technical Communication

Andrew Mara and Byron Hawk are putting together a really interesting call for papers for TCQ. Here's an excerpt:
According to N. Katherine Hayles, we have always been posthuman. Ever since the first social organization, the first use of fire, and the first development of language, humans have lived in and with systems. Even before its emergence as an academic field, professional and technical writers had been writing and living in organizational systems. Even when the profession is imagined as an isolated endeavor or end-of-the-process set of tasks, technical writers still must operate in larger, complex rhetorical situations. Many theorists have been trying to come to grips with this kind of situatedness from Michel Foucault's attempts to develop an archeological method to understand the human sciences to Bruno Latour's development of actor-network-theory to understand science's place within a complex social order. Professional and technical communication's emergence as a discipline has been marked by similar attempts to identify and articulate these systems perspectives. From Carolyn Miller's "Genre as Social Action" to Clay Spinuzzi's Tracing Genres through Organizations, the field has been trying to come to grips with the complex, and increasingly automated, systems a writer, text, and reader encounter, affect, and live in.
This special issue looks to extend the position that professional and technical communication has always been posthuman. By acknowledging this, we hope to open possibilities for thinking about rhetorical action in organizational, institutional, and technological contexts. As organizations become more complex, technologies more pervasive, and rhetorical intent more diverse, technical communicators need to develop multiple approaches to mapping and acting within these complex rhetorical situations. Philosophical, ethnographic, technological, or qualitative methods can all contribute to a larger understanding of the ways documents, technologies, and human actions affect/are affected by these larger distributed environments. Articulation theory in cultural studies, actor-network-theory in the sociology of science, GPS or data visualization in technical communication, and organizational theories in management are all posthuman rhetorics that enhance our understanding of the contexts in which writers think and act.
To submit:
Send inquiries, proposals, or completed manuscripts as .rtf or .doc attachments to the guest editors: Andrew Mara ( or Byron Hawk ( Proposals are due by July 17, 2008. For accepted proposals, first-draft manuscripts will be due September 25, 2008, and finished manuscripts March 12, 2009, for publication in Winter 2010. Please contact us as soon as possible if you would like to serve as a reviewer for this issue.
Looks like a terrific issue.

Facebook, facing its decline

I blogged recently that I thought Facebook might have turned the corner -- not in a good way. TechCrunch elaborates.

Facebook’s Glass Jaw
Blogged with the Flock Browser

An overview of OpenID

OpenID has been on my radar forever, but I haven't bothered to try it out because nobody was using it. Now that's starting to change. The Inquisitor has a brief overview.  

OpenID At The Tipping Point: What You Need To Know
Blogged with the Flock Browser