inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #51 of 97: Brian Slesinsky (bslesins) Thu 5 Nov 09 09:07
    
What types of communications are those?
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #52 of 97: Ari Davidow (ari) Thu 5 Nov 09 09:16
    
Anything that benefits from actual discussion, as opposed to being 
drive-by commenting opportunities. This would be true of support forums, 
online classes, political/cultural discussion boards, etc.

What I think we're re-discovering is that there is more to conversation 
than encouraging people to drive by and drop whatever lack of thought is 
at their fingertips at that moment. And, that the patterns to support such 
conversation include logins, often include verifiable (at least to the 
sysadmin) ID of some sort, ways to see what is new to you, ways to find 
things of interest to you, and ways to keep track of what you have seen, 
and where you have posted. Probably more, but in keeping with the current 
metaphor, this is what comes to mind on one foot.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #53 of 97: Gail Williams (gail) Thu 5 Nov 09 10:29
    
I see what you are saying.  Group conversation and context over time
(the key thing that makes authentic community) will be very different
without some continuity of identity of participants, and some
participants who persist in interacting.  

I think we often forget the differences between the metaphor of a
forum -- people take turns speaking and reply to one another -- and the
metaphor of a bulletin board -- people put notices up on a board with
a push-pin, and if you want to contact them you typically reply to them
via backchannel methods, like a phone call, since a conversation is
not the main goal of a physical bulletin board.

Currently site designers often choose tools that are fine for what Ari
calls "drive-by commenting opportunities" or bulletin-posting, but
weak for sustained conversation and keeping your place in an
asynchronous dialog over time. 
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #54 of 97: Christian Crumlish (xian) Thu 5 Nov 09 10:43
    
missed my van this morning ("by *that* much!") and so I'm falling
further behind in my reply cycle. Fortunately, last night I drafted a
skeleton mapping out the various points I wanted to grapple with from
Ari et al. Instead of these essay-like soliloquies, which I think tend
to ossify the dialogue, I'll try responding with a series of smaller
discreet observations, any of which may yield their own tangent. 

Just trying to break the logjam.

I actually think the question of whether things like forums (and even
blogs) that may seem like old hat next to the new shiny are being given
short shrift in a landscape like ours is a really good one. I know at
times that not only were we probably oversimplifying the issues around
tagging but that we were almost certainly only going to be able to
scratch the surface on topics like discussion boards and weblogs, each
of which could (and have) justify entire books. We probably retreated
there to the interface surface and may have to some extent missed some
of the more salient points or imposed a frame on things that obscured
older wisdoms. I'll have to look at that carefully.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #55 of 97: Christian Crumlish (xian) Thu 5 Nov 09 11:03
    
I've begun teasing out what I hear as the key aspects of Ari's initial
question in (25), parahprasing:

teasing out of the threads, then tackle each

1. The issue with material in the book proving "much more tentative
and incomplete than expected"

2. The question of the use or awareness of prior art (Design of Sites
in particular) and whether their insights have been lost or
bastardized, and to what extent we successfully absorbed the
foundational learnings of the past and used them as underpinnings for
our own work

3. The question of how tested are the patterns we present in the book
and what level of confidence we attribute to them, and a possible
ambiguity with regard to whether the book is a guide to Yahoo!'s own
interanl standards vs. a work of capturing broader web-emergent
standards. This might also be an opportunity to address <jonl>'s
followup about the influence and impact patterns (in the Yahoo! library
*or* in this book) have on Yahoo! product design.

4. Specific (thank you!) critiques of patterns and to some extent of
Yahoo! experience that may or may not exhibit those patterns (I'll bank
the occasional praise, back-handed or not): specifically 
problems with Yahoo sign in (I may ask someone from reg to talk about
this), in particular federation with Flickr ids.

4a. Specific questions about registration: the need to get more info
to put flesh on social bones, the idea of chaining the sign-up pattern
to the "encourage more completeness" engagement pattern, such as
LinkedIn's "profile is 60% complete" or "complete profile to unlock
feature x" strategems, the idea of optional upsell requesting more
information and explaining the value of doing so / funnel pattern

4b. Discussion of potential errors or problems or ambiguous levels of
confidence around forum patterns.

4c. Anything else Ari comes for with (you promised to call out
"blatantly wrong patterns" which for all I know may be misdiagnosed
antipatterns).

sorry, it's the IA in me that keeps wanting to sort this info to parse
it and respond. I'll next take on (1) above.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #56 of 97: bill braasch (bbraasch) Thu 5 Nov 09 11:04
    
Facebook is a drive by place, "I'm over here now" but I do see threads when
the topic starter steps into a role in the conversation.   I think the
difference is in the usage pattern which is motivated by the intent.

John Hagel does a good job getting conversations going.  He's looking to
instigate a bit of serendipity around his ideas.  Matt Scruby moves some
energy around as well, in a more (mattu) context.

Facebook itself seems to find it more interesting to divert the stream every
week or so with a new interface design.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #57 of 97: Christian Crumlish (xian) Thu 5 Nov 09 11:15
    
As far as "tentative and incomplete" goes, that definitely strikes a
nerve for me. On a work like this there is on the one hand the desire
to take forever and perfect everything and on the other hand there is
the urgency of trying to deliver a product to a marketplace in a timely
and cost-effective manner.

My primary way of trying to resolve this involves the "unbook" concept
<http://theunbook.com/2009/02/18/what-is-an-unbook/>. Basically it
suggests that there is a writing project intended to gather some body
of knowledge, that it had an author or authors but that rather than
being incubated in secret, delivered to the marketplace, and then
updated possibly at rare intervals, that it will be created in public,
with input from many interested voices, and published at regular
intervals when it seems that there is a useful snapshot of material
worth packaging for sale.

Dave Gray's Marks & Meaning <http://www.marksandmeaning.com/> is a
pure unbook. Jay Cross launched the idea
<http://jaycross.posterous.com/the-unbook>. 

As the web started growing up and I came over from publishing, I was
also working my way toward this sort of neverending spiral model with
the book no longer the central unit or focus of a writing effort.

Our book isn't a pure unbook, I think, so I don't mean to hide behind
this concept to excuse any error or any idea that might have been
amended on further reflection or with further input. Yes, we asked
people to review and comment on the wiki, and the manuscript underwent 
a more formal vetting process, reviewed by outside experts, and so on.
Of course there are blind spots and people skim and focus on things
they find more salient. 

Had we identified Ari as a reviewer at an earlier stage we'd probably
have been able to strengthen some of the patterns to capture areas
where he views us as having made oversights or to be giving wrongheaded
or half-baked advice. 

One of the terrors of publishing is the criticisms that are bound to
come in after the trees are felled that you would have given your right
arm to hear in time to address the first time around.

That's why I hope to unpack these questions and work with Erin to
figure out where we think we could improve what we've got so far,
annotate it on the wiki, where necessary submit formal errata, and then
in future editions make changes. 

That's why we also hope to drive readers to the website where they can
discover when ideas from the book have been undergoing review or
rethinking.

None of this undercuts the legitimate concern that there is an
implicit promise from a book that the contents are authoritative,
thoroughly reviewed, reliable, and so on, and I would be very concerned
about failing to justify or earn that confidence.

In fact, A Pattern Language gives each pattern 1 to 3 stars (or dots,
I forget), reflecting the degree of confidence the authors had in them.
If just 1 star, the implication was that the pattern is attested in
the wild (Alexander's threshold was low: he just needed to see three
real examples) but if 3 stars the patterns was actively recommended.

The Yahoo! Library similar denotes patterns as Beta, Working Solution,
or Best Practice. The book probably should use some similar scheme to
signal where something is strong and fierce vs. whether it's tentative,
emerging, subject to controversy.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #58 of 97: Christian Crumlish (xian) Thu 5 Nov 09 11:18
    
<bbraasch> FB does seem to combine its use of the "public
conversation" (public in the sense of Many Publics and not necessarily
the One Big Public) with its relative critical mass to foster some
interesting ad hoc discussion often cross-pollinating people from
different walks of the original poster's life. 

I'm not sure how this culture of ad hoc conversations will mature, but
its intriguing.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #59 of 97: Christian Crumlish (xian) Thu 5 Nov 09 11:22
    
As to the extent to which we stood on the shoulders of giants and
acknowledged it vs. somehow perhaps squandered the accumulated wisdom
and reverted to infancy, I don't know for sure how to answer that.

Erin and I have both read widely in this realm
<http://www.designingsocialinterfaces.com/patterns.wiki/index.php?title=Bibliog
raphy>,
books, blog posts, magazine and journal articles, and so on. We've
also studied and practiced in this area and logged thousands of hours
of conversation. We may not have internalized everything and we have
almost certainly forgotten things we used to know. We try to point
people beyond ourselves to the wealth of supplementary resources, both
to acknowledge the richness out there and its contributions to our work
and to liberate our project from having to retrace everyone else's
steps. 

We didn't write a "how to launch/manage" an online community book.
There are great ones out there, going back to Powazek and all the way
up to our sister title at O'Reilly, Gavin Bell's Building Social Web
Applications. We'd rather point to them and fill in some of the gaps we
found (with a systemic view of the entire landscape being one of those
gaps).
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #60 of 97: Christian Crumlish (xian) Thu 5 Nov 09 11:23
    
got to jump on some work, but will try to get to 3 and 4 and 4a. later
today.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #61 of 97: Ari Davidow (ari) Thu 5 Nov 09 11:25
    
I want to emphasize that I think that the effort to collect these patterns 
is needed. My comparison of what I think I know with differences in 
what I saw in the book raised questions, which I am sharing, but I 
obviously found reason to dip into the book quite a bit to explore.

There are some things in the book with which I disagree, or which I feel 
are flawed, but I don't want that to come across as unalloyed criticism. 
As I put together short notes for next semester's class, I am acutely 
conscious of how long it takes just to string together a few words and to 
try to articulate a few critical points.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #62 of 97: Erin Malone (erinmalone) Thu 5 Nov 09 11:31
    
Hi all - Am going to jump into the conversation and hope I can catch
up after a week of amazing questions and responses from Christian.

For folks who don't know, I am the co-author of Designing Social
Interfaces. I was formerly a Senior Director at Yahoo!. I was
responsible for founding the Pattern Library and bringing people like
Bill Scott and Christian Crumlish into the company to curate the
library.

I have been designing social applications and experiences since the
early days of AOL, working with early AOL Greenhouse partners
(Sweatnet, Nutribytes, Housenet, Astronet) back in 1994, 95 and 96. My
move to the web and designing community started in 1997 and has grown
from there with stints at Zip2, AltaVista, AOL and Yahoo! where my team
worked on the social platform for the company and had to approach many
of the social experiences in component contextless ways at huge scales
in order for the solutions to be applicable across Yahoo!'s many sites
regardless of context.

So I am going to try to address a few of the questions asked earlier
that Christian was kind enough to punt on (saying Erin will address).
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #63 of 97: Erin Malone (erinmalone) Thu 5 Nov 09 11:45
    
I want to go back to Ari’s question about the Registration pattern.
This pattern comes out of a couple of years worth of design and test
and design and test work done on Yahoo’! registration process. It is
well tested and in the context of Yahoo! it was extremely successful.
The sheer scale of the Yahoo! network gave us a lot of data to work
with to not only improve the usability of the form but also to see
measurable impact of the design improvements in terms of drop offs and
completions. Unfortunately, Yahoo! requires more security protocols
than many purely social sites which makes the sign up process feel a
bit long.

That said, this pattern was not written specifically for engagement on
the social site. I think that is a flaw and something I would very
much like to go back and rework if we are given a second opportunity to
do so. I have been thinking a lot about engagement and how we get
people into our social systems in the first place. Whitney Hess’s
sidebar essay
[http://whitneyhess.com/blog/2009/10/06/onboarding-a-sidebar-in-designing-socia
l-interfaces/]
touches on a lot of things that I think could be rolled into the sign
up process patterns. Joshua Porter, who wrote Designing For the Social
Web, has evolved many of his recent presentations to focus primarily on
Sign up and the usage lifecycle in the social environment. So it's an
important topic that I think designers are just now getting a handle
on. Something we could learn a bit more from marketers about.

I think there is definitely need to deconstruct some of these patterns
some more and add back into them techniques for progressive
engagement. I do talk about that as a consideration but it really is
just a bullet point and note pulled out with very much emphasis.

I won’t speak to the weirdness that is flickr and yahoo sign in and
sign out and I think in retrospect I should have chosen a better
example for the illustration of Sign Out (which I think is a valid
pattern despite the illustration).
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #64 of 97: Gail (gail) Thu 5 Nov 09 12:16
    
How great to see you here, Erin!  Nice pointer to the Onboarding piece
(with enhancement ideas WELL staffers covet acutely around here, with
our hard-to-change layered antique interfaces). 

Good point about marketers and social designers -- in fact most likely
each has knowledge the other needs.   
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #65 of 97: Christian Crumlish (xian) Thu 5 Nov 09 14:17
    
thanks, Erin, for dropping by and adding to the convo!

lesse, next item in my list was to do with the Yahoo!-ness of this
book. 

I follow twitter for mentions of this book or related keywords and
often see cool slices of life in which people report receiving the book
or reading it (on a train, a lot of the time, it seems) or even better
say they like it or quote some idea from it, but recently someone said
something like "I don't want Yahoo! social patterns. Give me Facebook
and Google." 

I didn't reply but could have told him that examples from Facebook
(and to some extent Google, although their social offerings are a bit
of mixed bag) throughout the book, and yes we use a lot of Yahoo!
examples and yes the book is part of the Yahoo! Press imprint we
produce with O'Reilly, and yes the work I do (and that Erin launched)
on the Yahoo! Pattern Library informs this project, but the book is
actually *not* intended to be read as a guide to how Yahoo! does social
design.

We deliberately took the social patterns project into an independent
wiki where we could incubate the ideas outside of the Yahoo! branded
space. 

Bit by bit, as appropriate, some of these patterns are beginning to
show up in the Yahoo! library I curate, which means that they are then
more formally embraced as Yahoo!'s recommended approach. 

Even then there will be cases (as with Your vs. My) where Yahoo! will
follow the recommended pattern in some instances and not in others.

So, yes, at times there is an element of "do as I say, not as I do."

If I see Yahoo! doing something and the rest of the web doing
something else and I think the rest of the web is correct, then my
pattern is  going to reflect that, and I'll engage with Yahoo! product
designers to try to move us toward what I think is the preferred
pattern.

This touches on Jon's question about process. The pattern library has
never had an enforcement arm. We have had pattern ratings that escalate
in terms of "level of adherence" but again it was mainly an honor
system and a challenge for the pattern curators past and present to
move people toward the preferred patterns through communication and
social governance.

Some of that is because of the federated nature of Yahoo!'s products
and platforms and the way design folds into product and collaborates
with technology and marketing in various complex ways.

In fact, we have a newer process inside Yahoo!, known as ONE (for One
Network Experience) and spearheaded by Luke Wroblewski and governed by
something called our Design Council. 

I sit on that in my capacity as the "patterns guy" but most of the
other participants represent functional groups for which they are or
represent the lead designer, or related disciplines with a stake in
these standards, such as front-end development or editorial.

The items in the ONE library can be viewed as components. They are
described in a pattern-like way but are specific to the Yahoo!-context
and don't have to do double-duty the way the pattern library once tried
to do. 

At the same time this frees that pattern library up to capture
Internet-wide best practices for interaction design and leave the
Yahoo! brand and consistency work to ONE. 

In that light, I am auditing the huge internal library (it has about
150 patterns) with an eye toward migrating many of them to the open
library (which currently has 50 patterns), and retiring the ones that
have become obsolete or superseded. 

We think we'll get better evolution of the patterns if we share them,
even when in an unfinished "beta" form, with the worldwide web design
and development community, instead of harassing a relatively small
number of very smart but very busy user experience designers who work
at Yahoo! to write and edit and review all the patterns.

Not sure I fully answered Jon's question but wanted to clarify a bit
the role of Yahoo! in the book and the larger social patterns project.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #66 of 97: Christian Crumlish (xian) Thu 5 Nov 09 14:22
    
Embedded in the question that inspired that last reply was one about
how or to what extent we test the patterns in the Yahoo! library.

Erin addressed that a bit when talking about registration, which we
test and measure obsessively, as it directly impacts engagement and
lots of other important numbers for us. As I may have said already,
without getting people signed up and logged in there is no one there.

Other patterns we test as much as we possibly can. It's not always
feasible to isolate a single aspect of a user interfaces and test all
the variations that might be candidates for patterns. Often we are
forced to make surmises based on tests that had other moving parts, but
we try to justify our patterns with hard evidence, research and test
results paramount among them, whenever we can.

Up to now we've never been able to share the research behind the
patterns, even the ones we've taken live. To some extent this is due to
confidentiality issues in research but that can generally be addressed
through summarizing, etc. (although then you're back to a level of,
sort of, "trust us - we tested this and it said x") but there's
probably also a bit of a sense that research is proprietary.

I probably shouldn't speculate in public about whether that might
change, so I won't.


3. question of how tested and confidence level of these patterns and
related to yahoo's own standards vs. the web's
(relates to question of what influence and impact patterns have on
yahoo product design), "like the flickr logout I detailed above are
simply inconsistent and don't connect to 
patterns", and others just feel incomplete--this is what we do--which
isn't 
the same, with the exception of the anti-patterns--as saying, "we have

good reason to believe that this is worth copying.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #67 of 97: Christian Crumlish (xian) Thu 5 Nov 09 16:30
    
ah looks like i copypasta'd in my notes to myself from 3. below above
there - comes out a bit like word salad!

anyway, i think the remaining threads from ari have to do with
possibly more on registration, such as the flickr/yahoo thing,
expectations of other possibly flawed patterns, and the more recent
comments on forum patterns, so to be fair to subsequent questioners, I
think I'll leave this bundle of ideas where it is for now, pending
further follow-ups, and continue unwinding the backlog.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #68 of 97: Christian Crumlish (xian) Thu 5 Nov 09 16:34
    
ok, feels like the hot open questions now are jon's

"You've been evolving this catalog of patterns within a company that's
constantly evolving new interfaces and interactions, so I assume that
your library is getting some use. Can you describe how developers
typically interact with you, Erin, and the pattern library to determine
the right feature set for a particular application or site?"

and Ari's question and subsequent fleshing out from jonl and others
about the differences between forums as places for ongoing discussion
and this kind of hit-and-run/drive-by commenting that has become
perhaps more the norm since the advent of blogging. Great topic that
I'll need to chew on a bit!
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #69 of 97: Christian Crumlish (xian) Thu 5 Nov 09 16:39
    
to continue addressing use of the library inside Yahoo! I'd say that
our goal (and we have some success with this) is that when designers
start work on a new project (whether it's a fresh product or a feature
or an upgrade or fix or whatever) that they consult the pattern library
(as well as the ONE component library) to see whether any of the
aspects of their work can take advantage of learnings we have already
captured.

So I often get people asking me if we have a pattern about x or for y.
When we do, I direct them to it and if asked discuss their project and
help them determine to what extent the existing patterns may inform
it. If several people come looking for a certain type of pattern and
don't find it, I bump up that requested pattern on my priority list.

Developers tend to approach this stuff via YUI, the code library that
is our sister resource, but sometimes they cite patterns (or
components) directly, pointing to them and saying "we are following
this approach." In my experience, developers are at least as positive
about the pattern library as designers are, even if they don't need to
read the patterns quite so closely to apply to their own work.

When possible we also make design stencils for our patterns (see
<http://developer.yahoo.com/ypatterns/about/stencils/ for kits in
several versions), and both designers and developers *love love love*
these, both inside and outside Yahoo!

Sometimes I'm also asked to sit in with a team or meet with them, look
at their current challenges, and tell them if I am aware of any
patterns that may help them decide between different approaches or
solve sticky problems. When I meet with folks like this I'm usually
also looking to see whether they have invented any solutions of their
own that may prove to turn out to be useful patterns for others.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #70 of 97: Jon Lebkowsky (jonl) Fri 6 Nov 09 07:41
    
Warm welcome to Erin; delighted you could drop by and join the
discussion!

Following up with Christian - I want to note some thoughts I was
having as I read these later posts about what happens inside Yahoo and
how it relates to the pattern library. Not really a question, but I'm
sure you'll have some thoughts to post in response.

I recall hearing a few years ago how Yahoo had through acquisition
evolved a crazy quilt of platforms that weren't integrated, and the
company was trying to create that integration through internal analysis
and restructuring. I guess the most visible aspect of integration for
many was the migration of Flickr from its own authentication process to
Yahoo's. Google's faced similar challenges (think Blogger
integration).

This got me thinking about how we've been organizing the contemporary
Internet post dotcommunism, post social software/web 2.0 genesis in the
early 2000s. We have a very few hypercapitalized giants like Google
and Yahoo who have grown through acquisition, and each acquisition
presents its own integration challenge.

This makes me think of my days in the late 90s as "Internet Guy" for
Whole Foods Market. Whole Foods is another company that's grown through
acquisition and faced the challenge of integrating other companies
that had different cultures and modes of organization. So there's that
aspect of growth, too - the absorption of whole other cultures and how
you blend those into your mix without killing vibrant and useful
aspects of the culture. 

I'm writing this as what I think is an important contextual note: I
think you're in a place where it's important to have standard thinking
about interface and social practice, but the "no enforcement" point you
made is interesting - you don't want to kill the creative goose with
mandates. I think that's the right decision, and I think there's
probably a whole other book that Yahoo could write about how you merge
cultures and platforms, how you collaborate your way into an approach
that's identifiably one company's without suppressing the identity and
the utility of the component companies. How you make Flickr part of
Yahoo without killing Flickr. (vs the Borg antipatterns.)

I think it's interesting to hear your remarks about how you make the
rest of the company aware of the library and its utility. I guess you
have to market persistently, internally, to ensure that the people who
could really be helped by your work are aware.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #71 of 97: Christian Crumlish (xian) Fri 6 Nov 09 08:17
    
jon, you nailed a couple of really interesting issues there. 

let me just say i'm not sure yahoo should write that book (yet?) - the
whole complex semi-intrapreneurial post-20th-century organization is
still figuring itself out as best as I can tell. mostly I can report
what it's like to be playing one small hand in a much larger game.

I sometimes wonder if Yahoo! needs to be so big. Or to put it another
way, does the Internet need a big Yahoo! to exist? Need may be too
strong a word. Does the Internet want big companies to thrive, and
possibly to drive policies and standards? I don't mean so much the
politics of slashdot and boingboing and reddit and so on but the very
grain of the web. 

The web definitely likes openness, for example, so I'm glad Yahoo! has
this steady drive to explore what can be made open and turned more
web-like in nature. The web, in the long run, doesn't seem to like
walled gardens, by contrast.

The case for a big Yahoo! is as a sort of helpful guide providing
ordinary people an integrated but ever enriching interface to the best
of the Internet (content, services, experiences). Not an intermediary
that prevents direct access to anything, but at best a helper that
gradually raises the skills and capabilities and
imagination-about-what's-possible of its users.

Bigness has some value. Scale, reach, uptime (SLAs) reliability, a
well populated social graph (one of the keys to Facebook's ascendance),
and of course an engaged, understood audience for publishers and
advertisers (both concepts taken broadly: people who earn their reward
by entertaining you or giving you information and people who want you
to know something).

Bigness also can be tremendously challenging, and hypergrowth is
painful on the joints. When I joined Yahoo! in 2007 it was already
limping a bit from strained ligaments, so to speak. Since then, some
major steps have been taken to really get the company technically on as
few integrated platforms as possible. There was always a desire to do
this but those centripetal impulses that Jon noted, and the
countervailing virtue of unencumbered experimentation, agility, even
rogue skunkworks type projects and so on tended to militate against
consolidation.

I think when they brought my grest-grandboss, Jay Rossiter, who came
from an enterprise background, to run the open strategy and ultimately
the entire consumer platforms group; and my great-great-grandboss, Ari
Balogh, also with an enterprise background, under whom Carol Bartz the
new kick-ass ceo consolidated the cto and product czar roles; I
realized that Yahoo! was getting serious about squeezing out the
wasteful redundancies and unclear lines and building a large muscular
platform for concierging the web (taken broadly to include mobile,
ubiquitous, and so on).

It's a bit of a matter of big agility vs. little agility. In the past
I think some small teams at Yahoo! maintained their agility by avoiding
entanglement with others as much as possible (contrast that with my
desire to constant cross-pollinate discoveries and plans). But this
little agility often, I think, hampered strategic or network-wide
agility. Consolidating systems and processes will enable in a concrete
way things like rolling out advances in media page designs to all of
the international versions of our pages with a single upgrade rather
than (I kid you not) a multiple-month rollout plan across scores of
platforms.

So on the engineering side I think there's been a need for
enforcement, a top-down commitment to doing things the right way and
eschewing expedients.

The Design Council process is a bit more negotiated and is a closer
parallel to the effort on the engineering side to agree about standards
and practices. We try to make sure every stakeholder has a voice, some
group takes the lead on a proposed component or package of components,
everyone gets to point out potential concerns, there is revision and
further review, driving toward a conclusion. 

If one group says "this pattern doesn't work for us because we share
have x y z unique constraints" the context can be specified to exempt
the scenario they're talking about, or they can simply be allowed to be
out of compliance. 

I tend to think of it as an 80/80/80 thing. If we get 80% of the
council to agree on a component, and 80% of the UED (user experience
design) teams hear about it and apply it, and if a team is able to
follow the pattern 80% then we're still so much closer to a consistent
experience for Yahoo! users that's consistent with our brand than if we
continued with a mishmash of occasional top-down brand initiatives and
an entirely emergent, curated pattern library.

and yes, I am evangelist outside Yahoo! and one inside as well. Every
couple of months I visit another office or team and preach the gospel
of pattern recognition. 

Notice something working in several places. Figure out why it works.
Write down what you know and tell everyone.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #72 of 97: Christian Crumlish (xian) Fri 6 Nov 09 09:04
    
Don't want to delve directly into conversation patterns but to step
back a bit to address this question of losing ground, the models
degrading.

To me one of the motions underlying the whole BBS/forum ethos and that
of disconnected, ad hoc, perhaps pseudo-conversations is a tension
between communitarian and libertarian impulses, seriously!

At various times and places online the favored context for promoting
community has been a persistent, asynchronous (or sometimes
synchronous), small-public, conversational discussion system. A BBS,
the Well, Usenet, the old Yahoo chat rooms before they were inundated
with spambots*, any number of other room / topic / conference / thread
services.

This is the communitarian model. Some benevolent entity provides the
space, all enter it equally, although senior participants may have
gained some status, etc.

The only real problem with it (unique, I mean, not present any time
people speak to each other) is that it has this more or less visible
owner, sometimes the person who started the mailing list, etc., and if
they ever are not benevolent the community can be damaged or destroyed
or individuals in the community may be wronged.

You could probably create a governance model that minimizes this
problem. They probably exist.

The other model, which ranges from comments on blogs (where there is a
primary participant in the conversation who starts every thread), to
the 2004-esque "everyone have a blog and ping each other when your
rebuttal is ready" to today's drive-by ephemeral conversations, seems
clearly inferior in that it does not tend to create longer standing
communities with persistent cultures and stories and dialects they
co-create, or at least not so readily.

What's been traded is the monarch. "Every man a king." It's the
techno-utopian solution that says each person will host their own
salon, write their own broadsides, participant as an equal in the grand
conversation. Very enlightenment. However, that was the demo and at
scale it's proven to be difficult to sustain.

Since LiveJournal at least the blog concept has been playing around
with subscribing or following, watching lists of people, all the stuff
twitter does now and Facebook is flirting with (Facebook flirts with
all of these models). Even in 2005 or so people were joking around
about the blogosphere rebuilding usenet badly. So the disjointed model
is still evolving an bolting-on features to address the inherent
antisocial isolation (every man *is* an island) of the libertarian
model.

Facebook right now seems to be a weird amalgam, thriving but
exhibiting some of the unhappy features of each model: the paternalism
and infantilization of the communitarian model and the disjointed
ephemeral small-public conversations of the libertarian model. 

Somehow it's working for them, this anti-dialectic.


----
* speaking of spambots - one early definition of social software was
"stuff that gets spammed"
<http://many.corante.com/archives/2005/02/01/tags_run_amok.php>
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #73 of 97: Gail Williams (gail) Fri 6 Nov 09 10:04
    
That's a good statement of some of the political assumptions embedded
in social code.  Those assumptions are there, like it or not, and often
are conflicting and not thought out.  Combine the rules and who is
granted permissions with the language used, and sometimes you see very
odd mismatches.

The thing I find odd about libertarian blog comment or founder-driven
group comment space is the lack of support for continuity that Ari
mentioned. How many times have people in the halls at conferences
bemoaned that devolution of function?  I think that core functionality
is the seed of the "blogosphere rebuilding usenet badly" effect, and I
think that people are going to start to address it.  (I believe that
<jef> has made a tool to make Flickr group discussions easier to
follow, for example, thanks to the Flickr API and policies ...  I
expect more of this is going on too.)
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #74 of 97: Christian Crumlish (xian) Fri 6 Nov 09 11:09
    
yes, that's why an open API is so important. I think twitter's
minimalism and relatively easy API has a lot to do with its uptake.
  
inkwell.vue.368 : Christian Crumlish, Designing Social Interfaces
permalink #75 of 97: Jon Lebkowsky (jonl) Fri 6 Nov 09 14:10
    
Bijoy Goswami, who favors a paradigm of simple personality types as a
shorthand for understanding teaming and collaboration, says that there
are three relevant types: evangelist, maven, and relater. Generally
those are driven by action, knowledge, and relationship, respectively.
He and I have done some collaborative thinking about how these apply to
communities. Most people are somewhere on a scale between two of the
three types - e.g. I think I fall on a scale between maven and relater,
closer to maven; Bijoy is probably a maven/evangelist. We realized
that online communities like the WELL tended to be strong relater/maven
communities - focused on relationship and knowledge, but not much on
action. I think this is another way of describing the communitarian
model, and of course it works well for people who are more focused on
knowledge and relationship, but it can drive action-focused evangelist
types nuts.

I'm wondering if the "drive-by" patterns we've discussed are more in
the action realm, creating a sense of activity rather than a sense of
ongoing conversation and relationship.

I think people are confused about 'social' and 'community.' They might
think a social network is a community, but it takes more than social
connection to make a community happen. Companies that talk about
building communities sometimes have a valid reason for doing so, but
sometimes they really want to build, not communities, but audiences.

Those are just some observations. I also thought of another question
while I was reading, and this goes back to the new vs old question.
Yahoo has a bazillion people using its Groups technology, but I'm
running into groups and organizations that need everything Yahoo Groups
has to offer, but they don't want to use the functionality, and their
only professed reason for avoiding it is that it's old tech, therefore
uncool. How do you think about making technologies that definitely work
as they are fresh so that they seem contemporary? Could that be a
matter of pattern assessment, followed by a reimagining of the platform
based on emerging patterns?
  

More...



Members: Enter the conference to participate

Subscribe to an RSS 2.0 feed of new responses in this topic RSS feed of new responses

 
   Join Us
 
Home | Learn About | Conferences | Member Pages | Mail | Store | Services & Help | Password | Join Us