Joe Martin Report
Sunday, May 14, 2017
Tuesday, March 16, 2010
SXSW: Is Canvas the End of Flash?
progressive enhancement from tables to pie charts.
Haase: The canvas api is good, but there aren't any tools for it.
It was interesting to see a few people that I've been listening to on podcasts for a few years. Admittedly it's been a while since I've heard Ben Galbraith's voice, but he looks and acts just as you'd expect. He's an evangelist for sure.
The takeaway is that canvas cool, but lacks the browser support to be a mainstream tool for traditional browser development. I also noticed that the cpu usage on my MBP goes way high when I fire up the canvas demos in FF3.6.
SXSW: Can You Run a 'Serverless' Business?
Know your application. The application characteristics determine if the cloud is right for you.
Ah-ha idea: cloud for government level security.
This panel was interesting and my notes don't really do it justice. I think the really thought provoking ideas were that your infrastructure should be transparent or a tool, not a hindrance. With some of the panelists representing sites that can have 20x traffic spikes, the takeaway quote was, "Don't let your best day become your worst."
This has inspired me to try to go further than I'd planned with our server monitoring and deploy automation. Trouble is that most website failures are a failure of code, not of hardware...
SXSW: Prototyping Web Apps - Nobody Loves a Wireframe
The guys at Google prefer prototyping. The ppt was pretty good so hopefully they'll post that soon.
hypercard, fireworks
Monday, March 15, 2010
SXSW: Augmented Reality - Gimmicky Trend or Market-Ready Technology?
Audience was split 50/50 on whether Yelp's Monacle is a good idea or not.
Good ideas: Lego, USPS, Google SkyMap
Demos: John Mayer's Heartbreak Warfare, DollhouseAR, Zugara's social shopping app
This panel only helped reinforce my opinion that AR is a gimmick in most cases. It can definitely be a great idea when the technology fits and it's executed brilliantly (like Zugara's apps).
That's all I have to say about that.
SXSW: Beyond the Desktop: Embracing New Interaction Paradigms
Minority Report and Iron Man were used as good example of futuristic interaction paradigms.
Industrial/Mechanical technology 1 input 1 output. Post industrial/digital technology.
This panel got really interesting as questions were asked and conversation developed. The crowd seemed a bit awed with Johnny Lee Chung (of WiiMote hacking fame), but to his credit, he did have some rather insightful slides. David Merrill also had some interesting comments. I'd see him before (at least I think it was him) on TED videos demoing his Siftables.
I don't think Johnny Lee is going to publish his slides, but the thing I found most striking about them was late in the deck, when he talked about the range of interfaces, from an iPod Shuffle, to a in-dash navigation system in a car. Any beyond a desktop computer we begin to use another, smaller device to control. Any device as small as a phone the trend is to use touch, so we can have input and output in the same space. Keyboard and mouse has been so effective for the space in-between a phone and TV that it's unlikely that a new paradigm is going to over take it. Where the challenge becomes more interesting is where you want to have more than one input to your device. Collaborative programming was the example used in the session. Food for thought.
SXSW: Revealing Design Treasures from the Amazon
Bring when the user is going to receive the item to the fore.
1. Engage through content: reviews. Target vs. Amazon (they're built on the same software). Target shoppers shop at target.com for things they would buy at Target. The users don't think to go back and leave positive reviews.
Reviews are integral to Amazon's success. The 'was this review helpful' question is estimated to be worth 2.7Billion to Amazon in 2009.
Ratio of reviewers to purchasers 1:1300. 20 reviews is the tipping point where reviews gain credence. 2% conversion of viewers to purchasers. You need 1.3Million viewers to get 20 reviews.
2. Don't fear new ideas:
Experimental critical to moving forward.
3. Eliminate Tool Time:
Don't have to authenticate.
Security Levels:
- Level 0. No cookie.
- Level 1. Amazon knows who you are from a cookie. You can
- Level 2. Amazon wants to reveal something only you should know - like address.
Goal Time vs Tool Time:
- Goal time is when the user is improving the outcome of the experience. ex: user is considering the products they wish to purchase.
- Tool time is when the user is moving forward without any improvement in the out come of the user experience
Changing the interface adds tool time. Amazon chooses to phase in interface changes. Took 12 weeks to roll-out updated navigation.
Not all use cases are the same. Salsa: Amazon vs. CD Baby.
Negative operating cycle. Amazon turns inventory in 20 days. BB turns in 74. Standard payment terms are 45. The days between day 22 when they're paid and day 45 when they pay vendors, they have cash float. Amazon can survive on that.
Playmobile security checkpoint, laptop steering wheel here.
Jared Spool did a great job presenting. Not only was the content inherently interesting, but he made it more compelling than it would have been otherwise. You can relive the experience here.
SXSW: Making Content Relevant To Me
I got to this late cause Dave and I bailed from the interview seminar.
How do the recommendation engines account for aging? They need to forget as the user gets older.
De-anonymizing Social Networks: article
The discussion in this panel was really interesting and therefore few notes. Twitter was popular with this crowd as well and I picked up this very good summary there: http://pattyryan.tumblr.com/post/450169096/making-content-relevant-to-me-sxsw
SXSW: Conducting Great Interviews
This was a bust. Too diverse a group in an open conversation and nothing was getting accomplished.
Sunday, March 14, 2010
Mashable Party
The Mashable Party was at Buffalo Billiards, a pretty cool, dual level sports bar. I think the nice foosball tables would make a few people in the office jealous. ;)
SXSW: Gmail: Behind the Scenes
Jonathan Perlow - Set out to make an email client that he would want to use: never have to delete mail, make the UI feel as sleek as desktop app.
Arielle Reinstein - leads the gmail blog, the public relations of gmail.
Edward Ho - Buzz lead, sense/strength/quality of team more important than the quality of the code. Have often code reviews to make sure people are executing as well as ideateing. The team sat unnaturally close to each other so that im & meetings were unnecessary. Also, there was less confusion, because everyone is aware of everything and able to contribute to decisions.
Todd Jackson - Google Gmail/Buzz project manager. Leads the group through consensus. Shit funnel vs. shit umbrella. "You have to be the shit umbrella."
Braden Kowitz - User experience at Google Ventues.
Google is a company of do. When someone has an idea, they do it (prototype, mock-up, etc.) Good example is the gmail un-send. That was a large internal debate and an engineer who wasn't even part of the core team implemented the feature into labs.
Google has on-demand build/deployment so they can test a full build of gmail with their real user accounts.
Latency is a critical issue. All new features go through testing to make sure latency isn't added along with the feature. Added latency can kill a new feature.
Listening to all the feedback from all the users is a challenge. Everyone on the Gmail team uses reader feeds to keep a pulse of the conversation about Gmail.
90%+ unit tested.
"We have this amazing technology called 'work really, really hard'".
Really extensive summary of the seminar:
SXSW:Experimental Design: Your User Interface Is Your Laboratory
When I heard that Kent Brewster was going to basically have a redo of last year's "Kick-Ass Mash-Ups", I decided to head to this panel on test driven design iterations. I'd hoped for some higher level discussions, but it kinda turned out to be an A/B Testing 101.
Example:
Do screenshots convert better than video?
A/B test for a couple days and results showed that screenshots converted better.
Interesting concept put forth by the FreshBooks and Google rep: The goal of all the testing is to keep more people in the funnel through out the process...basically to plug the holes in the bottom of the funnel. Google wants websites to have higher conversions (make more money from each entry to the funnel) so that they can get more ad business.
The moral of the story is to use funnel analysis to determine where you should do the A/B testing. Then, taking into account the user cycle, perform the test for at least one cycle. Also, testing is just a single component of the testing (personas, interviews, etc.)
I should have just stayed in Kick-Ass Mash-Ups.
http://my.sxsw.com/events/event/497
SXSW: Sunday Keynote - Valerie Casey
The Sunday keynote was a talk about the interactive community's role in sustainability. I figure I'll put up these notes and come back time permitting.
Why does a salad cost more than a big mac? NYTimes.com
Iraq burn pits
The Designers Accord:
personal accountability
collective accountability
FastCompany.com
a system is more than the sum of it's parts
feedback delays + bounded rationality = design traps
no such thing as a side effect
creating the right measurements of success
selecting the correct lever for change
enabling new models by recognizing the relationship between structure and behavior
Issue attention cycles: degree of awareness is inversly correlated to the degree of production action
Cannot get a new bahavior from a system without changing it's structure slinky vs. box
The Hub - place for social entrepenuers to collaborate
A system is a colleciton of elements and interconnections that are highly organized to achieve an overall goal or purpose
every profession bears the responsbility of understanding the circumstances that enabled it's existance
www.valcasey.com
SXSW: What If Your Phone Had Five Senses?
We're accustomed to a phone having the sense of touch, sight, sound, and location...temperature, voice?
Sleep cycle app, nearness nearfield.org
microvision: laser displays. Averch
Biddulph: the technologies that will takeoff are the ones with interfaces that are most familiar to us. Something foreign to human interaction isn't going to take off overnight.
http://my.sxsw.com/events/event/717
SXSW: Augmenting Maps with Reality
It was tough to get up this morning to make it to this session, but it was worth it. This is why I come to SXSW. Before the session started, I was browsing and found a few interesting sites for metro mapping:
Laura Diaz of Navteq: The base map is established, but as more geotagged content is shared the map begins to become a spacial
How about another axis? The z-axis? Z is unlikely to be elevation. It's more likely to be inside buildings or time.
Dennis Crowley of FourSquare: Indoor mapping is the next challenge, but likely most compelling: malls, stadiums, convention centers.
Kellen Elliott-McCrea of Flickr: We try to keep the map interface visual.
Ryan Sarver of Twitter: Difficult to represent non-specific or non-point data. Polygon is messy to render on a map. Pub vs. Yankee Stadium.
A challenge is to keep the map real-time. It's difficult to pull all the data together to get it into a public map.
The idea of dropping virtual things into actual space was a founding idea behind FourSquare. At SXSW, the ah-ha moment was to have the crowd shift embedded in a map. All the data is there, but the presentation layer could use a re-imagining.
There's so much temporal data, but it's typically not shown because it gets messy for an app to manage.
Considering the spotty accuracy of user generated geo-location data, at what point would maps become untrustworthy. For driving, this is important. In most cases, the mapping systems have nailed accuracy. For most other applications, the temporal information is more important. 8-bit Google maps.
Place data should be free. Place data is public data and in the public domain. Great Maps + great data should equal great social tools.
trendsmap.com
hot potato
Flook
app that'll remind you of something you wanted to do. Say it's 1pm. The app knows you haven't checked in for lunch yet and your phone knows you're active because of the accelerometer. It can remind you that there's a lunch place around the corner that you want to go to.
Geo fences. When you or a friend enter a space, the app can alert you.
Other links that came up on Twitter at the end of the panel:
http://my.sxsw.com/events/event/7270
Saturday, March 13, 2010
SXSW: Doing it Wrong: Recently Possible Technology
NYC Resistor & Garage Geeks:
Guitar Heronoid
Dove shitting Dove
#twittershiters
mp3 player grenades
breathing books
real need for speed console
sudo make me a sandwich
http://hackerspaces.org/
There were a lot of slides which I'll try to get a link to. This was an awesome unwind to the day.
SXSW: Playing with Place: Location-Based Games and Services
I got to this late...
Scvngr - Provides digital location based games for enterprises. Examples are the Smithsonian (take pictures in front of the shuttle, count teeth on a dinosaur, etc.) and the DARPA Balloon Challenge.
There were a few interesting ideas, but nothing really earth shattering. I pulled this summary from Twitter which gives more information than what I got: http://www.gearlive.com/news/article/q110-sxsw-2010-location-aware-games/
SXSW: Web Framework Battle Royale
http://seaside.st/ (SmallTalk)
http://www.sinatrarb.com/ (Ruby)
http://heroku.com/ (Ruby)
http://www.zope.org/ (Python)
This is pretty low level. I'm going to bail for Location Based Games and Services.
SXSW: CMS Admin UX
Steve Fisher. Success is when the technology blends into the background.
A site dedicated to the UX of Drupal 7: http://www.d7ux.org/.
Designing in the open vs. Just doing it.
Jane Wells. Works for Automatic contributing to WordPress admin UX.
http://my.sxsw.com/events/event/843
Subscribe to:
Posts (Atom)
