roundup: Future of Web Apps 2007 / Digg to support OpenID
fowa2007 - day #1
11.40 - Simon Wardley on commoditizing of IT: “there’s no sense whatsoever in setting up your own IT-infrastructure. maintaining your own web- or database-servers is what we call yak-shaving“. Zimki offers “pre-shaved yaks”, meaning a fully hosted, server-side javascript environment for rapid application development (not to forget Amazon S3 & EC2, which commoditizes web-infrastructure on a lower level).
13.45 - Matthew Ogle & Anil Bawa Cavia talk about attention data, tag cloud moderation and past, presence & future of Last.fm. they also shared some stunning usage numbers:
15mio tracks are scrobbled each day (that’s
1575 174 /sec or 6billion tracks since 2003). there are 190mio artists, 70 mio tracks, 700k tracks available for streaming, 145k artist wikis.
14.45 - CTO Amazon Werner Vogels sells us on S3, EC2 & SQS (Simple Queueing Service) and encourages companies to compete on ideas, not on technology. currently, the 30/70-switch rule demands 70% resources spent on the technological “heavy lifting”, while only 30% go into the product/idea. posterboy S3-implementations include SmugMug, SecondLife (just for client-downloads), YouOS; renderRocket (a virtual renderfarm) is one of few current EC2-users. interesting detail: Mechnical Turk’s not internationally available because of conflicts with labour-laws.
16.20 - according to this post, Kevin Rose is about to make an announcement regarding digg later today. stay tuned for real-time updates 
17.20 - Rose enters the stage… his talks’ topic is “the future of crowd generated media”…talking about the Digg-ecosystems, motivations for users to digg & submit content - Digg trys to buid incentives for users on every level (frontpage-exposure, traffic, networks of friends etc). he’s mostly summarizing what Digg has been doing the past two years. enhancing Digg (stuff they’re working on): how to manage a user-community of 900k, how-to implement fact-checking on stories (Digg does community-based, since a paid employees a la netscape.com won’t scale). Digg wants to enable/create connections between users based on several numbers - Diggs, buries, comments, word-strength etc. showing some slides on development of Digg-usage since 2005 + summarizing their flash-tools + announcing a “Flash Toolkit” later that year, which should help publishers monitoring their stories. commitment on APIs & export of user attention-data. finally the announcement: Digg is going to support OpenID in the future (also see TechCrunch)
that’s it for today…Wifi didnt really work very well, plus it’s not free 
maybe we’re more lucky tomorrow.
You're a kickass perl- or php-developer in the area
of vienna looking for a gig? get in contact!
Maybe you could explain me sometime
“15mio tracks are scrobbled each day (that’s 1575 /sec)”
15 million divided by 86400 seconds is about 174, not nearly 1575. Confusing.
of course you’re right, sorry for the typo.
I think 175 was the number they had on their slides…
Ah ha, no WI-Fi explains the lack of updates during the day. Hopefully tomorrow.