The Industry

Boulder Denver New Tech Meetup Report May,6 2008

This month’s New Tech MeetUp played to another overflow crowd. It seemed to start closer to 6:00 this time, as opposed to 6:20 last month - or maybe it was just that I had decided to come later, arriving closer to 5 minutes early instead of 15 minutes early.

Before the meeting there was held a paper airplane contest. “…the plane that stays in the air the longest is the winner” of “…bragging rights for the month of May.”

Well, you never saw such a set of un-flight-worthy things-with-wings… They all ‘tried’ to throw them at the same time, and that which would settle to Earth last would win. Personally, I think the wadded up piece of paper won, just because it could be thrown hardest and therefore travel the farthest and maintain a greater-than-zero altitude the longest. As it turned out, the guy with the cylindrical plane won. Good - he certainly had the most innovative design.

Then there was about 5 minutes of announcements - mostly job-wanteds and people-wanteds. I have a hard time hearing a lot of these - they are just regular people without a microphone in a very large room that is well dampened with sound-absorbing human beings. I did my best but please email or post any corrections you know of. sliceoflime.com is looking for people. 3DRadio is looking for volunteers to help write some Windows software. WebPartner.com is going into public beta. Mario Lurig announced that his PHP book is done and reminded people it would be a perfect gift for Mother’s Day. A site that sounded like ‘wayyagoingtobe’ is also now in beta. collectiveX.com is looking for UI Designers. And w3w3.com, one of this MeetUp’s sponsers, is doing something or another with some VCs [they have some interesting interviews with VCs here, in any case].

Test Common, Inc.

Presented by David Wipper
Test Common, Inc. delivers Testing as a Service (TaaS). Companies who need software testing services can access our global community of testers on-demand via an online marketplace and utilize a secure test environment to conduct tests in a cost effective manner.
www.testcommon.com

These guys are trying to be THE PLACE to go to for testers and people who want to get testing done. Right now they are still in the initial phases, mainly providing a match-making service between those who want testing done and testers. For example, this might be a easy way to add support for internationalization of your application, rather than trying to find and hire local people who speak German, French, etc.

There were some questions about the security of the code that would be released to ‘virtual’ testers - and although they do seem to provide a virtual secure deployment solution - there are just so many ways for testers to test that there seems to be more work done on enumerating all these possibilities and the solutions provided now and in the future. For example, explicitly delineating code reviews, test plans, writing regression tests, user testing, i18n, banging-on-it, etc.

I like this idea - one because it is Way Big, and I like way big ideas. And two, because testing one’s own software is like [bad metaphor alert] worrying a sore tooth. After awhile, it is just time to be over and done with it and have someone who has dedicated their life to doing it well just take care of it once and for all [until the next time :-()].

Trackvia

Presented by Matt McAdams
I’m going to demo our startup’s product, called Trackvia, which is a general-purpose online database platform designed for non-technical users
www.trackvia.com

Matt was a very clear, fast and articulate presenter. It was fun just watching him do the demo. Trackvia is a simple online database whose goals are to be very fast and very easy. The demo, and application itself, I think, focuses mostly on a single table view of things that are readily converted to and from speadsheets. Trackvia is perfect for HTML-centered websites without much of a server side who want to manipulate data of some kind - it is hosted elsewhere, backed up, and doesn’t require coding, which, face it, even the most simple PHP Fantastico apps do].

Some of the features are - Google-like search of the database, auto generation of HTML data input forms and data-output displays, audit trails, versioning, and support for geographically-inclined data (maps). The only time I have heard the audience go oooh and aaaah [in the all of two times I have made it to a New Tech MeetUp] is when Matt clicked on a button that generated HTML for a data input form [for a row in the database] and cut-and-pasted it into a sample HTML page for a sample website, and viewed the webpage displaying the nice, live, form [it is not the feature itself that is so interesting, to me, but people’s reaction to it - this being one of the easiest-to-implement features in all of the demos we have seen].

They used Real Estate agents as an example of people who need several types of data and easy ways to get at it [like lists of houses for sale and potential buyers]. We [I] really didn’t see how a potential user of Trackvia would be able to easily relate these to each other [i.e. which buyers were interested, or not interested, in which houses]. Five minute demos… they have their good side and they have their bad side.

In comparison with Intuiit’s QuickBase, Trackvia is easier, everything is just 4 to 5 clicks, faster, and more pragmatic [i.e. the stuff that 90% of the people need 90% of the time]. They achieve their speed by using lots of Perl and caching. They have 1000s of paying customers. They charge $10/month per seat [more or less, see their website for details].

Trackvia is VC funded and is actively hiring.

MapBuzz

Presented by Chuck Ray
MapBuzz is an online mapping site with a community wrap using web2.0 pipes to be used anywhere one may wish.
www.mapbuzz.com

MapBuzz did not make it to this MeetUp.

Yallery

Presented by Jennifer Ross
Yallery is the first Social Art Management product. Explore the relationships created by art between Collectors and Artists or Artists and Galleries when social networking solutions are applied to the world of art.
http://yallery.com/

Yallery is a social network site for artists, art collectors, and art galleries. As opposed to being exclusively commerce-centered like most of the competition, Yallery has features like listing an artists past (sold) works as well as those that are currently for sale. Currently Yallery is a Mac-only application and they are looking for someone to port the application to Windows.

Brightkite

Presented by Martin May
Location-based social networking
http://brightkite.com/

Brightkite is a location-driven social networking site that is in invitation-only/waiting-list beta mode. The demo showed a clean web 2.0ish interface that allowed people to be located by… location. In particular is their addictive ‘What’s Happening Around Me’ feature that finds other people registered at Brightkite [and events, like say, Happy Hours?] within some specified distance from one’s current location. They just added an iPhone app that has much if not all of the website’s functionality.

They offer privacy modes and the ability to label several different kinds of friends, as say either ‘trusted friend’, or the more ubiquitous, and suspiciously generic these days: ‘friend’. There was some talk and a screen displaying the ability to forward your location information from Brightkite to Twitter, if anyone might be interested in doing that.

Monetization possibilities included location-based advertising and what they described as Google-like analytics : the ability to generate statistics about where people come from and where they go.

Me.dium

Presented by Robert Reich
A prealpha demo of Me.dium’s search product
http://me.dium.com

Well, we didn’t get to see a demo - Robert finding himself password-less at one of THOSE dialog boxes, but we did get to hear about what they are doing - and it is pretty cool.

The current problems with search engines, as they see it, are that they:

1. Are static representations of the net
2. Are not people-centered
3. Are gamable
4. Require the building of very large indexes with very large numbers of servers

[I’ll get back to this list later.]

Then followed a lot of statistics that showed why this was worth getting into and why Me.dium has the requisite wherewithal.

[Hope I get most of these right]

Getting 1% [of the search traffic?] is said to equal $1B
60-80K of Me.dium’s 650K registered users make 14M webvisits per day, which is 25M minutes, using their tool, which then generates 25M various correlations between the users, their preferences and the websites they visited.

In this way, they are not using robot ‘crawlers’ to map the web, they are using people. They are measuring interest level verses number of links coming in [and all the other structurals that go into Google’s ‘rank’].

For example, search for Obama on Google and at the top of the list is the Obama Campaign home page. On Me.dium it would be [today] a video on Youtube. [Robert said that 50% click on the first link on a search results page. That is either a testament to how well Google does search [though I wonder if that stat is Google-specific] or a testament to how uncurious people are. My default Google results page has 50 results, and only about 15% of the time is what I want at the top].

Me.dium search is not keyword [and structure] focused, it is behavior focused.

In answers to questions, StumbleUpon requires a click and wants a review, Me.dium is invisible requiring no effort. Technorati is just blogs and Mahalo is just the top 100K websites [at some time in the past]. Robert thought digg and techmeme were the closest in spirit to the Me.dium approach.

OK, getting back to the list. Robert mentioned he wanted feedback during his presentation [well, now, maybe next time he will be more circumspect :-)], and this is a problem we have thought about some [like about a million other people], so here are some thoughts about this approach. Current search engines are:

1. Are static representations of the net

True, Me.dium will more of a ‘live’ view of what is happening on the internet - though this is asymptotically less important as Google updates information from ‘hot’ sites more and more frequently.

The problem with both of these approaches is that they reflect ‘lemming behavior’ by ‘tech savvy’ people to the exclusion of ‘excellence’ and ‘ordinary’ people. The cool, insider ‘car blog’ that true fanatics know is the best site for avid fans just doesn’t show up near the top of the lists in any of these search engines.

For example, Google could use this same technique, keeping track of what search results people click on when they search for ‘Italy Olympics’, and therefore be more ‘people-centered’. But, since most people only click on the 1st 10 links anyway, these links would just get more popular still and the search results would not be affected much [well, maybe WikiPedia would move even a little higher on the list :-)].

Google assumes people add links that point to their favorite websites, presumably by modifying their own website [or on Delicious, though that doesn’t really work for SEO, IMHO and now StumbleUpon - which does work :-)] which most [90%+? 95%?] people are not geeky enough [or have the spare time!] to do.

The good thing about the problems with these two approaches is that the door is still wide open for competition :-)

2. Are not people-centered

Well, both approaches are making assumptions about the meaning of what people do on the net. Links and content versus visits. Links and content can be gamed [really? :-)] and content can suck, even if it does have a lot of good keywords. Me.dium can be gamed just like Digg, and Me.dium’s users can visit pages they do not like, which will be recorded as a visit [I think], though hopefully less frequently than their visits to those sites that they do like [though some days do seem like bad-browsing days :-(].

3. They are gamable

[See 2) above.]

4. They require the building of very large indexes with very large numbers of servers

Seems like Me.dium is on the way to using some hefty servers with their kind of growth and data-collection habits, regardless :-)

My opinion, for what it is worth, is that Me.dium’s approach is cool, is worth doing, and I will use it - but that it is not a replacement for the Google-like approach, and that there is still room for other types of ’search engines’ out there.

© 2008 GUI.NET.