• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!

View
 

Anti-stealth projects

Page history last edited by PBworks 15 years, 8 months ago

A list of Next-NY'ers who are working on side projects and (following Charlie's inspiration) are choosing to go anti-stealth about their plans.

 

Some of these might be full-fledged startups; others just side hacking projects. Regardless, though, the participants are all inviting feedback and participation. If you have comments about their plans - or are interested in joining them - drop them an email.

 

And, of course, please feel free to add your project to the list!

 

Please note, by the way, that anti-stealth is something of a new practice, so even the participants here may not feel completely comfortable sharing all details of their work yet.

 

Summary

 

  1. David Rosenstrauch - Developing a scalable, distributed database
  2. Lindsay Watt - Trying to help people turn information into inspiration
  3. Laurent Kretz - Working on SubMate, Social commuting
  4. Lee Semel - Side project: CommandShift3.com - It's like hot or not, for web sites
  5. Eran Hammer-Lahav - Hueniverse building Nouncer

 

Details

 

David Rosenstrauch

 

Current project:

Former Amazon personalization guru Greg Linden wants a big virtual database ... and so do I!

 

Just about every major web business has found that the database becomes a major scalability bottleneck once the site reaches a large size. Caching helps the problem, but doesn't avoid it, nor does master-slave replication (since writes to the single master database remain a bottleneck).

 

Eventually all of these sites realized that ==partitioning (aka sharding)== was the only workable large-scale solution, as that enables the site to scale out instead of scale up But partitioning is a thorny issue with SQL databases, as it causes a number of other issues that must be dealt with. It also forces the use of other techniques that don't mesh well with the SQL database paradigm.

 

What's needed is a large distributed database, that can scale near-linearly with the site. A technology like this would help developers build - and scale - web applications much more quickly. Google knows this - and has already built one So has Amazon But they're not sharing. (Not that their code would be useful outside of their infrastructure anyway.)

 

And while there is an effort underway to create an open-source clone of Google's BigTable, 1) it's still in its early stages, and 2) it doesn't solve the problem properly, in my opinion. (e.g., requires Hadoop, can't use Amazon S3 for storage, doesn't implement some other cool advanced features like I have planned, etc.)

 

This provides a significant opportunity for someone to create a truly scalable database. It's my goal to be that someone! :-)

 

In the past, a large distributed system like this would have been infeasible for a single developer (or small team) to develop. (Or deploy, for that matter.) But nowadays a lot of the heavy lifting has already been done by Amazon and a number of open source projects.

 

Update:Yes, I know that Amazon has recently released SimpleDB to address this need. But there's a number of issues with their implementation and so I still think that neither they nor anyone else has solved this problem quite right yet. I have a different approach in mind that I think will make this technology much more usable (and useful!) to the average developer.

 

*Previous projects* (over the last year or 2):

ProjectDescriptionReason shelved/on hold
A new twist on a job search boardCan't be completely anti-stealth about this one just yet ... I'm still tinkering with it a bit on the side.On hold due to new focus on distributed database project.
An improved ORM frameworkCurrent Java ORM frameworks are good, but nearly all suffer from the same limitation: they only work with JDBC datastores (i.e., relational databases). They don't work with alternative databases or non-database storage engines and their API's aren't flexible enough to work with databases that haven't been invented yet either. So I wrote a better one.Completed a basic version of the framework, but shelved further development due to new focus on distributed database project. The API for the distributed DB library would largely remove the need for such an ORM framework.
A better spam filtering systemThunderbird's spam filter has become less and less effective for me in recent months. After looking at the spam I get, however, I found a very reliable technique to identify spam. I started coding this into an anti-spam system.1) Turning this into an internet service would require the development of (or adaptation of an existing) robust, scalable, mail processing/proxying server. This is a trickier thing to develop than a robust, scalable, web site.

2) Part way into the project, GMail beefed up its spam filtering capabilities considerably. With more and more people using GMail these days, market opportunity here is now small.

A music sharing siteA site where users can upload their MP3's, and then rather than allow downloads (which is a big no-no) allowed shared streaming.1) Saw no exact competitors at first; space is now very saturated.

2) Realized later that license fee issues complicates this idea quite a bit ... and also means less profit potential.

An improved data backup serviceA service where users can backup their data in an efficient, incremental manner, which existing services at the time did not offer.Again, saw no exact competitors at first; space is now very saturated.

 

 

Lindsay Watt

 

*Current Project:*

I've got a diverse set of interests and I believe they help me see problems in new ways. I've been frustrated with the lack of tools to help people manage their interests and ultimately be more creative. As a result, I'm working on a project called NeedTuNo and our goal is to turn information into inspiration.

 

We provide users with three services:

 

*1) An intuitive way to organize all the information you come across.*

 

  • Add tags, titles and summaries to any file or web page and store it with us. We index the content of the file/web page so that you can search across both your titles/tags/summaries and the content of each file/web page. You can also create notes simply by creating items without any attached files/web pages.
  • If you want to, you can organize your items into collections. These are like folders, except that an item can appear in multiple collections and collections can be based on queries (e.g., "Everything tagged with 'New York' and 'Entrepreneurship')
  • We also want you to be able to transfer all your info in and out via an API

 

*2) Discover new ideas by effortlessly sharing information amongst people you know*

 

  • You can share any item with another user, a group of users or email it to a non-user
  • You can share collections and use them to easily set up sharing relationships like "every time I tag something with xyz you get a copy - and vice versa."
  • We respect your privacy: you have to choose to share anything and you can mark items as not to be shared

 

*3) Synthesize new ideas by seeing your information from new perspectives*

 

  • We let you see all the items in your account in multiple ways
  • You can choose to see them in traditional ways like lists and tag clouds
  • More interestingly, you can navigate your account via a tag map (a web of your cross-referenced tags) and you can see all your items as thumbnails to rapidly skim to what you're looking for

 

*Where are we at?*

We're plugging away on our prototype. You can go to NeedTuNo and see a few screenshots from our current version.

 

*Sound interesting?*

I'm looking for a tech partner who wants to be an equal thought partner in the business.

 

For more on me, check out my LinkedIn Profile

 

Laurent Kretz - SubMate

 

*Our project: SubMate social commuting*

 

Commuting is a major part of our daily lives. It's not always fun but we believe it presents an amazing and enjoyable opportunity to network with other people and share common interests - in or outside the train, before, during or after your commute. I meet hundreds of people everyday going to work, to the gym, going out, ... and among them, I am sure there could be good friends, maybe dates, a guitarist for my band, why not a business partner or the guy that will push my resume to this or that company.

 

Also, all these people have interests, tips, recommendations about a good restaurant or bar in my neighbourhood, or a deli for lunch in my work area, an upcoming concert down the street and a gallery opening next block. I want to know what's happening on and around my daily journeys.

 

SubMate aims at allowing you to look for and then contact people that you see during your commute. Members can also share their tips and insider secrets on the best places to eat, drink, shop or just hang out at the beginning and end of their journeys - where they work, live, go out, study,.

 

http://www.socialvault.com/images/WhatIsSubMate-small.gif!

 

*How SubMate works*

 

Users create their profile and save both their regular and occasional trips: for example the trip from home on Astoria Blvd to their office on 57th Street, and the journey from their best friend's place on 96th Street to Astor Place to go out in the East Village ... We make it easy to find people who share part or whole of their journey and then get in touch.

 

Once they have saved their trips, they can give tips to other members about the great places they know best in their areas: their favourite lunchtime restaurant, their after-work bar, a great gym, that hidden park where they read in summer, and all the other hangouts they would recommend.

 

*Who we are*

 

We are three french founders - 2 in New York, 1 in Paris - Hanna and myself in New York, Olivier in Paris.

I am dedicated to go full time on the project as soon as possible, but for the moment I am still working at my day job.

We all three are urbanites, city lovers and enthusiats !

 

*Where we are*

 

We are working on the project since 2006. The development started last year but we encountered various problems with our developers (off shore, freelances, ....) and the architecture of the application (scalability). I will share details on our blog, but if anyone is interested please contact me.

Right now, we are working with a french team of rockstars developers to finalize the alpha version of the product - which should happen soon.

 

*Our next steps*

 

Once the alpha is finalized, we will test it in New York and Paris, France. If no major bug is found, we should (in parallel):

  • look for a technical partner/founder in New York for a role of lead developer or CTO,
  • raise an angel round allowing us to settle as a full time team (Lead developer + myself) and continue development and deployment,
  • expand to other major cities in Europe and the US.

 

If this sounds interesting and if you have any question, please do not hesitate to contact me ! Contact info on laurentkretz.com

 

 

Lee Semel - CommandShift3

 

At a recent Jelly coworking session, a bunch of us came up with the idea to create CommandShift3, a site that's like Hot or Not for web design. And we started that very night, working from 7pm until midnight. A couple more coworking sessions later, we're almost ready. CommandShift3 is the place to find the world's best web designs, rate web sites, find designers, and see what others think of your own design. Eventually this'll turn into a full fledged community for the design business. Also working on this project are Amit Gupta, Darrell Silver, Erin Sparling, Adam Varga and Dan Lurie. If you'd like to check it out right now and give feedback, just shoot one of us an email. It'll be launching in the next week or so.

 

Eran Hammer-Lahav - Hueniverse building Nouncer

 

Nouncer is a server-side web service offering developers an API to build real-time content solutions. The prime example of what Nouncer is about is building microblogging sites or functionality, similar to Twitter, Pownce, and Jaiku. Instead of building yet another (there are over 30) microblogging site, Nouncer is focused on the backend platform offering a sort of while-label solution instead.

 

There are many applications for the technology behind microblogging and real-time content delivery. The idea is to eventually focus on the corporate/enterprise world where email is becoming a problem rather than a solutions. With a platform like Nouncer, Hueniverse and other companies will be able to build new communication tools customized for the particular needs of smaller markets, something the big web players do not address.

 

Nouncer is currently a one person operation, and looking for another developer to join as an employee or co-founder. The biggest challenge is getting a critical mass of applications and developers using the product, and in order to do that, it is not enough to just build it, but to create a developer community and support. One of the ideas is to write a Nouncer-driven site and make it open source so that others can simply take it, change it, and make it their own. That is something that’s currently missing.

 

To learn more about Nouncer visit the Hueniverse blog