"I don't know the whole problem domain in P2P but I'm wondering why we couldn't start by using a very robust existing system and perhaps make smaller adaptions to fit our needs - rather than building something from scratch?" Absolutely! I see no reason to build a custom CMS (content management system) for this project. This would greatly increase up-front development costs and would be an unnecessary yoke on the project as we would be responsible for developing all site features alone. And it would be completely unnecessary, we would gain no additional flexibility by doing this, only forfeit functionality and support. I also agree with you that the Digg.com website has many of the features that we would like to see in the P2P project, and I think we have a lot to learn from them in terms of how to build a highly usable and fun community experience. They've put an awful lot of thought into their interface and design and it shows. That said, I am pessimistic about trying to use their code base directly for this project. Even if we could obtain it, unless it almost exactly met our needs out-of-box now and into the future it wouldn't make sense to use for these reasons: Functionality: From a design and usability perspective Digg.com is stellar. From a technology perspective Digg.com is not very interesting. Open source platforms such as Drupal can recreate 90% or more of the functionality currently seen on Digg.com out-of-box, and offer much more beyond that. While digg.com is afforded the features that are developed by the ~6 or so full time employees of the companies, communities like Drupal benefit from the literally hundreds of community code contributors. Flexibility: Building your site on technology provided by a partner which considers providing that technology as a secondary business is overly risky. - Support: For a startup like Digg, getting bogged down supporting a non-paying partner will be very tough to deal with. Digg will have limited resources and will naturally have to prioritize what they concentrate on. When push comes to shove primary business will come first and secondary will be supported as time permits. - Contingency: The risk of Digg as a company going out of business and taking their back-end technology with them are far greater than that of an established technology vendor or open-source product. - Technology: The Digg back end technology was developed primarily to meet the business needs of Digg, it was not developed as a reusable platform. This will drive up costs for you for customization and feature development compared to platform technologies built to be used by others. TCO: A maxim of IT managers everywhere is 'the cost is not what you pay for up front it's what you pay over time'. The pace of technology evolution and the adaptive requirements of technology owners usually generates greater 'hidden' costs than the price tag purchasers are quoted upfront. In all likely hood the cost of acquiring and setting up the Digg.com codebase would be far out- weighed by additional costs in building out the P2P web site over time. This means that the real technology investment question is not "what gets us from A to B cheapest and easiest" but "how much is this going to cost for the life-cycle of the project?". I am sure that the costs incurred by the inflexibility of the Digg.com code-base over time will be far greater than the costs to implement the P2P on a reusable open-source platform like Drupal. -Zack On Oct 28, 2006, at 3:08 PM, Steve Midgley wrote: > I've been thinking about the platform decisions for P2P a little > and am > wondering why this system doesn't just run on top of an existing tool > that does basically the same thing? Digg and Slashdot are essentially > P2P-like in their functionality. Slashdot's code is open source and > available (though ugly). Possibly Digg could be convinced to licensed > use for this specific project (esp. b/c Omidyar supports P2P and > Digg).. > > Here's my rationale (I'll use Digg because I think it's a slightly > closer fit - Slashdot is roughly equiv..) In Digg, contributors post > articles w/links and classify them within topics. Readers can > subscribe > (RSS) to articles according to topics. Within a topic, users can post > comments & links. The value of the comments themselves are also rated > by other users using a simple interface. Users themselves can be > tracked by the value of their overall contributions. Comments can be > "masked" to only allow the most highly rates comments to be seen - > they > can also be sorted by their rating.. > > In the P2P world, contributors would be patent holders who publish > their work for community review. Readers would be domain experts who > self assign to topics and get notified when new patents are posted. > Comments would be expert review and discussion. Experts can rate other > comments. > > I don't know the whole problem domain in P2P but I'm wondering why we > couldn't start by using a very robust existing system and perhaps make > smaller adaptions to fit our needs - rather than building something > from scratch? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osdl.org/pipermail/p2patent-developer/attachments/20061030/7ca54a6c/attachment-0001.htm
