Trip Report, OSCOM Sprint Zurich

Before the sprint, we struggled a little getting the initial momentum started. Specifically, we didn't get the teams and coaches lined up as we should have. This made me a bit concerned about the likelihood of success. For Twingle, we made a quick dash on IRC to get it usable enough to sprint while keeping Twingle quiet and unofficial.

I left on Wednseday by train, pretty long but not so bad. Finally found a TGV with electricity. I arrived in Zurich just before 23h00 and met Michael and Henri. We stayed at Michael's house.

We woke up Thursday and got to ETH early. Wireless kind of worked but only for HTTP, so we moved back to wires. Michael started the sprint with opening remarks and introduced the OSCOM board. Then everyone in the room introduced themselves. Some of the projects in attendence: Lenya/Cocoon, Wyona/Cocoon, Silva/Zope, Plone/Zope, Bitflux/PHP, Aegir/Midgard, Ariadne/PHP, AxKit, and Xaraya/PHP. I made some remorks about the idea of sprinting.

I had prepared a Twingle presentation to give to the sprint team working on the mini-project. Gregor said instead to give the presentation to everyone, so we spent the first 20 minutes going over Twingle and getting a demo from Christian Stocker.

We then discussed the various sprint teams: OpenOffice integration, SlideML, SoftwareML (matrix), Twingle server (DAV support), and Twingle.

We then reshuffled the room and wiring and started sprinting at around 11h30. As usual, the first hours of sprinting were all logistics. The sprint team consisted of Gidon and Gregoire from Zope User Group Switzerland, Silvan from Bitflux, Bernhard, Christian from Bitflux, and me. On the second day, Eric and Gregor joined us.

I hadn't tested Twingle on Windows, or specifically JSLIB on Windows. Most of the people on the Twingle sprint team were using Windows laptops, and thus we lost the first hour working around a Windows problem.

By that time it was lunch. Eric and I stayed behind as I tinkered with Windows.

At the end of the first day, we had made a little bit of progress. Everyone had a taste of working in a Mozilla chrome application, using the debugger, navigating around the Twingle directory structure, etc. Gidon and Gregoired had learned about drag-and-drop and had a small example working. Bernard and Silvan worked with Christian on adding HTTP DELETE.

On other projects, Wouter and Gijs from Ariadne made great progress on getting their server to talk DAV. I also worked with them a little, showing them Shane's tcpwatch tool which makes degugging DAV so much easier.

That night we had a nice dinner that was sponsored by IBM. OSCOM produces a really good atmosphere between projects that, in theory, compete with each other. Most are genuinely interested in how the others work and we enjoy each others' company.

Fortunately we kept it a reasonably short evening, getting back to Michael's place a little after midnight. When you are part of the sprint organization, you need your sleep. Otherwise it is a long day and you don't do a good job.

We made it back to the room the next morning on time. We started with reports from the sprint teams.

The Twingle team made a lot of progress during the day. Eric and Gregor added support for the RSD (Really Simple Discovery) standard from the blog world. To add a server to Twingle you just provide the URL to the site. (Later even this step won't be necessary, as sites will put the RSD location in a link element). Henri added RDF support and RSD support to Midgard. Andreas added RDF support to Lenya.

At the end of the day the Twingle team had a nice milestone. We gave a demo where four server types (Midgard, Lenya, Ariadne, and Apache/mod_dav) were listed as sites in the Twingle navigation tree. Henri then opened a page on the oscom.org server, where he had added Twingle support, and edited a real page in the WYSIWYG editor. He then saved it back to the server and viewed it in the Preview tab.

Afterwards some of us chatted briefly about the "go/no go" decision for Twingle. Specifically, did Twingle seem like a good enough idea to continue? We agreed it was and that the next step should include being a little more public.

That night we had an OSCOM board meeting, focused on planning for OSCOM 3 at Harvard. I left afterwards due to a headache.

The next day many of the sprinters went on a snow shoeing trip that had been organized by Christian Egli from Wyona. I left for a train back to Paris/Rennes, traveling with Roger Fischer from Bitflux. We had a really great conversation about the CMS business, OSCOM marketing, and using mobile phones ot originate content.

After the sprint I unpacked my camera, hooked up to iMovie, and wrapped up some QuickTime movies:

• Room overview (210 Kb)
• Michael and Gregor (337 Kb)
• Eric hacks Twingle (170 Kb)
• Paul intros Twingle (1.9 Mb)
• Henri demos Twingle (1.8 Mb)