It's always nice to see some visible results from your work, so I'm smiling broadly while typing this :-) We recently completed two new sites based on that great open source content management system Plone . The content is mostly Dutch, so English readers 'll just have to look at the nice looks before going back to reading Schlock mercenary or so.
Developing the website was lots of fun. Most of the time, two of us (me and Ahmad or Mirella) would work on location together with Just (from Triplep): a real, active, co-developing on-site customer in the best eXtreme Programming tradition! That meant instant feedback, lots of small change requests that could be implemented on the spot, etc.
"On location" in this case was something I liked a lot, too. Triple p is just a 30 minute bicycle trip instead of a 1.75 hour public transport trip to either Zest or one of our other customers :-)
In Dutch: about the triple p website .
Part of my work on the site involved making the UML model of the content types and the workflow, the core part was done by Ahmad, Mirella and Jean-Paul. The majority of my work was converting the old website, however. 1GB of content, some 1500 html pages, 2000 doc/pdf files, loads of images. Now that was a challenge! In the end it worked and it saved them a lot of time. It was my biggest estimation mistake till date, though. I thought to finish it in 5 days or so and it took almost something like 5 weeks in the end. Ouch... Milieudefensie was happy with the endresult, though :-)
In Dutch: about the milieudefensie website
Funny tidbit: Milieudefensie is an environmental pressure group and we made their website. Our office location is in the lower right of this google maps picture . The top half of the picture shows just a small part of Shell's #1 refinery complex :-) Perhaps its just my weird sense of humour that I find that funny :-)
We made both websites with a lot of help from ArchGenXML, which generates lots of the basic code needed for a Plone product from a UML model. This is especially valuable when creating the workflow, as a UML workflow picture makes it so much easier for the customer to understand and see what's happening.
ArchGenXML also helped us to work on the project with multiple people. The price you pay is a little bit of coordination to make sure you're not working on the same UML model at the same time. That was not really an issue in this case, it was somehow always clear who was the main person working on the model at any given time. If you needed an extra attribute you'd just ask that person to add it and it was done.
Where ArchGenXML helped us was that both projects had the same structure and the files looked the same (because they were generated). If you needed to change something by hand in the code, you had to do that in certain "protected sections" that preserved the changes when the project was re-generated from the UML model. So changes would always be in the same place, easily findable. And, also important, a large part of the project would consist of tamper-proof generated code. If something was wrong, you didn't need to look in those places. If a stupid mistake was made in those parts, it would be gone on the next re-generation.
Especially the ArchGenXML workflow generation is really really really great.
Triple P certainly looks nice. Congrats!
One quirk I noticed is duplicate search results that I have on my plone sites as well. For example, a search for kwaliteitsbewaking returns both: http://www.triple-p.nl/wij-bieden/publishing-services/kwaliteitsbewaking/?searchterm=kwaliteitsbewaking http://www.triple-p.nl/wij-bieden/publishing-services/kwaliteitsbewaking/kwaliteitsbewaking/?searchterm=kwaliteitsbewaking
I guess the first is the folder, the second the default file in that folder. This duplication is a pity, imo.
My name is Reinout van Rees and I work a lot with Python (programming language) and Django (website framework). I live in The Netherlands and I'm happily married to Annie van Rees-Kooiman.
Most of my website content is in my weblog. You can keep up to date by subscribing to the automatic feeds (for instance with Google reader):