Here are my notes from the Plone/Zope breakout session. Pardon my messy notetaking! -Jodi PS-I plan to post a version of these at code4lib.org eventually. If you attended and prefer I don't, please let me know. Vernon Chapman, NAL Kevin Clarke, Appalachian State University David Cloutman, Marin County Free Library Douglas Hahn, Texas A&M University Tom Ceresini, PALINET Heather Rolan, Yeshiva University Jodi Schneider, Appalachian State University Jon Stroop, Princeton University (Metadata Analyst) ================================ repose.bfg ---can run plone/zope right through apache ================================ http://library.tamu.edu 1.5 - 2 years (1 year in production) from Vignette ($ license, complicated end-user interface). Didn't have SDK's, hard to get content contributed. Content contributor user interface (Drupal easier for programmers, sucks for end-user.) 100-150 content editors 40,000 users Subject guides, finding aids, class guides--custom metadata types Pain points: caching, sending out the content, maintaining speed, getting locks on content (hard-coded to try 2 times--changed to 99 times) Run varnish (varnish-cache.com), pound (no longer needed if you run varnish 2.0), 2 zeo servers ZODB -- more complicated architecture than they would like Want load w/i 1 second 6 month ramp-out time, wanted display framework ================================ layers in front may be necessary (e.g. searching, load balancing)--except for Content mirror--dumped to DB, pulled to Django http://blog.kapilt.com/2008/07/02/introducing-content-mirror/ Solr--for searching add a writer to have a feed--Atom, RSS 120-150 LOTS of checks--wear and tear on performance--an issue for PUBLIC content ================================ plinket using plone repoze.org-- wsgi---web services gateway interface http://grok.zope.org/ ================================ Plone is the display layer for Zope ================================ http://www.amazon.com/Professional-Plone-Development-content-centric-applications/dp/1847191983/ ================================ Can deal with multiple levels: IA, web design, programming ZMI level/CSS level, python base code ================================ portal types/content types copy and paste--problem is when you upgrade, it goes away. ================================ Comparison to http://www.djangoproject.com/ framework ================================ ZODB compared to RDBS: easy to backup: text file. just do a diff takes anything (large stuff--videos ok) ================================ Where in the stack something is? ================================ What you like about Plone: inline editing versioning approve/publishing WSIWIG editors "just works" if it fits your content types "Spent all our time on making it pretty" -- "Didn't spend ANY time on content editors" ================================ ================================ Ask TAMU: Can they share their subject guide content type? (took 20 hours for them to do--e.g. 1 20 hr/wk --one ) class guide FAQ forms took longer--now has form builder product--complex forms with javascript checking UI work on display Goes out to solr--associated journals, etc. ================================ Tom went to online training from 7technw 4 hrs/day for 3 days, relatively introductory ================================ Clouseau: while developing, can have live python command line Firebug Web Developer Toolbar ================================ sole developer understand IRC