Introduction
On Wednesday 17th June 2008 I attended the Open Text UK Web Solutions Community Day - completing my trifecta of RedDot CMS conferences for this year. Unlike my previous reviews, I won’t be going into as much detail - the contents of the presentations (powerpoint plus audio) are available (well, the powerpoints are so far) on the UK WCM Community Site. For those of you without access to this site, each presentation could probably merit its own article - so I’ll just stick to the highlights for now. At the moment, I just want to get this out before the release of Version 10 on June 30th… Leave comments for anything you would like in more detail.
The conference was held at the Madejski Stadium in Reading, just a short and convenient hop from my place of residence. The day was broken up into four tracks - technical, partner, customer and Obtree. I stuck to the technical track and will be going through the audio and powerpoints of the other tracks as time permits (well, except for the Obtree track that is!)
The day started with the obligatory welcome and introduction - the main points of interest being the above mentioned launch of the UK WCM Community site (a replacement for the previous RedDot Community site) and that all sessions would be available there as powerpoints and audio. The obvious advantages being less note taking required and less concerns about missing the content of the other tracks.
Enhanced Template Design
First up on the technical track, enhanced template design - which showed how to use Render Tags and Pre-Execute scripting to create enhanced and multi-functional templates - ie single templates that handle what you may have previously considered implementing multiple templates for. Examples included alternative page headlines, styling the first item in a list differently, laying content out in different column configurations, pulling content up from pages nested below, detecting whether a list or container contains pages, using a single template for page links, documents or external URLs, splitting a list into two or more columns and A-Z listings.
If you haven’t done anything like the examples above, and/or you are not sure how to, then I would imagine this session (and the code samples) would have been beneficial for you. If the above sounds easy to you - well the benefits would have been limited. One thing to note - due to the limited scope and the introductory nature of the session, there are a number of issues not touched upon in the code samples, including placeholder escaping and code efficiency (one, you will probably want to use functions, two beware large amounts of logic in list processing - you can easily start creating Pre-executing script in the size of megabytes). Oh - and the Render Tag examples, while not requiring Navigation Manager, do require version 7+.
Two of the more advanced items to come out of this session - while you can use RQL in your Pre-Executing script - you should limit it to read options only, as write options during the page build are unpredictable and may cause issues. Also, you can switch from Pre-Executing script to RDExecute for debugging - though be aware that this will be easier if you use only a single Pre-Execute tag pair, and that the error and line numbers returned are specific to the generated script…
Implementing Web 2.0 Tagging & Voting using Delivery Server Web Components
An introduction to tagging and voting and how Open Text delivers them in Delivery Server, you can probably skip straight to page 17 of the presentation for the technical overview. Interestingly, you don’t need a license for the Tagging and Voting Web Components - but you do need the Delivery Server “RMVT” module license. In fact, the Tagging and Voting Web Components use standard Delivery Server functionality - you could write them yourself (though I am not sure whether this would include the integration into Delivery Server?) - they exist to get you off the ground quickly.
Tags exist in their own project so that you may share them between projects. Tagging can be assigned to pages or components of pages, and from Delivery Server or Management Server. A blacklist of unacceptable tags (including regular expressions) and a white list of suggested tags can be maintained. Future functionality (when?) includes managing tags from Management Server and utilisation of the BIRT chart engine for graphical views.
Having not implemented tagging, voting or Delivery Server recently - its not immediately obvious to me how good the implementation is - it will be interesting to hear some feedback in this regard. One note brought up was that there is no magic in the voting system that stops multiple anonymous votes. You can however limit votes for logged in users. In the meantime, the community site is using these components - so keep an eye on it to see how these pan out (its a bit barren to really test these components at the moment)
Navigation Manager - The Next Steps
Continuing on from last year’s Community Day Navigation Manager session (which I missed), this session goes into what has been learnt in the last year in terms of lessons learnt and best practice. If, like me, you have been through these “lessons” personally, then probably the explanation of Navigation Manager caching was the most interesting part of the session. For those that are yet to experience the fun of Navigation Manager - take heed - I can personally vouch for most of this session!
Some specific comments on some of the dos and don’ts:
- Use as few navigation areas as possible - 1 - 3 is okay, 7+ is not (on a single template that is)
- Use RenderTags for generating breadcrumbs and sitemap - the placeholder functionality may be deprecated in the future.
- Make use of detailed authorisations on navigation lists - restrict the users who can connect pages to the navigation system, reduce likelihood of connecting duplicate pages etc.
The only “Don’t” I would question is the “Do not use Pre-execute or RDExecute in the Navigation Templates.” - though I would certainly caution against it - especially if you are planning to execute RQL as well. I would suspect that this is more an issue with code efficiency (as mentioned in Enhanced Template Design above) - you need to be aware that your code size and complexity will be multiplied out by the number of navigation items rendered - which can easily increase you Pre-execute or RDExecute script size to unmanageable levels - especially if you have ignored some of the other best practices.
Code samples included recognising first and last links in a list, differences in Rendertag and page info element GUID representations and how to successfully compare them, for each loops (including counters), changing site colours and/or footers per section, global project values and the CurrentMasterPage rendertag.
Unfortunately I think this session did a better job of highlighting the restrictions, limitations, inconsistencies and hacks required to utilise Navigation Manager rather than showcasing the progress that has been made. My recommendation still holds - the most important best practice is to keep your site navigation as simple as possible!
I missed picking up the updated cheat sheet - and I haven’t seen a copy on the community site - yet.
Delivery Server and User Generated Content
A worked through example of an events management system using Delivery Server. As always - skip straight to the implementation for the “meat” of the presentation. I imagine this would make a good next step after the Delivery Server training.
Again, the example is bare bones due to the introductory nature and time constraints - but basically should give a good example of moving data in and out (ie CRUD operations) of Delivery Server - regardless of whether it is an events management system or even user generated. I don’t know if I would go as far as to tag Delivery Server as a rapid application development platform though (comments?) It should be noted that the session was Delivery Server focused, so there was no mention of integration with Management Server. Perhaps next Community Day (or perhaps here) an example of User Generated Content and integration with Management server would be useful (rather than replicating user and workflow management in Delivery Server)
Developing a Sharing Strategy
This session covered every (and I do mean every) sharing possibility in Management Server, from page instances sharing content classes to inter-project content sharing. For those who are no longer beginners I think you can safely skip to slide 10 - inter-project sharing techniques. The actual sharing strategy decision graph from slide 18 is probably the most helpful element - but it won’t be anything new to advanced practitioners.
Probably the only hole is that the session doesn’t specifically mention sharing of content classes via project or language variants - but these pretty much come under multi-sites - the decision is still whether the information architecture is shared or not. For language variant projects I have been on, I would suggest you want over 80% shared information architecture if you want to use language variants over master/child projects.
Version 10 Preview
The version 10 preview was not too different from the one I wrote about in the 4th European RedDot Usergroup Conference Review - Part 1 so I will only mention the differences here. This preview was running off of the servers back in Germany, rather than on a virtual server. Hence it was even closer to being the finished version. The obligatory apology for speed was changed from virtual server to network speed 
We thankfully didn’t have any long winded introductory powerpoint presentation though - so we got straight into the live demo.
The major drivers for the new version are usability and cross browser compatibility - the goal being to provide a “groundbreaking” step in Web Content Management. I guess we will all be the judge of that soon… Consequently, except for the PageBuilder improvement - don’t expect any developer enhancements this time around 
Instead, expect your users to have more convenient, faster, quicker, less clicks, less popups and more direct access to features, functionality and completing tasks (ie replacing all that RQL you have been writing…) I sense the Open Text sales team will be very excited - lets hope our users are (which will come down to stability and robustness of the final product in the wild I think) and us (who still have to develop for and support it).
The latest version of Telerik’s RAD Editor will be included - lets hope Open Text have integrated it more successfully (fully and bug free) this time - giving cross browser support, more options and pure XHTML code generation.
Search by page id directly in the search box. Management Server will remember recent searches and will show both the search results and a selected page - allowing users to update each page in the search results in turn without popups or having to re-search.
The changes are all in SmartEdit, so no SmartTree updates. Also, not all dialog boxes have been converted to the new look and feel - so expect some “flashbacks” occasionally. Other future items not yet implemented include combining tasks across projects.
I missed the recent webinar on Version 10 - so leave comments on anything last minute I’ve missed.
Conclusions
While I can’t speak for the partner or client tracks (yet) - the technical track was good, but not excellent. On the plus side, most of the presentations included real problems and solutions with code samples. On the minus side, most also included a lot of introductory waffle before getting to the technical meat of the session. All were limited by the time allocated and were necessarily simplistic in order to get the concepts across in the time frame. It is my experience that the real problems surface when trying to fill out the last 20%, not the first 80%.
The venue and catering were good. The main room (registration, lunch, partners, socialising) was pretty small, so there were a limited number of partner stands - but there were plenty of Web Solutions and partners floating about - including the Web Solutions support team. We got another Open Text branded netbook bag (similiar to the Content Day) with another pad, pen, pencil and partner parenphenalia (including a red dot stress ball from one of the partners).
In conclusion it was a good day - though I personally could have used the content of the presentations a year or so ago, better late than never! It was a good opportunity to catch up with past clients, Web Solutions consultants and RedDot Usergroup members in a much cosier setting than the Content Day. I would certainly recommend it over the Content Day if your dealings with Open Text are limited to their Web Solutions arm (and you can only get to one)
