Re: CMS migration

Hi Jaakko,

<disclaimer>I work for a software vendor (CambridgeDocs) that has tools to streamline this conversion process.</disclaimer>

I think that everyone that's responded so far has made good suggestions...
1) It's not going to be easy right out of the box
2) take multiple passes through the content
3) make efforts to tidy it up first
4) find a services provider to help
5) take it straight from the database (storyserver) if at all possible.
6) decide on a schema to represent the "fresh" content

I would add:
7) Spend time analysing content up-front
8) Anticipate future uses for your content

In my experience, on of the first things that needs to be done is an analysis of what kind of content you really have, and what you're going to doing with it in the new CMS.

>From that analysis you arrive at at least 2 data points: 1) Indentification of "sets" of data and 2) some "prioritization" of those sets.

This is important for a couple of reasons.

One is to identify any content that's ready for archiving (perhaps there's none).

Another is to set a project plan for the conversion process... we all know that one of the big CMS killers is "adoption", and if important content isn't ready on launch, or when you're going into "user acceptance testing" your project has just created another big hurdle to overcome.

Thirdly, by identifying "sets" of content, you are more likely going to be able to automate sections of the conversion process, and perhaps decide on specfic meaningful XML formats that suit each datatype better then a single generic format like xHTML... 

For instance you may decide that your user manuals are best managed in DocBook or DITA, your press releases in some subset of NewsML, and your annual reports in XBRL... Maybe there is a class of documents that should be managed as xHTML.

For a company like Vodafone, I'm assuming that some of the content that you'll want to manage are things like your vendors documentation outlining the various Alarms that your Network Operations Center manages on a minute-by-minute basis.  In cases like that you'll want a DTD that can isolate and return the exact Alarm Response for the particular hardware/situation.  Subsets of standard DTDs might work for that, but you should spend time analyzing your ideal scenario before starting your conversion process.

When you're considering scenarios for document re-use, the selection/modification of the correct DTD is key, not just for search/retreival, but also for the tailoring of whatever authoring/editing environment you might be using... you would not, for instance, want your users trying to fit a maintenance manual into a NewsML format.

Regarding your questions about CMS vendors having tools to do this type of thing, there is an interesting (and aging) survey at:
http://iainstitute.org/pg/the_problems_with_cms.php

Again, in my experience, and as supported by this survey, the attention paid to the actual content migration/integration during a CMS sales cycle is easily overshadowed by sexier features and promises of what the CMS can do... We've worked with a number of clients who have realized that to get everything they want out of their CMS they need to be able to get at granular levels of information within their existing content.

Hope this helps... 

Michael
-- here's a link to a great white-paper on "The 7 deadly sins of XML Publishing" which has several relevant point for a project like yours:
http://www.cambridgedocs.com/resources/whitepapers/sevendeadlysins.htm


-------------- Original message -------------- 
From: "Anttonen, Jaakko, VF-Group" <Jaakko.Anttonen-ANTagKRnAhdWk0Htik3J/[email protected]> 

> Hi all, 
> 
> We are undertaking a project to migrate our content management platform from 
> Vignette V6 to a new solution which is yet to be decided. As part of the general 
> project plan we have started investigating our options to migrate our content. 
> 
> Our biggest challenge is (beside that we haven't decided on the target CMS yet) 
> having to export content stored as HTML, tidy it up and convert it to XML. 
> 
> Do CMS vendors provide tools in this area? What we need is a tool that could 
> spider the live website, store the content as a repositary that can be edited 
> against and finally be loaded to the target CMS. 
> 
> Best Regards, 
> Jaakko 
> Vodafone Group Services Ltd 
> www.vodafone.com 
> 
> _______________________________________________ 
> cms mailing list 
> [email protected] 
> Subscription controls: 
> http://lists.cms-forum.org/mailman/listinfo/cms 
> Netiquette FAQ and related CMS lists - [cms-forum], [cms-pr], [contentmanagers], 
> [cmpros] 
> http://www.cmsreview.com/NetiquetteFAQ.html 
> http://www.cms-lists.org