Remember Mambo? How about the “rhuk-milkyway” template? Take one fair-sized site which to the public eye looked manageable, but beyond held a behemoth of an intranet with hundreds of registered users and six versions of templates pointing to odd morphed versions, each spaghetti-ing between versions, transparent gifs, shaded borders of images, css files and hard-coded JavaScript routines of slideshows... you name it, it was there. It was no small job to make any change at all, let alone move this whole thing to Joomla 3.

This was my assignment though and when my client called me and demanded that the time was now, I silently drew my breath to stifle a huge anxiety attack that was to keep returning over the next few months.

Planning

After our first meeting, most of my thoughts focused on how to get out of this. But truthfully, this client was a good client, one of my best, and I didn't want to pass them off or say no. I knew they had to do this. The hard-working web administrator had to take care of this mess and it was becoming more of a burden and security risk. All the reasons to bite the bullet were there staring us in the face every time we opened the administrator or looked at the file structure in ftp.

Still our first meeting left me with so many choices of approaches and concerns that I was paralyzed with indecision for the first few weeks, and looked at my puny notes with a sense of dread that I had met my match, my limit of adaptive competence, the level of incompetence.

The deadline for a plan was looming and my client's management team was waiting for the PLAN so taking a day, in a place ... my local library has desks and Internet ... no phones ... we made the decision on A) the approach and B) the major steps that would be our plan to get from point A to point B with little or no disruption to our users.

Let me say here that we first familiarized ourselves with the big picture of migration and had done some homework on Joomla.org. We were not foolish about the technical milestones that needed to be accomplished. Inventories of 3rd party extensions and conversion software were made and backup plans included. To become familiar with how the migration tool worked we thoroughly reviewed the documentation and actually did a practice run with a backup so we were comfortable with the tool.

Also important to note was the decision to keep the current "look" for both the public templates and the intranet templates, and to maintain the current functionality. The basis of this decision was rooted in our feeling that the site was at risk security-wise and time was critical. This included keeping the navigation the same. All design enhancements and new functionality would be added as a second phase after the current content had been successfully ported.

Third party extensions

Speaking of which, the current content consisted of 15-20 custom forms written in an outdated version of the old Facile Forms, now Breezing Forms, in the “classic mode”. These we knew would have to be hand-coded with updates and tested due to the significant chunks of code added to the submit piece of each form.

Additionally, there were a handful of 3rd party plugins and modules mostly outdated and some no longer identifiable which needed substitutes. This was accounted for in our plan and scheduled into the list of tasks to take place immediately after the conversion of the data.

Gotcha!

Then came the big decision handed down from management to split the public and the intranet off to two separate sites. At this point, it was another half-day to redo all the plans and identify 3rd party components and templates that were used publicly and those used by the intranet. Back to the library for another day... devoted to planning and research and no phones.

A month later...

In the end, the decision to split the site turned out to be a good one. We actually copied the current site twice and then removed the registered content from the public site and visa versa for the intranet. We tailored the public site, the smaller side first and got through a lot of the template issues on a smaller scale, almost another practice run in preparation for the larger intranet site. We were also able to tackle and simplify the public template bringing it up to date while becoming familiar with the different look and feel of Joomla 3. When we got to the intranet it wasn't nearly as scary because our familiarity with the issues was still fresh.

As for the actual conversion, we ran it twice: once for the public and once for the intranet. We had cleaned out all unpublished items on the split so running the migrator against the cleaned up content went well except for a few duplicate aliases. It was only after we had all the content and new 3rd party extensions installed that we started working on the templates which took about three days for the public site (one template) and about two weeks for the intranet (4 templates). For the forms, we had to do those by hand, but after doing the first couple, we had a pretty good idea of the changes that were required so these also went pretty fast after the first few.

Looking back...

In retrospect, making the decision to keep the current look on both the public and intranet sites allowed us to concentrate on the backend in a timely manner. Of course, to the users all that work was hardly noticed but we are now in a position to redesign the site and take advantage of the bootstrap styling in our own time without the pressure of the migration needing to be done also. We can also now make decisions on added functionality in the stable environment of Joomla 3. And just knowing that we're running a current, stable version of all our software brings a new level of confidence to everything we hope to do with the site and more. New ideas are explored with imagination and creativity unencumbered by doubts on whether the site can handle our big ideas.

Finally a word of advice...

To those that are faced with the task of migration, having a good plan is really the key to whether you’re going to be successful or frustrated. Taking those planning days off the grid and devoted to uninterrupted thinking through all the steps gave us a lot of confidence for the coming steps. Also, any research or experimentation that you can do ahead of time will be helpful to smoothing the process. Don’t forget to do your “due diligence” and get those 3rd party extensions in order beforehand. It’s worth the upfront effort for your peace of mind in the end. And lastly, don't be afraid to scale back to the basics if there are overwhelming circumstances that call for a more simplified approach.