AMD’s Technical Documentation team recounts their positive transformation after migrating to IXIASOFT’s DITA Content Management

Prior to DITA, our old toolchain delivered minimal content reuse. We used unstructured tools and even though we implemented an ad hoc single-sourcing system, we found that content was inconsistent, quality varied from document to document, and the level of reuse was poor. With DITA, topics can pretty much be anywhere you want them to be. It is a far better architectural model for content reuse.

By 2006, my technical publications team was facing a growing number of document deliverables using a toolchain that wouldn’t scale. But with MadCap Software's DITA CCMS solution, AMD has cut localization costs in half, more than doubled writer productivity, and has improved the quality of our docs.

Knowing our short product life-cycles, when we performed an internal audit of our existing content we ended up migrating only the latest-and-greatest content to DITA, saving us a lot of unnecessary time and expense.

For example, something like changing a simple header in one chapter would cascade throughout the document, requiring an arduous review of the whole thing and then make manual adjustments to the header hierarchies,” said Keith. “IXIA CCMS's publishing engine took this headache away completely, leaving writers to do what they do best: write.

Reducing the amount of time spent on formatting should translate into an increase in production, of course, but the 200% increase in our output surpassed all of our expectations.

I remember when a simple 200-word document would be hit with a DTP surcharge that was larger than the actual per-word translation charge. Ridiculous!

Under our old system, different choices of fonts within the same language would actually yield different translation results. One of my localization coordinators trying to figure out the various types of character encodings being used within a single, one-page document in Japanese. There were five different character encoding schemes being used, and it took her several days to sort it all out.

Many engineering documents use upper-ASCII symbols, one being the Omega (Ω) symbol which represents electrical impedance (Ohms). To get this symbol to appear in one font, one could simply hit Shift+W. But if this same document was printed from someone else’s machine that did not have that font, that person would literally see a “W” instead of a “Ω”. “W” equates to Watts in electrical terms, which would not only be wildly inaccurate, but would make us look like we didn’t know what we were doing! Of course, MadCap Software's DITA CCMS solution uses Unicode enforcement, overcoming such errors.

Using our old toolchain, we could not consistently make these simple, yet important changes throughout all of our documents without investing considerable time and effort. IXIA CCMS allowed us to implement an entity-based system, where all such symbols are linked to a centralized lookup chart, and a single change automatically gets pushed into all subsequent versions of a document.

We used a now infamous test to push the limits of Unicode compliance. There is this silly phrase ‘I can eat glass; it cannot hurt me’ which has been translated into many languages using major segments of the Unicode plane–more than we were actually going to need. We began with what we already knew to be accurate translations of this phrase in about a hundred languages. Next, we had this phrase rendered in Unicode and then–as part of our test procedure–gave it to the vendors to import into their CMS to see what came out. Out of the top 10 DITA compliant CMS vendors, eight flat-out failed that test. IXIA CCMS delivered.

IXIA CCMS's solution has an incredible ability to search topics quickly. It searches over 16,000 topics and swiftly produces useful results. We can instantly determine whether something has already been written about a particular subject.

DITA CMS has a very nice drag-and-drop interface for DITA maps. If somebody thinks, ‘actually, this needs to go here,’ DITA CMS does not even require a copy and paste operation but rather an effortless drag-and-drop of the topic from one part of the map to another.

Generating output is a breeze, initiated by a few clicks of the mouse and the output server takes care of the rest. Behind-the-scenes format changes are instantly updated on the central output server, ensuring output consistency.

From a manager’s perspective, I can generate on-the-fly productivity stats—something I never had before using the old toolchain. I can now tell who is really busy by looking at how many new or revised topics an individual writer has made within any time frame.

I can easily report to my Director and VP, ‘Here is what we’re publishing, here is our progress status, and here are all the languages we’re publishing in. I can even measure my topic re-use rates, predict future resource needs based on productivity numbers and graph it all using the Business Intelligence Reporting Tool (BIRT) that works seamlessly with DITA CMS. Other managers dream of having that level of detail and the ability to produce meaningful statistics. Such reports make it easy to justify requests for more writers when my forecast shows that we will be overloaded.

The implementation went smoothly. IXIASOFT worked with us to first set up a program where writers could go in and get acclimated to the CMS but not be expected to produce any documents with it. This gentle introduction to the system allowed us to build up content in the repository for later use as well as getting our writers used to doing things in DITA.

MadCap Software's DITA CCMS solution system combines the efficiencies inherent in topic-based authoring with industrial-strength processing tools. I can’t imagine any other way of using DITA more effectively.

Without disclosing exact figures, let’s assume that our localization budget was quite sizeable and growing at an alarming rate. After implementation, with the same volume of material, we were able to reduce that by 50% within the first year alone. And subsequently we have tackled additional documents in more languages and only use about a quarter of our original localization budget.

AMD has measured a 2.3x improvement in writer efficiency. In other words, for every 10 documents that a writer could produce under the old toolchain, he or she can now produce 23 documents in the same amount of time.

We are very happy with IXIA CCMS's product support response time. When one writer on our team reports a bug, it gets fixed, sometimes within the hour. With on-time deliveries and regular support meetings, we really do appreciate their first-class service. IXIA CCMS's solution is robust, scales well, and has delivered faithfully for over three years.