I had the pleasure to join the user experience event @AMIS in the Netherlands. It was an event organised together with Oracle UX Team.
As a general impression: very well done. Parking, event location, different themes and very good content are the main positive points that jumps into my mind.
Unfortunately were the sessions programmed without buffer time, so that when you followed one session you are certainly to be late for the next session in another room. Luckily sessions were repeated, so I didn't have to miss anything and I was able to visit the different Oracle stands explaining the diversity of UX.
For a full list of the sessions, take a look at
http://www.amis.nl/UxEvent/Lecture_details.html.
Here are some impressions I took from a couple of sessions:
UX Directions with HTML 5
This session was presented by a surprise act, in the sense that it was not Noel Portugal as presented on the site, instead it was a Belgian Oracle UX team member, namely Mark Vilrokx. He showed us what they are working on. It is very interesting to see how Oracle is investing into new technologies. The things this guy showed us, was even new to our UX designer. He focused on new possibilities with browser interactions in the mobile devices. Knowing that a server can actually track your whereabouts, how bright it is in the room, whether you are using the application or not, is astonishing. Not only passive information was gathered, also actions were triggered on the device. Things like activating the vibrating functionality, playing a movie without user interaction. This is a clear step forward for the browser in the war against native/hybrid applications. Which made the off-line capability the only remaining drawback of the browser app in the mobile environment.
The architecture he used was a standard one: HTML5 -> Ajax calls -> REST Services. To develop the mobile app, he used the Fuse Bootstrap, which is an adapted version of the Twitter Bootstrap.
Not presented in this session, but linked to the presenter is the
Oracle Voice product. It is a standalone product that will be released with the version 9 of HCM Cloud application. It is an app that you can install on your phone or tablet, allowing you to speak to the application. Not only in a question-response mode, where you answer to the question of the device, but also in entire sentences.
Example: Create a new appointment Follow up meeting at Contribute next monday. Not only will the solution interpret the sentence, it will also automatically fill in the fields of the new appointment. For the missing information, extra questions will be asked. When you have multiple opportunities for this account, a list of the possible opportunities for that account will be presented.
What is even more amazing is that the application does not need to learn. You used to spend an hour saying/pronouncing bizarre and difficult words, this is not needed anymore. Which indirectly means that you need to speak American English very well, preferably with a Californian accent :-).
A nice site note for my Belgian friends, this product uses Nuance for the Speech-to-Text and Text-to-Speech functionality. Nuance bought the voice recognition software from our West-Vlaamse creatives L&H and is using their solution as one of the possible resolution software for the Speech-to-Text challenge.
Oracle's Strategy for Cloud User Experience
This session was presented by Jeremy Ashley, what a presenter. It is not the content that struck me, instead his way of presenting and the feeling of clarity he gave us about this topic. A must seen presenter!!
Back to the content. The three main topics for Oracle UX are Simplicity, Mobility and Extensibility.
Simplicity is referring to the completeness of your solution. It must be no too much, neither not too little, it must be exact for the job at hand.
Mobility is not about the mobile devices, instead it focusses on the way we work. How we work differently now and in the future, then a couple of years ago.
The audience of these type of applications are people who will be using these not too often. Hence the importance of keeping the distance between wanting to do something and doing it, must be a as small as possible. The process that is the main driver, listens to the expression: Glass - Scan - Commit. Resulting in an application giving the user an high level overview of the data, allowing him to drill down for more detail and then close his action. This approach is very closely linked to the idea of 10-90-90, meaning that 10% of the tasks are performed by 90% of the users in 90% of the time. These 10% are the main focus of these applications. For full feature functionality, the user will be brought to the actual application.
Based on their experience with the Oracle Cloud solutions and fusion applications, the UX team delivers a set of patterns for you to use. These can be found at
http://www.oracle.com/ux.
While most of the development and testing is done through HTML5 and then pushed to the ADF team, APEX is gaining momentum here. In the "near" future, the APEX team will also include these patterns, just like ADF does.
UX Today with ADF
A presentation done by Sten Vesterli. This session was really focused on ADF, like the title mention. From the first moment it was clear that Sten was a guy with a lot of experience. Here are some of his main points during the session:
 |
| From http://www.vesterli.com/2014/03/20/good-user-experience-at-low-cost/ |
- A clear graph on the different types of applications you can build:
- Use GUI mockup tools to create samples of your UI which don't have a high fidelity. Reason: possible no feedback from end-users due to the hard work already done AND the possible feedback you get is that it could be finished tomorrow. Conclusion: make sketches. Possible tool: Balsamiq
- Another great point made by the presenter is the difference in type of application you can build between Forms,Apex and ADF. While Forms and Apex applications have a 1-1 relationship with the underlying data model, ADF doesn't. With ADF you have more a n-n relationship, allowing for a far greater UI experience. Conclusion here: do not generate your UI screens.
Don't generate your user interface, design IT!
Session given by Lonneke Dikmans from Vennster, ACE Director in SOA & BPM.
The problem according to Lonneke is:
- Using BPEL/BPM Human Tasks
- too fine grained
- too much data in process
- need to adapt the process to changes in the UI
- inflexible for use -> ex. no bulk operations possible
- not enough data for the UI -> only task data is available
The solution for this is to take an alternative approach:
- separate both solutions => BPEL/BPM and UX
- take the following actions in your project
- Define the different personas. Who will be using the application, how often, from which type of devices, do they have knowledge of the app, ...
- Define different scenario's for different type of personas
- Design the interaction process
- Ex: while the HR persons define the process, it should be the users themselves who define the UX/App
My question for Lonneke: how should we build BPM applications then? From the BPM flow or build a separate application that calls on the BPM API's?
Answer from Lonneke: take the latter one. This way the UI is really build for the end user and you have more flexibility.
Leveraging FMW for UX
Another session from Lonneke Dikmans. This session was a simple overview of the main components of the FMW stack of Oracle. Some points worth remembering:
- How to test the experience:
- Do Usability testing
- Do SCA testing => use mock responses
- Do Load testing
- Tip: learning by doing : learn whether something works or not
- ex: google class: 1st version done in 2 days
- Monitoring
Oracle and Mobile: From Design to Device; The tools that make it happen
Session by Luc Bors. Simple overview of ADF Mobile.
There seems to be a misunderstanding in the comparison between ADF mobile or hybrid solutions and native solutions. According to Luc and Oracle(Oracle is saying this also, I know I did it also before), the effort in making a native app is a lot bigger and more difficult then making a hybrid solution for multiple platforms. It is true that you need to make an app for each platform and that reuse at the level of the mobile app is almost zero. Nevertheless, making these mobile apps is only a small part of the total project. The major effort resides in building the back-end services, including security.
Take into account that the hybrid solution from Oracle only supports 2 platforms, while there are 4 in the market.
According to me is Oracle ADF great at the data oriented apps, while building great looking apps are better done in native solutions. One of our main references in this area is the Royal Belgian Football Association, who are really looking at their target audience and decide on the technology uses, like Oracle Forms, ADF, ADF Mobile and native mobile applications.
Conclusion
