I had previously contracted for Infinera working on their NMS suite. My contribution was a PoC running on D3 and Angular, using Semantic Topology that allows users to navigate any node or link in their network with zero clicks (scrolling and panning gets them anywhere à la Google Maps).
The hardest, most time-consuming part was prototyping because the tools out there (Framer, Principle, Quartz Composer, Adobe XD etc) are all optimized to work with lists and grids, not visualisations. I ended up creating three fully-functional prototypes in HTML and JS before finalizing on one. It was hard to get good feedback from the teams without live, interactive prototypes. There is also nothing like a standard library of well-designed components that can be grouped together to quickly mockup custom visualizations, like we have in Bootstrap for HTML frontends.
Another problem was performance; Angular, out-of-the-box, could not keep up with the amount of data coming through their pipes.
All in all, it was challenging and very, very fun to do. When it finally worked and people got their hands on it, it blew everyone away. If I were to do it all over again, I’d focus a lot more on explaining the challenges and setting realistic expectations. Custom visualisations are not 5-minute WordPress installs.
The work is under NDA but if you feel the need to discuss anything, my email is in my profile.
Some people I follow on twitter who are actively involved in DataVis:
Lisa Charlotte Rost (@lisacrost), Micah Stubbs (@micahstubbs), Jim Vallandingham (@vlandham), Nicky Case (@ncasenmare), Nadieh Bremer (@NadiehBremer)
to your point, non-technical should be taught to understand that this is challenging data and graphics programming (which means there is a whole dev cycle associated with it), not photoshop or indesign style work where the tool is itself a program that you instruct to do things by pointing or scripted actions.
The hardest, most time-consuming part was prototyping because the tools out there (Framer, Principle, Quartz Composer, Adobe XD etc) are all optimized to work with lists and grids, not visualisations. I ended up creating three fully-functional prototypes in HTML and JS before finalizing on one. It was hard to get good feedback from the teams without live, interactive prototypes. There is also nothing like a standard library of well-designed components that can be grouped together to quickly mockup custom visualizations, like we have in Bootstrap for HTML frontends.
Another problem was performance; Angular, out-of-the-box, could not keep up with the amount of data coming through their pipes.
All in all, it was challenging and very, very fun to do. When it finally worked and people got their hands on it, it blew everyone away. If I were to do it all over again, I’d focus a lot more on explaining the challenges and setting realistic expectations. Custom visualisations are not 5-minute WordPress installs.
The work is under NDA but if you feel the need to discuss anything, my email is in my profile.
Some people I follow on twitter who are actively involved in DataVis:
Lisa Charlotte Rost (@lisacrost), Micah Stubbs (@micahstubbs), Jim Vallandingham (@vlandham), Nicky Case (@ncasenmare), Nadieh Bremer (@NadiehBremer)