First, if you missed my public apology to Fluke Networks, you should read it. Besides giving some backstory to this post, it’s not very often you will see me eat my words. Wait, nah, that’s not true, I do it all the time.
A quick synopsis. In an event hosted by Fluke, I asked the question “When are you going to release a Mac client?” The response I received struck a nerve, and while I do not remember exactly what was said, it was something like “Why would we ever do that?”
So, this post will lay out the many problems I have with the current version of Fluke Networks solutions for the wireless industry.
The expense – AirMagnet Survey, Spectrum XT, and WiFi Analyzer are expensive products! I realize that they are complicated to build and maintain, but the cost is exceptionally high.
I have spent plenty of money on professional level tools without complaint, and yet every time I spend money on “Yellow and Blue” I can expect to be yelled at by my finance person and bruised by the purchase process.
This is further exacerbated by the fact that many organizations simply won’t spend the money. I spent four years working for a major global manufacturer with hundreds of sites, and many thousands of AP’s and I could not get them to purchase AirMagnet. In my current role, there is still no budget for the software. That means I end up spending my own money for software. I could get over this, except:
The software is OLD! AirMagnet Survey Pro is especially old code. It is clear in so many ways that the software hasn’t been refactored in many many years. I made a harsh comment at WLPC regarding the “Walking Man” animation when performing surveys, but the comment stands. The little walking man is wasting CPU cycles on a laptop that is running on battery, and doesn’t even perform it’s primary function. I need a set of crosshairs to indicate where I stop, nothing more.
Spending money on AirMagnet feels like I am spending money on software that will be discontinued momentarily by a company that no longer cares about it.
Most importantly, AirMagnet only supports Microsoft Windows. I won’t rant about how much I despise, abhor, and generally hate Windows, really I won’t. A quick survey of their users would show Fluke that a surprisingly high number of users are Mac’s. We use Mac’s for many reasons and most of us only ever boot Windows to use AirMagnet products. Many have tried to use VM’s, and most have found issues with the USB sharing which makes it difficult to do our jobs. Even if we are able to make a VM work, we are now eating through our batteries much faster than we should be. If we are surveying a large facility, we are wasting our time, and our customers money waiting on devices to charge.
Based on conversations with Fluke during their session and after, I was given the impression (not told, simply given the impression) that Fluke is looking at an OptiView-like device as a future AirMagnet tool. I understand their thoughts. Control both the hardware and software, and you have fine grain control to make the most of the solution. I can only hope that Fluke Networks hears and understood from the feedback they received at WLPC that we want a single device for ALL of our work. We have that device, our laptop. We do not need a uni-tasker to drag through the airports and risk losing, stealing or breakage. The other concern with this solution is the OptiView is INSANELY expensive.
Fluke Networks repeatedly asked “What would you like to see?” to the audience. I’ve also spent some time thinking about that question. Here is what I hope could resolve many of the issues that are occurring on both Windows and potentially Mac and Linux clients.
Build the intelligence into a Docker App. That’s right, I’ll say it again. Build the intelligence to run on Docker. Immediately, you can now move the app to Windows, OS X, and Linux.
The most important code base can now be ran in a custom environment, easily reproduced on any piece of hardware thrown at it. Wait, you might say. Then I would have to install Docker on my laptop. How is that better than a VM?
I’ll answer that question in two ways. First, look at all of the redistributable apps that get installed along with AirMagnet. Imagine all of that going away. Just imagine…
Next, the resource utilization for Docker should be less than 1/4 of the utilization for a VM. Plus, there is no underlying windows OS to babysit. No updates to validate and install. No weird driver issues. No licensing issues do to a minor change in the VM.
Once the important code is running in Docker, build a GUI for each client that includes the hooks for the hardware (Spectrum Analyzers, USB NIC, etc.) and presents them to the Docker app in a standard consistent way. The GUI would include all of the OS customization, visuals, hardware hooks, but none of the intelligence.
I think I am most disappointed that Fluke Networks became comfortable as the market leader and chose not to push forward with new ideas. I have a ton of respect for the company and I own a lot of yellow and blue tools. Now that they have a serious market contender in Ekahau, I hope they take a serious look at their current situation, and choose to focus on the customer, rather than attempt to force the market to their will.