David Haynes to GeoHipster: “Learn the back end…you’ll be viable forever.”

David Haynes
David Haynes

David Haynes II is an Assistant Professor with the Institute for Health Informatics at University of Minnesota-Twin Cities, and a health geographer who uses cutting-edge spatial analysis methods to advance knowledge of health and cancer disparities. 

David was interviewed for GeoHipster by Mike Dolbow.

Q: You got your undergrad degree in Biology, then a Master’s in GIS shortly after. I’ve met people who have taken all kinds of different roads to discover GIS, but I think a biology degree is a new one. So tell our readers, how did you get started in geospatial?

A: I went to a small liberal arts college in Cedar Rapids, Iowa, (Coe College). When I originally went to school I thought I was going to be a medical doctor, so I took a lot of biology courses. I was generally interested in human and environmental biology. It was in an environmental biology course that a professor offered me a summer research position. She had heard that I was good with computers and needed some help analyzing GPS locations. That was my first real experience with “GIS and Geography”. We were running Arc 3.1 and I started playing with AML / Avenue. It was a very cool experience and led to me going to St. Mary’s in Winona, MN for my master’s in GIS.

Q: You and I crossed paths by working with some mutual colleagues in the Health disciplines. We both know that where someone lives, commutes, and works can have huge impacts on their health, but sometimes I’ve found the area to be relatively slow to adopt spatial technologies. Have you found the same, and if so, do you feel like you’re constantly selling the business case?

A: So the medical field like every field goes in waves. 20+ years ago there was the idea of the environment being a factor for causing negative health outcomes. However, GIS was just at its beginning for the broader community and much of that literature showed that the scale of analysis was critical for determining that. So health went away from that to personal behaviors and thinking that was the main cause. We are at a point now, we know that if we control for many of these personal demographic characteristics (age, race, sex), we still see large gaps or disparities. We hypothesize that the environment could explain the disparities. However, the environment now needs to be more specifically defined which is causing a headache for everyone. Most researchers approach this from the traditional epi point of view and try to add spatial on at the end. So I spend much of my time on studies that want to add spatial to them. Not many studies start with spatial as the primary focus.

Q: It does seem like the field of ‘health geography’ is growing. Can you tell us what it’s like, in your experience?

A: Yes, I think the ability to use smart devices is going to make Geography extremely important in designing interventions. I am thinking of designing a study for smoking cessation that would send you SMS notifications if you are in a business that sells tobacco. 

Q: How would you describe “health geo-informatics”? Is this just another way to say spatial (or GIS)?

A: Yeah, pretty much. I think my focus is to make health spatial and to integrate more sophisticated spatial analyses that most researchers wouldn’t. 

Q: You were once a rugby player. I don’t know much about the sport, but it appears to me to be a lot like soccer and American football: where an awareness of space, angles, and boundaries is an advantage. Did you ever think of it like that? Ever map your games?

A: Funny, Yes. I want to teach a class one day call the “Geography of Sport”. Actually all sports are about Geography. You have a limited defined extent in which you have to operate. In many classic team sports (i.e., football, basketball, hockey, etc.), the goal for the defense is to limit the space and time for different players. If a QB is throwing the football they need time and the receivers make the pass easier by creating distance between themselves and the defenders. Staying with football, every defense seeks to limit potential areas of the field while leaving other areas open. The offense tries to exploit these areas. I could go on all day about this as I really enjoy watching sports from a geographic perspective. One last thing, this is why the prevent defense in football gets no credit. The prevent defense is to prevent a touchdown on the hail mary pass not prevent the Tom Brady 5-10 yard passes. Defensive Coordinators need to develop new defenses that use a mixture of man to man with zone to be more effective in the 2 minutes offense.

Q: In Minnesota, we have a saying that there are only two seasons: winter, and construction. But you’re into aquaponics and sustainable gardening. Isn’t there a gardening analogy to that, like, “weeding and canning”?

A: Yeah, aquaponics is a labor of annoyance. It always starts out great and then something breaks two weeks later. But mostly I’m dreaming farmer. I read the book “5 Acres & a Dream” and was hooked. One day, that’s what I want to do. I’d be more of a hobby farmer in the day and do some serious programming at night.

Q: What do you think of some of the indoor, urban, industrial agriculture systems that are cropping up, sometimes as CSAs? Given the rising environmental costs of shipping, and unpredictable climate we’re facing, this seems to me like something we need to invest in more as a society. But can aquaponics really save the world?

A: Realistically, I think it’s a part of the solution. It won’t save the world, but it would improve some things like the food supply chain. Every year there are 10 e-coli outbreaks related to vegetables. This isn’t going to change. But this might be an area where aquaponics could help. I have grand ideas of how aquaponics could be used to provide benefits to society. Mostly, I do it to help my kids understand where food comes from and what waste is. I like aquaponics because it recycles the water and is kind of a closed-loop system, although I feed the fish. But I think getting people familiar with the idea of how an ideal nature system could operate would help us. It is an opportunity to learn more about ecosystems and how plants, animals, and people can all interact in a sustainable way.

Q: I’m not sure how much you know about GeoHipster, but you’re a big PostGIS user in a nascent field – which makes you different than a lot of people I know, but more like some of the “geohipsters” I’ve met. So what do you think, might you be a geohipster?

A: After re-reading your definition, I would definitely be a geohipster. Geography and using GIS in the broader health world isn’t new, but it is really in demand right now and this trend will likely continue for at least 5-10 years. I’d also say, like many of the geohipsters you have interviewed that I am a big advocate for spatial and try to help people understand why spatial is important. 

I’d say I’m a big spatial database advocate. I tend to use a variety of tools that fit the need of the project. Which I think is the mantra of your GeoHipster definition poll. I think industries and the medical world tend to be OK with the installation of a validated commercial software. However, that can take weeks or months. They always seem to have a database around and if you can move data into that they seem to roll with it. Databases just seem safe. Plus I think the ability to scale analyses out in databases is easier than programming. But you need to know how to program if you’re working with big data.

Q: You’ve got a year or more of teaching under your belt. What career advice would you give to your students – or to our readers?

A: If you are into Geography that is great, and if you are just coming into GIS I feel sorry and excited for you. The GIS field is really changing, which can be daunting at first. When I came into Geography there was a general feeling that you could learn ArcGIS and get a job. You could learn a second proprietary software and be set for life. Programming was something you could do, but wasn’t necessary. That has all changed for me and any future students.

Google Maps, MapBox, OpenStreet Map, Uber, Lyft. etc have all changed this. We are truly embracing the big data and computational science revolution. This means that you need to have mid-level understanding of computer science. You need to know how to program in an Object Oriented Language for either front end or back end. I would tend to recommend students or new people to the field to learn the back-end over the front-end. Because there are a million web designers out there that can make a map better than you. They won’t know what they are doing. They won’t know what a coordinate system is and how it matters, but they can stick it together fast. The benefit of the back-end is that you will be viable forever if you apply your spatial analysis skill within a programming framework. Be flexible and adaptable to the programming platform. I write code in R, Python, SQL (PostGIS) and Scala. Some language may be your favorite, but keep your eyes open. One resource, I’ll point people to is Packt. They have a lot of good books that I’ve purchased online.

Making GeoHipster an independent business

When a door opens

Making GeoHipster an independent business

By Mike Dolbow, GeoHipster CEO

Mike Dolbow
Mike Dolbow

Someone somewhere, with a similar addiction to being busier than humanly possible, said that when a door opens, you should walk through it. In other words, when opportunity knocks, if you’re at all interested, you should pounce. I guess that’s what I was thinking about this time last year when Atanas Entchev reached out to the GeoHipster advisory board to see if anyone was interested in undertaking an effort to make GeoHipster a business independent from his previous ventures. I immediately said yes, and convened a hangout with several other board members to go over the options.

Fortunately for me, two other board members, Jonah Adkins and Amy Smith, also expressed interest in taking on new duties, and Atanas agreed to stay on once he knew he wouldn’t have to run the entire operation himself. It took a while for us to figure out the optimal formal business structure: a sole proprietorship LLC registered in Minnesota, which allows me to take over most operational and financial duties while the others focus on communications, editorial duties, and creative efforts. And yes, I fully realize and enjoy the irony that drips from the phrase, “CEO of GeoHipster, LLC”…and the fact that our fiscal year will start on Groundhog Day.

On the outside, however, very little will change about GeoHipster as a website and a collaborative effort. Our mission remains the same, we still rely on volunteer authors to help us generate content, and our editorial policy is unchanged. By undertaking this transition behind the scenes, we hope the result is a more sustainable GeoHipster, so we can continue interviewing interesting geohipsters from around the world, and our readers can learn from their experiences.

A few of my family members and colleagues have asked me why I decided to do this. Perhaps I was inspired by my good friend and fellow dad Justin Bell, who holds down a day job, plays in two bands, owns a side business, and teaches classes at night. I figure if he can make time for all those things plus family time, I can make time for something that I enjoy. And ever since that first interview I conducted with David Bitner, I’ve very much enjoyed my involvement with GeoHipster. It’s a major change of pace from my day job, a place where I can promote my tutorial on REST endpoints, and probably the only way I’ll ever be able to use a basin wrench as a metaphor.

Or maybe it’s all just a ploy to score another GeoHipster t-shirt. Might as well look stylish when walking through that door that just opened.

Michael Terner: “Choice is back”

Michael Terner
Michael Terner

Michael Terner has been working in the geo/GIS industry since 1985, initially in state government where he was the first manager of MassGIS. In 1991 he co-founded Applied Geographics/AppGeo, where he remains a partner and Executive Vice President.
Michael was interviewed for GeoHipster by Atanas Entchev, Mike Dolbow, and Randal Hale.

Q (Randal): So Michael, you are Executive VP at AppGeo. AppGeo has been around for around 25 years. You’re one of the founding partners, correct? What’s the history of AppGeo? ArcINFO was still command line at that point, and I’m pretty sure Windows NT hadn’t made a strong appearance in the market place. Plus I still had hair.

A: Yeah, I hate to admit it but I’m increasingly feeling like one of the “old guys” in this industry. I got my start in GIS in 1985, straight out of college when I got an internship with the Massachusetts state government in a small environmental agency. My task was to see what this new “GIS technology” was all about and see if it might help Mass with hazardous waste treatment facility siting. Long story short, that internship led to 7 years in state government where I had the privilege of helping to get MassGIS started, and was the first Manager from 1988-1991. In that time I took my ARC/INFO (correct spelling of the day) training on a Prime 9950 mini computer and ARC/INFO 3.2. Our first disk quota was 600MB, and the system administrator said “you’ll never fill that up.” We did in 3 months. I also helped Massachusetts buy its first copy of ARC/INFO to run on a new VAX computer at version 4.0. I have no nostalgia for the bad old days of command line, 9-track tapes, and needing to start projects by table-digitizing the data that you needed. I do miss AML a little bit.

I left state government to co-found AppGeo with two partners in 1991. My partner for 24 years, David Weaver, retired late last year. Our president Rich Grady joined us in 1994, and we’ve built a strong, internal management team. In hindsight, the one thing I think we’ve done best these past 25 years is anticipate and willingly change as technology evolves. We started AppGeo with one UNIX workstation that ran ARC/INFO for 3 people using terminal emulation on PCs connected to the workstation. We ran ArcInfo on Windows NT, and we’ve evolved through ArcView, ArcView IMS, ArcIMS and ArcGIS Server and ArcGIS for this and that. We’ve seen a lot of technology and devices come and go, and beginning in 2008 we began pivoting from Esri as the sole solution for all problems. Initially with open source, and now increasingly with newer web platforms from Google Maps to CartoDB alongside open source. Again, I have no nostalgia and have never had more fun in this industry than now. Choice is back, and innovation is flourishing. Everywhere. I still have some hair, but as my daughters remind me, my forehead has grown considerably since then.

Q (Randal): So what does AppGeo do?

A: We’re geospatial consultants, plain and simple. We help customers solve geospatial problems and we help them plan and implement geo. We both spec and create data. And we build a lot of applications. Nowadays, almost always on the web, and increasingly what we build is optimized for mobile device access. Sometimes our customers want our ideas; other times they need extra capacity, and sometimes they need special skills such as programming or project design. We really believe in “dogfooding” and “eating what you cook.” As such doing things like creating data and maps as well as applications helps us be more confident in the kinds of recommendations we put forth in our strategic plans. Now we also resell some technology, and we have our own software as a service (SaaS) offerings that we serve out of the cloud to many dozens of customers. Pretty much everything we do has geospatial in it, but as geospatial — or location — has gotten more mainstream, increasingly our work involves integrating with non-geospatial business systems or tying geospatial technology into traditional IT infrastructures. Which is good. As our development team will say: “spatial is not special, it’s just another column in the database.” It’s a bit of an exaggeration, but with SQL extended for spatial operations, not by much.

Q (Randal): One of the things I’ve noticed about AppGeo is that you have several business partnerships. Two of the most interesting were Esri and Google. One of the two decided they didn’t want to partner with you in the GIS realm anymore. Which one ran?

A: As a small business, we have always been open-minded to partnerships. In addition to geospatial supplier partnerships, we have partnered with a wide variety of other consultants on projects. From big engineering firms to well known IT consulting firms to firms that are highly specialized in a particular market such as airports. We add the geo/location expertise.

And we’ve always been very open to partnering with the geospatial software providers. In addition to the two you mention — Esri and Google — at various times (and to this day) we’ve been partners with Intergraph, Bentley, FME, and CartoDB. Our loyalty is to our customers, and we want the expertise needed to help them solve their problems, and we need to understand the tools that are out there very well to provide that expertise. Partnerships help us to do that.

So to your question: Esri kicked us out of their partner program after almost 20 years in the program. I blogged on that topic in 2014 and provided a good amount of detail on what we think happened, and what it means. Many, many people read the piece, and as I’ve traveled around, several other people have sought me out to tell me “their story.” We’re certainly not the only ones who have met this fate, but not many have talked about it openly. As I wrote, it was not a healthy partnership, and in the end it was good that it happened. We have never been stronger, and there are many other firms — including Google — that have welcomed us as a partner, and respected our non-denominational outlook on partnership. We still use a ton of Esri and feel very comfortable as a customer of theirs. As our clients know, our expertise in Esri didn’t disappear with our partner status. We greatly respect the company and Jack Dangermond as a strong and tough businessman. And, in our “best of breed” outlook on the geospatial landscape, Esri is the best at many things. But, in our opinion, not all. And that’s probably why we parted company.

Q (Mike): Any regrets about publicly airing all of those details? Do you think AppGeo would be different today if you had been able to stay in the program?

A: No regrets whatsoever. In fact, we have been a bit surprised at how many people were interested in the story. In the end, we had heard that Esri was telling the story to some of our mutual customers in “their terms”, and we felt it was important for people to have the ability to also hear that story directly from us. Quite honestly, I don’t think things would be much different for us if we had stayed in the Esri partner program. Business remains good, and we would still be using a variety of technologies, and we would still have our primary loyalty to our customers. Really, the biggest difference is in the posture of our relationship with Esri. Now we’re an Esri customer and user instead of a partner. And thus the kinds of conversations we have with Esri are somewhat different.

Q (Atanas): How is partnering with Google different than with Esri?

A: As you might expect, it’s an enormous difference. Google’s program is certainly not perfect, but Google is very clear with their partners on the role of the partner channel and Google’s expectations. It took some getting used to, but we have hit our stride and the partnership is very productive for us. Here are a few of the biggest differences:

  • Google has many, many fewer partners than Esri, and the partners are selected/recruited based on their qualifications. And there is not an annual fee to be in their partner program.
  • Google’s program is re-selling oriented. We do a lot of related services (e.g., application development), but that is between us and the customer; we work very closely with Google on providing the right subscription-based products. Unlike Esri, Google allows their partners to sell any of their geo products, not just the lower-end subset of products.
  • Google has other, non-geo product lines (e.g., Gmail/Google for Work; Google Cloud Platform; Search; etc.) and many of Google’s geo partners sell, or even specialize in these other product lines. Google’s partner conference (which I just attended in March) mixes all of these different partners, and it’s a really interesting and diverse ecosystem. There’s a specialized track for each product line (we followed the geo track), but you also get to see the whole cloud-based vision of the larger company and interact with, and learn from the non-geo partners.
  • Probably the biggest difference for us is that there is a very active exchange of leads and joint selling. We got more leads from Google in our first month in the program than we did in the entire life of our Esri partnership which spanned almost 20 years. Fundamentally, Google and their partners work together on sales which was not the case for us with Esri.

Q (Mike): While we’re drawing comparisons, you’ve been working with customers around the country. Are you noticing any regional differences in the way GIS or mapping technologies are approached?

A: Honestly, I don’t see fundamental “technological approach” differences across the country. Pretty much everywhere I go Esri remains the dominant player, but also I see people’s eyes and minds being ever more open to new approaches like open source (e.g., QGIS) or cloud-based platforms (e.g., CartoDB, Fulcrum). There may be slight regional differences in the rate of uptake of new technology, but everywhere people are more curious than I’ve ever seen. People are also increasingly interested in open data across the country, and even in Canada, which does not have the same public records laws and open records history as the USA.

The biggest regional differences are in governmental organization and the priority of particular issues. The things that vary on a regional basis are more like, “Do you work with more counties vs. cities/towns?” Or, “Is the drought, or agriculture, or public lands a big issue?”

Q (Mike): I’ll ask you what I asked you in Duluth last fall — can the open source community band together to make sure the Yankees never win another World Series?

A: Wish it were so. But as a fellow Red Sox fan I feel good about where we stand relative to the Yankees in the 21st Century, i.e., 3 titles Red Sox to 1 for the Yankees.

Q (Atanas): Hippest commute mode: ferry, train, or bike?

A: I’m a big public transit fan, mostly because the downtown Boston driving commute is terrible. Usually I’m on the commuter rail. But during the window from mid-May through the end of October there’s a commuter ferry from Salem (the neighboring city to my hometown of Beverly) into downtown Boston. So my favorite, and by association hippest, commute is the 1-2 days/week during the summer I get to bike the ~3 miles from home to Salem for a wonderful high speed ferry ride into Boston (and then back). This is the morning “entering Boston” view:

'Entering Boston' by Michael Terner
‘Entering Boston’ by Michael Terner

Q (Randal): So with all these questions behind us … do you feel geohipsterish? We did a poll way back in the beginning days of GeoHipster to define a geohipster, and the best we could come up with are they shun the mainstream, have a wicked sense of humor, and do things differently. Do you feel like one?

A: Yes, I hope so. I’m not sure I “shun” the mainstream, but I don’t believe there is a mainstream that lasts very long in technology. If you stand pat, you die. We’ve lasted 25 years so at a minimum we’ve bobbed in and out of the ever-changing tech mainstream fairly effectively. In 1985 when I started in this business, Esri was not mainstream. I appreciate humor (especially Randy’s) greatly, and I hope I’m occasionally funny (even if my family might disagree). Yes, I think we often approach things differently, and aren’t afraid to “try different”, and that’s been a great asset.

Q (Randal): I usually leave the last question up to you to say whatever you want to say to the world, and I’m going to do just that … BUT with a twist. Something big is coming to Boston in 2017, and you and the Geo community did a tremendous amount of work to make this happen. So what is coming to Boston in 2017?

A: Yes, the Global FOSS4G Conference is coming to Boston in August of 2017 as the three-continent rotation returns to North America after a successful stop in Seoul, Korea in 2015, and the upcoming conference in Bonn Germany in 2016.  See our nascent web-site to mark your calendars. The Boston geo community rallied, and I am extremely proud to have led our awesome Boston Location Organizing Committee (the BLOC) in generating the winning proposal to host that conference. We had incredibly tough competition with really strong proposals coming from both Ottawa and Philadelphia, and we are committed to putting on an awesome conference and rewarding the faith OSGeo has put in us. We are also excited to support the upcoming FOSS4G North America that will be held in Raleigh, NC in just 4 weeks. Please show your support for FOSS4G and learn lots and have fun with us in Raleigh.

The Special Tool

Mike Dolbow
Mike Dolbow

Mike Dolbow is a GIS Supervisor for the State of Minnesota and the operations manager of the Minnesota Geospatial Commons. He has served on the GeoHipster Advisory Board since 2014.

This past summer, I put a new set of stairs on the end of my deck. In the grand scheme of home improvement, it was a small job, but I’d never done anything like it before, so there was a lot of cursing, achy muscles, and extraneous trips to the hardware store. But that’s how I roll: chuck the manual and learn by doing. I’ve found that understanding your own learning style is a surprisingly underrated secret to success.

In a way, I learned that secret from my dad, who also made sure I had the skills to figure out how to get a job done with the tools available. He told cautionary tales about how my uncle would often dismiss a potential project because he thought “you need a special tool”, or it simply couldn’t be done. The way my dad saw it, that was a poor excuse for not doing the job. Sure, having a “special tool” would make it easier, maybe even faster. But if you had the will, you could find a way to get it done.

My uncle passed away years ago, but my dad and I still honor his memory with a running joke about “the special tool”. You see, now that he can afford some of those tools, he takes a sick pleasure in buying them for me. That way I don’t have any excuses when it comes to my own home improvement projects.

For example, the first Christmas I was a homeowner, he gave me a basin wrench. I looked at it and was like, “Dad, what the heck is this thing?” He replied, “Mike, that’s the special tool!” He proceeded to explain how important it would be in the upcoming faucet replacement jobs I had planned. And once I looked at the old, rusty supply-line nuts under the sinks of my 1915 St. Paul home, I knew he was right: the basin wrench would save me tons of time.

The original special tool: a basin wrench
The original special tool: a basin wrench

But if I hadn’t had it, I would have found a way.

Years later, he gave me his old carpenter’s square shown below. Could I have drawn that angle on the 2×4 without it? Sure. Did it save me time because I had it? Absolutely. And I’m sure there are also a bunch of hyper-specific tools that might have saved me time in adding those stairs. But if I had waited around to acquire every single special tool that would possibly aid in the process, I’d probably still be working on it now.

//platform.twitter.com/widgets.js

So, what does this mean in my day job at the intersection of IT and geospatial?

Well, if it’s not obvious by this point, I’m going to be wary of anyone who says they absolutely need software X,Y, or Z in order to get a job done. Listen, I’m going to tell you my requirements for a finished product. I’ll try to get you the tools you need to get there, but ultimately I don’t care how you get there, you have to figure it out.

With developers, I cringe if I hear them say stuff like, “I need a fully loaded Eclipse IDE with a local JBoss server and plugins for Git, Maven, Spring, and Hibernate pre-configured.” Well, for one, I barely know what some of those things are, but for the small web apps we build, I’m thinking you’re only going to need about half of them. Heck, if you really know what you’re doing, you ought to be able to write a decent app with Notepad++. But I simply can’t guarantee that you’re going to have the exact suite of tools you had at your last job, and you’re going to have to adjust.

On the geospatial side, I get concerned if someone says they need the full Esri stack to get anything done. “No, I don’t just need ArcGIS Desktop, I need the Advanced level. I also need an ArcGIS Server install with enterprise ArcSDE at 10.2 and an AGOL subscription.” Well, that adds up to some pretty hefty maintenance fees very quickly. If I just need a map of our office on our website, most of those tools are overkill.

//platform.twitter.com/widgets.js

Ultimately, I guess I can’t blame people for wanting to work with the best tools available: whether you’re talking Esri, Microsoft, Oracle, Mapbox, or Google, there’s some amazing stuff to work with out there in the IT and spatial worlds. Having access to so many amazing tools is a tremendous modern luxury. And if you’ve had a tool before, I know it can be hard to go without it. But sometimes, your favorite “special tool” won’t be available, and I might not be in a position to change that, even if the reason is simply “bureaucracy”. So let’s make sure we know enough about the problem to fix it with whatever we have at hand. After all, we’ll never run out of problems to solve.

Besides, the best tool you’ll ever have available is the one you’ve been using since you were born: your brain. So let’s put it to work, get the job done, and learn something along the way.

Maybe next time my dad will give you the basin wrench.