In the 18+ months since I stepped back into the world of geospatial consulting and services, one of the most striking things I have noticed is the seemingly increased tribalism in the geospatial space. Maybe I shouldn’t be surprised, given the increased tribalism of society in general, but that’s a different post that won’t get written by me.
What I mean by that is that I very quickly learned that I was perceived as an “open-source guy” by many. I heard variations of this characterization a lot in the first few weeks – “We don’t really do a lot with open-source” or “We’re looking to expand our open-source footprint.”
I’m fine with that association – the open-source world is a can-do community and who doesn’t want to be lumped in with that? But, given that I was finishing a seven-year stint at a proprietary SaaS platform, I found the quick association interesting. And, because I was an “open source guy,” I was, by default, “not an Esri guy.”
I’ll admit that I skew towards open-source in my personal projects, in addition to its prominence in my consulting work. It’s got such a low barrier to entry that I can chase pretty much any crazy idea I have. Naturally, I end up writing about that a lot here, so this content skews more toward open-source stuff. Also, I’m not sure if anyone’s noticed, but most of the major programming languages (Javascript, Python, R, Go, Rust) are open-source now. Hell, even .Net and C# are open-source these days. If you write code, you are an “open source person.”
But what does it make you if you use Python to customize ArcGIS? I submit that it makes you normal because most environments are heterogeneous – multi-vendor, proprietary and open-source, etc. If you use current darling Databricks, you’re using a heterogeneous source environment. If you use ArcGIS, you are as well.
I actually spend a good bit of time in ArcGIS each week, but nothing I’ve done there recently has been interesting enough to write about. That’s not a knock on ArcGIS, it’s just the nature of the work.
Regarding ArcGIS vice open-source, I think people can have reasonable concerns about heavy-handed marketing and customer-retention practices that obscure plusses and minuses of software and increase lock-in. They can have equally reasonable concerns for how to obtain sustainable lifecycle support for open-source tools in an organization that does not want to ramp up internal software capabilities.
In the former case, it can be hard to get past marketing-tech-generated content designed to emphasize the strongest features and the most successful use cases to figure out whether ArcGIS can really perform a specific job and, if so, how well. In the latter case, it can be a daunting task to survey the landscape and find the organizations that are providing lifecycle support for open-source software and it can be equally difficult to figure out how to procure that support. Technology of all kinds is getting more difficult and there is currently no panacea for that.
So, given all of that, I’ve been trying to position myself as a “user advocate” to help my customers navigate the landscape based on their needs. I can’t make Esri change their marketing, but I can do a decent job of interpreting it and getting customers to people who can answer their questions. To my knowledge, there’s not a central clearing house of organizations offering lifecycle support on various components of the open-source GIS stack and I am not inclined to make one. But I can connect my customers with those companies I know about so they can make informed decisions.
Ultimately, I’m interested in helping my customers find the best solutions to meet their needs. I am far more interested in the problem being solved than the tech being used to solve it. If proprietary offers the shortest path, then so be it. For example, recent experience has led me to believe that Esri’s native mobile SDKs are the strongest out there. If you are interested in building an app that uses spatial processing to solve a problem and you are not interested in doing a lot of customization of the spatial behaviors themselves, Esri will probably shorten your time to market. On the other hand, open-source solutions generally provide more efficient spatial data management than ArcGIS and they are iterating and adding advanced capabilities faster.
The beauty is that, in a world where reasonably good open standards exist, you can mix and match. I am finding increased interest in open-source among my customers and the customers of my customers. I can’t say if that’s a trend, but it’s what I’m seeing from my vantage point.
As I told someone the other day, I have been tilting at this heterogeneous-environment windmill in one way or another for a long time. Start here and work forward to see for yourself. Setting any tribalism aside, the one constant I’ve seen is that people with problems to solve will seek the best way to solve them, regardless of its source.
Header image credit: Acroterion, CC BY-SA 4.0 https://creativecommons.org/licenses/by-sa/4.0, via Wikimedia Commons
